Social Complexity and Religion at Rome in the Second and First Centuries BCE by Andreas E. Bendlin Brasenose College, Oxford A thesis submitted for the Degree ofD. Phil, in the University of Oxford (Faculty of Literae Humaniores) Hilary Term 1998 < DEPOSITED o x Social Complexity and Religion at Rome in the Second and First Centuries BCE by Andreas E. Bendlin (Brasenose College, Oxford) Thesis for the Degree of D. Phil, in the Faculty of Literae Humaniores ABSTRACT This thesis contains 96,000 words. This thesis studies the religious system of the city of Rome and its immediate hinterland from the end of the Second Punic War to the emergence of autocratic rule shortly before the turn of the millennium. The Romans lacked a separate word for >religion<. Scholars therefore hold that modern notions of religion, due tc their Christianizing assumptions, cannot be applied to Romai religion, which consisted in public and social religious observance rather than in individual spirituality. The first chapter argues that Roman religion can be conceptualized a* a system of social religious behaviour and individual motivational processes. A comparative definition of >religion<, which transcends Christianizing assumptions, is proposed t< support this argument. In chapter two, modern interpretations of Roman reli gion, which view Republican religion as a >closed system< in which religion is undifferentiated from politics and froi public life, are criticized. It is argued that these inter pretations start from unwarranted preconceptions concernin< the interrelation of religion and society. Instead, I sug gest that we should apply the model of an >open system<: th religious system at Rome was interrelated with its environ ment, but at the same time it could be conceptualized a being differentiated from other realms of social activity a Rome. 11 Chapter three refutes the view that the identity of religion at Rome can be described by models of political or cultural identity. Instead, religious communication in Late Republican Rome was characterized by contextual rather than by substantive meanings. The fluidity of religious meaning in Late Republican Rome, a metropolis of nearly 1,000,000 inhabitants, implies that normative definitions of the constituents of Roman religion fail to convince. In relation to colon! ae and municipia it is attempted to show that the religious system of Rome, a local religion geared to the physical city and its immediate hinterland, was not capable of becoming a universal religion. In the fourth chapter, the parameters organizing Roman religion are discussed. My thesis is that Roman religion in the Late Republic was decentralized in that religious authority was diffused and religious responsibilities were divided. In the city of Rome, there existed a market of religious alternatives, which was characterized by the compatibility of different deities and cults in a polytheistic context. IV ACKNOWLEDGEMENTS It is my pleasure to acknowledge my manifold debts to various institutions, teachers, colleagues and friends. For financial assistance in 1993-6 I am grateful to the Trustees of the Rhodes Trust, Oxford. The Craven Committee supported a visit to the British School at Rome in September 1995. Two Oxford colleges, Corpus Christi and Brasenose (listed in chronological order), deserve my sincere gratitude not only for their financial support but also for granting me the privilege of membership of their respective treasured institutions. My debts to individual scholars are acknowledged in their place. C. Robert Phillips kindly shared his ideas with me and offered some valuable bibliographical information in the final stages. Antonia Barke read the whole thesis; her perceptive comments have vastly improved the final result. My greatest debt, however, is to Simon Price, who supervised this thesis in the best way possible, not just enduring but unflaggingly supporting its author. The sole responsibility for my stubborn refusal to take full advantage of his or indeed of other people's knowledge and for any residual errors or imperfections is mine. V CONTENTS CONVENTIONS AND ABBREVIATIONS vii .................... 1 ........... >Religio< and >religion< The modern notion of religion ......... ............. Religion and culture >£eligious behaviour< versus >belief< ..... Defining religion ............... Towards a comparative concept of religion ... 1 8 11 20 31 38 .......... 43 1 INTRODUCTION 7.7 7.2 7. 3 7.4 7.5 7.6 ............ 2 CONCEPTUALIZING »ROMAN RELIGION* ....... 43 Roman religion - local religion? Old paradigms ................. 51 New paradigms ................. 65 ...... 74 A Roman model of >civic religion<? 2.4.1 Varro and civic religion ......... 74 . 85 2.4.2 Priests as mediators of civic religion? 2.5 Ancient views on religion and society ..... 92 2.6 On the impact of modern theories of >religion and 103 .................. society< 103 2.6.7 Social theories of differentiation . . . 108 ............. 2.6.2 Functionalism 110 2.6.3 A symbol theory of culture ....... 114 . . 2.7 >Closed< versus >open< religious systems 115 2.7.7 Religion's role in a closed system . . . 125 2.7.2 The openness of Roman religion ..... 135 2.7.3 Orthopraxy or orthodoxy? ........ 141 2.7.4 Conceptualizing an open system ..... 2.8 Social complexity and religious differentiation: an 148 ................ introduction 150 . 2.8.1 »... me et Cottam esse et pontificem« 2.8.2 Variability and stability in religious 154 ............... behaviour 2.1 2.2 2.3 2.4 VI 3 >ROMAN RELIGIONc: PROBLEMS OF DEFINITION ..... 3.1 Rome as an >imagined community < ....... ....... 3.7.7 Ethnicity and citizenship 3.1.2 On the >Romanness< of Roman religion . . 3.1.3 Addressing the pantheon of the city of Rome 3.1.4 Priesthoods and aristocratic competition 3.2 Cultural self-consciousness and Roman religion ..... 3.2.1 >Omnis populi Romani religio< 3.2.2 >Romanus ritus< and related concepts . . ......... 3.3 Excursus: >Romana religio< . . 3. 4 Roman religion and Late Republican Italy 3.4.1 Intervention and laissez-faire ..... 3.4.2 Coloniae and municipia ......... 4 RELIGION AND SOCIETY IN LATE REPUBLICAN ROME ... 159 160 160 174 189 196 203 207 211 222 234 235 245 260 4.1 Organizing local religion at Rome ...... ........ 4.1.1 >Sacrum< and >publicum< 4.1.2 The diffusion of religious administrative ............... authority 4.1.3 Decentralizing financial responsibilities 4.1.4 The limits of state interference .... ............. 4.2 Public and private ........... 4.2.7 >Privata religio< 4.2.2 Roman religion, families and individuals 4.2.3 Collegiate associations and >publica sacra< 4.2.4 Roman religion: public or private? . . . .... 4.2.5 Public places, private concerns 260 261 ................. 4.3 Polytheism 4.3.1 »Si deus unus est ...« ......... 340 343 4.3.2 The constitution of belief systems at Rome 5 CONCLUSION 6 BIBLIOGRAPHY 273 281 293 299 301 303 310 323 333 353 .................... 363 ................... 368 Vll CONVENTIONS AND ABBREVIATIONS As this thesis focuses on the second and first centuries BCE, all dates are BCE, unless otherwise mentioned. If dates are explicitly noted as BCE, this is only to avoid confusion which might otherwise arise. Abbreviations of periodicals and works of reference follow those recommended in the Oxford Classical Dictionary3 , xxix-liv, and in the volumes of L r Annee Philologique. For Latin and Greek authors and their works, the abbreviations of the Oxford Classical Dictionary3 , supplemented where necessary by those in the Oxford Latin Dictionary and the Greek English Lexicon 9 by Liddell, Scott & Jones, are used. I have, however, diverged from these conventions in a few trivial and, I hope, self-explanatory cases. The following works throughout the thesis: ANRW FLP HrwG LFlav LIrn LTUR LUrs are cited by a short title Aufstieg und Niedergang der romischen Welt (Berlin & New York 1972-) E. Courtney (ed.), The fragmentary Latin poets (Oxford 1993) H. Cancik & al. (eds), Handbuch religionswissenschaftlicher Grundbegriffe, vols 1- (Stuttgart 1988-) Cf. Llrn Lex Irnitana, cited from J. Gonzalez (ed.), »The Lex Irnitana: a new copy of the Flavian municipal law«, JRS 76 (1986), 147-243 M. Steinby (ed.), Lexicon topographicum urbis Romae (Rome 1993-) Lex coloniae Genetivae luliae Ursonensis, cited from RS, no. 25, pp. 393-454 VI11 MRR RAC RE Roma medioreppublicana T. R. S. Broughton, The magistrates of the Roman Republic, 3 vols (New York 1951-86) Reallexikon fur Antike und Christentum (Stuttgart 1941-) A. Pauly, G. Wissowa & al. (eds), Real-Encyclopadie der Klassischen Alterthumswissenschaft, vols 1(Stuttgart 1893-) Roma medioreppublicana. Aspetti cultural! di Roma e del Lazio nei RS secoli iv e Hi a.c. (Rome 1973) M. H. Crawford (ed.), Roman statutes (BIOS Suppl. 64), 2 vols (London 1996) INTRODUCTION »sicut aequum est homines de potestate deorum, timide et pauca dicamus« Cicero, De imperio Cn. Pompei 47 1.1 >Beligio« and >religion« In the introductory pages to his The ancient economy, Moses Finley wrote that the ancients »lacked the concept of an >economy< 7 and, a fortiori, that they lacked the conceptual elements which together constitute what we call the >economy<«. Finley went on to describe the ancient economy as of a system low productivity and little market orientation which did not possess the economic rationality characteristic of ferentiated world-economies. 1 dif- Finley's argument, whose im- plications will receive critical discussion later, 2 has been chosen because it illustrates a more general dilemma. For students of religion in the Roman world find themselves in a position which is that methodological similar to Finley's there is no ancient concept of their area of research: the Romans, like the Greeks and, modern incidentally, other insight pre- and/or non-European cultures, lacked a separate word for >religion<. Yet, is it a necessary consequence of that that ancient cultures a fortiori lacked the ability to conceptualize the constituents, or indeed to address 1 2 in Pp. 17-34, at 21. See below, 1.3. the distinct category, of what we are used to calling a >reli- gion<? By way of introduction, the lack of an ancient lin- guistic concept of >religion< becomes all the more revealing when one considers that the Greek language, for instance, has a number of words at its disposal to circumscribe the different qualities of the sacred 3 or to denote proper human behaviour towards the gods. 4 Yet, none of these terms was used exclusively in relation to the divine, and none of them was supposed to summarily represent human attitudes, either individually or communally, towards the gods. In short, none of these terms could seriously be considered to be a Greek rendition of the term »religion«. The Greek material figures as an appropriate preface the Roman evidence. For Latin may possess a more restricted vocabulary for the description of the sacred, whose is to altogether different from its Greek equivalent. 5 Never- theless, the Latin language, too, terms to possesses a variety of express the manner in which due attention is paid to the divine world, 6 none of which can meaning meaning >religion<. lay claim to the This becomes apparent once we consider the employment of the Latin sacred vocabulary in those texts 3 4 5 6 'lepoe, dyvos, CXYLOQ, OCTLOQ with CONNOR 1988; DIHLE 1988, 2-16. SepeaSau, VOUL£ELV, euaEpeua, euXdpeLa, depaneia, ETILlieXeLa, dpriCTxeia with PLEKET 1981; BURKERT I985a, 268-75. Sacer, >consecrated to the gods<, to be distinguished from sanctus, >inviolable<, and religiosus: Aelius Gallus ap. Festus 348-50 L; Trebatius Testa ap. Macrob. Sat. 3,3,2-5; Gaius 2,4-8. Cf. FUGIER 1963, sub vocibus; SCHILLING 1979, 54-70; DIHLE 1988, 16-23; CRAWFORD 1989, 95-7. See further below, 4.1.1. E.g. religio(nes), pietas, sanctitas, caerimoniae, deos colere, sacra facere with SCHILLING 1979, 30-53; DURAND & SCHEID 1994. which, from the first century BCE onwards, define Roman re- ligious terminology. These texts, philosophical treatises in particular, which deal with the domain that we routinely associate with religion, are entitled De natura deorum, ITepu decov, IlepL et>ae3eiaQ or De pietate, ITepu De 6eLai&aLy.ovias superstitione. As these titles suggest, their authors do not investigate a terminus technicus the that could designate religious realm as a whole. For by received philosophi- cal dogma, the discussion of different aspects does not constitute a of religion confined subject of investigation. Instead, these aspects are assigned to the respective of or physics and areas ethics. Accordingly, ancient authors start from the exploration of the existence of the gods which none of the ancient philosophical schools denied. 7 It is the rather more exploration controversial second step, the of their nature, 8 which provides the parameters for the inquiry into proper and improper human behaviour towards them and, even more important, the ethical dimension of such conduct in human life. In ND, Cicero describes the study of the nature of the gods as an undertaking which only provides the but, following such pleasure not of intellectual investigation philosophical enlightenment, prepares individuals to conduct their religio. 9 Since only the virtu7 Cf. SCHIAN 1973; FASCIANO 1982. 8 Cic. ND 2,12: quales sint varium est f esse nemo negat. E.g. LONG & SEDLEY 1987, 1,139-49 (Epicureanism) and 1,323-33 (Stoicism), for testimonia. Cf. BEAUJEU 1969; GUILLAUMONT 1989. 9 ND 1,1: quaestio ... de natura deorum, quae et ad cognitionem animi pulcherrima est et ad moderandam religionem necessaria. It is in this educational context that ND forms part of a broader inquiry which also comprises the treatises De divinatione and De fato; Cic. Div. 2,3. ous can properly conduct their lives, it should not surprise us that elsewhere Cicero presents religio as a virtue (In- vent. 2,53; cf. Part. Or. 78), just as in ND pious behaviour towards the gods, pietas, is a virtue subordinate to justice (ND 1,3-4). In the same dialogue, Cicero defines the subject of his discourse about the gods as follows (1,14) : ... quid de religione pietate sanctitate caerimoniis fide iure iurando, quid de tempi is delubris sacrificiisque sollemnibus, quid de ipsis auspiciis ... existimandum sit - haec enim omnia ad hanc de dis immortalibus quaestionem referenda sunt. This passage neatly illustrates that any one of these connotes just one particular aspect of due human behaviour towards the gods and thus covers only one domain that terms we fraction of the routinely associate with religion. None of these terms, however, provides a terminus technicus for that realm itself. Instead, pietas is explained as versum deos; iustitia ad- sanctitas stands for scientia colendorum dec- rum; religio is defined as cultus deorim?. 10 Furthermore, the principle on the basis of religio and sanctitas, which Cicero assembles amongst all the other terms in his list, appears to be merely additive: 11 there is no tive pietas, qualita- relation between them which would allow us to take any one of them as being more privileged and thus coming closer to a rendition of the concept of >religion<. Such a restrictive use of Roman religious terminology is not limited to the Ciceronian corpus. In fact, words like religio or pietas 10 ND 1,116-7, 2,9. 11 For the merely additive assemblage of these terms, see further ND 1,3-4, 3,5-6. This is the reason why they can be replaced by cultus f honores, preces in ND 1,3. are employed in a similarly restrictive way by other Roman authors of the Republican and Imperial periods. 12 Notwithstanding this evidence, historians of Roman religion have invariably tried to distill the essence religion from has Roman words like religio, pietas or superstitio. 13 Religio in particular, with its various gies, of been used contested etymolo- as though it represented an adequate terminus technicus for >religion< in a modern sense. Defined in such terms, the philological analysis of the word religio has been taken to provide both an insight into the religious mentality of the Romans and a comparative perspective con- cerning the differences between a Roman and a modern conceptualization of religion. For instance, religio, the cultus deorum according to the Ciceronian signify the scrupulous definition, performance can indeed of ritual acts, while superstitio, according to the definition routinely provided by the same ethical treatises, denotes the excessive worship of the gods that unbalances the human soul. Therefore, Roman religion has been thought to represent a ritualistic system by means of which the purely legalistic, allegedly prosaic Romans 13 excesses, as these two words, religio and superstitio, could sum- marize the respective positive and 12 a or formalistic, relationship with their gods, while being dismissive of any emotional though formed negative embodiment of Religio: DSRRIE 1974; IRMSCHER 1994; SALEM 1994, demonstrating how Lucretius rather exceptionally parallels religio and superstitio on polemical grounds. Pietas: KOCH 1941; MURR 1948; DORRIE 1974. Cf. below, 2.2 and 2.3. Recent expressions of this view are too numerous to list in detail. See e.g. ROLOFF 1953; KOEP 1962, 45-6; WLOSOK 1970; MUTH 1978; DORRIE 1980; WAGENVOORT 1980, 223-56; LIND 1992, 5-15. religious behaviour. If any religious feeling was ascribed to to the Romans at all, it was a feeling attached orderly ritualistic performance. Such an interpretation succumbs is what a only of fraction to illusion the that the linguistic evidence can attitudes represent the definable essence of Roman towards or indeed is capable of sufficiently circumscrib- religion, ing the realm of religion itself. 14 In fact, the division of religio and superstitio in Roman elite writers of the Late Republican and Imperial periods 15 is in itself influenced by distinction the between euaepeia and 6eLaL6aiiiovi(X in Hel- lenistic philosophy. Stoic doctrine defined the latter as a of the emotion of cpopoe which would disturb the sub-species definition, composition of the human soul. By adopting this Latin literary accounts introduced the normative concepts of moral philosophy into the discussion about proper and im- proper religious behaviour. 16 Furthermore, authors were etymologies of religio was etymology that linked the correct. 17 word again in thought« (ND 2,72), while 14 15 16 17 as to which one of the undecided themselves Republican Late Cicero favoured an to relegere, »to go over Lucretius or Livy ex- 39-49; For a critique, cf. ULF 1982, 151-63; FEIL 1986, PHILLIPS 1986, 2698-9; DURAND & SCHEID 1994. E.g. Cic. ND 2,71; Varro RD fr. 47 Cardauns; Sen. Clem. 2,4,1, Ep. 123,16. Cf. SVF 3, 394, 408-11, for the Stoic definition of 6euaL&aLVLOVia. Latin adaptations: Cic. ND 1,117 timor inanis deorum; Varro RD fr. 47 Cardauns; Sen. Ep. 123,16; Sen. De superstit. frgs 36-8 Haase (furor); LAUSBERG 1989, 1888-98. For a discussion of the evidence, see PEASE on Cic. ND 2,72; LIEBERG 1974; LlND 1992, 7-10. Unfortunately, these authors too try to reconstruct >the true meaning< of Roman religion from their respective linguistic analyses . plained it with reference to religare, »to bind (sc. oneself to the gods)« (Lucr. 1,931; Livy 5,23,10). What matters more than a necessarily arbitrary decision on this issue is that the etymological meaning of the word religio, itself unclear to the native observer, does not give an how the essence of indication as to Roman religion was perceived by those Romans who used the word. The word itself retains an elusive character which is impossible to link to any substantive definition of religion; for Cicero, it sufficed elsewhere to define religio as iustitia ergo deos, thus again giving a strong ethical connotation to the performative aspect in due human behaviour towards the gods, while assimilating the at the same time term to pietas (Part. Or. 78). The inter- pretation of a native linguistic system does not, in any case, help us to define das Nesen of the religious system of the Romans. The issue is further complicated by the fact that Cicero elsewhere, in the context of political theory, demanded ternal involvement in religious ceremonial. His understand- ing of proper religious behaviour, informed by the ideas moral philosophy outlined feeling elite in- on the other. 18 Those scholars who regard such passages as attempts on the part of a cated of above, penetrates the exclusive dichotomy between ritual observance on the one hand and ner in- to replace emotional involvement too philosophically edu- a Roman externalized formalism by readily accept the traditional dichotomy of externalized ritual behaviour in Roman religion 18 Cic. Leg. 2,19: ad divos adeunto caste, pietatem adhibento, opes amovento with Cicero's own commentary, ibid. 2,24-5. 8 and true religious commitment. 19 I shall return to this di- chotomy in the course of my argument. However, it noting worth that this dichotomy, absent as it is from Ciceronian political theory, is already clearly for is whom a expressed by Seneca, moral life guided by Stoic doctrine alone suf- fices to please the gods. The external practice of religious cult in general thus becomes superfluous warranted by the obligation customs of society. 20 This and can only be to comply with the norms and attack on the foundations of ritual behaviour in private and in public cult was applauded by Augustine, who thought that Seneca's critique of reli- gious practice was far more radical than Varro's critique of the theology of poets. 21 7. 2 The modern notion of religion The very word »religion«, used to denote a separate sub-system in modern society, stems from the Latin religio and be found in all can European languages at a relatively early date. French religion occurs as early as the eleventh century. Religion in twelfth-century England is a straightforward adaptation from the French. In Germany, where a native 19 20 21 ter- E.g. DGRRIE 1978a, esp. 247: »Einsicht in die Innerlichkeit«. Sen. Ep. 94,22, 95,50, 110,1; De superstit. frgs 38-9 Haase. Cf. ATTRIDGE 1978, 66-9; MAZZOLI 1984, esp. 978-9; LAUSBERG 1989, 1895-7. Even Christian apologists like Tertullian (Apol. 28,1) were ready to concede that pagans showed personal commitment in ritual; it is modern scholarship that construed a dualism of pagan cult routine and Christian faith; cf. P. STOCKMEIER, ANRW 2,23,2 (1980), 875, 888ff.; STROBEL 1993, 336-8. August. CD 6,10 (p. 269,11-3 D-K). See 2.4 below, for the thesis that Varro's theologia civilis in fact anticipates Seneca's critique. minology came into existence only after the Reformation, Religion is documented since the sixteenth century. In this period, however, despite interpretation by the intellectual re- Medieval scholastics and Renaissance hu- manists, religio - and its modern cognate religion - did not yet signify the same as »religion« in stead, it encompassed society. In- such diverse concepts as »religious service«, »religious behaviour* or, word modern rarely, »belief«. The complemented, rather than superseded, terms like fides or lex. Religio thus reflects, via the early Christian terminology of the Church, the indeterminacy which can be en- countered in the ancient pagan vocabulary. 22 The development of »religion« into a terminus technicus culminated, it seems, in the age of the enlightenment. While the humanists of the sixteenth and seventeenth centuries were only partially able to redefine »religion« as an cally informed belief system - a meaning which anticipates the word's modern connotations -, it was tradition which ethi- the enlightenment finally succeeded in presenting »religion« as a culturally and historically distinct system in its right. The reasons for this are too numerous to discuss in detail here. Suffice it to say that eighteenth tellectuals such as David 22 23 24 century in- Hume 23 or Immanuel Kant 24 draw their lessons from the discoveries made by tions. own earlier genera- Despite ecclesiastical opposition, these discoveries For documentation, see SMITH 1962, 15-50; FEIL 1986; RUDOLPH 1994, 131-4. Cf. The Natural History of Religion (1757); Dialogues concerning Natural Religion (1779). E.g., Die Religion innerhalb der Grenzen der bloBen Vernunft (1793; 2 1794). 10 documented, once and for all, the existence of and religious allegedly primitive - cultures that were different from the form of religion prevalent in contemporary Western civilization. At the same time, the differentiation of contemporary society, with the decreasing impact tradition of ecclesiastical as a side-effect, necessitated the - often apolo- getic - redefinition of the boundaries between the sacred and the secular and the formation of some new kind of »natural religion« on the basis of contemporary rationality and ethics, just as society as a whole developed an increasing reflexivity in relation to all matters religious. The next and, in our context, most influential step, however, was taken by Friedrich Schleiermacher at the beginning of the nineteenth century. In an unhappy marriage of German Protestantism and the Romantic age, he defined gion as a Gefuhl schlechthinniger Abhangigkeit (»feeling of absolute dependency*). Hume, Kant and had reli- their contemporaries defined religion as but one fragmented part of society. Now, by introducing the subjective and existentialist component of »feeling«, Schleiermacher reduced religion to a predominantly individual spiritual experience of the divine. Religion is, for the first time, exclusively presented as an internalized belief system detached from any religious in- stitutionalization. 25 This concept of religion theology but nineteenth and 25 not only informs Protestant also influenced historians of religion in the early twentieth centuries. For Schleier- On Schleiermacher's concept of religion, see NOVAK 1986. 11 macher's definition of religion as an internal feeling is easily compatible with the Judaeo-Christian notion of faith and salvation on a purely individualistic basis. This is one prominent reason why Schleiermacher's definition, once the eighteenth century distinction between culture and religion had been internalized, was readily adapted to their needs by theologians and historians alike. Schleiermacher's concept of religion, however, has also been adopted by who sociologists define the essence of religion as feeling. Rudolf Otto, for instance, portrays religion as a sense of >the holy< >the numinous< experience. which he views as central to any religious Transcending Schleiermacher's or definition its of Romanticizing religion belief system thus appears to have achieved origins, as an individual wider accept- ance. His notion nevertheless remains Eurocentric and at the same time draws on a Judaeo-Christian tradition as presented since the nineteenth century. 26 7. 3 Religion and culture Obviously, such a notion of religion, being Eurocentric informed by Christianizing assumptions, is not capable of providing a truly comparative perspective. has shown that cultures, a experience not conform to notion are perceived as deficient in that they seem to lack what Western culture 26 Past both so-called primitive and/or non-European ones, where the evidence does such and would define as »the reli- On the impact of Schleiermacher's concept of religion on religious studies in the nineteenth and twentieth centuries, cf. FLASCHE 1989; ROPKE I996a, 241-6. 12 gious«. The failure of anthropologists and historians to apprehend societies where the religious realm is presented as a ritualistic system rather than an individualized belief system is well documented. For that reason, some historians of religion have proposed to dismiss altogether of >religion< as the notion a tool for comparative research. Drawing attention to the lack of a distinction between the religious and the secular or civic domain in many pre-modern and European societies, Dario Sabbatucci advocates the abandon- ment of >religion< as an independent area its non- repatriation of research and into the realm of >culture<. 27 Sabbatucci thus proposes to undo the modern differentiation between culture and religion which took place, as outlined above, in the eighteenth century. His thesis is relevant to this argument there appears to generation far as of historians of and Roman religions. Jan Bremmer, for instance, seems to me very close to the implications of tion so be an inadvertent, though close, rapport between Sabbatucci and a new Greek in when he Sabbatucci's posi- talks about the Greek TioXie in the Classical period: »Whereas most Western countries have gradually separated church and state, the example of other societies, such as Iran and Saudi-Arabia, shows that this is not so everywhere. In ancient Greece, too, religion was totally embedded in society - no sphere of life lacked a religious aspect ... Indeed, religion was such an integrated part of Greek life that the Greeks lacked a separate word for >religion< ... Embeddedness went together with the virtual absence of private religion ... cult was 27 SABBATUCCI 1988, 57: »Die Geschichte der Forschung fiihrt mehr oder weniger of fen zur Auflosung des religiosen Spezifikums im kulturellen Allgemeinen, man spreche nun tatsachlich von Kultur oder von Gesellschaft oder von kollektiver Mentalitat«. 13 always a public, communal activity ... This public character also meant that religion was strongly tied up with social and political conditions ... Embeddedness also influenced the conceptualization of the sacred. In modern Western society the sacred is limited to a direct connection with the supernatural and sharply separated from the profane, but the situation was rather different in Greece.« 28 The concept of >embeddedness< allows Bremmer to construe undifferentiated picture of Classical Greek society, where religion is indistinguishable term an from what Sabbatucci would >culture<; Bremmer prefers to call this realm >state<, >society<, >life< or >public and communal activity< without discriminating further. >Embeddedness< is also the key-concept of account Mary Beard's of Late Republican Rome. Instead of using Sabbatuc- ci f s >culture< or Bremmer's >society<, Beard employs tics< 7 >military activity<, >poli- >public life< and >the city<, again without apparent distinction: »Roman religion had its centre in politics, military activity and public life ... Religion was not principally concerned with private morality, ethics or the conduct of the individual Roman citizen ... The more fundamental cultural alienness of Roman religion lies in the degree to which it was undifferentiated from the political sphere. In modern world religions there is frequently considerable influence, in both directions, between religion and politics; but they remain separable and usually separate (if interacting) spheres of activity. In Rome, by contrast, ... religion, as in many traditional societies, was a deeply embedded element within public life, hardly separated as a separate sphere of activity or intellectual interest until the very end of the Republic.* 29 These passages are representative of the comnunis They 28 29 30 had to opinio. 30 be cited in detail in order to illustrate the BREMMER 1994, 2-3. For similar sentiments, see e.g. ZAIDMAN & SCHMITT PANTEL 1992, 92-101; PARKER 1996, 5-7. BEARD I994a, 729, 734. For comprehensive documentation, see below, 2.3, passim. For critique, see below, 2.5 and 2.6. 14 dilemma which appears to lie at the very heart of rent the cur- debate about religion in Graeco-Roman city-states. For as a result of the approach shared by these they dismiss three authors, what they rightly consider a modern construc- tion of >religion<. Instead, however, they propose to replace the modern differentiation between religion hand and on one culture, society or state on the other by a simi- larly problematic model in which religion is an tiated the and inseparable undifferen- aspect of that culture, society or state. Sabbatucci adopts this strategy because he intends to regain a comparative perspective for both pre-modern and modern cultures; Bremmer and Beard because they consider the modern notion of religion to be incompatible with the »fundamental alienness« of Graeco-Roman culture. Incidentally, the latter position faces an intrinsic difficulty in that the model of the embeddedness of religion in ancient society is not without problems. For both Bremmer and Beard implicitly acknowledge that the undifferentiated system they portray is unstable as it disintegrated, becomes increasingly once the differentiation of religion and so- ciety as separate spheres occurs. As both Bremmer and Beard make clear, this is exactly what happens in the post-Classical TioXuc in the former case and during the final years of the Roman Republic in the latter. In argument, both periods course of their authors employ a highly evaluative language which suggests that they those the perceive religious evolution in as homologous with the dissolution of tradi- 15 tional religious structures. 31 Religious evolution understood as a not easily allow Beard acknowledge the religious roles in for clashes Their exclusivity integration. Both Bremmer and importance of individualization of the course of societal evolution. Yet, according to these scholars the roles here struggle between ideal types of religious choice which are logically incompatible. does is individualization of such with the overriding concern for the unity of civic religion and are therefore marginalized. 32 For Bremmer and Beard, the embeddedness of religion in is prior, both ancient society historically and logically, to any form of differentiation. It is a necessary logical consequence of this model, from which its advocates cannot escape, that any subsequent religious change whose structure does not comply with that model's criteria can only be perceived as deterioration. 33 And it is a further consequence must appear to that any change be >subsequent< to this model's adherents, who thereby tacitly embrace a concept of cumulative and linear evolution which in itself is highly 31 32 33 problematic. As I BREMMER 1994, 91: »a ... movement away from the ordered public sphere«; 92-3: »religion as embedded in the polis had become religion as choice of differentiated groups*; 94: »the emphasis on public cult was shifting to private religious practices*. Cf. BEARD 1994a, 755-68. BREMMER 1994, 94: »the traditional structure was still fairly strong and would only slowly be transformed by new elements*. BEARD 1994, 761: »a threat to the traditional forms of religion at Rome, to the largely undifferentiated amalgam of religion and politics*; 763: »the clash between traditional and differentiated forms of religious organization and experience*. Bremmer's discussion of maenadism is illuminating in that he employs a traditional structuralist model of »order« and »antiorder*: the »temporary disorder* of maenadism serves to emphasize the subsequent reintegration of women and the restoration of the ordered structure of civic life; BREMMER 1994, 19-20, 78-80. For further documentation, see below, 2.3. BENDLIN 1997, 47-8. 16 shall argue below, religious differentiation in the ancient city-state can be conceptualized without the underlying no- tion of decline only when we seriously modify that model. 34 The strategy of replacing an admittedly problematic no- tion of »religion« with the category of >culture<, >society< or >state< appears to be methodologically unhelpful for yet another reason. Although one cannot deny that there are cultures, including Graeco-Roman antiquity, where religion society are in much closer contact than they might be in contemporary culture, there are nevertheless and spheres and certain ideas in any society where what one might reasonably call »religion« is clearly differentiated from other ideas and spheres. 35 Without anticipating my argument, the dissatisfaction with the model of the »embeddedness« of religion in society originates from the realization that the extremely vague concept of »culture«, »society« or »state« is unable to describe, let alone account processes in for, any dynamic the course of which some realms of life might become distinguishable from others. this model's vagueness the For instance, due to adherents of that model cannot explain why the realm commonly associated with religion, rather than any other realm or, for that matter, »culture« and »society« themselves, should stand out as the subject of societal communication in the Graeco-Roman world as undeniably did. Therefore, prominently 34 35 in it while perhaps rightly reacting against the radical differentiation of religion and society as ancient past scholarship, the model's adherents seem to See further below, 2.8. See below, 4.1.1 and passim. 17 have gone too far in replacing differentiation by a holistic concept of societal uniformity in the Graeco-Roman city-states. As we shall see below, a much more flexible gical methodolo- framework, rather than simplistic dichotomies, is re- quired to describe the status of the religious system in Graeco-Roman society. Finley's reluctance to primitivist concept accept anything other than a of the economy was mentioned above. It resulted from the assumption that the ancients' inability to conceptualize an economy a fortiori reflected the absence of a differentiated economic rationality. Yet, the level of rationality which Finley took as representative was based on modern world economics. Exposing Finley's dichotomy of primitive subsistence economy and market-oriented capitalism, economic historians have been able to reconceptualize a system which, while lacking the concept nonetheless characterized by a of the economy, was level of productivity and trade interest significant enough to qualify as reflective of a differentiated economy. 36 Economic historians have thus resisted Finley's primitivist strategy of de-differentiating the domains of economy and of society as a whole. I would suggest that the peculiar differentiation of economic roles enables us to draw an analogy. The ancient lack of a concept of religion must not result in the conclusion that the an- cients were not able to conceptualize 36 religion as a dif- Productivity: e.g. LOVE 1991, esp. 93-8; PURCELL 1994, 668-70; Id. 1995, 162-73; MORLEY 1996, 108-74; PARKINS (ed.) 1997. For the differentiation of economic roles and for upward mobility achieved through economic rather than political means, see e.g. RAUH 1986; Id. 1989; AuBERT 1994. See below, 2.6.1, 3.1.1 and passim. 18 ferentiated sub-system. Though no ancient linguistic category of »religion« as such exists, the existence of a reli- gious system which somehow resembles »religion« cannot modern categories of be ruled out on a priori grounds. Before turning to the revaluation of the religious system at Rome, the question to be answered is: how do we define the category of religion which our comparative analysis requires? It is no doubt apparent that for above the reasons the traditional Schleiermacherian concept of religion is methodologically unhelpful. Yet, rather the outlined notion of »religion« than as an umbrella-term altogether, various sociologists have tried to formulate religion which a concept it is advisable to or reconstitute, the notion of »religion« as an um- brella-term at the meta-level. Otherwise it would be sible of meets the demands of a comparative perspec- tive. For it seems also apparent that retain, dismissing to impos- formulate any comparative notion of »religion« at all. Of course, this means that one has to account for very different forms of »religion«, ranging from the clearly conceptualized to the barely recognizable, at the object-level. Retaining »religion« as an umbrella-term fulfils still an- other important heuristic function: constituting the category of »religion« at the meta-level amounts tool for the the providing a analysis of a sphere which would otherwise a priori disappear behind a diffuse concept of versely, to lack culture. Con- of any such category means a priori ab- andoning any possibility of conceptualizing religion as differentiated from other societal domains. Once conceptualized as an umbrella-term, the idea of a differentiated sub-system 19 of religion can thus be without making any investigated the the object-level prior assumptions as to whether such a differentiated religious Often at investigation domain might does indeed materialize. prove that this is not the case. 37 When investigating the relation between society and re- ligion, modern scholarship has realized the cultural bias of its traditional categories of analysis: the historical phe- nomenon which past scholarship had confidently labelled »religion« has under scrutiny turned out to be fundamentally different from the model applied; that model of »religion«, as we nowadays realize, was a culture-bound in a Schleiermacherian concept coined tradition. Moreover, historians of religions emphasize that at the object-level »religion«, far from being a stable entity, is open to change and adaptation to its social context. Therefore, more recent research has understandably concentrated on those constant factors which, at the meta-level, would render »religion« a trans-cultural phenomenon that can be analysed in - and despite - its different historical realizations. focuses on the question as to what demands of many This recent discussion are the methodological a trans-cultural definition of religion towards the end of the twentieth century. However, approaching their subject with a comparative perspective, religious studies touch on several concepts that are also relevant to religion in the Graeco-Roman world, such as the dichotomies between »external« and »internal« - or »public« and »private« - and the differentiation and integration of religion in society. 37 Cf. SEIWERT 1987; RUDOLPH 1994, 135-9. 20 7. 4 ^Religious behaviour* versus ^belief* culture through which they wider a of It has become an axiom that religions are part are rooted in concrete social axiom structures and constrained by social conditions. This locates religious activity within a broader context of sothe overcomes cial activity. Such an approach successfully traditional Schleiermacherian notion of religion as a transhistorical phenomenon that solely focuses on the individ- ual's emotional state. Yet, while rightly dismissing such a notion, scholars have gone too far in the opposite direction and either described religion's place in society as organi- zational, with an emphasis on the organization of religion, the role of religious functionaries or the concept of »civic religion* in society; or they treated it as merely functional, focusing on the correlation between religious authority and social control. On the traditional view, »belief« is an exclusively »internal« matter, notwithstanding the fact that one's religious behaviour follows »external«, organizational and functional, patterns; and the recent emphasis in religious studies on external aspects of religious behaviour is certainly explicable by the distrust of that traditional result, however, recent As view. a religious studies have shown less willingness to re-address the issue of religious experience. Instead, in order to overcome any Schleiermacherian tions, the dichotomy assump- between »religious behaviour« on the one hand and »religious experience* or spiritual »belief« on the other, between »the external* and »the internal*, has 21 been pursued in the opposite direction. To be sure, this dichotomy informs various competing sciences: models in the social for instance, when explaining individual and com- munal behaviour, historians of religion are rather disinclined to embrace the Weberian idea of ing« and »intuitive understand- get trapped in the ensuing hermeneutic problem of how to construe a truly cross-cultural category tion«. of »intui- Instead, they favour a structural history (»histoire des structures*) which focuses on quantitative tically and statis- inductive data. This inevitably leads to the inves- tigation of the behavioural aspect of religion in its objectifiable social context rather than the subjective of »religious experience*. category Arguably, this seems to be the only strategy, if we try to escape from the notorious her- meneutic circle outlined above. The functionalism, however, which informs such a model has obvious limitations: content as it is to describe institutionalized regularities in the correlation between reli- gion and society, it cannot account for any phenomenon that, like individual »experience« or »emotion«, may be immanent to a particular religious system without figuring prominently in societal communication about the relation between re- ligion and society. 38 Furthermore, while historians of religion for good reasons doubt the rightful application of a Christianizing concept of religious faith to other cultures, there is a tendency to abandon the category of »religious experience* altogether. To give a preliminary historians 38 of illustration, Roman religion pay close attention to ritual For a critique, see H. J. SCHNEIDER 1979. 22 behaviour. The only area in which these scholars have always dealt with emotional processes and belief more freely is the area of mystery or oriental religions. Concentrating on subjective impact oriental and the foreign cults made on the individual, they follow the traditional differentiation be- tween what is seen as proper behaviour in acknowledged religious acts and >true experience< in marginal religious sub- systems . Few would nowadays concur with the once fashionable view that the emotional nature of these marginal religious cults was to supersede the formalism inherent in traditional pagan religion and to pave the way for Christianity. Only few, on the other hand, when investigating traditional pa- ganism, would seriously disagree with the heuristic perspective common among ethnographers and students of religion that meaning ought to investigated in the public and collective sphere rather than in individual thought and emotion. 39 In fact, as we have seen in the respective cases of and Beard, many studies Bremmer of Graeco-Roman religion readily establish an a priori separation of emotion and cognition, and dichotomize (private) individual experience and (public) collective representation, with their investigation predominantly focusing on the domain of civic religious behav- iour. 40 39 E.g. GEERTZ 1973, 10-3: »culture, this acted document, ... is public ... Though ideational, it does not exist in someone's head.« See further below, 2.6.3. 40 Notable exceptions include LIEBESCHUETZ 1979, 39-54; PHILLIPS 1986, 2697-711; CANCIK 1994; WlSEMAN 1994, 49-53; ROPKE 1995a, 453-71, 593-628; FEENEY 1998. See further below, 2.3. 23 However, the underlying schematic separation of iour« and »experience« »behav- is questioned by recent studies in the field of psychology. Psychologists define the realms the »external« and the »internal« / the latter comprising motivational processes such as emotion or feeling cognition, not as separable tions. To psychologists, causes which these realms are not also unrelated could generate clearly distinguishable exterPsychology thus strongly that the difference between public »behaviour« and individualized »belief« should not be defined at of but but as interdependent opera- nalized or internalized effects. suggests of differentiated logical or the level motivational, »external« or »internal«, operations. 41 To be sure, such a model of behaviour does not make any a priori assumptions about the actual interrelation between »behaviour« on the one hand and vidual belief systems, between »external« still be and described emotion or cognition on the other, »internal« in behaviour« sumption that these and processes; these must relation to the constraints of the social system. But psychology »external indi- replaces the separation of »internal experience« by the as- entities are themselves inextricably linked by attributional mental processes. These studies have wide-ranging implications which to be need taken into account. Individual »belief« and societal »behaviour« are not categories that could ever be separated. Rather, internalization and externalization are but two ments in an on-going dialectical process of mutually influ- encing, of implicitly reaffirming 41 mo- Cf. HECKHAUSEN & WEINER 1980. or overtly questioning, 24 received perceptions and public behaviour; and it would be wrong to locate the perceptual sphere of emotion or tion in motiva- the individual's personal world, and to construe a contrasting realm of external behaviour. 42 This theory casts serious doubts on the assumption that general, and social behaviour in religious behaviour in particular, can be de- scribed as if it took place solely at the public level, or that religious behaviour can be distinguished from the individual agent's internal motivations. By implication, it also questions the assumption that emotion and individual belief in religion are internal realms that could ever be from external affairs. Taking the separate realm of motivational processes in that relational sense may help us to question the assumption that the very category of »emotion« and individual »belief« is but a modern theological construct which could not be applied to pre-modern societies. The implication of these psychological studies any living is system must be described as a complex synthesis of different >materials<. Social studies have adopted observations that and formulated analogous these models of societal behaviour: social systems, too, are constituted through several elements - symbols and myths, material resources, genetically determined programmes, emotions and feelings, communal interactions and institutionalizations. 43 quence, any definition of As a conse- the organization of society in general and of religion in particular has to take account of both the externalized and the internalized factors of social 42 Cf. BERGER & LUCKMANN 1966; BELL 1992; WHITE 1992. 43 Cf. BUHL 1987, 70; WORTHMAN 1992. 25 re- behaviour. If they are prepared to accept this analogy, studies ligious must address both the institutional realm religious and an internal, psychological, dimension, as the arguably reacts to a multitude of organisational and sphere reli- a individual issues. As regards religious evolution, system is thus related to several changing functional gious references, and to the developing needs of both institutions that sion religious conclu- the and individuals. 44 This insight should lead to studies must not focus exclusively on the functional aspect of religion, addressing issues such as priestly roles, »religion and politics* or »civic religion«. Such an approach must be complemented by investigation the interrelation between external religious behaviour into and the internalization of religious it the structures. Moreover, an important implication of the foregoing discussion is that individual belief systems do not simply mirror a public model or civic belief system, as the advocates of the civic when assume they acknowledge the existence of individual interac- beliefs only in this limited sense. For the mutual of the perceptual and the behavioural aspects of human tion life entails that beliefs can never be stable. It is highly fore questionable scholars argue, a linear and tween whether Such Scholars of one-directional simplistic ancient models religion which are capable of accounting 44 exists, as these there relation be- civic behaviour and the beliefs it engenders in indi- viduals. public there- »behaviour« need to be abandoned. ought to develop frameworks for the possibility that and internalized »belief« are organized Cf. KAUFMANN 1989, 59-69; WELKER 1992. 26 not by a relationship of structured dichotomy but by a »performative structure* whose meaning is itself shifting, as it internal is constrained by several different external and factors. 45 Here, the discussion returns to its point of departure. the very fact that the traditional notion of »religion« For is a modern creation has resulted rash the in conclusion it should under no circumstances be re-applied to pre- that non-European or modern Graeco-Roman religions religious in systems. particular currently doubt the relevance of such value-laden categories »belief« as analytical of Scholars as »religion« or tools: allegedly, personal commit- ment, moral value systems or spirituality in the modern sense were non-existent in the lives of ancient Greeks and private mans; worship was an expression of civic religion, albeit on a smaller scale, but did not provide distinct ligious Ro- re- biographies. To be sure, this must be understood as a justified reaction against earlier scholarship. For preceding generations of scholars had based their assumption that the Romans did not believe in their religion tionable modern on ques- the dichotomy between true belief, to be found in individual religious experience, and the externalized cult acts of Roman religion. 46 By contrast, this new generation of scholars attempts to overcome such a dichotomy by doubting the very 45 46 category validity of the of »belief«. By stressing religion's role as I shall return to this suggestion below, 2.8 and passim. For a selection of references and critique, cf. PHILLIPS 1986, 2697 56 . 27 being preservative of the community as a whole - a simplis- tic functionalist definition to which we will return shortly -, these scholars assume that it would be possible to unmask the traditional notion of »religion«, an internalized system of belief grounded in individual emotion and faith, as being specific to the Western world steeped in the Christian tra- dition. Such a notion of »belief« would not be applicable to char- Graeco-Roman culture. Instead, ancient belief systems acterized the individual's ideological commitment to the system of public religious behaviour: »To be euaepTiS . . . was to believe in the efficacy of the symbolic system that the city had established for the purpose of managing relations between gods and men, and to participate in it, moreover, in the most vigorously active manner possible. « 47 These scholars would certainly concur who, starting with Rodney Needham from the concept of the contextual nature of the meaning of words, questions »the received idea that th[e] verbal concept [of belief] corresponds to a distinct and natural capacity that is shared by all human beings«. 48 Unfortunately, the revaluation of »religion« and »be- lief« which results from such a critique of past scholarship is itself inconsistent. Despite the fact that the traditio- nal Schleiermacherian notion of religion is the eighteenth century, a creation of noone has to my knowledge doubted its validity, when it is applied to Christian belief systems in the Middle Ages. Apparently, it is only with 47 48 respect to ZAIDMAN & SCHMITT PANTEL 1992, 15 (my emphasis, AB), discussing the Classical Greek city-state; cf. BREMMER 1994. For a similar argument with reference to Roman Republican religion, see below, 2.3. NEEDHAM 1972, 191, explicitly approved of by PHILLIPS 1986, 2696; Id. 1997, 130 with 131 7 ; FEENEY 1998, 12. 28 cultures non-Christian that scholars routinely make the a priori assumption that those internalized motivational procthe esses commonly called »emotion« or »belief« must not be study of religion. This assumption provides the of object the the justification for their neglect of the internal and focus their of research on the external aspects of human to behaviour. The underlying idea, however, seems non-Christian societies like Graeco-Roman that be culture do not accord to religion such an emotionally charged place because by »belief« is in itself not an entity that could be shared all human beings. What these scholars do not realize is that doing in they, so, attack only one particular, namely the Schleiermacherian, notion of belief, which they their discourses. Rather than truly lib- direct to permit unwittingly erating themselves from a past intellectual tradition, operate in the they framework of that very tradition: for when no- reacting against the application of a Schleiermacherian of tion belief to Graeco-Roman antiquity, the underlying parameters of their evaluation remain those >Christianizing assumptions< which these scholars claim to have overcome. How can the crude dichotomy between a belief and the Schleiermacherian mere endorsement of public religion be re- solved? It is necessary to find a more objectifiable category of »belief« which does not entail any Christianizing un- dertones of faith or internalized religious dogma. Once such a disinterested notion is available, the neglect of purely individual motivational processes becomes lematic. even more prob- For with a view to contemporary society, noone but persistent opponents of psychology could reasonably deny the 29 relevance of internal motivational operations - our emotions our belief and general and cognitions - for our societal behaviour in systems in particular. When making the a priori assumption that with respect to religion Graeco-Roman in society one can disregard such emotional processes, scholars implicitly postulate that the ancient mind was fundamentally different: not only in that the ancient meaning of religion differed from ours but also because the ancients did a prio­ ri not conceptualize belief systems in the way we do. 49 The problem with such a view is that it results in the assertion the that world ancient culturally but also in terms proof different from ours not only was of psychology. onus The of for this view lies with these scholars. On methodolo- gical grounds, however, their view is just as problematic as the position they react against. For they replace the traditional »Christianizing assumptions* about the importance personal commitment in religion by the mere, if emphatic, assertion of its relative unimportance in the case of religion. Either of Roman position starts from preconceived percep- tions; either adheres to circular thinking which is inter- twined with its respective interpretative preconceptions. So far, 49 the response to this ideological battle has been the The a priori assumption of a profound difference between »the Romans«' and »our« religious behaviour dates back to the early nineteenth century. It was made explicit by Mommsen and Wissowa, both influenced by Hegelianism; cf. WISSOWA 1912, viii, cited below, 2.2. For a recent methodological justification of the neglect of >belief< in the study of Republican Roman religion, see NORTH 1989, 605-6: »The theoretical problem is whether the elements of religious life can be postulated a priori for any society, or whether they are different and specific in different cultural situations ... the Romans' religious experience was profoundly different from our own and ... it is impossible to postulate what elements it should or should not have contained.* 30 explana- repetitive assertion, rather than a methodological tion, of the lack of belief, spirituality, emotion or meta- physical systems in Roman religion. 50 At this point, an intermediate position might suggest itself. This position acknowledges that internal motivations form an important part in the process of constituting a so- cial system, yet argues that religious confronting studies and primitive societies cannot access other than pre-modern external data. This position thus redefines the focus of its interest as merely the external and objectifiable structural elements of religion and admits to the incompleteness of its results. 51 Yet again, one should ask whether such of providing a framework for other than external it data, and does not a priori exclude from its analysis the possibility of completeness. This discussion problem in- functionalist approach towards religion is capable herently whether an which clarifies the lies at the centre of religious studies: be- fore assessing our data, we require a definition of religion which does not a priori marginalize part of the evidence but is capable of providing an analytical tool which takes account into both the externalization and the internalization of religion in society. If such a definition reflected underlying assumptions on the of its framework and took into con- sideration the danger of circular thinking, it would be able 50 51 For further critique, see below, 2.3 and 2.6. E.g., with exemplary caution, GEERTZ 1973, 123: »[S]uch questions cannot even be asked, much less answered, within the self-imposed limitations of the scientific perspective*. PHILLIPS 1986, 2710-1 argues that one ought to abandon »belief« as a heuristic strategy because it is too value-laden: »It can only serve the cause of obscurantism to talk about >belief< and >empty cult acts< for the Roman world.« 31 to reintroduce to the discussion the categories of »emotion« and »belief« as operational analytical tools beyond any cultural over-determination. 52 7. 5 Defining religion No historian of religion could nowadays claim that ject is his ob- »the holy« or an »ultimate reality« (Rudolf Otto) , or define religious studies as the »recreation of religious experience« (Mircea Eliade). These are phenomenological models which, while providing religious surrogates for their creators, do not separate between the external observer and what is to be observed at the object-level. 53 In an attempt to overcome such overtly approaches, the discussion has focused instead on two com- peting definitions of religion. These are »functional« and respectively the the »substantive« approach. »Substantive« definitions discuss what religion is, definitions phenomenological concentrate on what whereas »functional« function religion has in society. The former explains religion in terms of its essence, the latter with a view to what it say, does. 54 Needless to these are ideal types: contemporary definitions of re- ligion routinely combine a »substantive« with a »functional« 52 Here I am in general agreement with the heuristic aims of the so-called »cognitive-processual archaeology*; cf. RENFREW 1994b, esp. 9-11. Unfortunately, Renfrew's own approach to defining religious behaviour in terms of cognition (RENFREW 1994a) is compromised on purely methodological grounds; see 1.5. 53 54 Cf. FLASCHE 1989. For discussion, cf. DOBBELAERE & 1988, 15-27. LAUWERS 1974; KEHRER 32 element. In fact, a combination of both can already be found tions, defini- these in the writings of the »founding-fathers« of Edward Burnett Tylor and 6mile Durkheim, respective- ly. 55 An investigation, however, of these facilitates the discussion ideal two types of their respective advantages and shortcomings. The main objection to functional definitions of religion has repeatedly been raised. 56 It runs as follows: any specification of religion's function in society may fit also other institutions or operations. While Durkheim, Malinowski or Evans-Pritchard viewed the function of religion as pro- viding a bond of social solidarity among the members of a community, their functionalist successors have modified this approach and stressed the more active role of ritual and ever, How- system. 57 belief in regulating and governing a social they have failed to solve the basic problem: religion may »integrate«, »interpret the world« to or, take Durk- heim' s own functional definition, 58 provide »a unified set of beliefs and practices relative to sacred things, that is to say, things set apart and forbidden, - beliefs and practices which unite one single moral community - all those who adhere to them.« but one can easily imagine non-religious institutions and operations, such as »tradition« or »morality«, that are able to achieve the same effect or fulfil a similar function; and it is no doubt one of the shortcomings of such a functionfunction for 55 For an account of Tylor's and Durkheim's positions, MORRIS 1987, 91-140. cf. 56 57 58 PENNER 1971-72 and LUKES 1975 provide critical accounts. E.g. RAPPAPORT 1971; GEERTZ 1973, 87-125. DURKHEIM 1964(/1912), 37. alist analysis that it focuses on religion's 33 collective rather than on the meaning religious structures does however, not render functional definitions obsolete. Yet, to save his argument the functionalist function a find objections, other. In meet to order functional definitions tend to define as the religious any equivalent which is capable of fulfilling same to have would which could be related exclusively to the religious sphere but not to any these objection, This action may have for the individual agents. function as religion. Theoretically, the scope of what is described as religious can thus be extended ad infinitum. dif- However, such a solution is problematic. For it is not ficult for the external observer to constitute several functional criteria of what ought to be regarded as fulfilling a religious function. Yet, such a definition appears seriously for under-determined several reasons. For example, the obnot ject level, the society under consideration, may apply the same functional criteria when describing what it regards as religious. As a matter of fact, societal communication very rarely defines religion Moreover, a functional is purely functional terms. definition which explains religion as, for instance, providing what in social cohesion, may specify merely an unintended consequence of religion's ex- istence in society. A functional definition of religion thus leaves a definitional gap which it cannot fill in by continuously adding potential substitutes. Durkheim's definition of religion accords the merely sacred a integrative function but does not conceptualize this definitional gap. A functional definition like Geertz's, on the other hand, altogether abandons any category which might 34 provide a more nuanced account of the religious domain's distinctiveness. 59 More recent functionalist approaches have realized this dilemma and attempted to fill in the defini- tional gap. Unfortunately, however, they have fallen back on unhelpful conceptualizations of this gap: according »reli- gious experience« a central place in their models, both R. A. Rappaport and Colin Renfrew reintroduce a concept that is in itself, as outlined above, culturally over-determined and does not appear to provide a comparative perspective. 60 The missing link, in the form of a substantive which characterization transcends the merely functional while accounting for a comparative perspective, can, it may seem, be provided a modified substantive by definition of religion. Melford E. Spiro defines »religious belief systems* as: »beliefs concerning the existence and attributes of [superhuman] beings and of the efficacy of certain types of behaviour (ritual for example) in influencing their relations with man.« 61 With reference to Tylor's original definition of »belief spiritual in beings«, Spiro seems to provide a useful distinc- tion between religious and non-religious belief systems. However, his approach proves unsuitable for a different reason. For his monothetic definition with a particular sub- 59 GEERTZ 1973, 90: »a religion is a system of symbols which acts to establish powerful, pervasive and longlasting moods and motivations in men by formulating conceptions of a general order of existence and clothing these conceptions with such an aura of factuality that the moods and motivations seem uniquely realistic.« For a critique, see HOFSTEE 1986. Elsewhere Geertz himself warned against the poverty of a schematic functionalist approach: GEERTZ 1973, 122-3. 60 RAPPAPORT 1971; RENFREW I994a, 47-51. This shortcoming seriously diminishes the laudable attempts of the socalled cognitive-processual archaeology of RENFREW 1994b. 61 SPIRO 1966, 98. 35 stantive content (»superhuman beings«) provides only a very narrow categorization of what counts as »religious«. Spire's definition assumes that what he differentiates as being re- ligious and non-religious, has the same meaning at ject level. In ob- a society which worships superhuman beings this may indeed be so, in which case Spiro would particularly the fine description of provide a that religious system's phenomenology. But such a description could certainly not be applied to a culture in which there is no such worship of superhuman beings. The methodological Whereas functional dilemma thus becomes apparent. definitions seem too vague, substantive definitions appear too specifically centered in one historical notion of religion to be truly comparative. 62 In other words, functional definitions may provide numerous necessary but no sufficient criteria, as too many phenomena are included which do not constitute what would commonly be called a »religion«. Substantive definitions, on the other hand, specify a sufficient but may not a necessary condition: they exclude several conditions that could equally define a religious system. Is there a character of response to the over-determined substantive definitions? To fill in the above mentioned definitional gap, a formulation could possibly conceived which at the same time be provides a sufficient condition without being over-determined. Consider, for in- stance, my following formulation: »Religious systems differ from other cultural systems in that the activities associated with their elements gen62 Cf. LUHMANN 1977, 83. 36 erate a certain validity which is traced back to permanent and absolute authorities*. 63 This definition comprises a necessary ty«) and a functional (»validi- sufficient substantive element (»permanent and absolute authorities*). However, even this latter definition can not entirely escape the hermeneutic circle. For the criteria which constitute the necessary functional condition (»validity«) and the sufficient substantive condition ^permanent and absolute authorities*) still owe their formula- tion to the beholder's intuition. For intuition remains arbiter the in the question of what constitutes a necessary and a sufficient condition of religion. For reasons of princi- ple, this methodological dilemma cannot have a solution, and any definition of religion can only be an approximation. Yet, since this dilemma is unsolvable, the criterion for a definition's usefulness must be its theoretical potential. This potential can only be exploited if implicit assumptions are made point where the latter definition a explicit. 64 justifies definition's It is at this its temporary existence. It has several advantages: firstly, it provides a sufficient condition in that the category of »absolute au- thorities* , in a comparative perspective, makes it harder to force a particular phenomenological categorization upon ligion. Secondly, re- once an over-determined phenomenological category has been abandoned, the notion of »validity« pro- 63 My formulation is modelled on, and expands, SEIWERT 1981, 85: »Religiose Nomisierungen beanspruchen Giiltigkeit, wobei als letzte Giiltigkeitsprinzipien absolute Autoritaten gelten.* 64 Cf. RENFREW 1994b, 10-1, on the methodological problems of circular thinking faced by »cognitive-processual archaeology* . 37 vides a necessary functional condition which, while refocusing on the function of religion in society, replaces the purely integrational functional analysis model. 65 »Validity« is a Durkheimian a function which can become opera- tional in the public and the definition of private spheres alike. This thereby overcomes the traditional dualism of ex- ternalization and internalization inherent in other func- tionalist definitions. Thirdly, this definition draws our attention to phenomena which have hitherto been dealt with as religious substi- tutes of institutionalized religious concepts. When applying this definition to contemporary society, its utility becomes at once evident. Religious institutions and ideas must cope with a massive secularisation in society in that institu- tionalized religious structures and churches become increasingly marginal; parallel to the devaluation of religious institutions, the privatisation of religious feeling increases. In consequence, there is not less religion in contemporary society, but it can no longer exclusively be institutions. society has to more than one theological construct and more than one mental operation which is seen as religious. Such a tion in This does not mean that religion forfeits its societal importance. However, contemporary offer found defini- therefore attempts to account for what has been label- led a new contemporary polytheism in the wake creasing cultural complexity of ever in- towards the end of the twen- tieth century. 66 However, its value for describing the poly- 65 66 Cf. SEIWERT 1981, 71-9. LEMERT 1974. 38 theistic nature of Graeco-Roman religion will also become apparent. For taking contemporary religious complexity as an analogous case, I shall argue that the modern emphasis on Republican Roman religion as an institutionalized system, in the form of either Georg Wissowa's >Staatskultus< or of new paradigm of >civic religion< embedded in the socio-po- litical structure, is a contrary, we modern misrepresentation. On the are dealing with what was essentially a dere- gulated system of religious behaviour. importance the lay in Religion's societal the fact that it penetrated both public and private areas in a way which these models cannot por- tray, since the evidence comprises many aspects which transcend simplistic categorization. Moreover, I shall argue that studies of Republican Roman religion need to revaluate religious complexity and the differentiation choices. Viewed along such lines, the wider religious meaning of religious dissipation of and the diffusion of religious authority in a polytheistic context call into question the validity of previous models and shed new light on religion's place in Roman society. 67 1.6 Towards a comparative concept of religion How does the potential of this definition being materialize when applied to concrete historical religious systems? The answer to this question entails a further ramification to my argument. For the problem of how to provide one particular definition of religion is unsolvable for yet another reason. 67 For further illustration, see below, 3.1 and 4. 39 For if the axiom holds true that a social system is consti- tuted through its normative sub-systems - sets of orienta- tion, rules, and guidelines - and if it is further true that religion is one of these normative sub-systems, then pro- viding a definition becomes almost impossible. For as is a correspondence between there religion and society, and as society is continuously constituted and altered through ligion and other normative elements alike, we should expect religion itself to undergo dynamic changes in the course this re- of process, adjusting itself to social realities and con- stantly altering its referential frame. 68 Accordingly, we should expect religion to have more than just one stable and unshifting function in society. In constantly changing so- cial systems, religion is always open to societal re-definition, both communally and, as emphasized above, individual- ly. Accordingly, a few functionalist approaches have rightly emphasized religion's multi-functionality and its adaptability to changing social constellations at both the communal and individual levels. 69 That is the very simplistic functional »superhuman as »ultimate realities* religion is perceived »superhuman of those beings«; reconstructed and redefined within a changing society, so is the referential frame for ception both beings«, are unsuccessful. Religion may in- deed focus on subjectively yet, why and substantive definitions of reli- gion, with religion being directed at or reason the per- »superhuman beings« in society. The more flexible category of »permanent and absolute 68 Cf. D6BERT 1973a, 75-82. 69 Cf. LUHMANN 1977, 84; KAUFMANN 1989, 59-69. authorities* 40 seems better suited to describing these entities at the me- ta-level. At the same time, religious studies must react against still another trend: the disentanglement of »religion« as an area of research and its repatriation into the realm of »culture«, »society« or »state«. The formulation that a ligious system is culturally distinct from other realms of society, suggested above, because it is linked back to manent and re- per- absolute authorities is an attempt to provide a theoretical frame which could respond to that trend. It is the function of religion »to provide types of haviour that are communally performable and relevant to the individual, types that are reconcilable with ties be- social reali- and adjustable to individual circumstances, types that are able to generate and communicate not only one integra- tional set of orientation but also several other individually decomposable systems of meaning.« This functional definition of religion, paraphrasing Michael Welker, 70 appears to provide a comparative perspective for a post-Schleiermacherian age. Indeed, once it is agreed that a comprises universally religious system shared norms and individual beliefs, and that it interpenetrates both the social world and the individual sphere, the traditional separation of »behaviour« and »emotion« has disappeared. This comparative perspective embodies two prominent features which ought not to be ignor70 WELKER 1992, 370: »universalisierbare und zugleich individuell performierbare Typiken bereitzustellen [und diese] Typiken, die hochintegrative Orientierungsanspriiche und zugleich individuell dekomponierbare, multiple Ordnungen generieren, zu erkennen und zu kommunizieren.« 41 ed by the student of religion: firstly, it no longer tomizes the categories of »external« and »internal« motiva- tional religious behaviour, but them dicho- enables us to understand as complementary elements of one and the same process. Secondly, it implies the notion of religion as a separate and individualized cultural entity that ought to be perceived as interrelated with, though at the same time differen- tiated from, society as a whole. Finally returning to the Graeco-Roman way religion's role city-states, in society is described by Bremmer or Beard is fundamentally incompatible with the model have proposed the which I in the foregoing discussion. The distinction between »external« and »internal«, or the dichotomy of »public« and »private«, as well as the issue of religious dif- ferentiation in society lie at the heart of the debate about the unity and subsequent dissolution of civic religion in the Graeco-Roman city-state. It might at first appear futile to apply the foregoing considerations to these issues; as will become yet, clear, their revaluation in the light of a comparative perspective seriously challenges the very of civic religion itself model and finally leads to a modified understanding of religion in Graeco-Roman antiquity in eral and in Late Republican Rome in particular. 71 Whereas many historians of Graeco-Roman react religion 71 still appear to against an obsolete Schleiermacherian notion of reli- gion, my argument is that, with a view to Late gen- Roman religion in the Republic, the categories of »external« and »in- I shall return to this discussion in more detail chs 2 and 4. below, 42 ternal«, or »public« and »private«, and the ferentiation of issue of the religious domain, along the lines of a comparative model of religion, provide new insights into old an discussion. With some justification, Roman religion may be characterized by a »fundamental cultural alienness«, it dif- exhibits, as I but shall try to prove, some aspects which might strike the beholder as not too unfamiliar. CONCEPTUALIZING ROMAN RELIGION »It is the nature of models that they are subject to constant adjustment, correction, modification or outright replacement ... The familiar fear of a priorism is misplaced: any hypothesis can be modified, adjusted or discarded when necessary. Without one, however, there can be no explanation; there can be only reportage and crude taxonomy, antiquarianism in its narrowest sense.« Moses Finley, Ancient history. Evidence and models (1985), 66 2.1 Roman religion - local religion? One of the most promising changes which have occurred in the study of Graeco-Roman religions in the last two decades lates re- to the attention that is being paid to the »localiza- tion« of religious structures and experiences. With particular reference to Greek religion, scholars have observed its microstructure - increasingly its pantheons and cults as well as myths at the level of the local religious rather than paying excessive attention to its Panhellenic macrostructure as communicated through the of the static one pantheon Homeric poems and the Hesiodic theological specula- tions, or as reconstrued on the sole basis of of system particular the evidence city-state. Instead, local histories of religion focus on the peculiar composition of a pantheon and the (un-)availability of cults and mythologies in a strictly 44 local context. They thereby investigate the which frastructure« »in- religious constrained concrete religious choices and experiences in the different poleis throughout the Hel- lenic world. 1 With regard to the religion of Rome, applicability the of the notion of local religion is hardly controversial. For religious the the physical city for Late of significance reaction Republican or Augustan audiences, witness Cicero's to Pompey's plans to abandon the city in 49: >non est< inquit >in parietibus res publica. < At in aris et focis. Livy 7 another Roman of non-Roman origin, would subsequently exploit the theme of the physical city imbued with religious and tradition by inhabited the gods in Camillus' speech against abandoning Rome in favour of the site of Veii. 2 deed, the topography of Rome could be conceived of as a sacalendar, cred landscape: it was inscribed in the religious whose In- festivals, with rituals in temples or theatres and processions throughout the city, were year after year embedded in the urban space. Moreover, this sacred landscape permanently by the positioning of its sanctua- constituted ries or the creation of visual templa. 3 Through his was links between its augural antiquarian research into Rome's res divinae (comprising the priesthoods, sacred places, the festive calendar, the rituals 1 and divinities of the city), E.g. HENRICHS 1987; MORA 1995; PARKER 1996. 2 Cic. Att. 7,11,3; Livy 5,51-5. Cf. AMPOLO NORTH 1995, 145 & ISO 49 . See below, 3.1.2. 3 CANCIK 1985-86. 1991, 117; 45 Varro intended to revaluate the meaning of Roman religion for his fellow-citizens within this topographical context. 4 On the other hand, in the the second period under and first centuries BCE, Rome had long outgrown the confines of a nuclear city-state. By tion, consideration, way of illustra- the reported census-figures of 70/69 amount to nearly 1,000,000 Roman citizens who got registered. In all hood, likeli- those figures do not represent the full number of Ro- man citizens of that time. Even so, they exceed the corn-recipients in 320,000 the city of Rome that are reported as a result of the Lex Clodia of 58, in the course of which free corn was distributed not only to the free-born and to freedmen but also to a large number of newly manumitted slaves, and possibly to migrants as well. Caesar in 46 reduced number to 150,000, while that 200,000 recipients are recorded under Augustus. This reduction was achieved possibly by stricting admission to the plebs frumentaria to the free- born only. 5 These figures strongly suggest tionally large re- that an excep- number of Roman citizens lived too far away from Rome to profit from the corn dole. In the ever expan- ding Roman Empire, the idea that the city of Rome could provide 4 5 a physical focus of identity would become increasingly Varro RD I frgs 2a and 3 Cardauns. For Varro's critical intention, see below, 2.4.1. Census of 70/69: BRUNT 1987, 91-9, 376-84. Corn dole under the Lex Clodia: MRR 2,196. Abuse in the 50s: Cic. Dom. 25; Cass. Dio 39,24; cf. Dion. Hal. Ant. Rom. 4,24,5. Reduction under Caesar: Suet. lul. 41,3; Cass. Dio 43,21,4. Cf. VlRLOUVET 1991, esp. 48-55; MORLEY 1996, 35-7. On grain supply and consumption in the city of Rome during the final two centuries of the Republic, see GARNSEY 1988, 191-217; Id. 1998, 226-52 on malnutrition. 46 unreal. 6 As a corollary, the idea that the history of religion, when defined Roman as the religion of Roman citizens, could be written solely in terms of one particular local religious system seems most problematic. Consider, for instance, the Tullii town of Cicerones from the Arpinum which had gained full Roman citizenship in 188, eighty-two years before the birth of the orator M. Tullius Cicero. Although the family had moved to mid-Nineties, Cicero continued to Rome was fifties. 7 aware of the problem that a municipalis had two patriae, one by nature or origin (naturae, ortu, loci), other, Rome, the by citizenship and law (civitatis, iuris). In his conception of the ideal state, expressed in De municipium, the the assert his relation to Arpinum as his germana patria as late as the late Cicero in while legibus, representing a realm of life in its own right, was subordinate to the political domination of Rome. 8 Yet, this concept of political philosophy appears not to have seriously devalued the importance Cicero attached to sacra of his family at Arpinum. The actual differentia- the tion between the sacra familiae on the one hand and the sa- cra publica of Rome or those of the res publica municipii of Arpinum (Cic. Fam. 13,11) leaves one wondering whether religious practice was as easy to pigeonhole as the ideal of political philosophy would demand. On the 6 7 8 contrary, it is Cf. CORNELL 1991, on the Roman city-state as an anachronism and an ideology. Leg. 2,3: hinc enim orti stirpe antiquissima sumus, hie sacra, hie genus, hie maiorum multa vestigia, set in Cicero's villa in Arpinum. For the Tullii Cicerones and Arpinum, see NICOLET 1967; DORRIE 1978b. Roman citizenship in 188: Livy 38,36,7ff. Leg. 2,5. Cf. Acad. Post. 9: the city of Rome as Cicero 's domus . 47 the distinctiveness and social significance of one's own sacra familiae which is emphasized in such a passage. 9 A new man in the Roman political elite like Cicero might be expected to minimize the differences between tural tradition and a and Transpadane in the first century, achieved in the aftermath of the respective enfranchisement of these regions, must ceal cul- Roman political identity. Yet, the political Romanisation of peninsular Italy Gaul local the continuing not con- cultural diversity of Late Republican Italy. In southern Italy as in Campania, local no constitution. doubt adopted a Roman Romanized administrative coexisted with Greek municipal structures in these indeed >Italianization<, achieved by a Romans, was mixed communities of population counter-balanced these communities, partly of by Romanization, Greeks, Italians Hellenization and the continuing appeal of Hellenism to increasingly fluid upper-classes. If the But magistracies, even if the latter may have become purely honorific in function. The or communities anything, of local elite culture in Southern Italy increased in the first century CE. Further to the north, Samnite art and architecture also betrayed the direct impact of 9 Greek culture. Yet, whereas Samnium preserved its own Leg. 2,5 with SHERWIN-WHITE 1973, 159-73; NICOLET 1976, 57-68. In Fam. 13,11, written in 46, Cicero expresses his particular interest in the financial and administrative upkeeping of Arpinum's sacra ... et sarta tecta aedium sacrarum locorumque communium. The year's local aediles responsible for that task were M. Caesius, M. Tullius M. f. Cicero and Q. Tullius Q. f. Cicero, elected at Cicero's recommendation. RAWSON 1991, 450-1 links the phrase const! tuendi municipi causa (ibid. 13,11,3) to the reorganization of municipal government in Italy under Caesarian legislation; on the latter, cf. SHERWINWHITE 1973, 167-72. 48 language and traditions well into the mid-first century BCE, the Marsi, while insurgent in the Social immediate influence of Roman religious Latin as early as the second century. By long War, endured the culture and used way of contrast, into the imperial period, the peoples of the Appenines served as a stock-example of undiluted rusticity and Italian traditionalism. 1 ° The implications of such cultural diversity are illus- trated by the Caecinae from Volaterrae. Their wealth derived from remote rural estates, but the family, supported by the Roman Servilii Isaurici, also had a domus in Rome. Cicero defended the elder Caecina in 69. The younger Caecina became a friend. Neither father nor son entered the Roman Senate, though the latter f s Principate. 11 The children held consulships under the younger Caecina pursued business in Asia instead. At the same time, however, he was a famous expert on the disciplina Etrusca - an expertise that, together with an interest in Etruscan antiquities, he had inherited from his father. It is noteworthy that he was not only an orator of some renown but that his literary production and his correspondence with Cicero were written in Latin - the very language which he used for his great work on the Etrusca. 12 In contrast, disciplina remote Volaterrae appears to have been relatively unaffected by the contemporary economic agricultural and changes in the Italian countryside. Field work 10 Central Italy: CRAWFORD 1981, 158 (Marsi); Id. 1996, 414-24. Southern Italy: LOMAS 1995. Italian elites as mediators of Hellenistic culture: WISEMAN 1987, 297-305. 11 R. SYME, JRS 56 (1966), 58-9. 12 Cic. Fam. 6,6,3. Business interests: ibid. 6,6,2, 8,2. For the correspondence, see Cic. Fam. 6,5-6,8 with HOHTI 1975. 49 in the Volaterran territory, directed by N. Terrenato but as unlike yet unpublished, suggests that the area in question, areas, neighbouring villa economy but was retained a traditional pre-Romanized At any rate, Volaterran funerary urns and Etrus- structure. can inscriptions are still century. significantly exposed to the not At attested mid-first the around home, families like the Caecinae no doubt preand religious served their traditional funerary practices, cultural customs and the Etruscan language. 13 Their Volaterran environment was in striking contrast with their life in Etruscan opposition against Caesar in the Civil War, for which he was subsequently exiled. With a view to context, Catherine Edwards has the led the city of Rome. Furthermore, the younger Caecina related a remarked on the conflicts which arose between different regions of Italy and Rome in the course of the Civil Wars of the first century and stressed Italian reservations which inform the works of a trium- viral poet like Propertius from Umbria. 14 religious In these cases, choices would have been even more differentiated, and the attribution of a »Roman« religious identity, based 13 For the family tombs at Volaterra, see J. P. OLESON, Latomus 33 (1974), 870-3. Cf. FRIER 1985, 4-20, esp. 18-20; RAWSON 1991, 289-323, esp. 296-9. CRAWFORD 1996, 424-30 gives a general account of the use of the Etruscan language and the preservation of Etruscan funerary practices into the final years of the Late Republic. For the abandonment of Italian local funerary practices by elite families transferring to Rome in response to the new political regime in the capital, starting in the 1962-67, early Augustan period, he cites DEGRASSI 3,155-72 on the Salvii of Ferentis who abandoned their family tomb and moved to Rome in 23 BCE. 14 EDWARDS 1996, 55-7. 50 on parameters developed with a view to the local religious system at Rome, even less appropriate. 15 Republican Italy in a history of Roman religion, which rou- tinely and unwittingly focuses on disproportionality between the capital. history Late similar the Republic, history reli- system of the city of Rome become the universal reli- gious gion of all Roman citizens? As will become evident below, is only one set of pressing questions; the category of >Roman religion< defies a straightforward classification the of that seeks to construe an analogy with its political history? To what extent could the local this polit- of the Late Republic has been written. 16 How- ever, how valid is a treatment of the religious the A the history of Rome and that of Italy has been addressed in relation to the way ical Late of These examples illustrate the precarious place basis of ethnic, on political or geographical criteria. >Roman religion< is a modern umbrella-term for a large num- ber of cults, religions and belief systems, whose real value seems to lie in its imprecision. Attempting to apprehend the >identity< of such a polytheistic system generates problems of various kinds, rather than providing comprehensive ans- wers . 1 7 15 For traces of local Italian religious customs under the Empire, see Marcus Aurelius ap. Pronto 4,4 (66-7 Naber = 60 van den Hout), discoursing about the survival of local religious tradition in the Hernican town of Anagnia in Latium. 16 MILLAR 1995, 237-8. 17 For further discussion, see below, 3.1.1, 3.4 and 4. 51 2.2 Old paradigms very In past scholarship, the history of Roman religion has been often the history of one particular local religious system, namely the city of Rome. Starting with J. Har- A. Die Religion der Romer, published in 1836, the ques- tung's tion of how to describe the elements »Wesen«, or »character« the constituted that »identity« of Rome's religion has never ceased to attract scholarly interest. To Hartung as to reconstruction his immediate successors, this meant the the of authentic religious feelings of the archaic inhabitants vanished times historic of the city of Rome. Rome's religion had in beneath indigenous ritualistic formalism, deprived of any cognitive significance and lacking any intellectual elaboration in terms of mythology, and had subsequently been diluted by the inflow of foreign deities and rites. 18 »alt- Georg Wissowa accepted Hartung's premise that the romische«, i.e. Romans' authentic, religion should be the recovered in the form in which it had existed prior internal to any deterioration or falsification through outside in- fluences. 19 To Wissowa as to his contemporaries, Roman religion was closely bound up with the needs of community. It was an tainted by contact with with ethnic beliefs Rome's archaic religion which would become that were irreconcilable the Romans' prosaic and legalistic character. This was a peculiar local religion that would lose its communal reli18 P. ix: »es ist ein alter Tempel von einem Uberbaue verhullt worden, sodann sind beide eingestiirzt, und wir haben nun die Trimmer des ersteren Gebaudes unter dem Schutte des zweiten hervorzugraben.« Cf. SCHEID 1987, 304-7; RUPKE 1997, 4-7. 19 WISSOWA 1912, 1-2. 52 gious identity, once it became exposed to the outside world. As such, it was unsuitable for an city-state. 20 expanding Wissowa owed this view to his teacher Mommsen, to whom Roman religion was a national religion whose character, ethnic and legalistic, reflected the legalistic Roman nation-state. 21 The ultimate goal of re- Wissowa's was the reconstruction of the Roman »Volksreligion«. search In the introduction to his Religion ( 1 1901, und the Romer der Kultus 2 1912), however, Wissowa wrote about the difficulty religion the of recovering reliable information concerning of of the foundations People of Rome; what the evidence allowed him to do was to present a fairly complete picture of kultus«. 22 »Staats- Roman The category of »Volk« and the idea that the re- ligion of a people is an immediate expression of »Volks- geist« goes back to categories first developed by Herder and Hegel. 23 Yet, unlike Hegel, who described Roman religion as utterly dreary idea of state (»geistlos«) and utilitarian, 24 religion is not entirely negative. Herder's and Hegel's respective accounts of the the nineteenth Wissowa f s national states of century idealized the state as an organized realization of the »Volksgeist« - a quality the Romans, ac- cording to Hegel, totally lacked. To Wissowa, state religion 8-10. 20 Cf. RUPKE 1997, 21 Cf. MOMMSEN 1907, 390, for a summary: »Die Religion des romischen Gemeinwesens ist ... wesentlich national und in der That nichts als die ideale Wiederspiegelung des Volksgefiihls, die Religiositat der in sacraler Form zu Tage tretende Patriotismus.« WISSOWA 1912, 7-10, 15: »das letzte Ziel der Forschung, von einer Betrachtung der romischen Staatsreligion vorzudringen zur Erkenntnis der italischen Volksreligion.« 22 23 MANUEL 1959, 291-304. 24 This view was adopted, among others, by Th. Mommsen. See SCHEID 1987, 316-20. 53 (»Staatskultus«) may perience and belief of gion«. But in this not a have true, respect provided the emotional exSchleiermacherian, »reli- it reflected Roman religious feeling and behaviour. In a truly Hegelian sense, therefore, »Staatskultus« / the religious idea manifestation of the Roman of the state, was an institutionalization, or codified abstraction, of the »Volksreligion«. 25 Wissowa accepted Theodor Mommsen's view of old Roman religion as an exterior and formalistic affair informed by a native sober legalism rather than a spiritual dimension. 26 In doing so, however, he tried to rehabilitate gious behaviour previous generations, perspectives which and mutually inves- were different from Wissowa f s, they still followed, as we have seen, different super- their cultural conceptions of »religion«, had sup- posedly been unable to grasp. 27 While other scholars tigated reli- as an expression of the »Volksgeist« whose distinctly prosaic character imposing Roman his incompatible suggestion native that conceptuali- zations lie behind ancient and modern religion. Distinguishing »religion« and »cult«, or »religion« 25 26 27 and »magic«, and Wissowa's ultimate goal of penetrating beyond this codified abstraction and searching for the religion of the Roman »Volk« has invariably been accepted. Cf. e.g. LATTE 1960, 11-3; KOVES-ZULAUF 1978, 189. Cf. ULF 1982, 145-63, on Mommsen's legalistic approach as underlying Wissowa f s conception of Roman religion. WISSOWA 1912, viii: »Wenn man ... an meiner Darstellung eine gewisse Verausserlichung der religiosen Vorstellungen und Formen aus dem Gesichtspunkte des ius pontificum oder eine wenig Sinn fur Religiositat verratende einseitig juristische Betrachtungsweise tadeln zu mussen meinte, so wird die Frage berechtigt sein, ob denn >Religiositat< wirklich ein vollig feststehender und fur alle Zeiten und Volker konstanter Begriff ist, und ob nicht, was man an dem Buche als Mangel rugte, vielmehr dem Gegenstande der Untersuchung zur Last fallt. 54 defining »religion« in purely Schleiermacherian terms meant that these scholars could categories for the now create supposedly suitable religious phenomena they wished to de- scribe with reference to Rome. At the same time, they disregard would cognitive categories such as »belief« or »subjec- tive feeling« with regard to Roman religion, since these categories were regarded as belonging to the domain of religiosity, and not to a system of ritual performance. Whatever their perspectives on Roman religion, to Wissowa as to his contemporaries, deeply concept influenced by a Schleiermacherian of Christian religion which stressed individualiza- tion, personal belief and redemption as religiosity, the fundamental the cultural system of ritual performance and formalism essentials of alienness of Roman was apparent. 28 From this perspective of Roman religion as a deficient stage in the religious evolution, it was only a small step to HerUsener's explicit teleological praeparatio Christiana. mann Starting from the common premises that religious was a passage from the concrete-primitive to the abstract, from animism or polytheism to monotheism, history of ancient the most he regarded the polytheistic religions as a discipline culminating in the study of the origins of theism, perception Christian mono- complete, and (supposedly) truest, reli- gious option available. 29 Only a few dissenting voices, prepared to be accused of applying misleading cultural 28 29 precon- For the cultural background which shaped the assumptions of these scholars, see PRICE 1984, 11-16; PHILLIPS 1986, 2697-711. See above, 1.4, for a methodological critique of the dualism of externalism and internalism. Cf. SCHEID 1987, 310-1; SCHLESIER 1995, 334-6, for critical discussion. MANUEL 1959, passim provides numerous illustrations of this view from the seventeenth and eighteenth centuries. 55 ceptions, made themselves Glaube der Hellenen (1931-2) , applicability of heard. Wilamowitz, stubbornly in his Der insisted on the an emphatically Romantic category of »be- lief« to Greek religion. 30 In his correspondence with Wissowa, he defended his position with regard to Roman religion as well. 31 Yet, the appreciation of Greek religion as a truer religion than the Roman embodiment of mere ritualism was commonly expressed in the nineteenth century. To sophers Romantic philo- like Herder, who had in mind the recently rediscov- ered Germanic mythology, myth was the quintessential expression of the character of a »Volk«; and ancient Greece, its mythology, with could be thought of as the original creator of such an idea. Rome suffered the fate of being denied such myths. Wissowa's notorious dictum that Rome lacked a logy merely mytho- radicalized the nineteenth century view of the Roman People as being too prosaic to be capable of possess- 30 HENRICHS 1985, 290-305. 31 Wilamowitz to Wissowa, 11 March 1896 (ed. in BERTOLINI 1978, 191): »0b man mit dem, was mir in Religion die Hauptsache ist, dem subjektiven Empfindungsleben, schon viel machen kann, ahne ich nicht: dazu miisste ich selbst intimeren Verkehr mit den Italikern haben.« Id., 31 December 1901, on arrival of the first edition of Wissowa's RuKdR (BERTOLINI 1978, 191-2): »Sie haben die religio der Romer behandelt, und die Religion, die subjektive Empfindung, bei Seite gelassen. Das ist durchaus berechtigt, zumal in dieser Behandlung. Es soil auch nicht gesagt werden, dass man davon wirklich etwas wiisste. Aber der Neugierige fragt danach, zumal wenn er diese Fragen aufzuwerfen pflegt - die die griechischen Mythologen ja auch meist fortlassen.« Id., 11 May 1912, on arrival of the second edition of RuKdR, (BERTOLINI 1978, 196) : »Wir diirfen gar nicht immer bei den Festen auf die Cotter den Hauptwert legen; ich betrachte die originalen Gotter iiberhaupt lediglich als Exponenten eines Gefiihls, das aus der Tiefe oder von aussen kommt ... Ich glaube, auch bei den Italikern.« 56 ing the mythological imagination of the Hellenes. 32 Instead, German philhellenism in the generation of Herder, Heyne, F. A. Wolf and Winckelmann idolized the stereotype of the lenic genius' originality, and Hel- their definitions of myth were pointedly Hellenocentric. As a corollary, the persona- lized conception of the divine, culminating in the figure of Zeus, as expressed in Greek myths, represented to their contemporaries an early stage in the passage to monotheism; the Romans, lacking myths, lacked the very notion of a monotheistic principle. 33 This disregard of all things Roman furthermore had a markedly nationalistic nineteenth century ramification. scholars in France and Italy always fo- stered a cultural identity which focused on Rome, but how failed For some- to produce studies on Roman religion that would exert influence on an international level. By way of con- trast, in Britain and Germany, the latter particularly after Humboldt's educational reforms, the professionalization of Classical Studies resulted in manuals which gave an influential though, due to the idolization of Greek culture, highly negative assessment of Rome and its religion. The negative view of Roman religion prevalent in nineteenth and twentieth century scholarship must be seen as an indirect result of the marked preference for Hellenic culture. Yet at the time, it same reflected the disregard, motivated by nationalist concerns, in which France and Italy were held by the acade- mic circles in Britain and Germany. 34 32 WISSOWA 1912, 9. Cf. PHILLIPS 1991a, 31-4. 33 Cf. MANUEL 1959, 291-304, esp. PHILLIPS 1991a, 143-8. 34 Cf. CANFORA 1989. 302-4. 149; GRAF 1993a, Hellenocentrism: 57 This implicit antagonism towards Roman culture plain why may ex- it was in Britain and Germany in particular that the theorems of nineteenth century anthropology were applied to Roman religion. Scholars in these countries replaced Wissowa's explanation for the lack of myth in Rome by a tivist primi- view, namely that Roman society had been representa- tive of a predeistic and animistic stage in religious evolution; the Romans, lacking the tion, had not known Greek mythological imagina- an anthropomorphic conception of the divine but worshipped impersonal spirits and numina, »Augenblicks- und Sondergotter« or merely functional entities Deubner, (L. H. Usener, K. Latte, W. Warde Fowler). At the same time, while the Cambridge Ritualists developed a paradigm of »ritual and myth« with reference to Greece, a related school was not established in the field of Roman religion. The supposed absence of a Roman mythology reduced the stage of ligious evolution re- which was thought to be visible in Roman religion to the status of an ethnologic »survival«. To be sure, an exclusively aniconic, and thus early abstract, phase in Roman religion, spanning the first 170 years from the foundation of the city, had already been postulated by M. Terentius Varro. This first phase was anthropomorphic conception of deities devoid of and not in need of complementary myths. Mythology was a later invention of poets. to was presentations and, the monumentalization of sacred architecture and the introduction phase the Over the following generations, there was a passage, first, to simple temples without iconic later, any of anthropomorphic images. This latter dated to the year 584/3, when Tarquinius Priscus 58 statues invited the Etruscan artist Vulca to fabricate clay of and Hercules. 35 Given the fragmented nature of luppiter the material record, this thesis is as difficult to prove as ter­ it is to disprove. Material finds corroborate that the ante minus for the existence of anthropomorphic divine im- ages in Rome, resembling the Hellenic pantheon, is the sixth facie prima evidence archaeological the century. Yet, although seems to support the Varronian chronology in this re- spect, it does not a priori exclude the representations thropomorphic of an- of possiblity divinities in Rome at an earlier period, particularly since the material infiltration before of early Latium by Hellenic culture had started long the sixth century. The underlying problem is that Varro as well as scholars Wissowa such as Varronian or the interpretation primitive authentic and in the followed this respect, assumed that an religious Roman existed culture to its contamination by external influences. Such na- prior tivistic models representing always view the »closed a evidence historical system« tentative reconstruction in an idealized past that may never the have existed. The case is further complicated by agenda behind as a stage in the cultural evolution which is al- ready in decline and focus on the of who primitivists, Varro's moral theory. His belief that there was a period when an intimate worship of the divine in its ab- stract form was not yet in need of, and thus not yet tainted by, the visual impact of cult statues relied on Hellenistic 35 Varro RD frgs 18, 38, 235 Cardauns; Plut. Num. 8,14; Pliny NH 35,155-7, the latter presumably depending on Varronian material. Myths: Varro RD frg. 19 Cardauns. 59 the adhered Varro moral philosophy. As we shall see later, Stoic theology that the gods existed in immaterial form only. As a consequence, he criticized any representation the to in divine or mythological form as deviation material Tak- from true religion, unworthy of the gods' real nature. bias this ing into account, it may be safer to admit that ours. On the contrary, the coexistence of iconic and aniconic worship at Rome in being historical aniconic of primacy problematic on these latter the times, understood as a survival by Varro and his modern suc- cessors, rather shows that any the complete more Varro's material evidence need not have been than of over methodological conception evolutionary of anthropomorphic worship is grounds. 36 In reaction to primitivist tendencies in scholarship, Franz Altheim, in his Griechische Gotter im alten Rom, pointed out as early as 1930 that the existence of Hellenic deities and religious ideas at Rome in the archaic and presupposes the prior early Republican conceptualisation terms of anthropomorphism, thus making any periods of the divine in primitivist as- sumptions untenable. Later generations, supplying additional material evidence for the »Hellenization« of early Latium that can be traced back into the eighth century, ther corroborated intertwined with the cultural as well as religious evolution in the Mediterranean world as a whole from the 36 fur- Altheim 1 s criticism, and illustrated how the rise of the Roman city-state was socio-political, have eighth century See below, 2.4.1. TAYLOR 1931, accepting Varro*s thesis, interprets aniconic representations in Late Republican Rome as survivals of an early aniconic phase. Contra ULF 1982, 158-9; METZLER 1985-86, providing comparative material on the simultaneousness of iconic and aniconic representations. 60 onwards. As a result, the notion of Roman primitivism has rightly been replaced with a model which perceives of gious evolution reli- in Roman religion from the seventh century onwards in terms that are strikingly similar to those in the rest of the Mediterranean. 37 However, the rejection of the primitivist Roman view of the gods as impersonal entities has further momentous im- plications. For this rejection must be complemented by an insight into the conception of these anthropomorphic divinities as personalized entities. I shall argue below that, once we accept the implications of the concept of lization<, this >persona- has repercussions for our understanding of the way in which worshippers made sense, communally as well as individually, of their pantheon. 38 The current discussion about Roman mythology is illustrative in this respect. The modern rediscovery of Roman myths started a long time ago. In opposition to Wissowa and the primitivists, Carl Koch, in Der romische luppiter (1937), accepted the existence of myths and aitia in early Rome as an expression of rooted anthropomorphic concept construed a crude antithesis of postulated the 38 the »myth« deeply divine. Yet, Koch and »history« and deliberate »demythicization« of Roman reli- gion in historical times by its 37 of a anti-mythical state cult. Cf. conveniently CANCIK 1994, 363-70; WISEMAN 1994, 23-36, 124-127; CORNELL 1995, 81-150, 162. For a critique of the primitivist position, and for the primacy of an anthropomorphic conception of the divine, see KOCH 1937, 9-32; RADKE 1965, 10-38; DUMEZIL 1970, 18-31; NORTH 1989; BSMER 1990; CORNELL 1995, 159-63. For a critique of Usener's idea of »Augenblicks- und Sondergotter«, see GLADIGOW 1981, 1208-10, 1213-4; ULF 1982, 156-7. But the primitivist view of Roman religion is still alive in many quarters; see only LIND 1992, 5-15. See below, 4.2.5 and 4.3.2. 61 The of existing mythological alterna- suppression state's tives would on such a view be a result of the orchestration of human relationships with the gods under the city-state in general; the state control over religious beliefs aim its manifestation its ing; the authoritarian cult theology of luppiter optimus maximus. 39 Only a few years and a complementary aniconic early an between the public extinction of myths at Rome at stage and earlier, Lily Ross Taylor suggested a link Koch, by unacknowledged myth-mak- deviant which would otherwise become apparent in phase in Roman reli- gion. 40 Each of these two scholars implicitly drew on Dionysius' testimony that Roman religion lacked myths because the first legislators had banned mythology as being perception true of divinity. of distortive a It is clear from Dionysius' argument that posterior suppression presupposes prior existence of a mythology at Rome; and Taylor's attempt to combine this argument with Varro's construction of a passage from aniconism to anthropomorphism, and from the absence of myths to their existence in the theologia fabularis, is unsuccesson ful methodological grounds. Dionysius' theory itself is of course not without problems. As Emilio Gabba has Dionysius' out, that it was which had vision of Roman religion entailed overall reflective pointed of Hellenic religious traditions declined in the Greek TioXeLC. This view ran into difficulties once Dionysius realized that, unlike Greece 39 KOCH 1937, esp. 121-34. The school of Angelo Brelich further developed the idea of »demitizzazione«; cf. GRAF 1993a, 35-8. It has also met with approval in post-war Germany, e.g. from GLADIGOW 1981, 1213-4. 40 TAYLOR 1931. 62 Hellenic its with Rome myths, lacked a distinctly Roman mythology. Dionysius 1 answer to this gap was that civic suppression of an element which was regarded as detrimental to proper religious behaviour must be held responsible for this phenomenon. 41 Dionysius' assertion that Roman religion lacked a mythology resulted from his distinctly Current distance approaches perspective. Panhellenic themselves such an ap- from proach. Instead, they suggest that our concept of gy« 7 developed with »mytholo- the Hellenocentric perspective of the nineteenth century in mind, is not capable of explaining the the lack of independent Roman cosmologies and theogonies on one hand or the relatively recent date of most, if not all, of Rome's aitiologies (which presumably originated in a climate of expansion and self-definition from the century onwards) third on the other. In order to overcome a Helconstrue lenocentric perspective, scholars therefore mythology early Roman as a domain distinct from the Hellenic mythologi- cal experience. 42 By investigating a mythological narrati- ve's function in Roman society rather than its origin, they tend to stories, concentrate as on the relevance of »traditional tales«, for the evaluation of the meaning of ritual behaviour in society. This emphasis on the 42 new symbolist (old) paradigm of »ritual and myth« views ritual and myth as »parallel symbolic 41 aitiological processes*. Myth is Dion. Hal. Ant. Rom. 2,18-20; Euseb. Praep. Evang. 2,8; GABBA 1991, 18-38. E.g. GRAF 1993, esp. 5: »um weiterzukommen, mu£ der romische Mythos von griechischen Modellen getrennt werden«; FEENEY 1998, 47-75, esp. 74-5. Cf. BREMMER 1993, esp. 173-4 on the relatively late date of most Roman aitiologies. 63 seen as an exegesis of ritual in a particular historical situation so that both components complement one another yet intertwined cognitive systems differentiated supplying in for the explanation of behaviour. This approach has undoubtedly produced stimulating results, not least that fact the for question of a myth's origin or age can be exposed the as as an unsuitable category: as soon myths cosmological both Greek ancient and comparatively recent Roman aitiolo- gies can have a similar cognitive value in society, the en- tire discussion about Roman mythology must be seen in a different light. 43 Incidentally, this insight own assumption about the implicitly deconstrues its supposed difference between the respective use made of mythologies in Roman and Greek societies. For it is unclear whether a Hellenocentric concept could mythology with Hellenic world either. diverts any justification be applied to the Rather, zing, like myth-making in Rome, adaptation of and perspective was characterized by the a flexible mythical canon to different local contexts, and that the stories Panhellenic a from the fact that Greek mythologi- attention our of creation aitiologies of individual foundation which accompanied ritual activity was the norm rather than an exception. 44 Yet, there is still another problem with this new functionalist perspective. By 43 Cf. GRAF 1993a, 25-31, 38-43; VERSNEL 1993, passim, for the »ritual and myth« paradigm. Parallel processes: BEARD 1987, 2-3; SCHEID I993a, 126-7, at 127: »deux pratiques paralleles, traitant toutes deux a leur maniere des problemes fondamentaux qui preoccupaient les Romains«. C. R. PHILLIPS, CP 90 (1996), 283-6 conveniently outlines the further potential of this approach. 44 BENDLIN 1995, 266. 64 turning mythology into a merely cognitive exercise of making collective sense of ritual behaviour, scholars domesticize the motivational world of myth-making into endeavour constrained a rationalizing by local religious knowledge and ex- pectation. This must no doubt be the reason why these stu- dies emphasize the communal relevance of ritual and myth and the »Romanness« which these two domains are said to have defined in the memoire collective of a changing Roman society, rather than investigating any (potentially deviant) in- dividual responses. 45 Yet, there is a difference between the restricted communicative code of ritual and the often speculative communication through mythological narratives. This means that ritual and myth are not esses«. On simply »parallel proc- the contrary, myth-making exploits the restric- tions of ritual communication by providing explanations and exploring possible tensions at a level where cognition is no longer directly related to ritual behaviour. Therefore, the socialization of religion through mythologies does not merely mirror the religious knowledge of society at large, but complements this with the internal world of one's own personal myth-making, however constrained by societal expecta- tions it may be. 46 Romans would encounter personalized gods through material representations, statues, paintings, coins, at various ritual events, through drama, literature or tra- ditional tales. It may be impossible to know for certain how these different genres shaped, cross-influenced 45 46 or contra- E.g. BEARD 1987, 9-10; Ead. 1993. For documentation and critique, see below, 2.7.3. See BLOCK 1992, esp. 99-100, for the theoretical framework. 65 individual dieted perceptions of the divine, but the ques- tion is nevertheless worth investigating. 47 2.3 New paradigms So far, the discussion of key concepts of Roman religion nineteenth and early twentieth century scholarship may give Hartung the impression of an utterly fragmented discipline. tried unearth to in the religion of Rome's first native in- habitants; Wissowa attempted the reconstruction of authentic Roman »Volksreligion« in its of realization institutionalized state cult; and the primitivists thought they had isola- ted the survival of a primitive predeistic period evolution. gious Yet, these veil one fundamental concurrence the on focus in reli- obvious differences must not in the namely approach, reconstruction of the origins of Roman reli- gion. The paradigm was the religion of an idealized stage in Roman history when an authentically Roman religion<, ned in defi- century terms of ethnic purity, had not nineteenth yet been superseded by mere cultic formalism and the corrupcentury tion of religious structures. 48 By the second this religion would BCE, have degenerated into a lifeless rit- ualistic construct, an imposture used by the elite as a convenient instrument in political conflicts, as the manipula- tion of omina in the political assemblies proved. Widespread scepticism among the aristocracy as a result of the recep- tion of Greek philosophy and the impact of 47 48 See further below, 4.2.5 and 4.3. See below, 3.1.1, for a critique of ethnicity. the new cults stereotype from of 66 the East on the populace, the collapse of temples and the neglect of indigenous rites, and, above all, the formalistic nature of traditional religious practices that were unsuited for providing an emotional experience which met new individual needs - all this was seen as decline clear indication of of Republican Roman religion. As a corollary, these scholars, while appreciating Augustus 1 restoring traditional Roman alleged religious attempt period as at cults and institu- tions, perceived of the religious developments of the rial the Impe- a continuation of the Republican crisis to which Christianity would provide the ultimate answer. 49 As we have seen above, this picture has come tack more under at- recently. The impression of formalism and a lack of vitality in Roman religion is, it has been argued, due to the application of modern notions about place »religion« to the of* ritual in ancient societies. These traditional no- tions are often described as »Christianizing assumptions* by their critics. 50 »A Schleiermacherian tradition of >religion<« is, as I have outlined above, perhaps a more precise, since less under-determined, phenomenon. 51 As defining description of this many of these critics have pointed out, a certain punctiliousness over ritual detail does not, as the 49 The line of argument is familiar enough from innumerable portrayals of Late Republican and Imperial history. Accounts include WARDE FOWLER 1911; WISSOWA 1912, 60-72; TAYLOR 1949, 76-97; LATTE 1960, 264-93; DORRIE 1978a, 247-8; LIND 1992, 10-5. MANUEL 1959, passim, gives an illuminating account of how the idea of neglect and formalism in pagan religions re-emerged with the critique of religion in the age of the enlightenment, subsequently biasing modern perceptions of Roman religion. 50 NORTH 1976, 9-11 and PRICE 1984, 7-19, for of this particular term. See above, 1.2 and 1.3. 51 the coinage 67 Schleiermacherian dichotomy between dreary ritual and proper religion might imply, preclude a cognitive comprehension of ritual. As it is capable of conceptualizing in one's position the social world, »ritual« should, the argument convinc- ingly runs, be revalued in terms of its function in rather than society with reference to its origin or an alleged un- changeable meaning. 5 2 The elite's religious behaviour has also attracted considerable scholarly interest. To earlier generations, the interpenetration of religion and politics appeared to prove the former's abuse by manipulative politicians. Religious institutions in the city-state were regarded as an imposture of detached intellectuals. On the new consensus, the inter- relation of religious and politicial spheres ought not to be explained using modern conceptions about the both separation domains. Rather, these were inextricably linked reali- zations of one and the same cultural phenomenon. Since were of both magistrates and priests, they members of the Roman elite would feel that their religion was an integral element in the definition of their position in society; and we would underestimate the importance these people attached to their religious roles, if we understood their behaviour and thinking solely in terms of intellectual detachment. On the con- trary, the active, and sometimes 52 aggressive, appropriation For powerful critique of the view that Roman religion was formalistic and unimaginative, see DUMEZIL 1970, 102-12; NORTH 1976; Id. 1986, 251-3; ULF 1982, 145-63; PHILLIPS 1986, 2697-711; Id. 1997. For the revaluation of ritual, see PRICE 1984, 7-11, drawing on studies by the anthropologists C. Geertz and D. Sperber. Cf. BEARD 1987, 1-3; VERSNEL 1993, passim; FEENEY 1998, 115-36, among many others. 68 of rituals and ideas in the political arena only religious part proves that religion was an accepted a use to habitus, of elite's the terminology coined by Pierre Bourdieu. Their behaviour attests to their agreement on the centrality were of religion in public life. In this respect, they not different from the rest of the population at Rome. 53 this Could revaluation of Late religion Republican represent a paradigm shift as defined by Thomas Kuhn? Unfornot. For current research, while no doubt avoiding tunately many of the shortcomings of earlier fails scholarship, to fully disentangle itself from past traditions. For instance, in attempt an to escape the misleading dualism between a Schleiermacherian notion of »religion« and meaningless in malism ritual, the for- significance of categories such as »belief« or »disbelief« themselves have been challenged. The argument that Roman religion did not meet the needs of individual religiosity is contested on a priori grounds with the argument that heuristic categories such as personal commit- ment or individualized religious feeling should no longer be applied Roman religion. The notion of individual belief to is now only permissible as long as it refers to the individual's commitment to, and his participation in, a religious collective belief system which becomes homologous with civic religion in the city-state. In other words, individual religious cognition and behaviour were warranted only when were 53 related they to the welfare of the community which guaran- For the elite's interest in religion, see JOCELYN 1966; Id. 1982, 159-61; LlEBESCHUETZ 1979, 1-54, esp. 15-20; BEARD I994a, 739-42. 69 teed individual security. 54 However, I above monstrate have view de- with a view to the respective positions of Sabbatucci, Bremmer and Beard in what sense tionist to tried such reduc- a of »religion« / almost exclusively attacking a Schleiermacherian, and thus a priori limited, tradition, does not live up to its own methodological premises. 55 Jerzy Linderski's overtly functional description of religion's role at Rome is representative of this new paradigm: »Roman state religion was not interested in individual salvation; its only concern was salus publica, the security of the Roman state, or, in Roman terms, the preservation of pax deorum, the peace between the gods and the state. The goal of the cult was to keep the gods pleased and well disposed toward Rome.« 56 The underlying notion will by now be familiar to the reader. For few scholars would disagree with Linderski when he regards dis- any internalized motivational processes and defines Roman religion as an external contract between the gods and men based on purely rationalist terms. 57 It is therefore not surprising that those areas which earlier scholars had de- picted as alternative realms where true religious experience could be found in separation from the religion of the state namely private worship, family rites, the cults performed by vici, pagi or collegia, or the agrarian rituals described by Cato - that those areas on the new paradigm mirror public religious patterns. Whereas a few scholars would maintain 54 E.g. NORTH 1976, 1; PRICE 1984, 7-15; BEARD & CRAWFORD 1985, 26-31; BEARD 1986, 34; SCHEID 1985, 12-5; NORTH 1989; LINDER & SCHEID 1993; BEARD 1994a, 729-34; DURAND & SCHEID 1994. For similar views relating to the Classical Greek city-state, see above, 1.3. 55 See above, 1.3-1.6. 56 LINDERSKI 1995, 610-1. 57 Cf. LINDERSKI 1995, 621 1 : »... once we accept the premises of Roman state religion, it appears as a rationalist system, as a scientia (as e.g. the augurs used to describe their discipline)«. 70 that individual morality was an element in the religious life of the Romans, 58 the majority view is Roman religion did that Republican not provide distinct religious biogra- phies, individual moral value systems or spirituality in a modern sense. Instead, it is held that in a system of public religious symbolisation of civic identity, religion mani- fested itself in performance of cult, whereas belief was only recognisable, and did only matter, on the level of proper, i.e. public, religious behaviour. 59 The particular function of assigned to religion For as I argued this, preserving status the by such a view proves disconcerting. above, on purely methodological grounds like any functionalist, definition would at best out- line an unintended consequence of society. 60 As it happens, methodological religion's existence cautious as to which meets the a and nature of Roman religion« is revealing; and Denis Feeney's view of »[t]hese eminently practical and busy ...« of criteria of this new paradigm. In this respect, Henk Versnel's aside about »the practical juridical their statements - than Linderski to divulge their opinion about the supposed psychological conditioning people in most current scholars of Roman religion are less willing - or more ple quo may be more than peo- an unintentional slip of the tongue. 61 Yet, even though the proponents of this new para- 58 E.g. LIEBESCHUETZ 1979, 39-54; WlSEMAN 1994, 49-53. 59 To the references given above, add BEARD & CRAWFORD 1985, 30-6; SCHEID 1985, 12-15; NORTH 1989, 598-9, 604-7; PORTE 1989, 8-16; RIVES 1995, 4-13. 60 61 Cf. above, 1.5. H. S. VERSNEL, OCD3 (1996), 1613; FEENEY 1998, 4. Cf. RAWSON 1985, 321, contributing to what she herself calls »this over-familiar generalization*: »But on the whole 71 digm that issue, the conclusion which must impli- sidestep citly be drawn seems inevitable: a parameters are system religious whose organized with a view to ritual performance rather than to individualized internal motivations, with the result that individual religion is homologous with the religion of the city-state, such a religious system entails with regard to its agents the complementary view that they are as practical, juridical, functionalist and formalistically minded as their religion. Mommsen, Wissowa, Warde Fowler, Deubner or Latte, not to mention their successors innumerable past and present, would emphatically nod in agreement. 62 This amount of unexpected contacts between old paradigms has generations alleged whose supposed absence lack had bemoaned, has been solved by ques»belief«; tioning the validity of the very category of the new gone largely unnoticed. The discussion about the place of belief in Roman religion, earlier and and of primarily individualized personal commitment in religion has been turned into a virtue. The notion of >formalism<, which on the old paradigm had entailed negative connotations, is now, under the disguise of a nationalist system<, again a central category of Roman ligion. The re- mercantilist language which these scholars use in describing »contracts« between gods and men through vota or the contractual principle of »do ut des« once more expose 62 no doubt the Romans were practical enough in their ends.« ULF 1982, 145-63 gives a critical survey of the stereotypical characteristics traditionally applied to Roman religion. Most of them could also be used within the framework of the new paradigm. Cf. WEILER 1974 and BARGHOP 1994, 41-52 for a discussion of the stereotypes commonly used by historians when talking about supposed character differences in the ancient world. 72 legalistic formalism raison d'etre of Roman reli- the as gion. 63 Yet, it is difficult to maintain that idea the of reciprocity in the relationship of gods and man, although no doubt imitating cultural stereotypes of social reciprocity, can be reduced to a pseudo-mercantilist rationality. For the contract, if that is how one should call it, never bound the deity the way an economic or legal contract would have bound humans. The stock characters of Plautine comedy, after or Cicero when in exile, realized to their sacrifice, fuse pro- profound disappointment that the gods did not always respond as the worshipper had anticipated. 64 reli- The new emphasis on the close interpenetration of gious and socio-political realms echoes, as I have shown above, a in the Roman Republic tendency general in the study of Graeco-Roman religions to de-differentiate religion culture. 65 and The thesis that religion almost exclusively focused on the political Roman city-state, and military 64 65 of the that it was more or less >undifferentia- ted< from the political sphere, amounts 63 activities to its >embedded- »Do ut des« and its corollary, votum, seem accepted principles of Roman religious rationality: WISSOWA 1912, 381-5; LATTE 1960, 46; W. EISENHUT, RE Suppl. 14 (1974), 964-73; NORTH 1989, 593; PORTE 1989, 14; H. S. VERSNEL, OCD 3 (1996), 1613. E.g. Plaut. Poen. 449ff.; cf. Rud. 22ff.; Stich. 393ff.; Cic. Fam. 14,4,1: neque di quos tu castissime coluisti neque homines quibus ego semper servivi nobis gratiam rettulerunt; GWYN MORGAN 1990, 30-1. Reciprocity was anticipated rather than firmly expected: Cato Agr. 141,4; Cic. ND 1,116; CANCIK 1994, 393 with n. 61. Reciprocity in religious communication is not characteristic only of Roman religion: ULF 1982, 155-7; K. HOHEISEL, HrvtG 2 (1990), 228-30. The phrase do ut des itself is not attested but rather seems to be a modern coinage by analogy with phrases like Livy 10,19,17: si duis, ego . .. voveo. See above, 1.3. 73 ness< To in the socio-political administration of the city. 66 state Wissowa, reflection Roman of cult had been the institutionalized >Volksreligion<. On the new paradigm, civic religion is the religion of Rome. Local become thus religion has homologous with civic religion. As these exam- ples elucidate, at stake is not the existence of a public system of organized and administered religion in the city of Rome, but the normativity attached to the modern model of civic religion. Yet, before asking for ancient authorization of such a model, let me quote from John North's account of civic religion in Mid-Republican Rome (a passage of exemplary caution): »To put the point in its most extreme form, what we have might be an artificial historiographic construction, expressing a kind of official religion which never actually represented the religious life of the Roman People. « 67 66 The concepts of >interconnectedness<, >undifferentiatedness< and >embeddedness< are also applied to Greek religion in the Classical period: e.g. OUDEMANS & LARDINOIS (>interconnected<); SOURVINOU-INWOOD 1990, 322 1987 (>undifferentiated<): »... in the Classical period polls religion encompassed, symbolically legitimated, and regulated all religious activity within the polis«; BREMMER 1994, 2-4 (>embeddedness<, cited above, 1.3). For this concept of >polis religion< or >civic religion< in the Classial Greek city-states, see e.g. SOURVINOU-INWOOD 1988; Ead. 1990; BURKERT 1995. S. G. COLE 1995 is more cautious. 67 NORTH 1989, 582. 74 2.4 A Roman model of >civic religion*? Citing the authority of ancient writers Cicero, such Polybius, Varro or Livy, the modern advocates of the model of »civic religion* can claim to represent concerning a Roman viewpoint the role of religion in Late Republican society. Taken at face value, these indigenous to as interpretations seem prove the concept's validity beyond doubt. However, many students of Roman religion share the fallacious view that it is legitimate to attribute superior veracity to these exegeses as regards the »real meaning« fortunately, of Roman religion. Un- as I shall argue, these portraits of Roman re- ligion are just as subjective and as biased by conscious or unconscious motives as any other interpretations. 2.4.1 Varro and civic religion Consider Varro f s famous tripartite conception of theology prefaced his account of the origins of Rome's res di- which vinae, a selective account of Roman priesthoods, the sacred locations, its festivals, three books each. In his preface, into with divided theologia and drama; a genus physicon religious issues in terms of physics and philo- sophy; the genus civile quod Varro and deities in three aspects: a genus mythicon portrays the religious discourse of poetic fiction deals rituals city's addresses the religious practice in urbibus cives, maxime sacerdotes, nosse atque admi- nistrare debent. This entails a local religious knowledge as regards the deities that ought to be worshipped and and rituals sacrifices which must be performed by anyone in a civic 75 context. The first theologia has the on rejected be to that it misrepresents the character of the gods; the ground true nature, ought not to be conducted in pub- its vealing re- while second contains discussions of the divine which, lic; only the theologia civilis should be disclosed as use- ful to the civic domain. 68 Varro, via Augustine, attributed this per- tripartite to the pontifex Mucius Scaevola (cos 95). Unfortu- spective nately, Scaevola's actual contribution to the theologia tripertita is, due to a lack of reliable source It scure. has ob- material, therefore been suggested that Varro devised Scaevola as a persona in the dialogue Logistoricus Curio deorum cultu de to express his own views. On that hypothesis, Scaevola became the exponent of the genus physicon in a philosophical attack on traditional religious practices; and such a position would more easily befit a dialogue like the Logistoricus Curio than the RD in which revalue Varro set out to traditional Roman religious practices. 69 Yet, while Augustine appears to have known not only the RD but also the Logistoricus Curio, on purely philological grounds tribution of at- the passage in question to the latter has met with criticism. 70 Furthermore, it is clear the the that throughout first book of the RD Varro discussed the immaterial na- ture of the gods along the lines of Stoic philosophy, thereby implicitly compromising traditional and 68 69 70 religious practices beliefs. Varro may still have considered it appropriate Varro RD I frgs 6-12 Cardauns. Scaevola ap. Varro jRD I frgs 7 and 9 Agahd = Varro Log. Curio frg. V Cardauns with CARDAUNS 1960, 34-40, passim. Cf. esp. HAGENDAHL 1967, 619-20. 76 to use the persona of Scaevola in order to appeal the maiores as well as to pontifical authority (in to port sup- for ND HI/ Cicero pursued a similar strategy), this even but assumption (which in itself is unprovable) does not point to dialogue his Logistoricus Curio as the only possible place where this could have happened. In fact, Augustine remarked the respective arguments presented by >Scaevola< (whom that were Augustine took to be the historical pontiff) and Varro easily compatible. 71 As mentioned above, Augustine thought of Seneca's critique cizing Stoi- of cult activity as being far more radical than Varro f s critique of the theology of the poets. 72 In De superstitione, Seneca regarded ritual as a superfluous practice unwarranted by the philosophical demands on a virtuous and moral life. Religious cult could be the constraints of Jewish cults in Rome, but only by of complying with the norms of society. In the course of his argument, Seneca practices justified not attacked only the religion or of the so-called oriental directed his critique at traditional Roman religious practices, too. 73 On Augustine's view, Seneca's criticism of cult, which called into question the very foundations of religious activity, penetrated into main 71 72 73 of the do- civic religion - a liberty which, according to Au- August. CD 4,27, p. 179,21-181,3. Cf. HAGENDAHL 1967, 609-16; LIEBERG 1973, 101-2, 104-6. BEARD 1994a, 757 accepts Cardauns' view of Scaevola as Varro's persona on the ground that the historical Scaevola would not have been capable of providing the intellectual discourse characteristic of Varro's generation - a circular argument . See above, 1.1. MAZZOLI 1984; LIEBERG 1989, 1888-98. See further below, 4.2.5. 77 gustine, Varro had not dared tine's interpretation, to scholars take. 74 have Accepting often Augus- assumed that Varro, while reprimanding the religious discourses of poetry and drama, provided a theologia civilis, philosophical justification of the or even gave a systematization of state religion. To be sure, Augustine, who openly attacked pagan cult practice, presented Varro as a defender of civic religion. 75 Yet, the bias of Varro f s suspicious as to should make us his ulterior motives. Furthermore, as an inevitable consequence of the state presentation thesis that Varro defended religion these scholars have to surmise a tension be- tween Varro's philosophical views, which religious ritual, were critical of and his defense of traditional religious practices. The same scholars do not make entirely clear how Varro should have resolved this contradiction, on whose paradoxical existence already Augustine commented. Perhaps the image of Varro the antiquarian has further added to the mo- dern neglect of ulterior motives on his part. 76 His testimony has also been employed to lend ancient authority to the modern concept of >civic< or >polis religion<: »It may be reassuring to state that there is an ancient concept and term of polls religion, theologia civi­ lis. « 77 74 75 August. CD 6,10, p. 269,11-3 D-K: ... hanc libertatem Varro non habuit; tantum modo poeticam theologian reprehendere ausus est, civilem non ausus est, quam iste concidit. Cf. LAUSBERG 1989, 1895-7. Defense: CD 6,6, p. 256,30-257-15; 6,9, p. 265,13-8 D-K. Justification: ibid. 6,8, p. 260,31-2 D-K. 76 Tension: August. CD 7,17, 23. Positive views about Varro's aims: e.g. LIEBERG 1973, 82, 100-1; CARDAUNS 1978, 94-101; RAWSON 1985, 312-6; DORRIE 1986; CANCIK 1994, 395-6. 77 BURKERT 1995, 201. 78 It cannot be doubted that Varro the civic domain. addressed religion Yet, the philosophical context in which the notion of theologia civilis is introduced does not rant in war- such assurance as to its appropriateness as an ancient descriptive equivalent to the modern notion of >civic gion<. When introducing reli- his account of the deities of the city of Rome in RD books 14-16, Varro emphasizes that he going is to adopt the philosophical position of Academic scep- ticism. As to the literal truth of Rome's his entire account of divine pantheon, the author's judgement is emphatic- ally suspended. Moreover, in the final book of the RD, which attempted to interpret select deities in the Stoicizing theologia naturalis author's Academic suspension of light of the expounded in book one, the judgement is extended to those gods that received worship in the civic domain: de diis ... populi Romani publicis ... in hoc libro scribam, sed ut Xenophanes Colophonios scribit, quid putenif non quid contendam, ponam. Hominis est enim haec opinare, dei scire. 73 It must have been this use of Academic enabled Varro sophically practice to informed maintain argumentation which the tension between his philo- criticism of contemporary religious and his aim to provide a normative and educational account of the religious system at Rome. Moreover, for an understanding of Varro's conceptualizing of the theologia civilis it is crucial to note that he percei- ved the gods to be prior to the city-state, and civic reli- 78 Frg. 228 Cardauns. Varro's adoption of Academic scepticism in these matters is declared in frg. 204 Cardauns. The same author's work LL also follows Academic principles, using the argumentative structure of disputari in utrawque partem: Ax 1995. 79 gion to be its product. 79 In other religion was an words, to Varro invention of the city-state. According to Varro, an appropriate and comprehensive treatment gion civic of reli- as such would indeed require a wider approach. In par- ticular, it would include a philosophical discussion of the nature of the gods on the basis of the theologia natu- true ralis as proposed by the author earlier in the RD. Since religion in its civic aspect only was historically and logically posterior to the civitas which created it, the treat- ment of theologia civilis had to be included among a discussion of the institutions of the city-state. 80 The tripartite division of religion into the realms of mythologizing, philosophical speculation and civic cult is a philosophical model rather than the reality of civic admi- nistration. Attempts to trace this model back to one sophical philo- school or, indeed, one particular philosopher have proved futile, as it was widely used by Stoics, Sceptics and Epicureans alike. 81 As we have seen, Varro's the historical dependence Scaevola can neither be verified nor falsi- fied. Given the familiarity of the tripartite conception religion on of in first century philosophical thought in general, it is likely that Varro anyhow showed considerable independence of mind in relation to potential Roman or Hellenistic. predecessors, either Varro adopted Greek terminology for 79 80 Pace BURKERT 1995, 202. Frg. 5 Cardauns: si cut prior est ... pictor quam tabula pi eta t prior faber quam aedificium, ita priores sunt civitates quam ea quae a civitatibus instituta sunt si de omni natura deorum et hominum scriberemus, prius divina absolvimus quam humana adtigissemus. Cf. above, 1.1, for the parameters applied to the philosophical inquiry De natura deorum. 81 Cf. LIEBERG 1973; Id. 1982. 80 the first two genera theologiae but not in the case of the third. According to Augustine, Varro retained the respective Greek terminology for the genus mythicon and the genus phy- sicon, but translated the third into Latin le. 82 genus as civi­ What seems to be a minor philological point has, as I shall show below, wider consequences for the meaning of the Varronian passage. Firstly, the heuristic value of the Varronian concept of theologia civilis for our understanding of civic religion is no doubt undermined by its philosophical bias. Varro In RD, the acknowledged the existence of the gods, but perceived their nature on the basis of Stoic philosophy as immaterial. As such, the gods did not framework must regard require any cult. divine which pure conception aniconic, could was have refounded naturae formula. 83 the Since impossible, he chose the second best solution and tried to reconstruct the historically legitimized 82 idea institutions of Rome, he would have reestablished a true conception of the divine ex this the and thereby undiluted, worship of the gods in Rome's distant past. 83 If Varro religious of only the genus physicon could adequately express. This explains why Varro was committed to of theoretical contemporary cult practices or mythology as removed from the true and the This forms of Augustine subsequently translated the first two theologiae as fabulosum and natirrale respectively. August. CD 6,5, p. 252,17-253,4 D-K: ... tertium etiam ipse Latine enuntiavit quod civile appellatur. Deinde ait (frg. 7 Cardauns) >mythicon appellant quo maxime utuntur poetae; physicon quo philosophi; civile quo populi<. Immateriality: frg 22 Cardauns: dii veri neque desiderant [sacra] neque deposcunt, ex aere autem fact! ...; cf. frgs 23-8; De gente pop. Rom. frg. 18 Fraccaro; Dion. Hal. Ant. Rom. 1,77,3. Aniconic worship: frg. 18 Cardauns, discussed above, 2.2. 81 Roman religion in their most ancient form in order to ap- proach as closely as possible the true perception of divinity. His was a pseudo-restorative approach towards Roman ligion: Varro's cultural nativism re- criticized the complex present as an epoch of religious decline by introducing rameters pa- which postulated an allegedly authentic and primi- tive past in primeval times. 84 Secondly, since a Stoicizing approach accepted traditional cult practices as an imperfect, yet still commendable/ attempt to perceive the true nature of the gods and the universe, it followed that religion was perceived to be useful for individuals and for city-states alike; even if its im- perfect state meant that the true meaning of many of contemporary religion's beliefs and customs were unperceivable to the ordinary mind, while others were wrong. The category by Varro in the RD judged traditional religious practice which was its »utilitas«. Even the divinisation although clearly understood of viri fortes, as imposture, contributed to, rather than diminished, religion's overall utility. 85 This was of course a topical argument in the ancient debate about the justification of Euhemerism. 86 The socio-political utility of such ideas, stressed by Varro, belonged to another topos in the Late Republican and Early Imperial debate about religion in a civic 84 85 86 community. That topos entailed that Frgs 12-19 Cardauns. On Varro's theological premises, see BOYANCE 1972, 253-82; CANCIK 1994, 401. On the employment of nativistic models, cf. the general remarks in GLADIGOW I997b. Frgs 20-21 Cardauns: utile esse civitatibus ... ut se viri fortes, etiamsi falsum sit, diis genitos esse credant. K. THRAEDE, RAC 6, 877-90. 82 civic Roman religion had been first invented by monarchic and rulers for reasons of political expediency purpose of preserving the the socio-political status quo: rei publicae causa communisque religionis When served (Cic. Div. 2,28). Q7 relating these evaluative judgements to their historitradition cal context, their philosophical sufficiently is always not into account. As a matter of fact, when taken Hellenistic and Roman writers adopted an interpretative modinven- el which viewed the religion of the city-state as an tion of legislators acting out of socio-political necessity, they followed a tradition which went back at least to the later fifth century BCE. 88 I would suggest that we must also relate Varro's conception of the theologia dition. civilis to this intellectual tra- Varro himself paraphrased the genus civile as genus tradit[um] a principibus civitatis and attributed this definition to Scaevola. As we have seen, it is unclear whether this passage belongs to the RD or to the Logistoricus Curio. As I argued above, the distinction would not matter much. 89 For this paraphrase places the genus civile closely context in the of religion invented by legislators out of necessi- ty. As we have seen, the religious rituals and institutions introduced by these principes civitatis could be open to the 87 88 89 E.g. Polyb. 6,56,6-12; Cic. ND 1,118 (dixerunt ...); Livy 1,19,4-5, 21,1-2; PEASE on Cic. ND 1,118; LIEBESCHUETZ 1979, 29-34. Dion. Hal. Ant. Rom. 2,62,5 did not adhere to this doctrine: GABBA 1991, 123-4. E.g. ?Critias 88 B 25 D-K = ?Eur. TrGF 1,43 F 19; Democr. 68 A 75 D-K; Pi. Rep. 3, 415a-c; HENRICHS 1975; BORING 1978. For the importance of this topos for the critique of ancient pagan religion in the seventeenth and eighteenth centuries, see MANUEL 1959, passim. RD I frgs 7 and 9 Agahd = frg. V Card. ap. August. Civ. D. 4,27. 83 that criticism they were certainly convenient, utile, but not necessarily intelligible or true. translation Varro's Encountering of the third theologia into Latin, Eusebius and Augustine sought a Greek equivalent to civile and found it in the term TioXuTuxov or politicum. As a result, editors and commentators have that Augustine, at least, found assumed this rendition already in Varro's £D. 90 This hypothesis is unprovable. Eusebius the was first Christian apologist in the East to adapt the tripartite conof ception religion to his attack on paganism. However, since his use of Varro is questionable, it is more likely he adopted TioXiTixov via a different channel of trans- that mission. In the West, Tertullian instrumentalized the tripertita logia theo­ towards the end of the second century CE. Yet, while clearly using Varro, he changed civile to genti­ le. Later apologists offer further versions whose conception is close to Tertullian's. It is likely that these apologists adjusted nomenclature to the respective goals of their at- tacks: Eusebius and Augustine referred to pagan religion in its civic aspect in particular, whereas Tertullian, Arnobius and Lactantius reprimanded paganism as such in a broader sense. 91 A pagan tradition which instrumentalized partite 90 91 the tri- conception of religion materialized not long before Euseb. Praep. Ev. 4,1,2; August. CD 6,12, p. 271,5-7 D-K: tres theologias quas Graeci dicunt mythicen physicen politicen; ibid. 7,23, p. 302,20 D-K: politicum; PEPIN 1958, 284; contra LIEBERG 1973, 91. Tert. Nat. 2,1,8-15, at 10; Arnob. 3,11: ipsi in ritibus; Lact. De ira 11,16: ipsi qui deos colunt. On Tertullian's wider focus, see Nat. 2,1,1: nunc de deis vestris, miserandae nationes, congredi vobiscum defensio nostra desiderat; LIEBERG 1973, 89, 91. On the use of Varro among the Christian apologists, see conveniently CARDAUNS 1978, 91-4. 84 the Christian. philosophy, Arguably these representing writers employed Hellenistic Stoic the terms VOULXOV and voiiodeTcov I suggest that one ought to take the Greek VOJJLLXOV (Ae- tius, Dio Chrysostomus , Plutarch) , representative of a pagan Stoic tradition, rather than Eusebius ' or Augustine's TtoXt- TLXOV, as the rendition which gives a truer representation of the context of Varro's genus civile - a context which the Christian retranslation of civile into Greek as TtoXLTixov obscures. For the genus civile now appears to be version of the Latin the Greek yevoc VOJJLLXOV in the context of which law-givers, vouodeTCXL , and viri fortes had invented the im- perfect form of civic religion out of socio-political necessity. The conclusion seems inevitable: the fact that Varro, in the RD, judged institutionalized religion as a politically convenient, though philosophically compromises the status insufficient, system of the genus civile earlier in the same work. Should it not follow that the genus civile , regulating contemporary cult praxis with a view to that ought the deities to be worshipped and the rituals and sacrifices which had to be performed at Rome must be criticized on same grounds? 93 As a consequence, Varro's position is in fact much closer to Seneca's, who disapproved practices on religious norms of society Aetius Plac. 1,6, Dox. Graeci p. 292a,6-14 = SVF 2,1009 (VOULXOV); Dio Chrys. Or. 12,39-43, 44, 47 = Poseidon, frgs 368-9 Theiler (vojiodeTdi / vouodeaia) ; Plut. Amatorius 763 B-F 93 of philosophical grounds, but saw them warranted by the constraints of complying with the 92 the (VOVLOdETdL) . Cf. PEPIN 1958, 308-14, for argumentation along slightly different lines. 85 and by the usefulness of religion for the conduct of states. The in which the Varronian genus civile operates framework seriously compromises its heuristic utility in thority the to modern model of civic religion: to Varro, civic religion was an imposture, convenient philosophical perception au- lending as as long a of the divine could not be achie- ved. 2.4.2 Priests as mediators of civic religion? Varro's definition of the genus civile made particular tion men- of the responsibility of the state priests for the ad- ministration of religion in the city-state. 94 Disregarding Varro's critical perspective, this comes close to the normative position of Roman writers such as Cicero or Livy, who held that the city-state, represented by its religious functionaries, made authorized and religion, this that and amounted to a complete control of the organization, supervision and administration of its various public and private aspects. 95 With a view to these religious matters, both plicitly assigned ex- primacy to the civic priests rather than to the magistrates. According to Rome's historiographic tradition, Romulus and his successors, on the foundation of the city, not only ensured its military, stability, but also established political and social the sacra populi Romani. This entailed the prescription of the concrete means of worship and the nomination of priesthoods responsible for the maintenance of these religious rites. On this view, the ear- 94 95 Frg. 9 Cardauns: maxime sacerdotes. Cic. Leg. 2,19-22, 29-31; Livy 1,20. 86 ly the Roman state created Roman religion, of authorities just as they invented all the other norms and customs of the re- Roman People. Roman religion was no natural religion of velation artificial an but civic religion which was only It stitutions. intellectual tradition which the this was in- its imaginable in conjunction with the Roman state and vera Christian apologists attacked: according to Augustine, religio cannot be defined within the confines of one earthly political community. 96 In De legibus, Cicero emphasized the responsibility of the religious functionaries for this civic religious system, sacra the publica or sacra populi Roman!. Deities should preferably be introduced by public initiative; their temples needed to be dedicated publice; and sacrifices ought supervised priests by to be acted on behalf of the populus who Romanus. 97 This treatise (as well as other Ciceronian texts) reveals the author's concern with the public role gious functionaries pontifices, augures »sacerdotes populi reli- Rome, organized in the colleges of at and of XVviri. Romani« sacerdotes« »Publici are recurring phrases; and the in- priests' religious authority with respect to the public terpretation of religion on behalf of the community as a whole, even as far as sacra privata topical 96 97 98 motive in or Cicero. 98 were concerned, is a Commenting on the pontifical August. CD 6,4, p. 250,20 D-K. Cic. Leg. 2,19, 28, 30, 35. Cf. Mil. 85. E.g. Leg. 2.30 ut sine Us qui sacris publice praesint, religion! privatae satis facere non possint; Asconius 49 C: virgines pro populo Romano sacra facerent. Publici sacerdotes: Cic. Leg. 2,19-20, 22, 37; Dom. 132. Sacerdotes populi Romani: Cic. Verr. 2,4,108, Dom. 1. Publici augures: Cic. Leg. 2,20, 3,11, 43 (written shortly after Cicero's co-optation into that college). Cf. Cic. Font. 87 report on his house in 57, Cicero referred to the college as indices religionis and to the Senate as indices legis." narratives Scholars since Mommsen have linked these to various Imperial legal texts and have taken them as proof of a fundamental division of power and authority between secu- lar and sacred domains, between magistrates and priests. be sure, these scholars To realized the amount of necessary interaction of secular and sacred authority in Republican Rome. Yet, in their view the interpenetration or the instrumentalization of the sacred domain by politicians reflective of the blurring of traditional boundaries and the of decline of state religion in the Late Republic. 100 Con- versely, while the new paradigm of the civic was religion logical primacy of in Republican Rome also takes its lead from normative texts like those of Cicero or Livy, it is one of its great strengths that it no longer has to instrumentalize 47, a Vestal virgin as sacerdos; Senect. 30 and 61 for the sacerdotium of the pontifex maximus; Leg. Agr. 2,18 for sacerdotia referring to the lex Domitia of 104; Acad. 9; Rab. Perd. 27; Varro LL 6,21. Comparing ILLRP 62-3, 65-6 and 273, BEARD 1990, 43-7 (at 46) suggests that under the Republic sacerdos was used preferably as a designation of male and female religious functionaries of Hellenic and Eastern deities, and that it was »originally an external category for the Romans«. The latter suggestion is as difficult to prove as it is to disprove. Beard's further suggestion, however, that the alienness of sacerdos is reflected in the fact that the epigraphic record never uses that word when commemorating functionaries of the major Roman priesthoods is spurious. For as soon as the co-option into these priesthoods became a matter of constant popular concern (that is from C. Crassus's proposal of 145 to have new members of the major priesthoods elected by the People), the elite would no doubt be inclined to commemorate the precise designation of their priestly offices in analogy to secular civic offices. 99 Cic. Att. 4,2,4. 100 MOMMSEN 1887, 2, 18-73, followed by e.g. WISSOWA 1912, 479-80; BLEICKEN 1957, 446, 465-8; CATALAN© 1974. Cf. Gaius Inst. 2,2-10 with WATSON 1968, 1. 88 the stereotype of decline when describing the interrelation For of priestly authority and political power in Rome. to this new orthodoxy, religion was embedded in the cording the socio-political administration of view, ditional assuming categories de-differentiation A tra- of the sacred and secular as a weakness of the sys- of tem; on the view that such static Rome, in city-state. the categorical division of that spheres mattered, would see the stent ac- any divisions non-exi- were interpenetration is the unexceptional result of religion's embeddedness in political life; and the blurring of priestly and civic authority is inevitable rather than reflective of a crisis of the religious system. Though agreeing on this general picture, the new doxy ortho- is multivocal as soon as the particulars of the inter- relation of the domains of priests and magistrates are cerned. One theory, while rejecting Mommsen's fundamental separation of sacred and interdependence of secular priestly and domains, political maintains a relative differentiation of the mains con- emphasizes the authority but respective do- of religion and politics with a view to the independ- ence of religious power in the spheres of priesthood and sacred law. 101 The alternative position denies the independence of the religious domain in terms of power and authority and suggests that the priests were merely subordinate to the 101 E.g. SCHEID 1984, 259: »le droit sacre n'est pas coupe radicalement du droit public, c'est ce dernier qui assigne au ius sacrum sa place dans la cite, une place bien specifique, bien separee du droit public, mais non definitivement independant.«, ibid. 272: »deux provinces bien separees mais etroitement solidaires*. Cf. PAIS 1914; SZEMLER 1971; Id. 1972; SCHEID 1985; PORTE 1989; RUPKE 1996a, 252-8. 89 Senate and the People. This alternative view points out that a division of religious and civic authority would result in the assumption that the state was in fact in reality the priests secular, whereas actually did not provide the link between the city and its gods. While retaining their reli- gious authority/ they were expert advisers on religious matters due to their religious knowledge. Religious power rested with the Senate: it decided on religious matters like prodigies or the introduction of new cults. The too had magistrates religious functions in that they sacrificed before certain public events and took the auspicia assemblies and at specific political publica ceremonials. 102 To Mommsen and his followers, the bipartition of the te's authority into imperium before magistra- and auspicium reflected the exclusive existence of sacred and profane realms. In response to that view, it has been stressed that it would be mis- leading to conceptualize these two realms in terms of modern boundaries between religious life. This insight, however, scholars to assume that and has secular areas of public led some of the these due to their religious power the political functionaries in Rome, and not the religious functionaries, >mediated< between the citizens and the This concept of mediation is an unhappy coinage: using the term >mediator </>ii£CJLTTie<, which is Hebrews 6-8 inadvertedly based on and thus strongly invites the association with 102 For t ke senate's 1987. 1995, gods. 103 AUSpicium: 28-37. religious RUPKE authority, 1990, 41-51; see DETREVILLE KUNKEL & WlTTMANN 103 E.g. NORTH 1986, 257-8; BEARD 1990, 25-47; BEARD & NORTH 1990, 4-9. CONNOR 1988 discusses the inapplicability of the (Durkheimian) dichotomy of sacred and secular domains to Graeco-Roman society, pointing out the lack of such a division in Classical Athens. 90 Christian priestly charisma, blurs the distinction between a tenure magistrate's routine religious duties during his and office what is alleged to have comprised the much more encompassing responsibility for setting up between communication the and controlling Romans and their deities. 104 By contrast, it may be preferable to see the as of magistrate Roman more than the representative of the interests of the no Roman People (in their political capacity) before the gods. I shall return to this debate in the course of my ment. 105 For the moment, it suffices to stress the valuable contribution which this recent discussion has made to sing argu- expo- earlier stereotypes about the role of religion and pore- litics in Late Republican Rome. Many of its premises as gards the relationship of priesthoods and the political system or the important discussion about the inapplicability of dualism of »sacred« and »secular« to Roman society must the be accepted by scholars working on Roman religion. I am However, doubtful as to the underlying notion that religion at Rome can adequately be described within the parameters of model gives which logical primacy to the socio-political realm. For despite some conceptual differences fundamental agreement a among these there is a authors concerning the city-state's power in religious matters: the city-state provided the framework for and controlled religion, since its representatives, elite administrators, institutionalized and defined human contacts with the divine. In other words, 104 For a critique of the term >mediation<, see ROPKE 1990, 44 103 ; A> BENDLIN & al. , Numen 40 (1993), 92; ROPKE 1996a, 245-6. 105 See below, 3.1.4 and 4.1.1. 91 priestly authority and the religion is public institutionalization of indeed embedded in, and not fundamentally dif- ferentiated from, the socio-political domain. Yet, the assumption that the socio-political domain be the context in which religion is defined makes several other presuppositions which I find difficult To can to accept. 106 be sure, the identification of religion as a civic reli- gion, determined by city-state, is the institutions explicitly made, and ideals of the as we have seen, by Late Republican authors like Cicero and Livy. Both seem to sup- port the modern view that Roman religion was an affair which can adequately be understood in relation to the logical primacy of the socio-political realm. However, it will become apparent that these Late Republican texts cannot be disso- ciated from their contemporary cultural climate. Theirs is a holistic view that counters religious complexity with a normative model of civic control authorized by mos maiorum and retrojected into Rome's regal past. Although modern scholarship has readily succumbed to the illusion that such a holistic model accounts ancient model is for Republican Roman religion, this in itself compromised by an unacceptable vision of >religion and society<. 106 In 1.3 & 1.5-1.6, I have presented several cal reservations. methodologi- 92 2.5 Ancient views on >religion and society< Ancient intellectuals disagreed about the primary motive for social organization and the origins stotle of civilization. Ari- and his school advanced the idea that humans were by their nature (cpucei, natura) sociable and thus committed the formation to of the city-state as a means to promote what was good. The actual formation of the TtoXiQ was conceptualized by Aristotle as a CTUVOIWLOIJLOS. agreed. 107 The The Stoics and opposite view was held by Thrasymachus, the contract theorists from Antiphon and Democritus to and Cicero Epicurus Lucretius on the one hand, or by Polybius on the other: the prime reason for congregation and reflected was mere the weakness of mankind. 108 However, despite such disagreement there was concurrence on point. utilitarianism one fundamental Most of the »Kulturentstehungslehren« in the ancient ethnographic tradition, starting with Protagoras and Demo- critus in the late fifth century BCE, shared an evolutionary theory of successive stages of social congregation of in- creasing sophistication which culminated ment. This social theory in urban was developed by writers of the Hellenistic and Roman periods, among them Aristotle, rus, Epicu- Lucretius or Cicero. They perceived societal evolution to be a progress from the fragmented state of to settle- the collective embodiment of individualism social life in an urban settlement administered by laws and customs. In consequence, 107 Pol. 1252alff.; 1278b6ff. Stoics: SCHOFIELD 1990. Cicero: Rep. 1,39; Off. 1,157-8; passim. 108 Thrasymachus: Pi. Rep. 369 b-c; cf. Prt. 322 a-b. On the Epicurean concept of society bound together by utilitarianism, see LONG & SEDLEY 1987, 1,125-39. Polybius: 6,5,7. In general, cf. STEINMETZ 1969. 93 this theory presented the city-state, the contemporary realization of such a collectivity, as the primary form cietal of so- organization. As a corollary, individualism could be disregarded as a defunct stage in evolution. 109 This tradition took two directions. Varro in his On the as populus the a progress from Italy's pastoral origins to the foundation of Rome. 110 Yet, in his De re rustica following hand, De vita populi Romani, like other Republican writers before him, described the formation of Romanus one Dicaearchus, (2,1,3-5), he linked the progress of civiliza- tion to the topical argument of its moral decline and cated the return to advo- the virtues of the noble savage of a primitive pastoral society. By way of contrast, the portrait of the primitive Italian peasant soldier in Vergil's Aeneid, while drawing on the same ethnographic tradition of social progress linked to moral decline, was informed by the severe upheavals in the Italian countryside after the Civil War; and Vergil no longer provided an unqualified portrayal of the Italian warriors as noble savages. 111 A contrasting, and more influential, philosophical tra- dition was represented by Plato, Aristotle and the Peripatos or Cicero. In that tradition, the ancient city-state in its ideal form provided not only the most advanced form cial but of so- also of moral organization. This moral aspect was 109 Arist. Pol. 1252b 27-8: f| . . . XOLVCDVLO, TeXeuoc For documentation, see MOLLER 1972/1980; T. COLE 1990; NIPPEL 1990, 11-29. 110 NOE 1977, esp. 297-8. 111 Aen. 9,598-620. HORSFALL 1971 and DICKIE 1985 provide innumerable parallels for the ethnographic topos . 94 brought out in Cicero's definition of the res publica as res populi: »an assemblage of a large number of people associated by means of a common idea of what is right and shared utility.* 112 Aristotle defined the TeXoe of such a community as dyotdov TL (Pol. 1252al-7) - a good which any human being would wish to acquire. In the light of the Aristotelian definition of in- dividuals as £cjxx TioXiTLxd, the most likely interpretation of this passage is that any individual is sociable by nature, and that by their very nature humans have the innate capacity and the innate impulse for a life in the TioXis, since only the city-state can provide the opportunities for a virtuous life. The provocative Aristotelian theory that the city-state is by nature prior to the individual such a complements view. For the state's priority rests on the assump- tion that individuals can realize their virtuous life, capacities, and a only in the context of, and subject to, the city-state. 113 On the basis of the citizen's natural respect and striving for the virtues which the city-state it would thus guarantee the preservation of society as a whole and of its constituting elements. For nomic only eco- as a whole, the natural sociability of the TioXte would pre- serve individual property rights. 114 The bias instance, and legal affairs in the city-state would also be de- termined by the institution of the city-state since embodied, primarily ethical of this theory is apparent: since life in the communi- 112 Rep. 1,39,1 with SUERBAUM 1977, 1-37; SPRUTE 1983, esp. 153-5; WOOD 1988, 123-8; SCHOFIELD 1995, for discussion. 113 MILLER 1995, 27-66. 114 Cf. Panaetius frg. 118 van Straaten, Off. 2,73; ANNAS 1989. endorsed by Cic. 95 ty, and life determined by the community, was »according nature«, any conceptualization of to elements which were in potential conflict with that community was regarded as unnatural . To be sure, Hellenic and were not unaware of Roman political philosophers the distinction between »society« as composed of the inhabitants of a city-state and the as representative of the »state« citizens in their capacity as a political body. However, in their political theorizing disregarded that distinction. For instance, Plato prefaced his discussion of justice in the city-state with an of the TioXie they as a place of consumption inhabited by artisans economic and outline productivity and traders Rep. (e.g. 2,369 b ff.) . However, in his Nomoi he regarded the TIO\LQ as a political entity of citizens. 115 In Politics book one, Aristotle discussed the TioXts with respect to the its OLXICX as basic unit. There, the city-state was in socio-economic terms described as a place of production which included men, women, foreigners and slaves rather than just body. the citizen- In Politics 3, however, Aristotle gave a socio-polit- ical definition of the TioXiQ as a »XOLVCDVLCI of the citizens arranged in respect to their 7ioXiTEL<x«. 116 in his discussion of the different degrees of fellowship (societas) among human beings, Cicero perceived the city-state as providing a focus for the political, religious, legal and pects economic as- of social life. However, these realms operated in the socio-political framework of the civitas or res publica con115 SCHOFIELD 1993. 116 Politics: Pol. 3,3, 1276bl-2; society: Pol. Cf. HANSEN 1993, 16-8; OBER 1993. 1253 biff. 96 stituted by its citizens (cives) . For the the most significant res publica was of all human societates, just as the private realm was quasi seminarium rei publicae. This defi- nition implies an evolutionary process which led to the perfection of the individual and the family life in the polit- ical community governed by virtue. As a corollary, the individual's moral obligations towards the city-state were as more seen important than his obligations towards his family, friends, clients or mankind in general. The individual was, first of all, defined through his political affiliations. 117 As a result of their underlying moral bias, ancient litical theorists deliberately confused the two meanings of TioXie, civitas or res publica. They blurred between the city-state as »society« the difference on the one hand and »political organization* on the other. 118 It is not hard find an po- ancient »state«. Yet, terms linguistic like to equivalent of the modern term TioXie, noXLTELCt, KOLVcovua, res publica, populus or civitas are unsatisfactory renditions of the concept of »society«. They are not only closely linked to the city-state's socio-political organization but empha- 117 Off. 1,53-9, 160; 3,99 with WOOD 1988, 123-8; ATKINS 1990, 263-81. STRASBURGER 1982-90, 3,407-98 discusses De officiis as »Tendenzschrift« in its political context. Due to its politicized character, more general categories of duty are missing from De officiis, just as any obligations of guardianship towards clients or one's household are under-determined. For a very different picture of social obligations at Rome, see e.g. SALLER 1994, 74-132, esp. 105-114. Arrius Menander Dig. 49,16,4,15, in a discussion of absence from military duty without leave (emansio), specifies the duties towards the state as being inferior to individual concerns: ... datur venia valetudini, affection! parentium et adfinium, et si servum fugientem persecutus est vel si qua huiusmodi causa est. 118 MILLER 1995, 357-66. 97 size the community's collective nature and sociability (XOLVCDVLCX, societas) . It is in accord with their bias that underlying these terms do not conceptualize the modern dis- tinction between »community« and »society«. Rather, both realms become homologous. This holistic approach assigned primacy to the socio-political sphere, making it logically, prior if not historically, to other forms of social organization in the communi- ty. Individuality thus became subject to the superiority the social collective, whereas collective solidarity was legitimized with regard to its socio-political the moral benefits the of utility and individual would receive from the community. 119 From Plato onwards, this philosophical tra- dition perceived the plurality of the constituents of society in analogy to the plurality of the constituents of the human soul. Just as the the equilibrium of only be described soul could with a view to its necessary order in a unified and harmonious state of individual the psychological affairs, so morality and justice could be realized only in a uniform and peaceful socio-political community. 120 The ideal city-state was the best teacher of morality: such a tra- dition conceptualized differences between the individual and the city-state, between private and public only with regard to their necessary unity in the public domain. How did religion relate to this ethical theory? Aristotle omitted religious behaviour from his discussion. 119 STRASBURGER 1982-90, 1,423-48 and 3,9-127 provides an account of this tradition. 120 For this philosophical bias, see GADAMER 1978, 41-63. 98 Elsewhere in philosophical discourse, it was an element in the construction of a virtuous life in the civic Plato community. appears to be the earliest author in Greek literature who expressed the doctrine (aocpta, dv6pELa, of 6ixaio<j6vTi the four in Euthyph. 427 virtues and cjoxppoauvTi) . The virtue of oatOTTie is added elsewhere in the Platonic already cardinal corpus. 121 Yet, bl ff., TO Euaepes TE xai oaiov, defined as TO TiEpi TTIV TCDV dEcov dEpanEiav, is a subdivision of justice. This ethical connotation is the context in which the religious institutions of the Platonic city-state oper- ate. The subdivision of iustitia into its parts in Cicero followed the definition of the subordinate virtues of justice as ETTLCTTfJTJLOtL in Stoic doctrine which include EUCJEpEia or pietas. 122 Cicero defined justice as a assigned to mental habit which everyone his dignitas in relation to the cornm- unis utilitas. Religio, as one of the subordinate virtues of justice, belonged to the same level, being defined as a be- haviour in relation to what was useful with view to the communal realm. 123 In De natura deorum, this subdivision was also employed. To maintain fides, societas generis humani and, above all, iustitia, justice's constitutive parts, pietas, sanctitas as well as religio must be preserved. 124 This perspective represents a philosophical tradition which un- derstands religious behaviour not as a in any disinterested societal phenomenon sense but as a mental habit directed 121 Pol. 427 e33. For precedents, see Aesch. Sept. 610; Xen. Mem. 3,9,1-5. For the inclusion of OCJLOTTIC, see Prot. 329 c; Lach. 199 d; Men. 78 d; Gorg. 507 b. 122 Stob. 2,59,4ff. (= SVF 3,262-4), at 3,264) : EUCJEpEldV 6E ETILCJTfniTIV dEQV 123 Inv. 2,161-2. 124 ND 1,3-4. 2,62,2-3 (= SVF 99 towards the benefits of a virtuous life in the community. In this theory of collective mension of religious well-being, any analysis. di- behaviour such as personal belief or individualized cult acts are on an a priori from subjective basis excluded Indeed, in philosophical theory such a sub- jective dimension becomes homonymous with superstitio: it is not in accord with the ethical principle of human action as collectively informed ethical behaviour. 125 This intellectual tradition which saw society as tional form of political parameters, formed conceptual core of indigenous models of civic religion. As the law-giver of Cicero's De legibus created the laws his ra- organization, and which defined religion in relation to these political the a of state in accord with nature, and as nature entailed so- ciability in the city-state, the reflect religious law as the its col- identity. Greek and Roman writers variously expres- sed the collective character prosperity was due to of early Rome: its initial the collective achievements of the populus Romanus and the unity of its religious social populus a collective body, the original creation of its religious institutions needed to be reflective of lective must this exclusive focus on the socio-political domain. Since Livy set out to portray the formation of Romanus code beliefs and institutions. 126 When insisting on the city-state as 125 Cf. above, 1.1. 126 E.g. Polyb. 2,41,9; book 6 passim; Sail. Catil. 9. J. GRIFFIN 1985, 178-80 discusses these passages and surveys several modern studies which uncritically accept the stereotype of collectivity in Rome. The emphasis on an abstract Roman value system, including qualities such as concordia, pietas or fides, as the expression of a uniform Roman identity belongs to this tradition. For critique, see BARGHOP 1994, 41-8. 100 a form of political organization religion and the embeddedness of in this socio-political realm, historians of Clas- sical Greek and Roman Republican religions are close to cepting ac- this ancient model. In this respect, the difference is small between Wissowa's »Staatskultus« as the institutionalization of Roman religion and the modern notion of »civic religion* which had its centre in political life. Both cepts con- accept the refusal of the ancient political theorists to distinguish between the state and society or to assign place a to religion outside the socio-political domain. Oswyn Murray's position, while informed by a considerable theoretical awareness, represents this refusal in relation to the Classical Greek TioXus, but the reader is by now able to provide Roman parallels easily: »The polls as a rational form of political organization is the expression of the collective consciousness of the Greeks.« 127 I believe that this refusal is the result of what like to I should call the »Moses-Finley-syndrome«. Finley connected the absence of an ancient concept of the economy to the inability to conceptualize an economic system and the lack of a differentiated economic rationality in antiquity. Yet, as suggested economy was degree of I above, 128 Finley's own primitivist concept of the based economic on modernizing assumptions about the rationality which would be required to qualify as economy. In reaction to Finley's primitivist per127 MURRAY 1990, 19. This modern sentiment about the inseparability of society and politics in the Classical Greek city-state is shared by many scholars, ranging from Moses Finley to Christian Meier; for a critical assessment, cf. OBER 1989, 35-6. For the actual differentiation of the realms of politics and >society< in Classical Athens, see e.g. HANSEN 1991, 61-4. 128 See 1.1 & 1.2. 101 spective, which economic allow historians have developed an ancient economic rationality to reemerge. I would suggest that a similar primitivist in parameters strategy prevails many studies of Roman Republican religion. Scholars have accepted an ancient intellectual tradition of blurring the realms of society and state, as though this tradition gave a descriptive account of social reality. Yet, just as a de- veloped form of economy existed in the ancient world despite the absence of the concept of the economy, so the refusal of an entire intellectual tradition to as a societal conceptualize religion phenomenon outside the civic realm must not mean that religion existed only within the civic realm. As a matter of fact, modern tended to scholars have treat the ancient theorists as descriptive, even when these theorists themselves did not hide Plato's unwittingly their agenda. political programmes in the Politeia and the Nomoi, as the author himself made clear, were normative rather than descriptive. The same attitude informed ical Aristotle's polit- writings: they portrayed ideal collective constitutio- nal states which did not find their realization in the ferentiated present. 129 dif- Livy never claimed to describe the reality of Rome's religious system, but instead presented an ideal model of religious homogeneity in the regal past, which contrasted with the contemporary situation in Rome. 130 129 E.g. Pol. 1257 b 33-4; 1258 a 10-4; 1279 a 13-5; 1296 a 36-8; WINTERLING 1993, 182-3 and 205: »Aristoteles erscheint vielmehr ... als Theoretiker der stratifizierten Gesellschaft, der dieser dann freilich eine politischmoralische Neuintegration auf der Basis >politischer Tugend< vorschlagt und zugleich die Okonomie als Quelle der Desintegration wieder in den Oikos verbannen will.« 130 Livy Praef. 5, 9-13. 102 The >laws< in De legibus claimed to be intended Ciceronian for realization in a political context. 131 However, Cicero's const!tutio religionum was a normative account philosophy modelled on the mos maiorum. The author did not believe and that these laws stood a chance of becoming operative in- stantly. Instead, he thought that he wrote for people in the future (de futuris hominibus) whose education and instruc- tion (educatio et disciplina) was a necessary before any laws could be implemented. 132 As will be argued in detail later, these two Roman with prerequisite, social, authors, when confronted economic and cultural changes, responded with theories which attempted to reintegrate an increasingly complex society through an anachronistic »Sinntotalitat« was situated in the socio-political realm. However, their responses show that these authors perceived the Roman state and gruous, thus by which contemporary implication Roman (idealized) society to be incon- acknowledging the conceptual distinction between a particular political organization called >the state< and the wider, and more comprehensive, entity of >society<. 133 131 Leg. 1,17, 37, 57; 2.14 (non studii et delectationis sed rei publicae causa); 3.14 (ad usum popularem atque civilem) . 132 Leg. 3,29; cf. 1,58-62. Mos maiorum: ibid. 2,23. 133 For the distinction between organizational systems< and >social systems<, to which here I refer, see KNEER & NASSEHI 1993, 42-4. 103 2. 6 The impact, of modern theories of ^religion and society* It would no doubt be unfair to accuse modern scholarship blindly accepting this ancient Scholars often realize that this elite intellectual tradition of tradition. represented an perspective, a >little tradition< of upper-class dis- course, whose very stress on any form of political organization ideologically supported tion. 134 the elite's posi- Rather, I would suggest that ancient theorizing is thought to be compatible with, and thus forces, political conveniently rein- modern preconceptions about ancient society in gen- eral and ancient religion in particular. In a discipline which evidence in is seriously fragmented model-building that tries to take into account indigenous interpretations is not automatically a bad thing. However, in this case the result has been an unfortunate congruity of ancient and modern models. This congruity obscures the fact that the ancient in- tellectual models provide a welcome confirmation of received modern interpretative world. I believe that these modern preconceptions have been preconceptions about the ancient shaped by, and subsequently led to the endorsement of, three particular social theories of »religion and society*. 2. 6. 7 Social theories of differentiation Firstly, the insistence on the undifferentiated character of socio-religious life in the Classical city-state entails important, an though largely implicit, assumption about socie- 134 Cf. WINTERLING 1993. See below, 3.2, on >Roman religion< as an entity construed by the little tradition of elite thinking. 104 tal and cultural life in the ancient world in general. For this insistence amounts to the view that the differentiation of societal behaviour and cultural ideas, the experimenting with different and contradictory ability to choose belief systems, and the between different options in a >market- place< of cultural choices, characterizes modern rather than pre-modern society. The logical primacy of forms ical of polit- organization amounts to a >Sinntotalitat< which covers all aspects of life in the traditional city-state. viant social Any de- behaviour can then only be explained in terms of crisis of these traditional structures. 135 The parallel with contemporary sociological models of social and cultural complexity is illuminating. For sociologists define the political system in contemporary society as a subsystem of society as a whole. »Society« is structurally differentiated into interacting, yet independent, constitu- ents like the economic, the cultural or the tem. political sys- »Society« thus becomes an umbrella-term of social ana- lysis which describes the whole of its constituent parts. By way of contrast, sociological city-state as representative theory regards the ancient of a primitive stage in dif- ferentiation. 1 36 It occupies an intermediate position in the evolution of civilization: having succeeded the stage of primitive and segmented society, the city-state rationalizes increasing complexity through organizing social and cultural 135 Cf. above, 1.3. See below, 4.1.3, for the application of the >market-place< model to religion in the city of Rome. 136 E.g. PARSONS 1966; 1977; Id. 122-41. 1980-89; DOBERT HAHN I973a; 1986; Id. I973b; LUHMANN KNEER & NASSEHI 1993, 105 life in developed socio-political systems. Increasing com- plexity results in the differentiation of cultural and religious choices. However, this internal stratification: complexity its is controlled by processing is limited to the closed system of elite communication, while remaining sub- ject to the determinative frame provided by the socio-political system's primacy. The final stage of differentiation characterizes modern society: here individuals are no longer exclusively defined through stratification or through their belonging to a single socio-political entity. At this stage, individuals use several distinct interpretative models, just as society itself develops differentiated domains of social and cultural activity which are no longer determined by the socio-political primacy of the state or by stratum differences . These sociologists are not primarily interested evolution of pre-modern societies. their willingness to accept a evolution which seems, model of historical increasing complexity as a succession of discontinuous epochs. model the This fact may explain linear conceptualizes in However coherent this the smoothly linear increase of complexity in society as a result of the gradual detachment of social and cultural life from the socio-political realm entails a highly problematic teleological concept of history. 137 Yet, it is no doubt the linearity of such a model tractive that proves at- to those ancient historians who postulate the pri- macy of political organization in the traditional ancient 137 Cf. GLADIGOW 1995, esp. 24-5. The distinction of different historical epochs is in itself a modern construct; cf. GLADIGOW I997b. 106 city-state. As discussed above, the thesis that all forms of societal activity in the ancient city-state (economic, cul- tural as well as religious) can be understood as being negotiated with a view to the socio-political sphere entails the unspoken assumption that societal evolution is a linear de- velopment from embeddedness in a unified state of affairs to complexity and differentiation. Once more, the discussion about the ancient economy poses the weakness of this assumption. 138 As we have seen, Finley's primitivist position denied that the ancients ceptualized con- their economy as a differentiated system. It is acknowledged that the economies of the became ex- ancient city-states increasingly complex domains. However, their partial disintegration is held to have been counterbalanced by the fact that they remained integrated into and dependent on the socio-political system. Roman Empire promoted Yet, Similarly, individual the trade expansion and of the productivity. the development of trade and economic exploitation ap- pears to rather have been constrained by primarily political, than merely economic, parameters which were tailored to Rome's oligarchic elite. This view, however, entails danger of the over-primitivizing. For economic historians have come to realize that the model of a »political economy« cannot fully explain the existence of regionalized economic systems in Republican Italy and under the Empire. 139 Indeed, the concentration on the political institution of the city- state tends to overlook that the socio-political 138 Cf. above, 1.1 & 1.2. 139 NORTH 1981; WOOLF 1990; Id. 1992. elite did 107 at the same time develop a strong financial interest in dissociating itself from an urban market economy and, indirectly, from the determinative force of the urban centre. 140 Nor does the emphasis on the socio-political sphere sufficiently take into account the role of the ancient city as a market- place as well as a religious and recreational centre for a rural peasantry which was not much involved in the political life of the urbs. 141 Even within the physical city of Rome, the separation of the economic and the political life became more and more visible. From the third century onwards, trades like those of the butchers and fish-mongers were ex- pelled from the Forum Romanum which, by the was vile left first century, to civic representation - and to luxury tabernae. In its turn, however, the creation of separate market spaces for specialized sale (the macellum or the different cial fora) resulted commer- in the concentration of resources and the further internal differentiation of the city's economic life. 142 As a result, economic historians have reformulated nomic eco- parameters with a view to a mercantilist system which was characterized by a significant level of productivity and trade interest and, at the same time, developed differentiated economic roles which were no longer congruous 140 WHITTAKER 1995; BENDLIN 1997, 42-3, with po- 58-63. 141 GARNSEY 1998, 107-32. 142 DE RUYT 1983, 158ff.; FRAYN 1993, 1-37, 56-73; G. PISANI SARTORIO, LTUR 3 (1996), 201-3. For artisans and luxury tabernae on the Via sacra, see BOMER on Ovid Fasti 5,129-30, with PURCELL 1994, 659-67 on their economic importance. For an account of how the development of the city's public spaces reflected this process of differentiation, see N. PURCELL, LTUR 2 (1995), 325-42. 108 litical ones. 143 Going beyond the primitivist dualism of primitive subsistence economy and market-oriented capital- ism, it becomes thus possible to apply a more nuanced interpretation to the economy: true, it was not an example of >structural differentiation< in have that its structure could been entirely dissociated from the political organiza- tion of the city-state; yet, it certainly operated in terms of a >functional differentiation< in that it developed functional realms which were no longer homologous with the po- litical system. What is schematic at stake is the inappropriately dualism of a traditional model of differentiation which assigns logical primacy to the political sphere on a priori basis. As I shall demonstrate below, the distinc- tion of structural and functional differentiation will mit us an per- to develop a modified model in relation to religion as well, which explains the concurrence of interdependence and independence in the relation of sacrum and publicum. 144 2.6. 2 Functionalism The traditional emphasis on the undifferentiated character of pre-modern societies inevitably leads to a second assumption, which is shared by anthropologists and ancient his- torians. For when focusing on the collective nature of religious meaning in a political organization, ancient his- torians tacitly adopt the view of religion as an essentially collective phenomenon. This view was originally outlined by 6mile Durkheim in his Les formes elementaires de la vie 143 For references, see above, 1.3. 144 See below, 2.8, 4.1.1, 4.1.3. re- 109 ligieuse (1912) . Durkheim argued that religious ritual reaffirms and authorizes social facts in terms of collective human consciousness. It is noteworthy that Durkheim's tionalism was indebted func- to the idea that ancient societies found collective solidarity in the city-state and therefore modelled their religious identities according to the state's collective principles. Durkheim developed this idea through contact with N. D. Fustel de Coulanges, the author of La cite antique. Etude sur le culte, le droit, les institutions de la Grece et de Rome (1864). 145 Since Durkheim's anthropological hypotheses concerning religion's role in primitive societies have found little favour with his modern the continuing attraction of critics, his religious sociology for ancient historians must lie in the usefulness of its theoretical implications. A neo-Durkheimian position has utilized Durkheim's theory of collective solidarity without accepting his crude dichotomy of >sacred< and >secular<; and the civic model endorsed by historians of Graeco-Roman religion capi- talizes on the neo-Durkheimian functionalism concerning ligion's affirmative status re- in a political community. The concept that religion is totally embedded in a collective political community and therefore only comprehensible in its public form is already prefigured in Durkheimian thinking. However, the problems of this position are manifold. The methodological shortcoming of any purely functionalist definition of religion has been discussed above. In addition, the neo-Durkheimian perspective of portraying ancient reli- 145 For Durkheim's religious sociology, see conveniently MORRIS 1987, 106-22; JONES 1991; LUKES 1992, 450-84. On Fustel de Coulanges' influence, see WARNKE 1986; JONES 1991, 99-104; LUKES 1992, 58-63. 110 or political ritual as representation collective of a gious communal identity a priori postulates ideological at congruity the object level but excludes from its analysis dissent- ing voices. 146 2.6.3 A symbol theory of culture Thirdly, the prominence of a neo-Durkheimian perspective among ancient historians is also explicable by the interpretative models which a developed Durkheimian functionalism presents to our discipline as being compatible with our own preconceptions. We saw how religious studies have challenged the traditional notion of belief in ancient religion. As I have suggested, this view is based on »internal« and the dualism of the »external«. It denies the importance of internal motivational processes for the explanation of tural the cul- or religious processes in Roman society. Instead, the role of religion is perceived in (dualistic) terms of external activities or events, while individualistic emotions, thoughts beliefs or feelings become marginalized. 147 The affinity of such a perspective with Neo-Durkheimian goes largely thought unnoticed. For when placing Roman religion in the context of a culture which was determined by the character or public of its political organization, ancient historians accept the definition of culture as a »public symbolic sys- 146 See above, 1.5. Endorsing a neo-Durkheimian tradition, several ancient historians make the aspect of political collectivity central to their theoretical framework: e.g. ZANKER 1987; MURRAY 1990, 18-23; HOPKINS 1991. For a comprehensive critique of this tradition, ancient historians must still turn to sociological studies such as LUKES 1975 or BELL 1992. See further below, 2.7.1 and 2.7.3. 147 Cf. above, 1.4 and 2.3. Ill tem« which betrays the direct influence of neo-Durkheimian cultural anthropology. For instance, as we have already noticed above, the thropologist Clifford Geertz defines culture, including re- ligious culture, as a public system less than emotions in which thoughts not take place system« it in people's heads, but is located in public symbolic action shared by a social group. symbolic no are, by definition, excluded. Instead, culture is a public affair. Though it consists of ideas, does an- directs A »shared and structures cultural activity and thus preserves the meaning of communal communication. On this view, culture emerges as a cognitive system which the historical agents This Geertz calls »common train of thought betrays the direct influence of a Durkheimian transmitted action define and understand on the basis of shared value systems, which sense«. of perspective of organized solidarity as through the neo-functionalism of, for instance, the American sociologist Talcot Parsons. 148 As soon as religion is understood as such a »public symbolic system« which can be analysed in terms of one's public behaviour, the constituents of that system are merely intersubjective elements shared by this public community. As a result, this Geertzian perspective construes several dualisms - »thought« and »ac- tion«, »internal« and »external«, »subjectivity« and »social »private« experience«, and »public«, »individual« and »culture« - which become mutually compatible: »To undertake the study of cultural activity - activity in which symbolism forms the positive content - is thus 148 For a critique of this tradition, see conveniently KNEER & NASSEHI 1993, 35-7. 112 not to abandon social analysis for a Platonic cave of shadows, to enter into the mentalistic world of introspective psychology ... Cultural acts, the construction, apprehension and utilization of symbolic forms, are social events like any other; they are as public as marriage and as observable as agriculture.« 149 Geertz' definition has been criticized as reductionist by social anthropologists. To be fair to Geertz, his definition provides a heuristic perspective which excludes psychological processes from the anthropological analysis of »re- ligion as a cultural system« (the title of one of his es- says) . Yet, he acknowledges the fact that the study of religion is a twofold undertaking, and that the public, or sym- bolic, meaning of religious operations would, in world, psychological be constraints. related His to their provocative social definition and of an a ideal religious culture as public must therefore be seen as a disillusioned, if necessarily preliminary, reaction against earlier psychologizing interpretations of religious meaning. 150 Geertz, however, did not pursue his interpretations beyond the liminary status of this heuristic device; and this may be the reason why its methodological incompleteness acknowledged pre- is rarely by his followers. The neglect of the relevance of internalized belief systems for the analysis of religious behaviour, a heuristic perspective for Geertz, must the attraction of this explain theory to ancient historians, who categorically deny the significance of individual commitment as a field of research in the study of Roman religion. Once religion is defined as a »public cognitive system«, the cit149 GEERTZ 1973, 91. Common sense: GEERTZ 1975. 150 GEERTZ 1973, 124-5. For a critique of the incompleteness of Geertz 1 theoretical framework, see BELL 1992, 25-34; WHITE 1992. Cf. above, 1.4. 113 izens of the city-state share in the intersubjective symbolism of civic behaviour, civic secular or religious rituals of the festive calendar, or the religio-political rituals conducted by magistrates and priests. Religion has become an exclusively public affair. 151 This mistrust of internal aspects of social life is complemented by a semantic theory of communication which postulates the abandonment of textual culture of a people is subjective meaning. »The an ensemble of texts« is, again, Clifford Geertz' formulation. 152 Indeed, this view is a ne- cessary consequence of the definition of culture as a public affair. As soon as behaviour is publicly shared symbolic behaviour, the individual subject can his or be dissociated from her linguistic or bodily utterance which has become public symbolic knowledge. On this view, >meaning< is loca- ted in the realm of shared societal communication, which has become independent of the subjectivity voice. The definition of culture as public and the de-individualization aspects and symbolic action of communication are comple- mentary processes: cultural behaviour individual of the authorial is stripped of any bound to communal cognition; texts are subject to communally determined meaning. The critics of this position hold that such a semantic theory interrelation of reduces the self and environment - of subject and ob- ject - to the interrelation of sentence and textual fact and 151 See e.g. PRICE 1984, 7-11, at 8, acknowledging his debt to Geertz; HOPKINS 1991, mentioning the influence of the anthropologist Maurice Bloch; FEENEY 1998, 15, in a discussion of the contributions of Sperber, Gouldner and Foucault: »... the individual's psychology cannot be the ground for social institutions*. 152 GEERTZ 1973, 452. 114 thus curtails the individual voice which refers to its tus as subject. Culture, thus reduced to symbols of sta- society or the human psyche are public representation, but it becomes difficult to define what the objects of such symbolic representation - or its audience - are. It has therefore been suggested that one process of must reverse this interpretative decontextualization by reconstructing the self- referential subjectivity which a theory of intersubjectivity denies. 153 However, the model of popularity intersubjective meaning of the as the expression of a interpretative >Roman identity< located in rituals, texts, myths and indi- genous exegeses betrays the on-going influence of this theory among classicists. 154 2. 7 »Closed« versus >open« religious systems As a result of these various intellectual influences, recent studies of Roman religion have developed several theoretical compatible concepts to circumscribe the homology of »reli- gion« and »civic religion« in Republican Rome. The following list is not intended to be comprehensive; the most yet it important concepts: >undifferentiatedness<, >col- lectivity <, >the political organization's logical >the includes primacy<, externalism of societal activity<, and >the public na- ture of religious culture<. As we have seen, constituents entails any of these various methodological problems. How- ever, what is an inevitable consequence of these interpreta- 153 See HABERMAS 1981, 1,530-2; CRAPANZANO 1992, 297-301. 154 See 2.7.3, for documentation. 115 tive models, namely the portrayal of religion as part of a »closed system« / seems to be their most serious shortcoming. 2.7.7 Religion's role in a ^closed system* Prima facie this criticism might appear unjustified. For outlined above, the as revaluation of elite attitudes during the Late Republic does effectively combat traditional as- sumptions about the disruption and decline of Roman religion in earlier scholarship. Instead, as soon as religion is closely linked elite's interest in cult and ritual can be reinterpreted as to the political domain, the political being preservative of a thriving religious system. 155 At the same time, these scholars emphasize that civic Republican in Rome was not a static system of rituals and pre- scriptions whose preservation was left to while religion the aristocracy, other groups of society sought different forms of re- ligious experience. Instead, they illustrate how innovation, creative development forms allowed a and fluid reinterpretation centuries. 156 In traditional civic religious system to adapt to changing socio-political circumstances first of particular, in the recent second and scholars have stressed the extent to which the stability of ritual forms, its orthopraxy, was compatible with the fact that its »meaning« in society could change over time. The openness of 155 See above, 2.3. Cf. CANCIK 1985-86; LINDERSKI 1995, 608-25, for the vitality of religion in Late Republican Rome, and BEARD 1994a, 734-49, on the restoration of temples and continuing elite interest in traditional cults. 156 NORTH 1976; Id. 1980; Id. 1986, 251-4; Id. 1989, 616-624; BEARD & CRAWFORD 1985, 36-39; SCHEID 1985; BEARD 1994a, 739-55. On Greece, cf. e.g. BREMMER 1994, 94. 116 and Roman sacred law and ritual to creative change adapta- was a prerequisite of its preservation. 157 The adapta- tion cen- tion of civic religion to outside influences is thus a element of its vigour rather than a symptom of its de- tral cline. Nevertheless, these scholars present a in which, »closed to the nature of the concepts used - undif- due organization's ferentiatedness, collectivity, the political logical system« primacy, the externalism of societal activity, and the public nature of religious culture -, any religious phenomenon is classified on the basis ligious system or otherwise it must is an a priori dual it can either be incorporated into the local re- structure: there of a problem be marginalized. Yet, with such a view in that the a priori ex- assumption that identifiable civic religious identities ist and can be reconstructed from public communication inestends capably to exclude from its analysis any voluntary religious belief systems and religious activities that could to be regarded as being external and potentially disruptive the communal life of the city-state. Any such religious phenomena need to be marginalized on a priori grounds. To be sure, it is often held that state control over civic religion became manifest in the suppression of deviant religious behaviour. Even when disregarding the notorious case of the Bacchanalian affair of civic interference 186, 158 the record of with the cultural and religious life of 157 BEARD 1987, 8-10; HOPKINS 1991, 482-3. Sacred law ritual: RADKE 1980; NORTH 1976, 3-9; ROPKE I996b. 158 por an assessment, see below, 3.4.1 and 4.2.3. and 117 the city in the second and first centuries BCE seems impres- decree senatorial by sive. Philosophers and rhetors were expelled 161, although this measure does not prove a con- in sistent Roman attitude in the second century. For in 155 the Athenians decided to send three representing philosophers, three of the four major philosophical schools, on an embassy to Rome; the Athenian officials no doubt expected a positive to response that decision from the Roman authorities. 159 Jews Astrologers were expelled in 139. The expulsion of the year same was presumably caused by the fact that the factor Jewish community, while being a significant life religious in in the of the city, could be accused of forming an publicly illegal association and performing unacknowledged rites rather than by a direct Jewish interest in proselytizing. 160 The fifties and Capitoline early cult of Isis was suppressed in the forties BCE. 161 The philosophers and magicians expelled from Rome in the second half of the first century CE are also likely to have been regarded as illegally organized groups which threatened public order. 162 However, in any of these cases it would be wrong to forget the limits of elite should it control in Republican Rome. Nor be implied that the Republican elite agreed on a 159 Expulsion: Suet. Rhet. 1,2; Cell. 15,11,1. Embassy: Cic. De Or. 2,155-60; Rep. 3,8ff.; Plut. Cato mai. 22; Cell. NA 6,14,8-10. For the second century context, see GRUEN 1992, 223-71, at 232-3. 160 Astrologers: Val. Max. 1,3,2. Jews: Val. Max. 1,3,3: ludaeos .. . qui Romanis tradere sacra sua conati erant idem Hispalus urbe exterminavit arasque privatas e public-is locis abiecit; Serv. Aen. 8,187: ... ne quis novas introduceret religiones; GOODMAN 1994, 68, 82-3; NOETHLICHS 1996, 13, 154-5. 161 See below, 4.2.3. 162 Philostr. Vita Apollon. 4,41; KNEPPE 1988. 118 consistent body of rules and expectations when determining disreputable social and religious behaviour. 163 Moreover, it is not at all clear whether the aristocracy would have been genuinely interested in consistent shall repression when, as argue below, Roman culture and society were not mono- lithic entities but a fluid system of great and little ditions and their respective, is more infrequency of deci- to expel groups from Republican Rome suggests that it worthwhile repression to ask how individual which of responded to the a new cultural complexity posed to the socio- political system; and how they symbolized individual instances in Rome, however erratic and motivated by tempo- rary socio-political concerns they were, threat tra- if only loosely connected, interests. 164 Moreover, the relative sions I the attempts of members among the socio-political elite to rein- force the elusive primacy of the political realm in order to temporarily regain control over what had become an increas- ingly complex market-place of cultural options. 165 In retrospect, the influence of any official decision of expulsion and repression appears to have been limited. To be sure, domestic authorities which regarded of the establishing a permanent fire brigade as a potential threat to stabi- lity - »people who gather together into the same group soon 163 On the limits of elite control, see Livy 25,1,6-12 and, in general, WISEMAN 1994, 57-67; NIPPEL 1995, 22, passim. Cf. PHILLIPS 1991b, 269: »[N]either the legal, religious, or scientific systems had an interest in precisely defining unsanctioned religious activity.« 164 See below, 3.2. 165 For an illuminating study in these terms of two events of the early second century, see GRUEN 1990, 34-78 (the Bacchanalian affair of 186) and 158-70 (the burning of >Numa<'s books in 181). 119 become a political faction« 166 - also viewed religious group behaviour socio-political rather than as a religious a as securitas, problem. In these instances, it was not and a religious concern, that mattered. The case of civic policing the socio-political, rather than the religious, illustrates anxiety. dimension of the state authorities' and statutes, municipal models, prohibited coetus (illegal gatherings) coniuratio colonial Roman administrative on drawing The as well encroached upon sodalicia and collegia held and for that purpose - although sodalicia and collegia in selves not were as forbidden. 167 them- The urban authorities were clearly expected to prosecute illegal associations. However, the phrasing and positioning of these statutes entails civic that taking action was motivated by suspicion of organized groups. It is also significant that responded to individual the civic authorities indictments and often made ad hoc restrictions. The Senate's measures concerning religion in the Republic were a political, and not a religious, phenomeThe non. Senate administrative structures and restricted regulated the access to different religious choices, but it never interfered with the religious sphere as such. The situation may rather interacting with the be described as the political system's religious system's fringes, without ever penetrating to its centre. Furthermore, the subject of crisis and decline Republican religion, which of Late had featured so prominently in earlier accounts, seems to have 166 Trajan ap. Pliny Ep. 10,33-4 167 LUrs ch. 106; LFlav ch. 74. been shifted rather than 120 truly resolved by recent scholarship. For as the advocates of the model of civic religion surmise, it is the differentiation of religious a result choices in the Mediter- ranean world from the fourth century BCE onwards at the test that civic religion, of la- no longer sufficiently able to incorporate new choices, would itself become disintegrated. The exponents of this model must envisage these developments within the strict limits of civic religion in the ancient city-state; any development beyond these limits can only perceived in terms of the system's failure to integrate in- creasingly differentiated choices. According model, in the to the emergence of autonomous through religious which was the groups and practices and through the development of religious se, civic Late Roman Republican and triumviral period deviation from public religion became manifest increasing be experti- independent of traditional religion. At the same time, philosophical reflections on religion were marked by a highly rational and sceptical attitude. This increase in alternatives of religious and cultural behaviour resulted in a structural differentiation of the religious system, pointing to a religious pluralism, and a changed of religion's role in evaluation society, that the public system at Rome - and that is the crucial point - could no longer fully integrate. 168 Ultimately, religious development from was moving the embedded religion of the traditional city-state to the differentiated religion of a complex empire-wide and socio-politically fragmented environment. Religious develop- 168 E.g. NORTH 1979, esp. 96; Id. 1986, 253-4; CORNELL 1991, 59; BEARD 1994a, 755-63. 121 ment became synonymous with the disintegration of civic re- ligion. 169 On such a view, the history of religion in the ancient world would indeed become the history of the destabilization and eventual dissolution of civic religion. 170 This tension in the model of civic religion becomes fully apparent, the once model is no longer exclusively applied to the socio-po- litical local context of the classical city-state, but in- stead is employed with a view to a supra-regional context of change and differentiation in the Roman Empire, incidentally reminding us of Wissowa's local religion that, once would concept of Roman religion as a exposed to the outside world, lose its communal religious identity. 171 The termino- logy used can describe more complex religious phenomena only as deviations from the approved norm; and >religious pluralism< or the the category of notion of >differentiation< receive an inherently negative connotation. 172 It seems as though the model of religion in Late Republican Rome presented by these scholars is fundamentally incompatible with social structures that exceed a certain level of complexity. Their emphasis on religious homogeneity, based on ceptualization of Roman religion the con- as undifferentiated and 169 NORTH 1992; RIVES 1995. 170 BENDLIN 1997, 47, for a critique. 171 Cf. above, 2.2. 172 Cf., e.g., NORTH 1976, 11; Id. 1992, discussing >pluralism< in the context of the failure of the civic model. BEARD 1994a, 755-68 uses >differentiation< to denote the increasing fragmentation and disintegration of the religious system of Late Republican Rome. RIVES 1995, 245 parallels >pluralism< and >anarchy<: »the religious pluralism, not to say anarchy, of the empire reflected the absence of any organized system of official religion.* Cf. above, 1.3. 122 collective public activity, only works as long as it is possible to marginalize factors that introduce complexity. fails to work, the chosen When this strategy that kind of model of civic religion needs to be abandoned. From what has been said so far, it should that these scholars perceive different religious choices become clear an implicit tension between which are potentially incom- patible and see a religious system's survival resting on the exclusivity of a fixed set of religious options rather than on the cohabitation of alternatives. In a sociological lysis, ana- such a model of Roman religion would be described as representative of a >closed system<, whose organization complies with a linear causal principle. 173 What does the cription of des- civic religion in terms of a closed system en- tail? Any system, whether psychological or social, is determined by the input it receives. To describe this phenomenon, system theory has developed the notion of the In complex open >black box<. systems, the input entering the system as well as its output can be observed. Normally, however, input and output differ, and as soon as the input does not directly determine the output, the system's organization of the relationship of input and output cannot be observed. The box is >black<; and we must assume that the system is self- determinative. By way of contrast, in closed systems the box remains >white<: the input does directly determine the sys- tem's output, and their relationship follows a linear causa- 173 For a classic theoretical account (and critique) of the sociological concept of >closed systems<, see VON BERTALANFFY 1951. Cf. KNEER & NASSEHI 1993, 20-3. 123 lity which can be observed from outside, because it is not further determined by the system itself. In the case of the civic white: as religion model, the box is the model's advocates inextricably link religion to only one relevant input system, namely the socio-polit- ical realm, it is that particular realm's input which exclusively enters and thus determines the religious system. Such a closed system, based on a purely linear relationship be- tween input and output, aims for equilibrium. Such equi- librium, however, as we have seen, can no longer be maintained once this relationship is compromised: whereas the once determinative input-system, the political domain, ceases be in to control of the input into the religious system, var- ious other inputs, consisting of new forms of religious thority, lead to religion's au- differentiation; the box has suddenly turned black. Prior to this process of differentiation, the equilibrium of the closed system was homoeostatic: the civic model's emphasis on religion's adaptation to pos- sible change entails that the religious system may have been capable of adjusting to changing social circumstances. Yet, any change was dependent on external regulation by means of the political system's input. It is a crucial element of the civic model that the religious system itself is incapable of self-determination or organizational variability. contrary, any such processes of self-regulation are On the percei- ved as a form of dissolution by the civic model's advocates. The emphasis character of on the undifferentiated and collective Roman religion and the public nature of reli- 124 internal structure. its for gious culture as a whole has a further implication For as we have seen above, any private indi- religious cults or the religious rituals performed by viduals are understood as incorporated into the civic frame, expressions of a communally shared reli- miniature forming gious identity. Such an externalist view can haviour of be- religious course sidestep the concept of polytheism. of op- For in any society choosing between different religious tions does not present any difficulty, as long as these opdeter- tions are controlled by a collective identity, which mines individual religious behaviour. However, such a view leaves inexplicable the variety of different cults and divinities which existed at Rome. As a matter of fact, there a tacit contradiction (which has never been sufficiently addressed by these scholars) between the the closed system religions addressees polytheis- consisted of various differentiated forms of religious behaviour and a large number of divine potential of worship. 174 The failure to address this issue has to be linked to the modern conception of Roman as of religious homogeneity of civic religion on the one hand and the open principle on the other that ancient tic is religion a predominantly public affair taking place in the frame- work of a communally shared value-system, and to the denial of individual motivations which might have been deviant from such a civic frame. 174 For further discussion of this aspect, see below, and 4.3. 4.2.5 125 2. 7. 2 The openness of Roman religion However, before moving to these internalized processes it is important to stress that on externalist grounds theory cannot acknowledge alone this the extent to which religion at Rome was an »open system«. The Romans, like the Greeks and other ancient peoples, perceived the gods to be historically and logically prior to the city-state and not restricted to a particular ethnos or civic community. On the contrary, the gods were perceived to travel across the temperate Europe Mediterranean and constantly adopt new local identities in the places in which they received worship. time, they At the same brought with them a distinctly supra-local per- sonality and a biography which extended back into the thus and past, precluding total identification with their new father- land. 175 Pagan Rome continuously extended her local official pantheon through the addition fourth and of new divinities. In the third centuries BCE alone, the Romans witnessed the establishing of a large number of new civic cults in the city, including several transfers of foreign deities in re- sponse to portents. 176 Dignus Roma locus, quo dens omnis eat (Ovid Fasti 4,270) - under the empire, Rome could, with rhetorical hyperbole, be perceived to be the ETILTOUTI Tfie OLXOU(Athen. Deipn. l,20b), where all deities resided and 175 The gods prior to the city-state: e.g. Varro RD fr. 5 Cardauns; Cic. ND 2,5; Livy 1,19,4-5. A similar point is made with respect to Greek religion by S. G. COLE 1995. Divine supra-local identities: Livy 42,3,9: ... tamquam non iidem ubique di immortales sint; Apul. Met. 11,26; cf. BENDLIN 1997, 61-2. 176 See below, 3.1, for discussion. 126 received worship. 177 The prosperity well of the city-state as as of the Empire was directly linked to all those dei- ties which received worship at Rome. The ritual procedures following the deaths of members of the imperial family in the Early Empire serve to this mentality. The deaths of C. illustrate Caesar in 4 CE and of Germanicus in 19 led to iustitia, as a result of which the temples of all gods (templa deorum) were closed spanning the period from the death of the individual concerned until the end of his burial. 178 The Tabula Hebana stipulated the an- nual closing of all temples in the city and those within one Roman mile around Rome on the anniversary of the death of Germanicus. The temples of all deities were liability, concerned; any however, was passed on to those in charge of the respective shrines. 179 On the anniversary of the death of C. Caesar, sacrificia publica, supplicationes (for which the 177 Cf. Ovid Trist. 1,5,70; Luc. 3,91; Min. Pel. Oct. 6,1 (the pagan Caecilius speaking); HSA Aurelian. 20,5, and, for the Christian response to this claim, Arnob. Adv. Nat. 6,7; Prudent. C. Symm. 1,189. In later antiquity, Rome could be presented as the templum mundi totius: Amm. Marc. 17,4,13 with FOWDEN 1993, 45-50. 178 C. Caesar: ILS 140 (Pisa, 4 CE, adopting regulations set up for the capital): cunctos veste mutata, templisque deorum immortalium balneisque publicis et tabernis omni­ bus clausis, convictibus sese apstinere, matronas quae in colonia nostra sunt sublugere; cf. the Fasti Cuprenses (Inscr. It. 13,1, pp. 243-8). Germanicus: Tabula Hebana (RS no. 37) line 55: templa deorum clauderentur; WEINSTOCK 1966, 892-3. The editors of RS ad loc. compare a decree for Apollonis from Cyzicus from 25-50 CE, M. SEVE, BCH 103 (1979), 327, lines 41-4: xXeiadfivaL TE TO. uepd xai TO, TEUEVTI xat rcdvTac TOIJC vaouc. 179 Lines 57-9: ... templa deorum immortalium quae in urbe Roma propriusve urbem Romam passus mi lie sunt erunt quot annis clausa sint idque ut ita fiat ii qui eas aedes tuendas redemptas habent habebunt curent. For the validity of urban regulations in territory within one Roman mile around Rome, see the Tabula Heracleensis (RS no. 24) line 20; Livy 34,1,3. 127 temples of all deities in Rome would be open), marriages, convivia publica and ludi scaenici circensesve were ded. 180 The suspen- disturbance caused by the death of a member of the imperial family not only curtailed public secular ness, but closing of also all extended busi- to a divine plane. The temporary sanctuaries, entailing the artificial >scarcification< of contact between humans and gods, thereby came to dramatize a situation in which death had encroached upon human relationships. Under these circumstances, the Roman imperial authorities addressed all the deities of Rome and of its immediate hinterland. These imperial regulations, however, are also tive illustra- of a change: a funus publicum was apparently first or- ganized in 23 following the death of Marcellus. 181 Under the Late Republic, iustitia were held on certain public occa- sions relating to military disaster or to social upheaval in Rome. This could involve the closing of secular places for gathering such as baths or tabernae, whereas secular business, jurisdiction public and Senate meetings were postponed. The term iustitium was understood as a iuris quasi interstitio quaedam et cessatio on certain days when legal was suspended (Gell. 20,1,43). When describing iustitia, Late Republican authors addressed the suspension of business, juridical or business political, secular and the necessity of a military dilectus on the declaration of a state of 180 ILS 140,25-30. It is noteworthy that these only applied to the domain of publica. 181 Cass. Dio 60,27,4; KIERDORF 1980, 138. emergen- prohibitions 128 cy. 182 By way of contrast/ the closing of temples during iustitia, like the introduction of anniversaries in commemoration of the deceased, appears to be an non. imperial instance, a iustitium was presumably held as part For of the funeral of Sulla in 78. That occasion, pears phenome- however, ap- to have been necessitated by a civic desire for secu- rity in the aftermath of Sulla's death, permitted since the measure the policing of the populace of Rome. 183 Similar- ly, Tiberius Gracchus' closing of the temple of Saturnus in the course of a iustitium (Plut. Ti. Gracch. 10) ought to be understood as an attempt to prevent the running of the aerarium. These exceptions provide indirect confirmation of the thesis that Republican temples were in during iustitia. general not It is not before the imperial period that writers explicitly link the closing of all temples occasions. closed to such It is with the rise of autocracy that the iusti­ tium is imbued with a religious meaning tailored to the public funeral of an individual. 184 Returning addressing to the desire for comprehensiveness in the gods, Roman pietas extended beyond the realm of local religion. The letter of the praetor M. Valerius Messalla to the city of Teos in 193 BCE embodies this view. 182 E.g. Cic. Brut. 1995, Livy 3,27,2, 4,31,9-32,21, 9,7,6-15 (tabernae); Plane. 33 (auctions), Ear. Resp. 55 (aerarium), 304. Cf. MOMMSEN 1887, 1,263-6; KUNKEL & WlTTMANN 225-8. 183 Granius Licinianus 36,25-9 Criniti. In other respects, Sulla's funeral foreshadowed the fimera publica of the imperial principes, although the matrons' mourning for Sulla over a period of twelve months, reported by Granius, must not be linked to the iustitium: App. BCiv 1,105-7; ARCE 1988, 17-34. 184 E.g. Lucan 5,115-6: tempi! ... fruuntLrr iustitio. See further below, 4.1.4. 129 The document argues that out of their particular or pietas, toward the gods in general the Romans feel obliged to Dionysus, the divinity of the city of Teos. The good- will (suiXEveua) which Rome experiences from the divine is direct result of that particular pious attitude. 185 It was an axiom that the res publica of the Roman People »augmented by a the resources had been and councils of the immortal gods«. 186 What is remarkable about that document, the Lex Gabinia Calpurnia de insula Delo of 58 BCE, is not just that it states with imperialistic directness that the Romans owed their political and military successes to divine favour, but also that Roman was due to the involvement of di success immortales, whose identity transcended the narrow confines of civic religion. This principle can be observed throughout the later history of pagan Roman religion, from the SC de Cn. Pisone patre of 20 CE empire-wide to Caracalla's announcement supplicationes to the immortal gods in 212, the supplicationes requested by Decius in 249/50, nouncements of leading up to or the pro- the persecution of Valerian in 257, all of which were phrased in such terms as to demand the recognition of the gods as such. 187 185 SIG3 601,11-7. The Roman claim to be unusual, as regards their piety towards the gods in general, was not unfamiliar to a Greek audience, since the same topos had been expressed previously by others in almost the same words: e.g. SIG3 372,17-9 (Samothrace honouring Lysimachos in c. 288-1); SIG3 615,4-5 = FDelphes 111,2,89 (Delphi honouring the Athenian Apollodorus in 180). 186 RS no. 22, lines 5-6. 187 S.c. de Cn. Pisone patre: W. ECK & al. (eds), Vestigia 48 (Munich 1996), lines 12-5 dis immortalibus. Caracalla: P. Gissensis 40 I (ed. J. H. OLIVER, AJPh 99 (1978), 405), lines 3 TOLC d] EOLC [TOLC] dd[av]<XTOLc, / 7 TG>]V decov. Decius: ducov xat an£v&cov TOIC OeoCc 6i8TeXeaa vel sim. in J. R. KNIPFING, HThR 16 (1923), 353-4; cf. H. A. POHLSANDER, ANRW 2,16,3 (1986), 1838. Valerian: Euseb. Hist. Eccl. 7,11,7-9. 130 It is difficult to maintain that any of these particularly highlighted a link between civic authority and Roman religion. From the first were no instances longer a national crises, century supplicationes CE, means of appeasing the gods in times of but commemorated imperial success; and Caracalla's or Decius 1 instruction to supplicate to the gods celebrated the suppression of imperial opposition. Their intention was to display profuse gratitude toward the divine for the exposure of conspiracies. It betrayed a ality proportion- in religious thinking in that the degree of gratitude was proportional to the number of worshippers involved: more worshippers, the greater the gratitude displayed. For that reason, as many inhabitants of the Empire were required to pay worship identity was left unspecified. 188 script concerning the Di nostri as possible to the gods as such, whose Similarly, Trajan's re- Christians recommended that alleged Christian! ought to prove their nostril. 189 the innocence supplicando dis refers to the wide spectrum of tra- ditional gods as opposed to the Christian god, not to any particular Roman divinities. Neither Pliny nor Trajan needed to identify the gods in question. This attitude towards the divine did not befit only civic the realm. References to di deaeque omnes in the Plautine corpus are far more frequent than those to all the individ- 188 The common claim that P. Gissensis 40 I linked worship of the gods to Roman citizenship under the Constitutio Antoniniana of 212 may not be supported by the actual text; cf. J. H. OLIVER, AJPh 99 (1978), 403-8. Proportionality in Roman religious thinking: GWYN MORGAN 1990, 26-36. Supplicationes: G. FREYBURGER, ANRN 2,16,2 (1978), 1418-39. See also below, 3.2.3. 189 Pliny Ep. 10,97,2; cf. ibid. 10,96,5. 131 ual deities taken together. 190 Faced with a variety of di- vinities to choose from haviour, the at all levels of religious be- functionaries of city-states and private wor- shippers alike counterbalanced the danger inherent in con- fusing the divine pantheon with a particular concern for the incorporation of all deities, di irmortales, 01 possible deoi ddavotTOL TKXVTES xai nacrai, including those whose iden- tity was temporarily or continuously unknown, sive deus sive dea. The restoration of an altar to »the god or the goddess« by the praetor C. Sextius Calvinus following a decree of the Senate, presumably in 127 BCE (ILLRP 291), obeys the same set of ritual rules as both the private dedication sive sive dea by C. deo Terentius Denter ex voto, also dating, it seems, from the Republican period (OIL 6,111), and the agrarian prayer in Cato's De agricultural 9 ^ However, the religious system was »open« in a more fundamental sense, which the view that Roman religion lacked a deeper personal commitment must neglect. For I would argue that Roman religion represented an »open system« which, while no doubt receiving input from the civic sphere as well as from other sources, was self-determinative. To an externalist perspective is no doubt be invited sure, by the ritualistic nature of cult performance in the ancient world. When invoking the divine, functionaries responsible for maintenance of the civic religious acts such as the praetor of 127 and individuals like C. Terentius Denter 190 HANSON 1958, 62. or Cato made 191 Cato Agric. 139: si deus, si dea es f quoium illud sacrum est. For further documentation of this prayer formula, see below, 4.3.2. 132 use of standardized religious formulae and gestures. The proper observation of such ritual details was be an essential perceived prerequisite of positive divine response. This punctiliousness over ritual detail, and the of verbal precision, importance created what must be regarded as the basis of a general agreement on the »orthopraxy« behaviour. 192 of ritual The pontifical records preserved compilations of ancient and more recent prayers and means to ritual formulae by of which the college, if requested, could provide ex- planations and precedents, thus giving advice on matters religious concern to civic bodies like the Senate or to in- dividuals. 193 The prayers in Gate's De agricultura may been of identical to those preserved in the have pontifical books. 194 Yet, these written codifications ought blueprints which provided gestures be seen as information on ritual behaviour rather than served as canonical texts. and to The actual prayers performed by religious functionaries in civic cult acts remained largely just as they had been set down in public records or priestly codifications. public religious acts Prayers used in were first read out by scribes from the official records and then repeated by the magistrate who performed the ritual. The prayers uttered by generals before battle and those used by magistrates before comitia and contiones employed set formulae to ensure divine favour. The same phraseology was used by orators, historians and poets 192 NORTH 1976, 1-3. 193 On the character of these records, see ROHDE 1936. 194 Serv. Aen. 9,641 claims that the prayer in Cato Agric. 132 derived from the pontifical records. 133 to good dramatic and rhetorical effect. Prayers to the im- mortal gods that used formulae Romano ritu could be found in libris sacerdotum populi Roman! et in plerisque antiquis orationibus. 195 But prayer formulae were adaptable to dif- ferent contexts and occasions. They would be used by orators in the context of emotional appeals to their audiences, or employed in political disputes. The proviso that magistrates had to swear obedience to statutes by luppiter and the Penates was apparently deployed as a political instrument and later by Caesar in the case of his lex agra- Saturninus ria of 59. 196 Over the course of time, prayers on behalf the state could be altered, prayer on performing the sacrifice by shows. The of as the case of the censor's lustratio and the suovetaurilia original prayer expressed a wish for the expansion of Roman power, while the new version was me- rely praying for its preservation. 197 The flexibility of ritual frameworks, when their modation to individual needs was required, is well brought out by the ritual programme of the Ludi which accom- extended the saeculares of number of the deities addressed in the ritual sequence. Those responsible for the celebration rificed and 17, sac- prayed not only to the di infer! traditionally 195 Cell. 13,23,1. 196 Saturninus: App. BC 1.131. Caesar: App. BC Dio 38.7.2; Plut. Cat. Min. 32. 2.42; Cass. 197 Val. Max. 4,1,10 attributed this change to Scipio Aemilianus' public refusal as censor in 142/1 to use the standard phraseology. Given the scrupulousness in ritual procedures, this is an unlikely assertion, whereas the prayer text's alteration prior to a public performance was possible: NORTH 1976, 3. HARRIS 1979, 118-20, believes that alteration of the prayer's text occurred in the Augustan period and was subsequently attributed to the distant past. 134 associated with the Secular Celebrations, but newly intro- duced luppiter optimus maximus, luno Regina and the Palatine Apollo. However, the prayer uttered by Augustus, Agrippa and the Roman matrons was traditional, asking the respective deities in their turn to be XVvirum collegia, mihi, propitious domo, p. when similar for- addressing Mars pater in the context of agrarian ritual. As a matter of fact, public Quiritibus, familiae. 198 In the second century, the Elder Cato advised to use a very mula R. in different contexts, both and domestic, the very same formula could be used in addressing a wide range of individual deities, or in praying to the gods as such. 199 It would be misleading to assume that religious formulae were necessarily normative, or that religious behaviour characterized by was mere scrupulous performance of previously conceived patterns. Due to its formulaic character, there is a tendency among many scholars to overlook that ritual lan- guage was routinely accommodated to individual needs in contexts of praying, pleading, praising and expressing one's gratitude towards a deity. Silent prayers or peals, emotional ap- which could include establishing bodily contact with a divine statue, complemented such religious behaviour. A 198 PIGHI 1965, 114,96 and 99: volens propitiusque. Cf. PIGHI 1965, 157,12. On the Augustan changes, see conveniently PRICE 1996, 834-7. 199 Cato Agric. 141,2; cf. 134,2-3, 139. Cf. Suet. Div. Aug. 58,2. luppiter: Plaut. Amph. 935, Asin. 781. Dii: Serv. auct. Aen. 1,730 (at dinner in the context of domestic cult); cf. Cic. Div. in Caec. 41, Verr. 2,5,37; Petron. Sat. 60. Si(ve) deus f si(ve) dea: Festus 488 L (the dedication of a templum from the libri pontificis). Cf., among the Leges ararum, the Lex Narbonensis of 11-12 CE (ILS 112, numen Caesaris Augusti) or the Lex Salonitana (ILS 4907). See APPEL 1909, 122-3; HlCKSON 1993, 59-62. 135 whole range of differentiated, yet at the same time comple- mentary, forms of prayer is and attested, whether spontaneous informal or more formalized. Contingent upon the temple personnel's permission, worshippers would address a statue by formulaic prayer language; or utter deity's silent prayers; or touch and kiss the cult statue; or stand or in a sanctuary's cella sit and talk to the deity about their personal problems. 200 »Ein Blick auf die private Kultpraxis der Romer am Kultbild und in der cella ... konnte geeignet sein, eine schnelle Einordnung der romischen Religion als bloSe Kultreligion ... zu modifizieren.« 201 2.7.3 Orthopraxy or orthodoxy? Only if we assume the logical primacy of civic religion portray religion as a closed system whose response to input follows the principle of a >white box<, then individual ligious and behaviour merely mirrors the civic. An author like the Elder Pliny, when not describing religion, the domain of civic could routinely disregard any logical distinction between the religious rituals performed city-state re- on behalf of the and individual cult acts. 202 The explanation for 200 Sitting and talking: Prop. 2,28,44-8; LEWY 1928-29. Greeting, touching and kissing of statues: Cic. Verr. 2,4,94; Lucr. 1,316; Varro LL 5,58; Tib. 1,1,11-24; Min. Pel. Oct. 2,4; WEINREICH 1921. Worshippers were apparently prohibited from touching the statues of Virbius in the sanctuary of Diana in Nemi (Serv. Aen. 7,776) and Fortuna muliebris on the Via Latina (Val. Max. 1,8,4; Festus 349 L), but it is unclear whether ritual or administrative considerations lie behind these prohibitions. Silent prayer: WAGENVOORT 1980, 197-209, on insufficient grounds attacked by VAN DER HORST 1994. In general, see VERSNEL 1981; R. FLASCHE, HrwG 2 (1990), 456-68. 201 GLADIGOW 1994, 15-9, at 17. See further below, 4.2.5. 202 Cf. KOVES-ZULAUF 1978, 198-9. 136 this blurring of categories is not that distinctions did not matter. Rather, these different domains of religious practices used a standardized >orthopraxy<, thus inviting the holder to imagine an elusive homogeneity of religiosity as expressed through the stability of ritual. What of be- any notion religious homogeneity in the ancient city-state does not sufficiently take into account is the absence of a systematic >orthodoxy< Phillips and in ancient pagan religions. As John Scheid have emphasized, this distinction between orthopraxy and orthodoxy is a vital one. 203 would wish Robert What I to stress is that this distinction allows us to conceptualize potential differences in emotional or cogni- tive behaviour (doxa) while accepting similarities in stable ritual forms (praxis), when a multitude of different addressees (deities) is concerned. In Republican society, which lacked systematic legal moral or codes of normative social behaviour, a quintessential corpus of through conventional values was orally communicated telling, retelling, or enacting exempla of what was expected to count as right and wrong. 204 Elite education introduced young aristocrats to the value system of the aristocracy, as well as to its networking, through practical political instruction or through the service communal contubernium in of a military commander. 205 By contrast, reli- 203 PHILLIPS 1986, 2746-52; SCHEID 1992, 122-4; 116, 124-7. Id. 1993a, 204 For t he importance of moral examples, cf. Hor. Sat. 1,4,105-6; Plut. Cato mai. 4,2 with SALLER 1994, 109-10. Cf. LIND 1979 f 11-5 for upper-class exempla virtutis. 205 E.g. ILLRP 515, the consilium of Cn. Pompeius Strabo in the Social War, with E. BADIAN, Gnomon 62 (1990), 28: »It was contubernium that ensured the basic homogeneity of the upper class, political and non-political, and 137 gious socialization was far less homogenized. As Roman religion lacked sacred texts or holy books, religious education, as a means of perpetuating religious tradition, was achieved through the observation of and ritual; through the participation in transmission of traditional religious knowledge to younger and unexperienced members of societal religious families, groups, collegiate associations or the colleges of civic priests; 206 through the institutionalisation of gious a roles and the creation of cult functionaries; and, to limited priestly degree, records through the written cultures, preservation cannot, pre-eminently or its fixed comprises reli- meaning: the chronological frame within which tradition can be preserved more unchanged oral as research into oral traditions has de- monstrated, guarantee the unalterable preservation of tradition or rit- can be different for different participants. This is in accordance with which recent problematizes the theorizing among anthropologists, notion of meaning of ritual forms for communal interpretation in society. For it is to less not more than three generations; more- over, even within that limited period the meaning of a ual of of ritual stipulations. 207 However, these forms of rigidization, though common to gious reli- define the meaning of ritual as such. The stability of ritual forms may provide the which difficult rules and the framework in participants make sense of their behaviour. But while the active participation in ritual very often engenders or that made it possible to open the doors of the Curia to Equites and to promote some of them even to the consulship. « 206 VAN DER LEEUW 1939; CANCIK 1973; BREMMER 1995. 207 ROHDE 1936. For the limits of writing in Roman religion, see SCHEID 1990; BEARD 1991. 138 positively invites a wide range of different (and potentially incompatible) interpretations about the meaning of a rite, the ritual form on its own may well be meaningless. 208 It remains important to apply both a synchronic and a perspective when evaluating the process by means of which religious meaning is scholars diachronic created. As we have seen, accept that at a diachronic level the stability of ritual forms does not entail that the >meaning< of rituals in society could not change over time. Conversely, in a synchronic perspective the orthopraxy of ritual behaviour must not disguise that individuals made different sense of ritual even at one and the same point in time. By assuming that a in synchronic perspective rituals and their exegeses consti- tute a communal, if changing, Roman sense of identity, recent studies misrepresent this some synchronic dimension of contextual meaning. 209 When making such an assumption, these scholars fail to take account orthopraxy and orthodoxy. of the distinction between For they imply that ritual per- formance, exegeses like the Ovidian Fasti or mythological narratives, while not constituting orthodoxies, could never208 Cf. the discussion in Religion 21,3 (1991), 205-34, responding to the thesis of STAAL 1989 that ritual systems are meaningless per se. For a cautious application of that discussion to the study of ritual in Roman religion, see, in addition to the studies of Phillips and Scheid, e.g. RUPKE 1995a, 410-2; Id. 1996b; FEENEY 1998, 119-21, with further references. 209 E.g. BEARD 1987, 1, 7-12, at 12: »[The ritual calendar] offered a pageant of what it was to be Roman ...: to perform the rituals through the year ... was to discover and rediscover that Romanness«; Ead. 1991, 55-6; Ead. 1993, 55-6; WALLACE-HADRILL 1988, esp. 226; HOPKINS 1991, 484-6; EDWARDS 1996, 47. >Romanness<, a term that is used time and again among these scholars to circumscribe the native definition of a Roman identity, is criticized below, 3.1.2. 139 theless be interpreted as the direct expression of communally shared meaning. On such a view, these events operate in the context of religion as a »public symbolic system« in a neo-functionalist or Geertzian sense: they offer interpretations of societal life within a communally accepted context. Ritual and exegesis become parallel symbolic processes which define a Roman religious identity. The problem with such a view is that, despite the methodological awareness of scholars, these orthopraxy and orthodoxy in fact re-emerge as ho- mologous. For the underlying assumption is that a local ligious knowledge cognitive frame existed to which determine re- would provide a communal the meaning of religious rituals as well as their exegeses. However, these scholars fail to specify to ritual orthopraxy what degree or the playful Ovidian exegesis with its multiple interpretative choices, written for a small educated audience, and presumably unrepresentative of Roman society as a whole, could create, or at least reproduce, accepted and binding interpretations of religious meaning other contemporaries. For instance, it shared is impossible to construe a direct link between the religious realia ned in mentio- Ovid's Fasti and his individual exegesis, since the restrictive communicative code of ritual orthopraxy and often by speculative communication through the exegesis are not necessarily synonymous. 210 For the poet of the Fasti did not offer a religious orthodoxy which would interpret one level 210 Cf. PHILLIPS 1992; SCHEID 1992, 122-9, at 124: »... the Fasti ..., like other exegeses, were profitable if not indispensable for religion and society«; ROPKE 1994; Id. 1995a, 408-16; FEENEY 1998, 127-31. 140 of religious meaning, ritual performance, on a complementary level of religious meaning, that of textual exegesis. His portrayal of cultic reality is just biased by subjective and as conscious or unconscious motives as other inter- pretations. The poet which as engendered commented on a ritual performance, several different cognitive and emotional processes in himself and in his audience. For even in a synchronic perspective, the >meaning< of religious pends on ritual de- various interpretative operations, it is shifting and shows a >performative structure<. The reason why scholars repeatedly fail to unstableness of address the the meaning of a religious sign is the ap- plication of problematic models of how meaning is created in cultural communication. For in assessing the individual's response to visual as well as to religious signs, traditional linguistic models are routinely applied that are based on a linear relationship between signifier and signified. These linguistic models, however, cannot provide much more than a limited analogy, when the response engendered by religious signs is and signified is always under- as the problem of the meaninglessness of ritual forms mentioned above ilustrates. This gap of is or concerned. In the latter case, the re- lation between signifier determined, visual determination subsequently filled in by a >meaningful< response, whose relation to the signifier is often haphazard. 211 It possible to connect now this discussion to that of the short- comings of the civic model. In each case, the individual is exclusion of motives that could be deviant from the system of 211 HARDIN 1983. 141 public cognition has led to and to the a neo-Durkheimian perspective postulation of a white box. In each case, how- ever, the box is always black rather than white: it is im- possible to construe a linear and simple causal relationship between a particular and determinative input, be it that of the political realm, and the religious the form system's output in of the creation of a particular religious meaning solely determined by the political realm. To be sure, most scholars would not disagree with the view that ritual orthopraxy, and not an underlying systematic orthodoxy, at a public level characterizes ancient polytheistic religions. However, as a result of the assumption made about the nature of Roman religion as a collective operation at the public lev- el, they unwittingly reintroduce the tion. Despite an awareness notion orthodox dogmatiza- of the polytheistic nature of pagan religious systems, this notion heralds a perspective: of monosystemic local religious knowledge is defined as though it could become constitutive of societal behaviour as a whole within the allegedly normative framework of civic religion. 212 2.7.4 Conceptualizing an >open system* When responding to such a view, several scholars realize, as we have seen, multiplicity that of it is religious problematic to describe the meanings at Rome as though they constituted a defined system of Roman religion. Yet, how can the various processes by means of which religious meaning 212 The incompatibility of polytheistic systems with such monosystemic perspective is discussed below, 4.3. a 142 was created be organized by the native participant and the modern observer, without either embracing a notion of gious collectivity or altogether reli- abandoning a meaningful category of cohesive belief systems at personal or societal levels? In order to meet this methodological problem, I suggest not only abandoning the description of Roman Republican religion in terms of a »closed system«, but also questioning the primacy of the political realm. Instead, I propose a model of the »additive extension of an >open system<«. The basic assumption of this model is that religious choices are made with a view to the extension, rather than the replace- ment, of a religious system. This model tries to explain religious change not as the change of particulars within the limits of a closed system which is determined by the input of the socio-political realm only. Rather, a given religious culture is, I would argue, the result of the constant mization of a religious opti- system through various inputs of what is communicated as desirable options. In this model, the desirability of choices is not exclusively determined by the civic domain. Rather, we are dealing with an open proc- ess in that the individual agents continue to optimize religious configurations within the constraints of their cular environment and parti- of the polytheistic market-place of religious alternatives. 213 The notion of an open system of course does not entail that every religious choice would be acceptable or even possible. The idea is not, as scholars often imply, that open- 213 Cf. BENDLIN 1997, esp. 52-4, where this model is used for describing religious behaviour in the Roman Empire. 143 ness and plurality result in anarchy. As I suggested above, the output of all systems is determined by their input. Even if there is no linear causality as to the relationship of these two parameters, systems receive inputs which are com- municated within a framework of cultural and social expectations. In all systems, the constant relationship of input and output entails processes of interaction and exchange between system and environment. Therefore, a system can never be self-sufficient. Systems are open, however, when their output is determined by different forms of input. Unlike closed systems, they cannot be described linear on the basis causal links which solely determine their input. The openness of a system does therefore not imply that the tem of sys- is entirely self-sufficient. But it is autonomous in so far as it is the system itself, rather than, let us determinant identity situated at concepts of self-sufficiency and relative autonomy is crucial, since a system's relative autonomy a esses a the public level, which regulates an input. The difference between the absolute say, priori entails proc- of integration that are characterized by the combina­ tion of independence and interdependence rather than by a crude dichotomy of autarky versus complete embeddedness: »Systemintegration [ist] abhangig von Interdependenz, Differenzierung, relativer Autonomie und reflexiver Abstimmung unterschiedlicher Umwelten. Systemintegration kann dann definiert werden als Kompatibilitat relativ autonomer interdependenter Teile, deren Gesamtheit eine spezifische Systemidentitat konstituiert.« 214 This methodological framework allows us to conceptualize the odd blend of independence and interdependence Roman that informs religion's interaction with other realms of Roman so- 214 WILLKE 1978, 236-48, at 248. 144 ciety. Religion's relative autonomy is the reason why the religious choices realized by groups of worshippers or individuals, even if they resemble those of civic religious be- haviour, are not necessarily homologous with them. For within the availability of a large number of religious choices, it is the religious system's self-determining capacity which leads to a specific output of religious alternatives at the expense of another one which could find its another realization at time, while a third option may forever remain unre- alized. The box is always black rather than white. However, the replacement of a model which uses tion of the no- a closed system by one which advocates the concept of an open system is not just a convenient heuristic strategy. Rather, this replacement operationalizes the concept religion as of both an external and an internal affair, which has been formulated above. As the box is black, it is unwarranted to believe that the output of a religious system is a priori identical with the input of civic religious expect- ations. Rather, it must be assumed that the religious system is capable of displaying variability, both in terms of be- haviour and cognition or emotion, the input. 215 This when further processing strategy attempts to make the study of religion in Rome compatible with current interpretative models concerning the organization of personal and social tems. sys- What follows is a theoretical outline of two of these models. 215 See above, 1.5 and 1.6. 145 When developing models of individual haviour, sociologists must draw on and societal be- analogies with other disciplines. For instance, biology has contributed to modelbuilding concerning the organization of personal and systems by pointing social out the case of the human brain whose operations do not follow a simplistic hierarchized matrix. The human brain consists of at least three different levels, the reticular system, the limbic system and the neocortex. Rather than forming a stable hierarchical system, the particular control mechanisms of sensorial and processes, for which the reticular system is responsible, which motorial these adaptive three levels learning and overlap: cognition, are directed by the limbic system, and the discursive and reflexive abilities determined by the neocortex - all of these constitute a personal system through constant inter- changes between basic sensorial and motorial, affective, and reflexive stimuli. 216 The human brain, then, is not an entity which could be described as a closed system working on the basis of a hierarchical principle or determined by a linear causality. In instrumentalizing further drawn this insight, sociologists on the analogy with another scientific model which seems particularly apt for providing description have a more nuanced as to how »open systems* of personal and social behaviour are organized. This is the principle of >nonlinear fluctuating systems<. The notion of systems< >nonlinear fluctuating is adopted from thermodynamics where it designates chemical reactions of fluid and gaseous 216 LEGARE 1980; ROYCE & POWELL 1983. dissipative struc- 146 tures. Fluctuations appear when systems are removed from one of their original states of equilibrium. An orderly transi- tion to a new state of equilibrium is then achieved the wide dissipation of through elementary units and through the creation of flows and anti-flows which, as a consequence of their abundance and regular occurrence, are able to equalize system disorder and recreate a dynamic, though at best tem- porary, equilibrium. Order and achieved through organization are thus not stable structures which receive the input of a hierarchically organized centre. Rather, the adaptation to different forms of input is achieved through the uous dissolution of contin- structures, through heterogeneity and dissipation. 217 In analogy, system theory formulates theories concerning open systems which are based on the assumption that personal and societal behaviour can also described through models of >nonlinear fluctuating systems<. By using the analogy of the non-hierarchical and multilevel nature of biological and chemical system reactions, system theory replaces the static and hierarchical model of which the mechanical neo-functionalist models control fluctuation by a model of and the deregularization of personal and social systems. 218 We have seen that a model systems on of closed systems (as advocated by Parsons or Geertz) are based, nonlinear systems, of closed is in constant danger of misapprehending any change either as change within a stable environment or as a process which endangers the stability of a homoeostatic system. 217 PRIGOGINE 1976; NICOLIS & PRIGOGINE 1977. 218 E.g. BUSCH 1979; BUHL 1987, 59-87. The 147 alternative model attempts to overcome such a dualistic po- sition: change in the organization of personal and social systems is dealt with through processes of internal fluctuation by means of which a system adjusts its organizing principles to ever changing inputs. This is the reason why personal or social system fluctuation does not entail the creation of or chaos. For instability, anarchy it would be wrong to assume that >nonlinear fluctuating systems< are in a continuous process of fluctuation. These systems are syntheses of genetically or bio-psychological, of symbolic determined or interactive, and of institutionalized materials. 219 As they interact with environment, they must their have moments of different fluidity and different stability. Only the organization of these different elements, resulting in a range of temporary organizational structures from entirely open to nearly closed, bles a change system and ena- to cope with differing forms of input, with evolution, while still maintaining internal consistency. 220 In this sense, both personal and social systems always strive for stability. Any kind of stability is achieved through the instrumentalization of recurring of fixed meaning(s) and of standardized types of behaviour. We can see that social, cultural or often political systems are stable entities. However, this stability is not a re- sult of the fact that a given system possesses a tic codes structure or is homoeosta- determined by central control mecha- 219 Cf. above, 1.4. 220 Cf. SAHLINS 1985, 12-4, drawing on field-work, for an illustration of how prescriptive, or closed, and performative, or fluid, structures coexist simultaneously in a given society. 148 nisms. Rather, >macro level stability< is achieved through >micro level variability<, namely through the interaction of small and >fuzzy systems< in the process of regrouping: lo- cal transformations and subsystem changes, rather than tral organization, guarantee the overall system's exist- ence. 221 When various forms of input are dealt with subsystems, when the process structure in the of equalization is achieved through the subsystems' fluctuation, then the tem's cen- overall sys- seems to be virtually unchanged. Only mas- sive dissipation of elements can restructure or even dissolve personal or social systems. Normally, however, social evolution is not achieved through spectacular changes but by means of gradual, and often hidden, transformations of the social world. 2.8 Social complexity and religious differentiation: an introduction This interpretative model suggests a heuristic shift from holistic a description of society as a »closed system« to its portrayal as an »open system« of differentiated choices, from >hierarchy< to >autonomy<, from >collectivity< to >variability<, from the mono-causality to the multi-causality of historical situations. related concepts of At the centre of this shift lie the complexity and differentiation. Both categories have received a negative connotation in models of »closed systems* in which, due to the linearity of causal connections, there is no room for the plurality of 221 BOHL 1987, 68-71. choices. 149 In a changed methodological perspective, however, these ca- tegories denote the norm rather than an exception. and Personal social systems are complex, since the availability of a range of non-linear connections entails that assume more a system can than just one (temporary) condition. This com- plexity of choices necessitates selection. In order to with increasing complexity, personal and cope social systems create differentiations. 222 Complexity is not dealt with the at macro level, and thus escapes holistic models of closed systems, but assigned to the micro level. The >macro level stability< principle of through >micro level variability< applies here as well. I believe that this model is not only more adequate the portrayal of for modern societies, but also better suited than current conceptions for describing social, cultural and religious complexity in second and first century Rome. This Republican model can be applied to personal and social be- haviour in general. It is a problem of evidence rather the consequence of a categorical difference in behaviour that the elite in particular can be shown to the than have utilized evolution of ideas and developed alternatives of social and cultural behaviour as competitive options in a cultural market place. This was a result of the more dramatic social, economic and cultural changes which took place in the last two hundred years of the Republic, and from which the primacy of the political realm has attention. occasionally distracted our It is one of the paradoxes of the religious his- 222 Cf. conveniently LUHMANN 1980-89, 1, 9-71; KNEER & SEHI 1993, 111-6. NAS- 150 tory of the Roman Republic that the the vitality and stability justified emphasis on of religious institutions and practices in Late Republican Rome (their >macro level stability<) had to obfuscate >micro level fluctuation<. The son for this disproportionality in modern scholarship is that no adequate theoretical model has been available could rea- which give a more complete portrait of religious complexity without at the same time reintroducing a hidden notion of dissolution. 2.8.1 >... me et Cottam esse et pontificem* One example which addresses the differentiation of choices cultural towards the end of the Republic gives a preliminary illustration of this point. In his De natura deorum, set a in private context in the seventies BCE, Cicero has the Aca- demic C. Aurelius Cotta deny the intellectual sufficiency of any proofs concerning the existence and nature of the In the course gods. of the argument, Cotta is reproached by the Stoic interlocutor Q. Lucilius Balbus that Cotta's prominent public position as well as his pontificate ought to prevent him from adopting any form of Academic scepticism concerning such theological questions. 223 In defense, Cotta argues with regard to theological discourse as well as to philosophical dispute in general that there must be room for rational reasoning as a means of inquiring into the truth on of Academic the basis scepticism. As regards religious cult (religio and cultus deorum), however, strict adherence to ancestral 223 The juxtaposition of both roles in the dialogue emphasizes their differentiation: ND 2,168: et principem civem et pontificem; 3,5: Cottam et pontificem; 3,6. 151 custom and pontifical law is obligatory. On such matters, Cotta complies with his predecessors in the office of pontifex maximus and with the augur C. perienced Laelius. These men ex- a similar dilemma of allegiance to Cotta's, illu- strated by the personality of C. Laelius, augur and sapiens, the member of a state priesthood and a theless/ the philosopher. social behaviour of these men was regarded as exemplary. 224 Academic scepticism and the traditional of worshipping Never- the forms gods are thus made philosophically and intellectually compatible. 225 Through the persona of the Roman Cotta, Cicero, a himself member of a distinguished Roman priesthood and a consular as well as an Academic inquirer into philosophical truth, presents two options of social and cultural behaviour to his Roman aristocratic audience. These two options, namely religious practice on the one hand and the philosophical cri- tique of religion on the other, are attributed to differen- tiated domains. Cotta, the politician and pontifex, inhabits the domain of public religious tradition constrained by the mos maiorum. Cotta, the sceptic, where theology represents another world and the philosophical inquiry into religion are set apart from the religious life of the Roman Forum. The distinctions between these two options, and the apparent incompatibility of Cotta's two roles, are not dissolved in 224 ND 3,5. The combination of pietas and scientia, used by Cicero to distinguish the area of traditional religion on the one hand and these men's interest in philosophy or civil law on the other, is also exploited elsewhere: e.g. Dom. 136, 139; Plane. 20; De orat. 3,134; Leg. 2,52; Amic. 6-7; ND 1,115, 2,165; Off. 2,40. 225 ND 3,5-6; cf. 3,43, 65, 93. Compatibility: ARDLEY 1973; BURNYEAT 1982. For the consistency of Cicero's academic principles, see GAWLICK & GORLER 1994, 1084-125. 152 the course of the work. Gotta, the sceptic, like characters of this dramatic the other setting, does not give in to Cotta, the pontifex. Equally, the closure of the dialogue does not a clear case for establish either accepting or refuting the philo- sophical proofs which are presented concerning the existence and nature of the gods. To the Epicurean theoretical Velleius, Cotta f s refutation of the Stoic argument concerning the nature of the gods is most compatible with his own views. To Cicero, the Stoic theology presents a degree of probability which even the Sceptic can (with due caution) accept: haec cum essent dicta, ita discessimus ut Velleio Cottae disputatio verior, mihi Balbi ad veritatis similitudinem videretur esse propensior. 226 Cicero's cautious championing of Stoic theology concern with the reconciliation suggests a of the domains of philo- sophical theology and ritual practice. After all, among the arguments presented Stoic theology was most easily compatible with traditional religion. Yet, it is important to stress that the division of domains is not resolved but extended by the author beyond the closure of the dialogue. According to the traditional view held the past, such by scholars in sceptical discourse further contributed to the supposed decline of traditional Roman religion, since it implicitly attacked cult practices. While rebutting this view in its crude form, even the current revaluation of Late Republican religion succumbs to a dualistic perspective when 226 ND 3,95. Cf. GAWLICK & GORLER 1994, 1092-4, at 1094: »Cicero selbst ist fast angstlich um eine korrekte Formulierung bemiiht; der logisch anstossige Komparativ bleibt dem unsensiblen Epikureer vorbehalten.« 153 it comes to the relationship of traditional religious prac- tices and philosophical speculations about religion. 227 Such a perspective, however, is the logical consequence of a model which implies that the differentiation of cultural choices in first century Rome signifies the fragmentation of Roman society in the Late Republic. As a matter of fact, the link between differentiation and fragmentation is implicitly made, as soon as the theological discourses of that period are perceived as recent manifestations of cultural and tensions of subsequent disintegration, which earlier generations of Roman intellectuals had not imagined. The differentiation of cultural choices is thus connected to theme of the the the general socio-political crisis of the final years of the Roman Republic. 228 This dualistic model represents more no doubt mis- theological efforts of earlier generations. The beginnings of the Roman elite's evaluation of traditional religious practices on the basis of philosophical speculation can be traced back to the early second century. 229 Moreover, this dualistic model utilizes a concept of structural differentiation< whose a priori emphasis on structure sees differentiation as a process of the emergence of new 227 Dualism: BEAUJEU 1969; ANDRE 1975; BRUNT 1989. BEARD 1986, though more cautious, tends to adopt a similar position; 43: »the strategy of appeal to expediency and tradition does not entirely remove the tension«, 45: »[the] cultural clash between different systems of thought«. 228 E.g. MOMIGLIANO 1984b, 199: »it was in this revolutionary atmosphere [sc. the final upheavals of the Republic] that ... some of the Roman intellectuals began to think in earnest about religion*; RAWSON 1985, 298-300; BEARD 1994a, 755-61; GRIFFIN 1994, 728: »at the end of the Republic«. 229 E.g. Ennius Euhemeros: ROPKE 1995a, 363-5. Drama: Pacuvius inc. 366-76; CANCIK 1978, 332-4. Theological efforts: RAWSON 1991, 80-101. 154 structural entities whose relation to other structures is problematic at least. 230 As I suggested above, such a nition, originating from neo-functionalist thinking, is un- helpful, since the views defi- concept of structural differentiation structures as dissociated elements in a closed system in which the primacy of one structure is subsequently endangered by other emerging realms. Instead, the dualism of this traditional model should give way to the idea of functional differentiation in that an open system can develop new functional domains which are no longer congruent with a pre-existing structure, yet prove compatible and so do not endanger that structure's existence. 231 As I suggested above, the box is always spective black, system and need fluctuation at the level of the renot result in conflict or even dissolution. 2.8.2 Variability and stability in religious behaviour Any traditional dualism can only misapprehend this logic societal behaviour. Both individuals and social groups at large are capable of attributing social activity to mous subsystems. As the without autono- norm rather than an exception, a personal or social agenda is dealt with subsystem of in the respective major interaction with other subsystems, and the balance of the overall system rests on the principle that one area of social activity need by no means compromise or dissolve others. Rather than documenting the fragmenta- 230 Structural differentiation: HOPKINS 1978, 74-96 (with reference to T. Parsons); NORTH 1979; Id. 1992; CORNELL 1991, 59; BEARD I994a, 755-68. 231 See above, 2.6. Cf. conveniently KNEER & NASSEHI 1993, 35-44, for a theoretical discussion. 155 tion of Late Republican elite discourse, Cicero's ND de- monstrates how the complexity of cultural choices, which must have been the norm among Cicero's intended audience, is managed while the tension inherent in the interference of a philosophical critique of religion and the public role of statesman and priest is maintained. The result is a theological construct which attempts to fill in the gap between accepted tradition and the demands of a more sophisticated present. Social complexity is thus dealt with in a dialectic process: similar to Varro's discussion, and critical juxta- position, of the realms of theologia naturalis and theologia civHis (suspending any judgement by adopting the ative strategy of Academic scepticism), ND differentiates these realms into different roles so that the herent argument- conflict in- in these choices can be resolved through micro level variability by their attribution to different realms of cultural activity. In addition, with a view to macro level stability the inherent since very the logical contradiction is countered, production of such a theological construct creates a dialogue between distinguished provinces and thus leads to their reconciliation. The process of producing that construct is the moment when the system is in fluctuation. As a result of this indigenous interpretative process, macro level stability remains preserved. It is one advantage of the model of an social complexity open system of that it can explain how the system works and at the same time account for social and cultural contradictions and tensions. Moreover, such a model does not to postulate a need white box in which the civic domain solely 156 determines religious behaviour; nor does it the have to claim homology of societal communication and individual atti- tudes. On the contrary, the Ciceronian example that demonstrates the domain of public religio represented only one sub- system in societal communication about religion, as elite communication freely moved between the realms of philosophical critique and the public affirmation of civic religion. At the same time, we are able to replace the traditional notion of ancient Roman religion as a religion of participation in public ritual practice only. For it has become possible to attribute the realm of religio and cultus deorum to a subsystem in its own right. The functional differentiation of religion in Republican Rome entailed the existence of differentiated domains, of which the domain of public ritual behaviour or that of intellectual abstraction were only two. 232 Existing structuralist models, which view evolution in religious terms of linear differentiation, resulting in the subsequent dissolution of civic religious structures, are therefore based on misleading methodological premises. What, however, about the opposite extreme? In the past, several studies of Graeco-Roman religion have employed psychological models of »cognitive dissonance« to address the phenomenon that contradictory belief systems are applied in different contexts. 233 Further developing such models, Denis Feeney provides a post-structuralist perspective when 232 Cf. CANCIK 1994, 394-404; Id. 1996, 112; FEENEY 1998, 140, on the latter being an integral part of »Roman religion*. 233 Cf. VEYNE 1983; VERSNEL 1993. 157 describing the religious system at Rome as fragmented into competitive areas of religious knowledge: »... [T]here was no one Roman religious system existing essentially, inherently meaningful/ waiting to be participated in. Rather, what we call >Roman religion< or the >Roman religious system< was compounded of all kinds of different forms of religious knowledge, from the performative to the philosophical, literary or antiquarian ... [M]eaning was generated in the interaction between the various genres of belief.« 234 Feeney's description, however, is as under-determined as the view against which he would be reacts is over-determined. For it a mistake to rest content with a mere description of inconsistencies which entail the continous contextualization and fragmentation of meaning on the part of the participants. A native native participant like Cicero, while re- negotiating the boundaries between different belief systems, was nevertheless concerned, as we have seen, with their mutual harmonizing in a structural hierarchy. In a more general sense, to the native agents of communication >meaning< and >truth< are very merely often substantive, rather than the relational categories as which they are presented by the (post-)modern observer. attempted to present a The preceding discussion has model which accounts for both the variability of competing belief systems at a micro level and the desirability of stable macro level. beliefs and convictions at a In that model, utter variability and absolute stability of the religious system are ideal types of organization rather than belief systems realistic options. Instead, religious have moments of different fluidity and dif- ferent stability, resulting in temporary realizations 234 FEENEY 1998, 140-1. which 158 encompass a wide range of system conditions from entirely open to nearly closed. 235 235 See above, 2.7.4. 3 ROMAN PROBLEMS RELIGION: OF DEFINITION &OXEL Y<xp TiXeuov fi TITILCTU TOU eCvai f| ctpxfi, xai noXXd auu-cpavfi 61 5 TKXVTOQ aUTflS TCOV £T|TOinieVG>V Aristotle, EN 1098b 7-8 In the preceding two chapters, the dissatisfaction with past scholarship stimulated the inquiry into a new framework the for reassessment of religion at Rome. Employing the vocabu- lary of system theory, I suggested that religion in Late Republican Rome ought to be viewed as an open system characterized by its complexity and internal differentiation. In this and the next chapter, I shall investigate ters ruling the parame- the organization of that system and determine, to use the metaphor of system theory, its stable as well as its fluid moments. When talking about >Roman religion<, what inferences can one make concerning its constitution and its constituent parts? And is it at all possible to use the tion of >Roman religion<, no- as if we were presented with a straightforwardly identifiable entity? What does happen to this entity, which materializes as the local religion of the city of Rome, when it is transferred to coloniae or munici- pia? These questions entail a problem of definition as to what we mean be talking about Roman religion, which has only insufficiently been taken into account by scholars and which therefore will be addressed in this chapter. 160 3.7 Rome as an >imagined community* 3.7.7 Ethnicity and citizenship Rome's foundation stories commemorated how the Roman was a People synthesis of various ethnic groups, continuously ex- tending its citizenship and thereby assimilating foreigners, fugitives or slaves. Archaeological research this picture by has pointing out the »multi-ethnicity« of ar- chaic Roman society and its constant adaptation cultural influences. The ideological which suggested to Wissowa or the religion was a supported to outside issue of ethnicity, primitivists that Roman closed, and ethnically pure, system subse- quently diluted by foreign influences, has thus been rebut- ted more radically than even Altheim and other early critics could have imagined. 1 The Roman custom of extending its citizen body through conferral of citizenship on slaves upon manumission was conspicuous to outsiders already in the mid-Republic. In 214 (the earliest direct evidence for this practice), Philip of Macedon V. (mis-)represented franchise by manumission as a typically Roman strategy of increasing manpower in order to strengthen its military base. However, imputing such a grand strategy to the Romans would be misconceived. Under the Re- public, manumission and subsequent conferral of citizenship was left to the slave owner's initiative, but the regularity of manumission of urban and suburban slaves in Republican Rome has often been exaggerated. Moreover, manumission, both 1 AMPOLO 1981. Cf. MOMIGLIANO 1984a, 379-436; 1991, 63-5; Id. 1995, 48-80. Cf. above, 2.2. CORNELL 161 while the dominus was still alive and by testamentary fideicommissum, belonged to the domain of civil law. Apparently, Republican state did not influence individual decisions the creation the tail state's (which systematic did at cur- or or impose restrictions in order to either increase of Roman manpower. On the contrary, the interest citizenship controlling in any rate not commence prior to the Augustan period) showed in its curtailment of manumission as pre- a requisite of enfranchisement. 2 That is not to say that admission to Rome's citizen body specified the legal already Roman citizen when a of privileges Tables Twelve rested on haphazard principles. The having commerce with a foreigner (hostis) . This example suggests an awareness by the mid fifth century that gory the cate- of Roman citizenship possessed an element of juridical and political exclusivity, which those with that citizenship would be eager to defend. 3 In this definition those who lacked when these a most distinctive manner, of the boundary between Roman citizens and that became privilege institutionalized two status groups were juridically separated by means of introducing the office of praetor peregrinus in about 245/44. In the second and first centuries, this office 2 3 SIG3 543,4,30-5; Dion. Hal. Ant. Rom. 4,22,3-4 with CHANTRAINE 1972; SHERWIN-WHITE 1973, 322-7. Regularity of manumission: WIEDEMANN 1985; CHAMPLIN 1991, 136-8. For the legal requirements of manumission, see Gaius Inst. 1,17 with GARDNER 1993, 7-20. On the Augustan legislation, see Suet. Aug. 40,3-4; Gaius Inst. 1,18-9, 35-47; SHERWIN-WHITE 1973, 327-34. Tullus Hostilius' alleged attempt to curtail enfranchisement upon manumission (Dion. Hal. Ant. Rom. 4,22,5ff.) is an Augustan retrojection. E.g. Tabula 6,4 (RS no. 40, p. 660-1): aeterna auctoritas <esto>. adversus hostem 162 would take responsibility for jurisdiction with regard to the peregrin! in Rome. It cannot roughly the same time, be coincidental that at in 241, the thirty fifth and last Roman tribusr was established. Before that time, the ex- tension of territory in which Roman citizens were settled or native peoples received franchise was symbolized by the ad- dition of further tribus. In 241, was replaced by this additive the decision to include new territory and new citizens in those tribes which already existed. been plausibly principle It has suggested that one should see this develop- ment and the institution of the office of praetor peregrinus as parallel processes. The latter, apart from responding to practical necessities, provided the juridical institutionalization of the definition of Romans and non-Romans; the former may have signalled an awareness among the political elite that Rome's immediate territorial expansion grown the narrow confines had out- of the nuclear city-state, and that a merely additive principle of incorporating new territory and new citizens ought to be replaced by the controlled admission to Rome's civic community, whose boundaries became increasingly transparent. 4 It may not be purely accidental either that it was also in the second half of the third century that a distinct Lat4 Cf. NOETHLICHS 1996, 28. For the office of praetor per­ egrinus, cf. Lyd. Mag. 1,38 (date of introduction); Lex repetundarum (JRS no. 1) line 12: pr(aetor) quei inter peregrines ious deicet; Gaius Inst. 1,6, 4,31; Pomp. Dig. 1,2,2,28. In this context, it may be relevant that terms like imperium or provincia first received a territorial connotation in the second century, that is once a conceptual difference between Rome and her imperial subjects had been defined. For the limits of this conceptualization, see BENDLIN 1997, 36-8. 163 in literature first emerged, producing what has been described as an autonomous Roman cultural tradition that depended on, yet at the same time stood cultural models. 5 out against, Mediterranean These processes point to the latter half of the third century as the period in which not only a more restrictive approach towards the admission to Rome's citizen body became manifest, but also a more conspicuous awareness of a Roman socio-political identity was developed. John North has plausibly suggested that the history of cult transfers to the city of Rome in response to the recommendation of the Sibylline Books provides a roughly contem- porary parallel. As North observed, the transfer of the ter magna from Asia Ma- minor and her admission to the Roman pantheon in 204 was the last extension of the city's divine citizen body as instituted by the state officials in response to portents, notwithstanding that both the state and in- dividuals continued to introduce to Rome new the result of a divinities as military commander's vota or of personal initiative. 6 Moreover, the dates of introduction of cults to Rome by civic initiative confirm a striking tween 5 6 a large number of inequality be- temples built in the fourth and E.g. RAWSON 1991, 80-1. Cf. GRUEN 1990, 79-157; Id. 1992. BREMMER 1993 points out the »need for new myths« which led to the creation of literary aetiologies from the late third century onwards. The Roman dramatic and poetic tradition was initially created by members of the Hellenized Italian cultural elites: WISEMAN 1987, 297-305. NORTH 1976, 8-11. For civic cult translations in response to portents, ending with the introduction of the Mater magna in 205/4, see SCHMIDT 1909; VAN DOREN 1954-55. RAWSON 1991, 93-101 cautiously discusses the arguments for and against Scipio Africanus 1 supposed transfer of luno Caelestis from Carthage to Rome in 146 and arrives at a negative conclusion. 164 third centuries and a comparatively small number in the second and first. 7 It seems that by the beginning of the second century there was a elite that feeling among Rome's socio-political the city's religious system had reached a point of saturation with regard to any further extension of its pantheon. This is not to say that one should follow of argument North's line too rashly and interpret this development as a process of dichotomization, through which the civic system at Rome marked itself off from differentiated and alternative forms of religious experience that emerged from the sec- ond century onward. For whereas the elite undeniably applied the principle of expanding Roman religion by adding external elements with far greater caution from the end of the century, little third caution prevailed as to its expansion from within. On the contrary, the period between the end of the Second Punic War and the year 175 is the time when the principal extension of the number of days of the city's major religious festivals - the Apollinares, Megalenses Ludi plebei, Ceriales, and Florales - occurred. The prin- ciple of proportionality in reference Romani, religious thinking, to which has already been made, should be held responsible for this increase: more profuse worship could be paid to the gods once more days were earmarked for religious thereby festivals, attracting more worshippers from outside Rome. This pattern of development suggests that by the beginning of the 7 For the dates of introduction of civic cults to Rome, see WlSSOWA 1912, 594-6; ORLIN 1997, 199-202, the latter more accurately distinguishing between those cults introduced by civic initiative and those instituted by the state as a result of individual initiative. 165 second century the civic authorities at Rome took into ac- count that Roman citizenship was no longer definable through residence in the nuclear city-state, and that worship, here primarily defined in quantitative terms, could be maintained only if Roman citizens from farther away were given a istic chance to visit the Roman festivals. 8 From Gwyn Mor- gan's analysis of the proportionality of worship implication a to further emerges. At the beginning of the second century the civic domain, while becoming more restrictive gard real- the addition with re- of new deities to its pantheon, was equally concerned about extending the opportunities for worship as such, if they fell within the religious ture infrastruc- of Rome. As I shall argue below, it was partly through such quantitative expansion from within that civic religion stimulated the creation of an increasingly complex market of religious options in the city of Rome under the Late Repub- lic. 9 With a view to the socio-political came, identity which be- as far as we can plausibly say, more developed toward the end of the third century, »Roman religion« could be de- scribed as the set of religious behaviour shared by all full members of Rome's political community. The criterion for membership in the res publica - as the res populi (Cic. Rep. 1,39,1) in its strict political sense - would be full zenship and participation in the political and public life of the city of Rome in its military, financial and 8 9 citi- comitial GWYN MORGAN 1990, 26-36. For full documentation, TAYLOR 1937. On proportionality, see above, 2.7.2. See below, 4.1.3. see 166 aspects. 10 In addition to his military and political obligations and privileges of the civis Romanus, citizenship would on this view also define his religion. One obvious problem with this position is that it considers the adult male citizen only. By concentrating on one fraction of Roman society, it excludes from its analysis on a priori were basis those who politically and legally disadvantaged: women and chil- dren, freedmen, slaves without or foreigners who at Rome a full share in the city's political and legal pri- vileges. Political inequality and legal ever, lived did not curtail the religious disabilities, how- behaviour of these groups. The society in which they lived was characterized by a high degree of social mobility, both upward and and downward, (despite elite ideology to the opposite effect) the ab- sence of clear-cut stratification. These parameters entailed the participation of all status groups of society in socie- tal representation. For instance, in the case of freedmen restricted the privileges Roman civil which normally accompanied full citizenship by curtailing their testamentary rights as as obliging them to rights well perform certain duties for their ex- masters. While de iure no political law statutory limitations to their existed, freedmen were in effect excluded from a number of political functions. The Republican polit- ical elite's anxiety about the ever increasing prominence of a substantial group of Roman society which was of servile origin, but whose representational 10 For these aspects, 237-41. Neither of matters. behaviour imitated the see NICOLET 1976; MILLAR 1995, these authors discusses religious 167 elite, resulted in the marginalization of the freedmen f s political influence. For example, in 168 the censor Tiberius Gracchus proposed to deprive freedmen of their votes; compromise, freedmen were restricted the four crises were in probable that during the the latter half of the first century newly manumitted slaves, and possibly all freedmen in excluded limited urban, and thus least prestigious or influen- tial, tribus of Rome. And it is corn a from registering in more than one urban tribe. Subsequently, they to as Rome, were from the plebs frumentaria by Caesar and by Augus- tus, as free-born status (ingenuitas) became the criterion which entitled citizens to receive corn doles. 11 The political with, and marginalization of freedmen arguably was an immediate response to, their so- cial upward mobility. This was the indirect Roman contrasted expansionist policy result of the in the Mediterranean basin from the early second century onwards. The accumulation of wealth through trade and business in transactions the provinces and business in Italy and Rome on the one hand and the tra- ditional means of acquiring riches, namely landed as well as urban property, on the heterogenous group other of encompassed an agents, consisting of the urbanized land-owning Roman aristocracy and equestrians, (and later municipal) status as well as 11 a increasingly elites, large the Italian freeborn citizens of lower number of people of servile Tiberius Gracchus: Livy 45,15,1-7; MRR 1,423-4. Freedmen and corn supply: VIRLOUVET 1991, 48-55; see above, 2.1. In general, see TREGGIARI 1969. 168 origin. 12 For instance, although a significant amount of the economic activities in the Greek East was conducted by Rome's Italian socii, the survey of names of traders on Hellenistic Delos points to the presence of a large number of Roman freedmen on the island. For two reasons, this observation has interesting implications. Firstly, this evidence suggests that the non-Greeks who in the Greek and so-called demonstrably participated oriental cults on Delos were Roman citizens. Secondly, the fact that these freedmen bore senatorial nomina gentilicia entails that the Roman senatorial elite may have been involved in the trade, for which the island served as a centre. Rather than merely accumulating their wealth through landed property and leaving any commercial activity to equestrians and freedmen, these Roman senators, prohibited by legal regulations from direct involve- ment in commercial enterprises, organized maritime trade and other lucrative forms of provincial business through freed- men acting as brokers. 13 However, this is not to from the 12 13 14 that first century freedmen would establish themselves as independent agents of the economies of the deny Rome, Italy and Empire. 14 The wealth that all these heterogenous status Sources of senatorial income: SHATZMAN 1975, 47-74, esp. 53-67; JACZYNOWSKA 1976. Upward social mobility of Roman freedmen and their accumulation of wealth: PURCELL 1983 on apparitores; RAUH 1989 on auctioneers; cf. WISEMAN 1971, 65-94 on the rise of the nouveaux riches obscure loco natus. Roman names on Delos: SOLIN 1982. Economic activity and religious affiliations: BASLEZ 1977; RAUH 1993. On freedmen serving as brokers for senators, see Plut. Cat. mai. 21,5 ff.; Cic. Att. 6,1,19, 5,2 (Cicero's exploitation of Chersonese property through his freedman Philotimus); RAUH 1986. For legal sanctions on senatorial economic activity, see RAWSON 1991, 209-12. Independent freedmen: e.g. RAUH 1989; GARNSEY 1998, 28-44. 169 groups accumulated through such activities in was re-invested social representation and intended to generate, to use a phrase of Pierre representation Bourdieu's, of freedmen »symbolic capital«. As the in funerary reliefs, imitating upper class representational values, shows, freedmen in Late Republican and Early Imperial Rome participated in such public display of social status. 15 As we shall see later, in the second and particularly in the first century the representational and religious behaviour of freedmen is an inte- gral aspect of the social and religious history of the Roman Republic as a whole. Similarly, women in Roman society were unrepresented the political arena and legally would be wrong in concluding from in disadvantaged. Yet, one these disabilities that the social standing of women was such that they did not make distinct contributions to religion at Rome. If one accepted such a conclusion, women like the matron Publicia, the of Cn. wife Cornelius, who restored the sacred precinct and the altar of Hercules in Rome de suo et virei, would have to excluded from be one's analysis, and dedications such as that of lulia Sporis, the wife of the aedituus of Diana Plancia- to Silvanus ex visu would have to be marginalized. 16 On na, the contrary, religion at Rome which was a cultural system to different status groups contributed regardless of po- litical or legal differences. A position which attempts to define religion at Rome by recourse to exclusively political criteria 15 16 is thus ready to conflate interrelated, albeit at Cf. ZANKER 1975. See below, 4.2.3. Publicia: OIL I 2 ,981 = ILLRP 126. lulia 1971,31 with PANCIERA 1970-71, fig. 6. Sporis: AE 170 the same time differentiated, levels of analysis: society and politics. This instance further illustrates the methodological dilemma of a position which assumes these two do- mains to be homologous. 17 Investigating the modalities of enfranchisement confirms that further dilemma. Legally speaking, Roman citizenship could be acquired in different ways: by origo, manumissio, adlectio or adoptio. 18 Consider the case of adlectio: in the later second and first centuries, Roman citizenship was conferred both upon provincial communities in toto and groups of foreigners as well as viritim. For instance, full citizen rights or Latin status were granted by Caesar to a number of Spanish communities which had remained loyal War. Similarly, in the Civil military commanders could reward groups of foreigners with franchise for their military achievements on the battle field: Marius conferred citizenship horts of Camertes on two co- in the Cimbric War, and Pompeius Strabo granted citizenship to a troop of Spanish horsemen by decree in 89 under the Lex lulia of 90. 19 Viritane franchise was granted to individuals as ward for a re- their services to Rome. By the time of the Second Punic War, this institution appears to have been unexcep- tional. In second century historiography, the origin of viritane franchise on merit was projected back into Rome's ear17 18 19 See above, 2.5 and 2.6. Cod. lust. 10,40,7; cf. Ulpian Dig. 50,1,1. Caesar: Livy Per. 110; cf. Cass. Dio 41,24,1, 48,45,3, 49,16,1. Marius: Cic. Balb. 46-51, at 46-7; Val. Max. 5,2,8; Plut. Mar. 28,3. Pompeius Strabo: ILS 8888; Inscr. It. 13,1 pp. 85 and 563; App. BC 1,209-10. In general, see WOLFF 1986. 171 ly Republican history. 20 In the Late Republic, viritane enfranchisement was offered under a variety of utes. Under Roman stat- provisions of leges repetundarum, indivi- the duals received citizenship for successfully prosecuting ca- of repetundae. Furthermore, the phrasing of a preserved ses Lex repetundarum, which is presumably of Gracchan date, plies that before the Social War Latins who held a local magistracy had the benefits of provocatio exemption from im- at Rome and of military service as well as from compulsory munera. It seems very likely that after 90 magistrates in communities with ius Latii automatically received Roman citizenship per magi stratum. 2 ^ In addition, in the first century enfranchisement occurred through individual generosity. For instance, Caesar granted Roman citizenship to of Athenian a number magnates, Eastern potentates, philosophers and the Greek teachers who settled in Rome. The >rule< that Roman citizens could not claim dual citizenship, if it really had any binding legal status second in the and first centuries, appears to have fallen into de- mise by the final years of the Republic. Yet, already in the second and first centuries dual citizenship appears to been unexceptional to many Romans. In Pro Balbo, delivered in 56, Cicero himself acknowledged homines, obtaining have that nonnulli imperiti Athenian citizenship and sitting in the Areopagus, were unaware of the regulation that a Roman citizen should not hold 20 21 another citizenship. Whether such a Cato Orig. 1,26 Chassignet (= 25 P). Second Punic War: Livy 26,21,10-1; cf. Cic. Balb. 51; CORNELL 1991, 63. Cf. Cic. Balb. 53-4. Lex repetundarum (RS no. 1), lines 76-9; Tarentum fragment (RS no. 8), lines 1-7. Citizenship per magistratum: Asconius 2-3 C; BRADEEN 1958-59. 172 >rule< was acknowledged as normative by Cicero's contemporaries or whether it would have been regarded as a mere ana- chronism is, however, difficult to assess. One the one hand, it is noteworthy that Cicero could pretend that there was a rhetorical and legal point in insisting on the impossibility of dual citizenship. 22 On the other hand, under the Late Republic Romans were initiated in the mysteries or Samothrace, Romans in Chalkis -, in Delphi held To or as are most likely to have acquired honorific local franchise in addition to their zenship. who offices in the Greek city-states in the late sec- ond century - for instance as decopo66xoL uepOTiOLOL Eleusis even though local citizenship was not a re- quirement of initiation. By contrast, the religious in Roman citi- these Roman citizens, combining dual civic and religious obligations, the Ciceronian >rule< no doubt ap- peared to be insubstantial. 23 At any rate, by the triumviral period dual citizenship was not only a wide-spread practice, but seems to have become sanctioned on a more systematic basis. For instance, Marcus Antonius was both a Roman and an Athenian citizen; in this latter capacity, he performed duties required of the a gymnasiarch in 39/38. The Romans who held local priesthoods on Cos from 13 BCE onwards presumably had acquired Coan citizenship on a viritane basis. 24 22 23 24 Claims concerning the impossibility of dual citizenship: Cic. Balb. 27-9, 32; Caec. 100; Nep. Att. 3,1. Eleusis: Cic. Leg. 2,36 (Cicero and Atticus); Suet. Aug. 93 (Augustus). Samothrace: SIG3 3,1053 (L. Sicinius M. f. Romaics, after 90). Delphi: BCH 1921, 11 (M. Cornelius M. f., Aulus of Crete, L. Octavius, mid-second century). Chalkis: SEG 29 (1979), 806 (M. Marcius M. f., c. 100). Cf. ERRINGTON 1988, esp. 146, 157. I owe this reference to Prof. Hubert Cancik. Plut. Ant. 23,33-4, 57, 72. Cos: R. 1901, 484ff. no. 4. HERZOG, SB Berlin 173 At the same time, an unaccountable and number ever increasing of provincials shared Roman franchise while maintai- ning indigenous citizenship. If enfranchisement was without irmunitas, further local Sometimes, however, franchise was obligations granted in granted persisted. conjunction irmunitas from the burdens of local citizenship, as in with the case of Seleucus of Rhosus, who was not only given Roman citizenship, but also permitted by Octavian tween 42 and 35 to some time be- keep his local priesthoods and civic honours. The Roman citizens who held priestly offices in the Eastern Mediterranean under the Late Republic or the Eastern potentate Seleucus serve as paradigmatic confirmation that citizenship and religious affiliation were by no means linked or even regarded to be compatible. 25 Saulus of Tarsos, who asserted his Roman citizenship, may provide another lustrative example, although his claim to Roman citizenship has been doubted recently. 26 This is not to deny that citizens shared certain Roman citizenship Roman cultural expectations as regarded their religious behaviour. These examples making il- solely demonstrate that relevant to the study of Roman religion unwittingly treats a political and legal ca- tegory as if it were an exclusive criterion determinative of religious behaviour. 25 26 Seleucus: E-J 301,2,3. Imnunitas and local obligations: FIRA 1,68,3 = SEG 9,8,3, lines 55-62, the Third Augustan Edict to Gyrene. For discussion, see BRUNT 1982; MILLAR 1983; RAWSON 1991, 455-9. Cf. NOETHLICHS 1996, 173, for documentation. 174 3.1.2 On the *Romanness« of Roman religion In retrospect, it is not hard to see why the attempt to link religion to ethnicity and citizenship must fail. For it rives from the nineteenth century city-state ought to be described as view a that de- the Roman nation-state whose organization was based on a legal system which reflected its ethnic nature. Clearly, such a position misapprehends reli- gion's cultural role in what these scholars took ethnic community. to be an Yet/ I argued above that such a position is seriously under-determined as regards the functions it accords religion in Roman society. 27 However, its underlying connection of religious behaviour and communal identity provides a further point of contact with the neo-functionalist (or symbolist) position also encountered above. tion also representing symbolic rightly means the way in which the Romans conceptualized their religious system and construed their »Romanness«. while posi- appears to be under-determined when defining, as we have seen, rituals and their exegeses as of That rejecting For a political or ethnic category of Roman identity, that position substitutes a shadowy cultural notion of »Romanness« (often by reference to opposing cultural model of the allegedly Hellenism), but fails to make clear whose »Romanness« this is - the elite author's, the peasant's, or the Roman's of our imagination. 28 27 28 Cf. MOMMSEN 1907, 390: »... fordert die Ordnung der romischen Gemeinde von dem romischen Burger romischen Glauben und das diesem Glauben entsprechende Verhalten.* For the impact of this position, see further above, 2.2. See above, 2.7.3, for documentation and critique. 175 At the same time, the neo-functionalist position is seriously flawed methodologically. For the very scholars who assert the cultural alienness of Roman religion nonetheless believe that they are able to determine the motives which underlie the »Romanness« of native behaviour, although these motivations, with their alleged profound difference, should on a priori grounds be undeterminable to Furthermore, the modern theory about religion ought to be native discussion In fact, scholars. 29 that native communication interpreted as if it were the category of »Roman- a Latin linguistic equivalent which could appropriately describe that concept was not available to a about »Romanness« conceals the fact that native communication did not use ness«. these prior Late Antiquity. 30 Considering the absence of such a lin- guistic category in ancient Rome, a neo-functionalist posi- tion that views native communication about religion as if it were a communication about Roman religion is in constant danger of over-interpreting the data. As a ciple, it prin- must no doubt be granted that one cannot exclude the possibility that native ritual exegesis, heuristic behaviour, as well as often addressed issues of individual and communal priori identification with Rome. grounds native communication as if it committed treat any itself to making an implicit when it Yet, one statement cannot about on a »Romanness«, does not explicitly say so, as these scholars tend to do. Therefore, it seems not entirely unfair to impute these 29 30 scholars to a tendency to succumb to circular thinking: On this hermeneutic leap, see in general RENFREW 1994b. For the late antique concept of romanitas, see IRMSCHER 1986. Cf. FOWDEN 1993, 37-60, on the socio-political context instrumental in creating that cultural category. 176 the thesis that native religious communication asserted its own Romanness merely is the result of one's initial hypothesis that religious communication may have had that function, as long as that thesis cannot be fully vindicated by the data. 31 While this symbolist concept appears to be an of unsatisfactory »Romanness« therefore heuristic category, the search continues for a methodological frame for encompassing individual as well as communal the religious system at native Rome. identification with Consider Camillus 1 speech against abandoning Rome in 390 in Livy's history (presumably published by 25) , which Rome's famously emphasizes the city of religious significance for its inhabitants. 32 Follo- wing the recapture of Rome, destroyed by the Gallic invasion in 390, Livy's Camillus persuades the Roman 31 32 People not to In chapter 1, I determined »religion« as an umbrella term for the native conceptualization of religious behaviour, although there was not a linguistic concept of that term in native communication. While it could be argued that the creation of the category of »Romanness« provides an analogous case, there is in fact a difference. As it is employed by scholars, »Romanness« is not an umbrella term with which to conceptualize several concepts of cultural behaviour, as is the case with »religion«. Instead, »Romanness« is understood to operate at the level of native communication, to be instrumentalized whenever Romans referred to a particular Roman way of self-understanding. The implicit categorical distinction is crucial: »religion« is a category of behaviour which is expressed by a variety of concepts, but which can be conceptualized without one single linguistic concept. »Romanness«, in contrast, is to these scholars a category of communication. Yet, in order to fulfil a communicative function, the category of »Romanness« has to be expressed in linguistic terms (which it is not), or otherwise is only warranted as a modern heuristic device (which is not how it is employed). Livy 5,49,8-55. For valuable discussion, see LUCE 1971; LEVENE 1993, 175-203; EDWARDS 1996, 44-52. Livy's own religious agenda is sensibly discussed by LIEBESCHUETZ 1967. 177 abandon the city in favour of the site of Veii, which the Romans had destroyed in 396. Throughout his speech, Camillus links Roman religious piety to the physical site of Rome. That the gods approved of Roman piety is shown by their past support of the People's political and military activities. These, however, are inextricably linked to the city of Rome: proper religious rites can take place only within fines of Rome's urban the con- space (5,52,2, 5-7, 13-7); several ancient myths and rituals, such as the rites of Vesta, are closely associated with the foundation of Rome and cannot be transferred to another place; the spatial dimension of the city's festivals and the ritual duties of the Roman hoods are closely priest- linked to the city's sacred topography; Rome was founded in particular religious terms; 33 the significance of stressed, Rome's for religious boundary, the pomerium, is comitia, like any other official state acts, must take place auspicate ... intra pomerium. 24 When abandoning Rome, the populus would leave behind its political religious origins; the gods, both and ancient and those who chose to inhabit the city of Rome more recently - the Capi- toline hill is referred to as sedes deorum - would be deserted; just as their divine presence guaranteed Rome's great- ness in the past and would continue to do so in the future, so their divine retreat would leave Rome a place deserta ... 33 34 Livy 5,52,2: urbem auspicato inaugurate conditam habemus. Cf. Ennius Ann. 155 Skutsch; Livy 1,6,4-7,2. Cf. Varro ap. Cell. 14,7,9 on auspication preceding Senate meetings; LINDERSKI 1986, 2196-8, 2213-4, passim. Auguration as a means of constituting a sacred topography regarded the irrbs with the pomerium as the centre of its system of spatial organization: Cic. ND 3,94 (on auspication): dill gentiusque urbem religione quam ipsis moenibus cingitis-, CANCIK 1985-86. 178 ab dis hominibusque; however, this fate of the urban space would merely symbolize the fate of its citizens. It goes without saying that Livy's narrative is anachronistic. The archaeological record, Livy's assessment of Rome's rather poor withdrawal, suggests that Rome was politically than state after the Gallic neither materially nor too heavily afflicted by the military defeat of 390. 35 Rather, Livy's portrait of Camillus as tissimus confirming religionum the diligen- cultor (5,50,1), the second founder of Rome and pater patriae (5,49,7), over-stresses the depress- ing state of Rome in the year 390 for rhetorical effect: the fictitious Camillus' concern with religious and national revival is Livy's contribution to the contempory revisionist debate about the place of religion and the importance of the city of Rome in the years after the Civil Wars. 36 That revisionist debate started, as far as we in can say, the final decades of the Late Republic. The link between religion and the city of Rome was made by Varro in his Antiquitates rerun? divinarum (presumably published by cording to ought remind them gods citizenship, reli- practice and the physical site of Rome was understood by Cicero, who saw Varro's achievement in the of a 35 CORNELL 1995, 313-22. Contra Livy 5,55,2-5. 36 which to worship and what the appropriate context for worship was. The normative link between gious Ac- Augustine, Varro explained the religious system of Rome to his fellow citizens to they 47) . traditional, presentation if anachronistic, Rome to his elite au- For the cultural context, see conveniently 812-24. PRICE 1996, 179 dience, which was increasingly threatened by social and change. 3 7 Narratives such as the speech of the Livian Camillus have tempted past scholarship to regard the religious traditions inscribed in the stones and rituals of the physical city as an essential part of »Roman religion*; or to re- construe the meaning of Roman religious rituals and institutions from the Varronian account; or to deduce a core of Roman religious beliefs and sentiments from ments the pronounce- of the orators who called upon the gods in defense of their causes. Prima facie, this approach does not lack plausibility: Livy instrumentalized the which the physical site of religious connotations Rome carried as a rhetorical means of creating pathos among his audience. Varro's revaluation of religious institutions and rituals the cultural expectations lend weight audiences. 38 In respective their order to succeed in performing the desired communicative function of projecting conviction among 38 to argument; they would not have used such a strategy if they assumed these propositions to be unacceptable to 37 with of at least one reader, Cicero. The orators made religious propositions to their concurred audiences, their these pronouncements had to refer to August. CD 4,22 p. 172,12-30 D-K = Varro RD frg. 3 Cardauns: pro ingenti beneficio ... praestare se civibus suiSt quia non solum commemorat deos quos coli oporteat a Romanis, verum etiam dicit quid ad quemque pert!neat. Cf. Cic. Acad. Post. 9: nam nos in nostra urbe peregrinantis errantisque tamquam hospites tui libri quasi domum reduxerunt ut possemus aliquando qui et ubi essemus agnoscere . The way in which religious and cultural symbols, individual monuments or the topography of Rome were used to create pathos has been studied extensively; see JAEGER 1997 on Livy, and both POscHL 1983, 17-37 and VASALY 1993 on Cicero. Cf. CANCIK 1985-86. 180 shared cultural expectations. communication What constitutes meaningful is the existence of »structuring structures*, a horizon of cultural expectations and value systems, ural systems of »common sense« (to use a helpful Geertzian phrase referred to earlier), which create constraining cult- a understanding by potentially indefinite number of communica- tive choices. Meaningful communication constantly refers to these »structuring structures*. 39 However, the application of a linguistic theory of un- derstanding leaves the degree inexplicable to which the cultural system of »common sense«, which existed at Rome, was definable in terms of a socio-political or cultural ness«. In other words, what needs to be investigated is whether religious communication at able »Roman- Rome was in principle or unable to constitute a coherently defined system of »Roman religion*. Those scholars who believe that communication was able to achieve exactly that effect must assume that such communication expressed a binding Roman religion, religious whose core of propositions were widely shared, or that it defined a normative religious behaviour so as to set Roman religion apart from other religious systems. it is highly unlikely that more than few of their Roman contemporaries understood the contributions of or Cicero However, Livy, Varro as substantive propositions in this sense or in- deed as a reflection of a pre-defined Roman religion. For it is problematic to hold that the Rome 39 constituted a urban space of public forum for »Romanness« or deter- »Common sense*: GEERTZ 1975. 181 mined a Roman religious identity, when by Late Republic the urbs the end - but the was not only a focus of political communication - in the form of comitia, contiones stiones of quae- or also a huge and wealthy centre of juridical and economic life, of entertainment, recreation and of religious activity for nearly one are million inhabitants. Rarely the implications of this demographic development consi- dered by historians of Roman religion: during the final centuries of the Republic, the city of Rome turned into a metropolis whose demographic dimensions were exceptional any two by standards. It is not until the period of industrializa- tion in nineteenth century Europe that a similar urban concentration was degree of reached again. Furthermore, it is one of the most important implications of the city of Rome's demographic development that the exact constitution population was notoriously of its hard to determine. This was so because high mortality among the urban population on the one hand and large-scale migration ensuring the rapid on the other (the growth of the urbs under the Republic) entailed a constant change in the composition populace. Both of the city those who already inhabited the capital and those who migrated to Rome did so because the not latter urban centre only provided a huge market for various kinds of commo- dities, but also generated an economic and cultural ever production increasing demand for as well as consumption. They were not primarily attracted by Rome as the political centre of the Roman Empire. 40 40 Cf. MORLEY 1996, 33-54, for the Late Republican demographic development. See above, 2.6.1, on the economic differentiation of Late Republican Rome. 182 This demographic variability in population of Rome found the structure of the its equivalent in the Roman co- lonial constitution, which provided for colon! incolaeque hospites atventoresque , » (Roman) citizens of the colonia and resident outsiders, guests and visitors*; and the euergetism of local notables applied to all of these groups. 41 As was the case in Rome, incolae in expected to obey colon! ae municipia and the respective local statutes. They were subject to the authority of the local magistrates and local jurisdiction were in matters under of private law. 42 In these Italian communities, cives et incolae often collaborated local building activities, sharing resulting honours even- ly. 43 Yet, the expulsion of foreigners that from Rome suggests the actual cohabitation was not always seen as harmon- ious. Similarly, in Roman colonies superimposed on settlements existing tensions between the indigenous old and the new Roman elites were not uncommon. 44 In these phic in variability implied cases, demogra- competing political and cultural claims . In the context of the city pansion and of Rome ' s demographic its cultural differentiation during the second and the first centuries, claims of »Romanness« 41 42 43 44 ex- as well as LUrs (RS no. 25) ch. 126, line 31. Cf . ILLRP 617 (Interamniae) : Q. C. Poppaeei Q. f. patron (i) municipi et coloniae municipibus colonels incoleis hospitibus adventoribus lavationem in perpetuom de sua pecunia dant; sim. ILLRP 662 (Sassina) : ... municipibus sueis incoleisque loca sepulturae s(ua) p(ecunia) . Llrn chs 19, 28-9, 54, 83-4, 94. LUrs: chs 95, 98, 133. Cf. Gaius Dig. 50,1,29. E.g. OIL I 2 , 1514 (Cora); Inscr. It. 3,1,36 no. 51 cei) ; PACI 1980b. 103, (Vol- E.g. Cic. Sull. 60-2, describing the conflict between Oscan inhabitants and Sullan colonists in Pompeii; LOMAS 1997, 32-3. On expulsion from Rome, see above, 2.7.1. 183 definite statements about a core of religious and cultural identity would have been constantly challenged by an of competing forms of communication. As a matter of fact, from the late third century the backdrop to a influx plethora city of Rome provided of different communicative signs, both religious and secular, whose over-abundance in the ban space precluded a the ur- emergence of unchangeable ortho- doxies. Moreover, in a society whose lack of efficient con- trol mechanisms for scrutinizing the truthfulness of oral or literary propositions corresponded with its neglect of fac- tual accuracy, persuasion rather than >the historical truth< was the aim of forensic oratory or of historiographical other forms of literary pursuit. 45 As a corollary, aristo- cratic self-representation freely used gies or fictitious genealo- invented its own idiosyncratic >historical truth<, which would in its turn be challenged by other concerning and a family's propositions political or religious past. 46 Such was the resulting variability of communication about cultural and religious issues that the elogia in the 45 Forum Augu- Oratory: e.g. CLASSEN 1985, 5-8, passim. Historiography: WlSEMAN 1979, 3-53; WOODMAN 1988, esp. 73-4, 81-94; KRAUS & WOODMAN 1997, 1-9. 46 See e.g. RUPKE 1995c, 196-202 on the consular Fasti as the secondary products, rather than the sources, of historiography and aristocratic self-display, commencing in c. 173 in the temple of Hercules Musarum; below, 4.1.3. Fictitiousness of genealogies in aristocratic laudationes: Cic. Brut. 61-2; Livy 8,40,4-5; MlLLAR 1989, 140; FLOWER 1996, 133-50. Inventive use of a family's political past: Asconius 12 C. Invention of mythical genealogies in Late Republican and Early Imperial Rome: Serv. Aen. 5,704 and 389 on two works De familiis Troianis by M. Terentius Varro and Hyginus respectively; Dion. Hal. Ant. Rom. 1,85,3; WISEMAN 1987, 207-18. For the unscrupulous creation of fictitious genealogies and pseudohistorical data by political elites in primarily oral societies, cf. the comparative perspective of SCHOTT 1990. 184 stum, drawing on the Augustan compilation of historical in- formation in the annales maximi, painted a picture of Roman history that differed in many details from Livy's. 47 Varro's interpretation of Roman religion was not an explicit exposition of a widely accepted and internalized religious system, but entailed a critique of the religious behaviour of his fellow-citizens, thus implying the interpretations existence of competing and of differentiated belief systems. 48 Fi- nally, when invoking the deities of the city orators did not draw on a theological orthodoxy concerning their nature character. Rather in the absence of such orthodoxies, theo- logical issues were notoriously debatable. theological truth were just When acceptance claims of as fluctuating as the city's population, the character of a deity such as the her and Bona Dea, by Cicero or her refusal by Clodius and his supporters among the city population, could become the sub- ject of public discussion. 49 Consider for instance, the Mater magna, whose been had transferred to Rome in 204, and whose sanctuary on the Palatine, overlooking the Forum Boarium Campus cult and parts of the Martius, attracted the attention of visitors as soon as they entered the city through the Porta Trigemina. Scholars tend to treat the transfer of the goddess to Rome as assertion an of the city's national identity during the Second 47 LUCE 1990. Cf. KRAUS & WOODMAN 1997, 70-4, with further literature. Regarding the sack of Rome in 390, the elogia relevant to Livy's narrative are Inscr. It. 13,3 nos 11 and 61. 48 Cf. FEENEY 1998, 139-40. 49 Cf. WISEMAN 1974, 136-7 on the polemics concerning the (un-)chastity of the Bona Dea in the mid first century. 185 Punic War. 50 The statuary and coins present the Mater as magna a chaste mother goddess. Cicero describes her games, the Ludi Megalenses, as more institutisque maxime casti, sollemnes, religiosi (Har. Resp. 24). On the other hand, the sence of pre- the Metragyrtai and Galli in the city suggests an Eastern religion, whose allegedly orgiastic aspects alien- ated traditional religious sensibilities in Rome. 51 The goddess was associated with agricultural fertility: as early as 204, the exceptionally productive fertility of the Italian soil was autumn harvest and the related to the deity's arrival earlier that year (Pliny NH 18,4,16). In Late Republican statuary, dating to the fifties and presumably re- sembling the (lost) cult statue on the Palatine, the goddess tunica, is seated on a throne and wears a belted and sandals. graphy of a mantle She is adorned with ears of corn. This icono- fertility and nutrition has tentatively been linked to Pompey's policy of corn supply. By Stoic allegorization, Varro identified the Mater magna as luno, Ceres and Tellus. He stressed the aspects of fertility the deity and portrayed as the bread-winner of mankind. 52 By way of con- trast, among the votives found in the area of the Mater magna 's sanctuary figurines of Attis, of inexpensive quality, stand 50 51 52 poor out by their sheer number. They predate the first burning of the temple in 111 and Attis and thereby prove that was worshipped on the Palatine as early as the second E.g. GRUEN 1990, 33; ALVAR 27-33. Cf. BURTON 1996. 1994, 154; BEARD I994b, E.g. Dion. Hal. Ant. Rom. 2,19,3; LATTE 1960, 259-60; BEARD 1994b. Cf. Catull. 63,91-3; Varro Sat. frg. 142 Cebe = 133 Astbury; WISEMAN 1985a, 269-72. varro RD frgs. 64, 267-69 Cardauns; cf. Strab. 10,3,12, 469. Statuary: FUCHS 1982. 186 century. 53 Attis had been introduced to the Hellenic cult of Cybele in the fourth century, and had been part of the god- dess' cult at Pergamum, from where the goddess was transferred to Rome. At Late Republican Rome, however, Attis did not receive public worship by the state's functionaries. 54 The private worship of Attis on the Palatine the thus becomes all more remarkable. Arguably, the Mater magna and her cult were capable of evoking a wide range of responses from wor- shippers and observers alike. Given this over-abundant field meaning, native communication of competing about define an unchanged therefore preferable to treat core of religion cannot on a priori grounds be taken as a factual statement clearly forms of nor can it »Romanness«. It is religious communication in such contexts as a performance whose validity is situational at best. What does this notion mean? Religious communication refers to a cultural system of »common sense« which constrains understanding. It is with respect to that cultural system that religious communication may entail substan- tive claims. But these claims are debatable; they may evoke, as we have seen, different responses in the audience that range from emphatic approval to outright neglect. As I argued above, religious communication does stable and unshifting meaning not have a single which is determined by the meaning that the sender of communication attaches to it, as the received analogy with the linear transmission of meaning 53 ROMANELLI 1963. Date: 54 Attis is not represented in the Republican coinage. The reverse of a denarius struck by M. Volteius in 76, commonly thought to show Attis, in fact represents Bellona; TURCAN 1983, 13-4. COARELLI 1977, 10-3. 187 in linguistic theory might suggest. Rather, it may engender a vast number of different responses in its addressees. 55 Expanding that suggestion, I would religious propose that Roman communication is performative not only because it is capable of creating various responses. It is performative also because these responses are situational in that they do not entail that in a different situation the to commit addressee has himself to their validity. 56 This is not to deny that authors like Cicero, Varro or Livy deliberately created the category of political or Rome as one means cultural identification with of meaningful communication with their respective audiences. There is no reason to assume that they doubted the persuasiveness of that category in their respective communicative contexts. However, realize that their religious other, created only one it is communication, particular form of important to just as any communicative input into an over-abundant field of communication. It was a situational construct whose claim to making a binding statement about the true nature of Roman religion was constantly questioned by the input of competing communicative signs. A focus on the performative, rather tional, than the role of religious communication in Roman society is therefore more than just a heuristic device. In the of a absence shared orthodoxy concerning religious belief systems, religious discourse, shaping its content 55 56 proposi- according to its See above, 2.7.3. See above, 2.7.4. For »performative structures* in social behaviour, see the evocative account in SAHLINS 1985. With regard to ritual, cf. H. CANCIK & H. MOHR, HrwG 1 (1988), 147. 188 rhetorical needs, expressed partial truths at best and used single elements of religious reality without ever offering a substantially complete picture. Livy, orators Varro or the Roman created imagined, rather than represented real, au- diences. The communities they depicted are >imagined communities <; the Roman society they Roman society that is envisaged is an idealized a matter of ideology, but does not exist in reality or at the level of personal contact. It is only in communication that a meaningful face-to-face-society could be recreated, given the abundance of competing forms % of meaning in an urban space which confines of had in outgrown the Livy's Camillus the was past, the response of a late first century au- dience to his expositions must have been just as as the a nuclear community. 57 While the imagined com- munity of Roman citizens addressed by set long reaction of fragmented the addressees of Varro's work to his revaluation of Roman religion. The creation of imagined communities is most conspicuous in the case of political oratory. Although it created the impression of talking to, as well as on behalf of, the Roman People as a whole, its actual addressees were at any time during the second and first centuries only a of the tiny fraction Roman citizen body. Assemblies over-represented the plebs urbana to a disproportionate degree, whereas for merely practical reasons citizens from distant voting-districts scarcely attended contiones and comitia. In the first century, 57 the census was conducted in local communities and the For >imagined communities<, see B. ANDERSON 1991, esp. 15-6, used for the study of forensic oratory in fourth century Athens by OBER 1989, 31-3. 189 results delivered to the censors at Rome, so that Roman citizens did not have to visit Rome in order to get enrolled. 58 Moreover, the expulsion of peregrin! from Rome by the tribunus plebis M. lunius Pennus in 126 or by the consul C. Fan- nius in 122 must not simply be taken as reflective of a more general trend of expulsion in Late Republican and early Im- perial Rome. In these specific cases, the measures were presumably intended to address the interference of Italians in the Latins and political contiones of the latter half of the second century, at which the issue of Italian franchise was debated. 59 These incidents elucidate that the orators in the Forum Romanum did not speak in front of a pure citizen body, but addressed an agglomeration of plebs and female urbana, male migrants to the capital and short-time visitors with or without citizenship. 3.1.3 Addressing the pantheon of the city of Rome The variability of religious communication entailed that its scope was open to situational renegotiation. Addresses to Rome's pantheon are a case in point. Its gods were constantly invoked in the course of the city's life: at the begin- ning of a contio the presiding magistrate, before addressing the People, uttered a traditional and 58 59 formulaic prayer to Census: Tabula Heracleensis (RS no. 24), lines 142-56. Overrepresentation of the plebs urbana: e.g. PURCELL 1994, 644-6; MiLLAR 1995, 240-1. Scarcity of attendance at contiones and comitia: MACMULLEN 1980. Expulsion in 122: Plut. C. Gracch. 8,2-3, 11,2-3, 12,1-2. Cf. E. GABBA, CAH 8 2 (1989), 240-2; A. LlNTOTT, CAH 9 2 (1994), 76 and 83. Cic. Off. 3,47 is rather unclear about the events of 126. 190 gods; 60 the and prayers were topical in other public ora- tions as well, political or juridical. 61 Orators invoked the religious connotations of Rome's sacred topography by ing their audience's draw- attention to a nearby divinity or by addressing that divinity's potential intervention in the orator's case. 62 As noticed above, the oath by luppiter and the optimus maximus Di Penates publici was used as a political instru- ment by Saturninus and Caesar. 63 An epigraphic parallel that oath is for provided by the Lex Latina tabulae Bantinae, dating from the very end of the second century. This statute required both present and future Roman magistrates to take an oath pro aede Castorus palam luci in forum vorsus et eisdem in diebus quinque aput quaestorem ... per lovem deosque Penateis, sese quae ex hace lege oportebit facturum neque sese advorsum hance legem facturum scientem dolo malo neque sese facturum intercessurum esse quo haece lex minus setiusue fiat. Its binding character was emphasized by detailed tion of the specifica- consequences which a refusal to take the oath would have for the individual magistrate, amounting to a termination of any future political career. Furthermore, the 60 61 62 63 Livy 39,15,1: consules in rostra escenderunt et contione advocata cum sollemne carmen precationis, quod praefari solent priusquam populum adloquantur magistratus, peregisset consul, it a coepit ...; 39,18,3. On carmen preca­ tionis, »ceremonial prayer formula«, see SCHEID & SVENBRO 1989-90. Cf. Cic. Ad Her. 4,68; Pliny Paneg. 1,1, 63,2; MOMMSEN 1887, 3,390; APPEL 1909, 57. Serv. Aen. 11,301: nam maiores nullam orationem nisi invocatis numinibus inchoabant, sicut sunt omnes orationes Catonis et Gracchi. For a sample from the Ciceronian corpus, see p. Corn. I frgs. 1, 2; Verr. 5.184-9; Mur. 1-2; Dom. 144-5; Rab. perd. 5; Post red. ad Quir. 1; Arch. 1; cf. Div. Caec. 43. For a collection of the Ciceronian material, see HEIBGES 1969. Cf. VASALY 1993. Cf. above, 2.7.2. 191 oath had to be taken on the open platform of the sanctuary of Castor and Pollux, which had been built in front of the backdrop of the temple facade proper and was facing the most crowded and public place of Rome. 64 The same document re- quired Roman senators to swear apud quaestorem palam luci the aerarium per lovem deosque Penateis, that is in front of the temple of Saturnus, thereby ensuring in ad area similar publicity bordering on the north-west corner of the Fo- rum . 6 5 The emotional appeal to »the di patrii ac penates« was effectively instrumentalized by Cicero, who no doubt drew on the oath which magistrates had to take by luppiter optimus maximus and the Di Penates publici, as if the orator were required to take an oath by those deities that his defendant was innocent. The phrase should probably be taken as a tau- tology, since the di patrii (the ancestral gods rather the gods of the patria) comprised the Di Penates publici that were housed in the delubrum of Vesta and therefore came appropriate addressees of an 65 66 patrii penates as the guardians of Rome and the res publica, and no doubt expected proper 64 be- orator delivering his speech in the Forum. 66 Cicero portrayed the familiaresque than emotional response from the RS no. 7, lines 14-8. For the architectural design, see ULRICH 1995, 81-107. Ibid, line 24. Cf. the Tarentum fragment (RS no. 8), lines 20-3 of magistrates required to swear apud quae­ stor em Romanum ... per lovem deosque Penates; the Lex de provinciis praetoriis (RS no. 12, Delphi copy, block C) lines 10-15. E.g. Verr. 4,17, Sull. 86,1, Har. Resp. 37, Sest. 45; Nep. Themist. 7.4. Di Penates publici: LATTE 1960, 89-90. Illustration: TURCAN 1988, 1, nos 107-8. On the link between the Di Penates publici and Vesta, see Tac. Ann. 15,41,1; RADKE 1981. 192 pontifical college, his elite audience on that occasion. 67 Similarly, the connection of patria and Di Penates in emo- tional narrative was not sought for the sake of alliteration either. It denoted a core of possible religious identifica- tion in a Roman elite context. 68 However, as my emphasis on suggests, even such regarded about of essence elite Roman as prepositional which For in large statements creating proportion marginalized of the city populace. As they received sacrifices from Roman officials, the Di Penates public! came closely associated be- with the political domain. In the city's domestic cult, the Di Penates were generally, if invariably, an identified with the Di Penates of Rome, the orators implicitly took into account that they a already religion. Rather, these pro- nouncements are markedly situational. audience context supposedly straightforward pronounce- ments should not be the the not worshipped by the freeborn; and accounts of up- per class houses depict an affectionate relationship between house owners and their Di Penates. By contrast, their familiae, both slave in urban surroundings and on rural estates, showed a marked preference for worshipping the Lares and the Genius. 69 67 68 69 Don?. 144. Cf. Verg. Aen. 8,678-9. Patria - penates: Livy 22,3,10, 25,18,10, 30,33,11; Curt. 4,14,7, 5,5,20; Tac. Ann. 11,16, Hist. 3,84; Symm. Epist. 6,72. The archaeological evidence for this differentiating pattern comes from houses in Pompeii and Herculaneum: FR8HLICH 1991, 30-1, 40, 178-9, 261; Id. 1995, 205; GEORGE 1997, 316-7; below, 4.2.4. Upper class households and the Di Penates: BARKER 1994, 180; NEVETT 1997, 289-90; cf. SALLER 1994, 88-95. Contrast e.g. Cato Agr. 2,1, 143,2; Plaut. Merc. 834; Cic. Att. 7,7,3, LATTE 1960, 90-2, elucidating the attachment of the slave or rural familia to the Lar familiaris. Cf. ILLRP 196, the 193 The communis opinio, dating back to the nineteenth century, holds that Vesta, as the goddess of the hearth, was the archetypical Roman deity of domestic worship, whose prominent public representation served to symbolize the inex- tricable link between the spheres of private and Roman religion. Contrary public in to that view, the archaeological data drawn from the Lararia and the paintings of Pompeian households demonstrate the absence of the cult of Vesta from major realms of private religion and thus corroborate the suspicion that Vesta did not play a significant role in dividual or in- domestic worship during the Late Republican or Early Imperial periods. Vesta was the patron goddess of Roman pistores, which makes her omission from other areas of private cult only the more conspicuous. 70 By contrast, the goddess 1 cult in her delubrum on the Forum Romanum was south border of the inextricably linked, just as her impact was strangely limited, to the political domain; as such, the juxtaposition of luppiter optimus maximus, publici the Di Penates and Vesta provided only partial religious identifi- cation in a context which was tailored to the politicized discourses of the Late Republic. It is this limited connota- 70 dedication of an aedicula to the Lares familiares by a magister called Draco; ILLRP 197, a shrine for the Lares by the vilicus Philargurus de sua pecunia; ILLRP 199-201. The Lares publici were the recipients of civic worship: e.g. CIL 6,456 (4 CE); Ovid Fasti 5,129ff. with BOMER ad loc.; ROPKE 1995a, 61-2. Yet, the available evidence referring to their prominence in civic cult does not appear to predate the Augustan revival of the cult of the Lares: cf. e.g. Augustus RG app. 2. See RADKE 1981, 363; FROHLICH 1995, 208-9. Contra WissoWA 1912, 156-61. R. L. GORDON, OCD3 (1996), 1591, assuming that »in the historical period the state cult effectively displaced private cults (sc. of Vesta)«, succumbs to an implausible theory of religious evolution. 194 tion only which political oratory exploited in a field of religious communication that was, as I have argued above, over-abundant in different meanings by any standards. However, even in political oratory the spectrum pantheon could be of the extended. To be sure, Cicero humorously attacked those who believed they were well prepared to give a public speech, when they had only memorized the line lovem ego optimum maximum ... (Div. Caec. 43). luppiter optimus maximus was indeed an essential addressee of public Furthermore, the second and prayer. first century dedications by foreigners on the Capitoline hill to luppiter optimus mus maxi- and to the Populus Romanus documented the god's central role in the religious definition of the Roman state to siders as well as to Romans. Nevertheless, as we shall see below, the Capitoline triad also provided a focus vidual out- of indi- worship that was in striking contrast to civic reli- gion. 71 Yet, public oratory employed a wide range of different deities according to its rhetorical needs. As noted above, in order to create pathos among his pontifical Cicero, on his audience return from exile, exploited those deities who were accommodated in the political centre of Rome: piter 71 Capitolinus lup- (sic), luno Regina, Minerva, the Penates See below, 4.2.5. For an edition of the dedications on the Capitoline hill, see DEGRASSI 1962-67, 1,415-44; and for their date, LINTOTT 1978; MELLOR 1978. See ERSKINE 1994 for documentation of the evidence for cult of the Populus Romanus in the Greek East, which started in the early second century. For the centrality of the Capitoline cult to Romans, see e.g. C. Gracchus ORF F 58 Malcovati: Quo me miser conferam? Quo vertam? In Capitoliumne? At fratris sanguine madet. In general, see FEARS 1981. 195 and Vesta. 72 By way of contrast, the conclusio final of Cicero's speech against Verres addressed not only luppiter op- timus maximus, luno Regina and Minerva, but also Latona, Apollo and Diana, Mercurius, Hercules, the Mater magna, Castor and Pollux, Ceres and all the other gods, omnes di, in- cluding those indigenous deities whose rights had been vio- lated by Verres in Sicily. 73 The deities invoked were chosen with the case under consideration in view; but whereas their choice was never merely incidental, the underlying principle of including or excluding deities displayed a high degree of variability, which was constrained by contextual rather than by any unchanging substantive criteria. A historiographical sample illustrates this contextual!ty of choices. In Livy's history, it is often one particular deity to whom prayers and queries were made. 74 Or, depending on the circumstances under which a selection of was uttered, been discussed above, in most cases it is the gods as such, di imnortales or di deaeque, to whom individuals, the Senate, magistrates 72 73 74 75 a two or more deities was addressed. 75 However, it is worth noting, and has already that prayer or Cic. Dom. 144-6, effectively placed near the end of the speech as an essential part of the rhetorical conclusio. Verr. 2,5,184-9. Livy 1,16,3, 16,6, 3,17,6 (Romulus); 1,18,9, 4,2,8, 28,28,11 (luppiter optimus maximus); 2,10,11 (Tiberinus); 10,18,14 (Hercules); 29,14,13 (Mater magna). Livy 5,21,2-3: Apollo Pythicus and luno Regina on the occasion of the evocatio of Veientine luno; 6,20,9: luppiter and di alii invoked by M. Manlius, the defender of the Capitolina arx; 10,24,16: luppiter optimus maximus and di irnmortales addressed by P. Decius as consul in a contio; 24,38,8: a reference to the local deities Ceres and Proserpina in an address to the troops on Sicily; 34,24,2: luppiter optimus maximus and luno Regina, a reference to Roman military power and to the tutelary goddess of Argos respectively. 196 military commanders made prayers. 76 Livy portrayed Camillus as a diligentissimus religionum cultor who took note of large number of divinities living in Rome and demanded the preservation of their temples Caesar the and cults. Livy introduced Augustus as templorum omnium conditor ac restitutor, a sentiment which was later echoed in Augustus' own Res gestae. 77 Great individuals, both past and expected to present, could be take responsibility for all gods and goddesses at Rome. 3.1.4 Priesthoods and aristocratic competition Cicero famously illustrates the link between political authority religious and and thereby seems to support the civic model's thesis that religion was conceptually undifferentiated from the political life of Republican Rome: Cum multa divinitus, pontifices, a maioribus nostris inventa atque instituta sunt, turn nihil praeclarius quam quod eosdem et religionibus deorum immortalium et summae rei publicae praeesse voluerunt, ut amplissimi et clarissimi cives rem publicam bene gerendo religiones, religiones sapienter interpretando rem publicam conservarent. 78 Yet, since Cicero addresses the pontifical not surprising 77 78 it is that he wants to make such an explicit link between political administration 76 college, and religious expertise. E.g. Livy Praef. 13; 2,6,8; 4,13,14; 4,46,4; 5,18,11-2; 5,21,15; 5,32,9; 6,23,11; 6,29,2; 7,26,4; 7,40,4-5; 9,8,8-10; 10,13,12; 10,35,14; 21,17,4; 27,45,8-9; 28,41,13; 29,22,5; 29,27,2-4 and 9; 31,5,3-4, 31,8,2; 31,7,15 37,36,6; 38,51,10. The very same formula was used by non-Romans: e.g. Livy 6,26,6: dictator from Tusculum; 7,20,3: Caeritean envoys; 23,13,4: Hanno; 26,41,16, 36,7,21: Hannibal; 35,18,7: Alexander of Acarnania; 40,9,5 and 11,5: Perseus; 42,13,12: Eumenes. Camillus: 5,49,7-54,7; Augustus: Livy 4,20,7; Augustus RG 20,4-21,1. Cic. Dom. 1. See above, 2.4.2. 197 Yet, there is a more fundamental problem with the civic model's emphasis on the embeddedness of religious expertise in the political realm. As we have seen above, this model pre- sumes that religion had its institutionalized place in politics and public life, and surmises that it was the political realm that determined religion. postulates an abstract >Roman Such a model unwittingly state< represented by >the Senate and the magistrates<. The uniformity of the political realm, however, is a problematic concept. Political life in Late Republican Rome was not a stable entity that described could in terms which resemble the modern concept of in- stitutionalized government. It has repeatedly been that what stressed we call the Roman constitution did not mirror an institutionalized political system, but rather political be life that was characterized resembled a by informal power structures based on relationships defined by the exchange of officia among the elite. A distinguished nomen, dependent on the agnatic lineage of the family, could secure political success and office. Office-holding was of course not hereditary, but divitiae, imagines and the memoria praeclara of one's ancestors were the nobility's assets in elite competition. 79 However, the practical impact of agnatic lineages on the competition for political office was short-reaching. The nobility of through office-holding, and material understood to be the Republic was defined and symbolic capital had to be re-invested in each generation in order to achieve political 79 success. This is not to deny that some families Sail. lug. 85,38. Cf. ibid. 85,4: vetus nobilitas, maiorum fortia facta, cognatorum et adfinium opes, multae clientelae; WISEMAN 1971, 95-116; HOPKINS 1983, 107-17; FLOWER 1996, 60-90. 198 held offices over many generations. 80 But century from the second onwards inheritance of office was increasingly con- trolled through the elections, and genealogies, imagines and nobilitas the elusive concept of which though could not guarantee, succession. facilitated, were Many sons of office-holders are likely social to have constructs failed to succeed to the office that their fathers or grandfathers had obtained. 81 Given the fluidity of the political elite should be expected system, the Roman to advertise or commemorate its authority through priesthoods, if religious authority was entirely undifferentiated from other spheres of public life. Such an inextricable link between religion and politics, however, is not corroborated by the political elite's representation. Consider self- the laudatio funebris held in 221 by Q. Caecilius Metellus, pontifex from 216, in honour of his father L. Caecilius Metellus, the second plebeian ponti­ fex maximus (243-21), as transmitted through Pliny: Q. Metellus in ea oratione quam habuit supremis laudibus patris sui L. Metelli pontificis, bis consulis, dictatoris, magistri equitum, xv viri agris dandis, qui primus elephantos ex primo Punico bello duxit in triumpho, scriptum reliquit decem maximas res optumasque, in quibus quaerendis sapientes aetatem exigerent, consummasse eum: voluisse enim primarium bellatorem esse, optimum oratorem, fortissimum imperatorem, auspicio suo maximas res geri, maximo honore uti, summa sapientia esse, summum senatorem haberi, pecuniam magnam bono modo invenire, multos liberos relinquere et clarissimum in civitate 80 81 For this aspect, see BURCKHARDT 1990; JEHNE 1995. Cf. Cic. Pro Mur. 16; Asconius 12 and 23 C; MILLAR 1984; Id. 1986; H8LKESKAMP 1987, 204-9; YAKOBSON 1992. HOPKINS 1983, 31-119 and SALLER 1994, 12-69 provide an important demographic perspective. For the patterns of office-holding, see WISEMAN 1971, 1-12; BADIAN 1990; MlLLAR 1989, 141-4. 199 esse. Haec contigisse ei nee ulli alii post Romam conditam. 82 Pliny's own >preface< to this text mentions Metellus 1 pontificate and political offices. Augustan and Pliny clearly imitates the later imperial structure of a cursus honorum, which routinely listed priesthoods as a prominent part of an individual's career in reflection of the emperor's phasis on his priestly status. 83 - em- By contrast, Metellus' priesthood is missing from the - otherwise sive own very comprehen- list of his life achievements drawn together by his son, who was to become a pontiff himself. The Metellus' virtues employs classification: Metellus' two catalogue complementary sapientia was of patterns of manifest in the of domi militiaeque and showed in his civic and pri- realms vate responsibilities. Metellus was both a foremost bellator and an eminent orator; he had imperium and held the auspi- cium as magistrate and military commander; his maximus honor was the consulship; he embodied the ethos of an active and prudent participation in Roman political life (surnna sapien­ tia and surnnus senator); his wealth traditional means had been acquired by (through landed property rather than the exploitation of the allies or direct business in the Greek East, one might presume) and he begot sons that were to survive their father. A reference to Metellus' holding the of- fice of pontifex maximus is only implicitly contained in the claim that he was clarissimus in civitate. taken 82 83 This should be as referring to Metellus' rescuing the sacred objects ORF4 no . 6/ f rg . 2 ap. Pliny NH 7,139-40. Priesthood: SZEMLER 1972, 68-9. The text's competitive ethos is analysed by KIERDORF 1980, 10-21; WlSEMAN 1985b, 3-4. Cf. ECK 1984, 148-52. 200 of Vesta from the goddess 1 burning delubrum, the loss which led to of his eyesight. As an expression of their grati- tude, the Roman People bestowed on Metellus the right to use a cart when entering the Curia, an before. Metellus 1 honour never conferred of the sacra of Vesta must preservation presumably be linked to the office of pontifex maximus. vertheless, he clarissimus was Ne- due to the bestowal of an extraordinary honour by the People and not because he had held this particular priestly office. Apparently, in late third century Rome an individual's public fame and the assertion of his virtue were not contingent on religious affiliations or beliefs. 84 Moreover, the domain of religion, either in its civic forms or in its private manifestations, could be domain of societal excluded communication. entirely from Its absence is all the more remarkable as Metellus 1 civic and domestic virtues otherwise It domestic which began of of political power nevertheless interrela- and religious expertise suggests that sacred and civic domains seems secular- to differentiate secular from sacred domains. However, while Cicero's emphasis on the tion pro- would be rash to suggest that this list repre- sents a catalogue of secular virtues in an age ization, are meticulously listed as warfare and civic oratory, military and civic office, public renown and sperity. the misleading remained to interdependent, maintain on a it priori grounds that religious communication was conceptually embedded in, and undifferentiated from, other aspects of life 84 Pliny NH 7,141; cf. Dion. Hal. CANCIK 1994, 357-8. Ant. Rom. 2,66,4. in See 201 Republican Rome. Roman elite's contingent On the religious contrary, activity in the expression of the the city-state upon the context of communication. Religious au- thority did not universally have a place in public cation at plebeian communi- Metellus 1 laudatio was not exceptional in Rome. this respect. Livy's praise of Licinius Crassus third was Dives, the pontifex maximus (212-183), probably echoes the original laudatio funebris: Is ... bello quoque bonus habitus ad cetera quibus nemo ea tempestate instructior civis habebatur congestis om­ nibus humanis ab natura fortunaque bonis. Nobilis idem ac dives erat; forma viribusque corporis excellebat; facundissimus habebatur, seu causa oranda seu in senatu et apud populum suadendi ac dissuadendi locus esset; iuris pontificii peritissimus,- super haec bellicae quo­ que laudis consulatus compotem fecerat. 85 Again there is no direct reference to Crassus 1 but rather to his expertise in pontifical law. The preemi- nently public occasion of strikingly pontificate, omitted his the commemoration priesthood. This of of dead dissociation of civic and religious roles is also found in Cicero's sion the discus- eminent Roman public orators: he mentions the elo- quence, the consulship and the censorship of P. Scipio Nasica Corculum, but does not refer to Nasica's holding the of- fice of pontifex maximus (Brut. 79). To complete this sequence of plebeian pontiffs, the terary plebeian li- epitaph for P. Licinius Crassus Mucianus, the fourth pontifex maximus (132-30) indeed mentions his priestly office: Is Crassus a Sempronio Asellione et plerisque aliis historiae Romanae scriptoribus traditur habuisse quinque rerum bonarum maxima et praecipua: quod esset ditissi- 85 Livy 30,1,4-6. Priesthood: SZEMLER 1972, 105-7. 202 mus, quod nobilissimus f quod eloquentissimus, quod iuris consultissimus, quod pontifex maximus. 86 This example shows that under certain circumstances of the elite would members publicly document a priesthood. 87 But these are deliberate contextual choices. For instance, the famous inscription on the sarcophagus of P. Cornelius Scapula, dating to the second half of the fourth century, only gives the office of pontifex maximus, but omits any other honours. If this text refers to the consul of 328, the omission is significant. Presumably, the text should be under- stood as the assertion of superior status by a patrician family in the generation immediately preceding the Lex Ogulof 300, which opened the colleges of pontifices and au- nia gures to plebeian families. 88 Membership in the Xviri sacris faciundis is (ILLRP 316). of included in the elogium of Cn. Cornelius Cn. f. Scipio Hispanus, presumably 130 college dating to c. To the generations immediately preceding the Lex Domitia of 104/3, the issue of electing members of the major Roman priesthoods in the popular assemblies became part of public therefore could political become self-representation and controversy; 89 expedient assets and priesthoods in aristocratic political competition. It may have been this element of aristocratic ambition in the arena 86 87 88 89 which political led to the inclusion of the office of pontifex Sempronius Asellio frg. 8 P ap. Cell. hood: SZEMLER 1972, 121-2. 1,13,10. Priest- Cf. e.g. 'C. Memmius' (tr. pi. Ill, MRR 3,141) ap. Sail. lug. 31,10: (nobiles) incedunt per ora vostra magnifici sacerdotia et consulatus pars triumphos suos ostentantes. Text and commentary: PISANI SARTORIO & QUILICI GIGLI 1987-88, 259-60: P. Cornelio P. f. Scapola pontifex ma­ ximus. Date: BLANCK 1966-67; H. SOLIN, Gnomon 67 (1995), 613. Lex Ogulnia: HOLKESKAMP 1988. Cf. NORTH 1990, esp. 535-9; below, 3.2.1. 203 maximus in the laudatio of Licinius Crassus. The situational expediency of advertising one's priesthood in elite competition explains why some members of the upper classes would sometimes want to refer to their priestly roles. Q. Cornificius mentioned his membership in the college of augurs, as that priestly office had been bestowed on him in 47 presum- ably due to lulius Caesar's intervention (thus foreshadowing the custom of promoting the Roman nobility to priesthoods by imperial patronage). 90 By contrast, M. Porcius Cato the El- der, for whom a priesthood is not attested, that did not think civic priesthoods would necessarily matter in the pub- lic life of a Republican aristocrat: Cato primus Porciae gentis tres surnnas in homine res praestitisse existimatur, ut esset optimus orator opti­ mus imperator optimus senator. 91 3.2 Cultural self-consciousness and Roman religion The emphasis on the contextuality of choices rather than any substantivist parameters does not entail that the crea- tion of meaning in religious Yet, on communication was haphazard. the model provided by the Geertzian cultural system of »common sense« is too imprecise to be truly capable of defining one particular input factor which would have determined 90 91 ILLRP 439; RAWSON 1991, 272. For Cornificius' coinage of 42, showing him veiled and with the lituus f and carrying the legend aug(ur et) imp(erator), which linked his augurate to his military victory in Africa, see RRC 1,518-9, no. 509 with the discussion by FEARS 1975; RAWSON 1991, 281-6. Pliny NH 7,100. Cf. Cic. De or. 3,135; Livy 39,40,5-6; Corn. Nepos Cato 3,1; Quint. 12,11,23. No priesthood attested: Cic. Cato mai. 64 (in vestro collegio) with J. G. F. POWELL ad Joe.; contra (wrongly) MRR 3,170. 204 religion at Rome or which could rubric of »Romanness« in a be categorized under the socio-political or cultural sense. As I suggested above, this result is not surprising. For we are not dealing with a closed system characterized by a linear process of one particular input, which for us could determine the >meaning of Roman religion<. Rather, the religious system at Rome was open in that its output was deter- mined by different forms of input. It is important that to note the religious system itself, rather than a determinant socio-political or cultural identity situated at an ideological level, regulated and processed the input. 92 This is not to deny the input which the religious system received from an area which was generated by communal identification with individual the Roman city-state. Rome's military and political supremacy in the Mediterranean from or world the third century onwards and the roughly contemporary formation of a conscious discourse about her appear to be parallel cultural past processes. As mentioned above, the emergence of Latin literature toward the end of the third and the beginning of the second century can be understood as a process in the course of which a conscious attempt was made to create a national cultural tradition. In what was corollary process, Roman literature from Fabius Pictor on- wards provided an ethnographic inquiry into the origins different a and meanings of Roman social and religious custom. To the historian of Roman Republican religion, of a 92 See above, 2.7.4. the production Roman cultural consciousness with regard to the reli- 205 gious system is mainly visible in these literary accounts of the second and first centuries. 93 However, their epistemological status as a source of direct evidence about Roman religion is compromised, once we realize that the logical organization and classification of the religious data is constrained by the ulterior literary sources 1 motive of creating a programme of cultural identi- fication. It is therefore methodologically advisable to sume that they did not necessarily provide an undistorted picture of religion in society as a whole. It is place that as- a common- in Roman society, as in many traditional socie- ties (and in European society prior to the eighteenth century) , the literary reflection on religion extremely low degree was, due to the of literacy, an elite privilege, and that its products were tailored to the cultural expectations and needs of that very elite. 94 When describing the place of elite literacy in Roman society, the between of popular culture and the the »great tradition* received »little tradition* of upper-class culture is a ristic useful in religious 94 be- Roman society as a whole and the »little tradi- tion* of elite thought about the cultural meaning 93 heu- device. This distinction enables us to conceptualize the relationship of the »great tradition* of haviour distinction of reli- RAWSON 1991, 80-1 outlines the creation of a »new consciousness of the Roman religious tradition* among members of the political elite after the Second Punic War. Cf. PHILLIPS 1986, 2690-1. Literacy: HARRIS 1989, 149-267, esp. 248-67. 206 gion as presented in the literary sources: this relationship need by no means have been characterized by homology. 95 Moreover, the little tradition of elite discourse itself created a multitude of different, explanations and often incompatible, of Roman religion rather than one coherent and meaningful system of elite culture. Its precarious status as a literary cultural form implied that it was incapable of providing a cultural system that could have been constitutive for Roman society as a whole. Rather, its relation to the great tradition of societal religious behaviour must be de- scribed in terms of the coexistence autonomous areas of cultural of differentiated production. and Roman culture clearly defined itself was not a monolithic boundaries, a fluid system of different realms of cul- but entity with tural activity, which escapes a substantivist definition to as what its identity was. 96 And it is far from self-evident that the Republican elites, with their contempt for the other strata of Roman society, would have had an interest in consistently imposing their little tradition on society at large. 97 In what I have described as a which communicative abounded with different interpretations, »Roman reli- gious culture« was not one stable entity, was 95 96 97 field constantly under-determined. In and its meaning that field, literary Cf. LUHMANN 1980-89, 3,275,-with further literature, on that distinction. Other genres of literary production such as drama did become part of popular culture by transcending the limits of literacy; cf. GRUEN 1990, 79-157; Id. 1992, 183-222; RAWSON 1991, 468-87. BUHL 1987, 66-87 and SHENNAN 1994 provide useful theoretical discussions of the complex issue of »cultural identity*. Cf. PHILLIPS I99lb, 263. 207 assessments of a Roman religious identity could not become determinative of religion as a whole. 3.2.1 Omnis populi Romani religio As it is the little tradition that preserves the linguistic expressions of the elite's identification with its religious culture, the following discussion with that realm. When outlining the cult at Rome, Aurelius is primarily concerned organization of civic Cotta (the persona of Cicero's De natura deorum) divides public religion into the provinces of pontiffs, augurs and quindecimviri (adding the haruspices to the latter) whose respective responsibility lies with sacra, auspicia and advice on procuration of prodigies. This tri- partite scheme is topical: according to Rome's literary tradition, the proper functioning of these three areas guaran- teed the flourishing of the city-state since time immemo- rial. They constituted omnis populi Romani religio: sed cum de religione agitur, Ti. Coruncanium P. Scipionem P. Scaevolam pontifices maximos .. . sequor, habeoque C. Laelium augur em eundemque sapientem quern potius audiam dicentem de religione in ilia oratione nobili quam quemquam principem Stoicorum. Cumque omnis populi Romani religio in sacra et in auspicia divisa sit, tertium adiunctum sit si quid praedictionis causa ex portentis et monstris Sibyllae interpreter haruspicesve rnonuerunt, harum ego religionum nullam contemnendam putavi mihique ita persuasi, Romulum Romulum auspiciis Numam sacris constitutes fundamenta iecisse nostrae civitatis quae numquam profeeto sine summa placatione deorum immortalium tanta esse potuisset. 9Q 98 Cic. ND 3,5. The tripartite scheme of pontiffs, augurs and XWiri and their respective responsibilities is, for instance, also employed in Cic. Har. Resp. 18; Leg. 2,20, 2,30; Varro RD frg. 4 Cardauns. 208 But how much influence did this little on the tradition exert great tradition of religious behaviour at Rome? The Ciceronian definition of omnis populi Romani religio follows immediately after Gotta f s reference to the authority of eminent pontiffs and one augur as regards religio - a three strategy which allows Cotta to defend his cism in philosophy." The Academic scepti- three pontiffs listed by Cotta were all influential in the development of Roman ius civile. Ti. Coruncanius (cos 280) was the first to lecture in public on civil law and to issue memorabilia and responsa sacred and civil legal both matters under his own name rather than on behalf of the pontifical college. and on Like Coruncanius the second pontiff of Cotta's list, P. Cornelius Scipio Nasica Corculum (cos 162 and 153, pont. max. from 150), Publius Mucius Scaevola (cos 133, pont. legal max. from 130) gave responsa to the public. He was also the author of ten books De iure civili. 100 In the passage cited, however, Cotta refers to their authority in pontifical law, not civile; elsewhere Cicero stressed these two domains of legal science and pontiffs ought to ius the incompatibility of suggested that the to be followed in matters of sacred but not necessarily in those of civil law. 101 It is pontifical law which links these three pontiffs to C. Laelius. Cotta explicitly alludes to Laelius' collegiis of speech De 145, which defeated C. Crassus 1 proposal that 99 See above, 2.8.1. 100 Cf. WIEACKER 1988, 531-51; 0. BEHRENDS, ZSS 107 (1990), 587-99. The historical Aurelius Cotta (cos 75), too, had legal knowledge: Cic. De orat. 1,25, 3,31, Brut. 183, 201-4; Asconius 14 C. 101 Leg. 2,46-53. 209 all members of the priestly colleges (which I take to be the pontiffs, augurs and decemviri) should the be elected through comitia. 102 In that speech, Laelius appears to have de- fended the status quo by stressing the traditionalism of civic religion in Rome, and not only dealt with the election of priestly colleges but also discussed proper worship of the gods as laid down by pontifical law, mos maiorum and the regulations of Numa. 103 As noted before, the definition of the phrase omnis puli Romani religio po- employs a tripartite structure, which divides civic religion into the responsibilities of pon- tiffs, augurs and XVviri. This structural principle of classifying civic religion at once points to a literary tradi- tion of systematization as a possible source for phrase both that and its underlying tripartite structure of division. Given Cotta's reference in De natura deorum to pontifical law, it is very plausible that it was legal prose which provided such systematization. 104 Furthermore, Cotta's referen102 Although no ancient source lists the colleges associated with C. Crassus' proposal or the Lex Domitia of 104, references to quattuor amplissima collegia do not predate the Augustan period: e.g. Augustus RG 9,1; cf. ibid. 7,3. By way of contrast, Republican writers like Cicero routinely employ a tripartite structure listing the pontiffs, augurs and XVviri. I am therefore sceptical as to whether one should include the Vllviri epulones. Contra MOMMSEN 1887, 2,29 3 ; LATTE 1960, 395-6; SCHEID 1985, 68; BEARD 1990, 44 69 . 103 Cic. ND 3,43: ... meliora me didicisse de colendis diis immortalibus iure pontificio et more maiorum capedunculis his quas Numa nobis reliquit, de quibus in ilia au­ reola oratiuncula dicit Laelius; RAWSON 1991, 82-5. The speech was still read in Cicero's time: Rep. 6,2,2. 104 BEARD 1990, 44-5 and RAWSON 1991, 339-46 discuss the issue of classification (the latter with regard to Latin legal prose); both treat that issue as a phenomenon new to the first century and therefore fail to link it to the much earlier systematization of Roman law from the third century. Cotta's deference to the authority of 210 ce to the authority of the three pontifices maximi suggests that he uses the phrase omnis populi Romani was defined and religio as it employed in pontifical law. Moreover, the extent to which Laelius's speech, which used the tradition- alism of the organization of civic religion, is used in Gotta 1 s argumentation in the third book of De natura deorum lends some probability to the thesis that Cicero found omnis populi Romani both religio and its underlying tripartite structure in that speech. This is not to deny that such systematization could have reflected nor influenced the great tradition of reli- gious behaviour at Rome. However, as a product of tematic not classification the sys- of the religious data through legal texts, and through sacred law in particular, omnis populi Romani religio appears to us as part of the little tradition of elite discourse about the interrelation of the political organization of Rome and its upper class religious functionaries. This position, which by implication postulates a centralized system of religious life at Rome, can be acceptable only to those scholars who subscribe to the homology of ligion and politics in Republican Rome. Above all, it is noteworthy how rare this phrase is. In the plicit instance in re- only other ex- Republican or Early Imperial Latin li- terature that I have been able to find, contra populi Romani religionem et fidem, the phrase is significantly altered, Roman priests as regards religious matters, rather than to the Stoic philosophers, in Cic. ND 3,5 conceals the fact that the systematization of sacred law was indebted to Hellenistic, and particularly Stoic, logic. 211 and its meaning has become different. 105 There is therefore little justification in treating omnis populi Roman! religio as if it provided a religion. disinterested >description< Roman The religious system at Rome was not a monolithic entity, and omnis populi Romani religio, while the of interests of the representing little tradition, may fall short of adequately describing anything but those interests. 3.2.2 >Romanus ritus< and related concepts This is not to say that a conscious feeling of cultural distinctiveness did not exist in the great tradition, or that the great tradition's adaptation to cultural change in Roman society in the second and first centuries was unaccompanied by a feeling of irritation about foreignness. to The this feeling is a recurring theme of Roman comedy in the second century: the derision of alienus mos is at old least as as the comic tradition at Rome. This comes close to the little tradition's category of cultus way allusion Romanus, >the Roman of doing things<, which provided a normative account of Roman, as distinguished from foreign, cultural behaviour. 106 However, the question here is whether the tion's assertion of little trad- a Roman cultural identity through the 105 Cic. Balb. 10, a tendentious >comparison< of traditional Roman (i.e. pagan) religious custom with the religious behaviour of the Jews. At Livy 42,47,7, a corrupt passage, there is no reason to read religionis ... Romanae with J. Vahlen. On religio, see further below, 3.3. 106 Comedy: E.g. Plaut. Per. 212; Ter. Andr. 152. GRUEN 1990, 124-57 places such passages in the context of the emergence of a Roman national literature in the second century. Cultus Romanus as >Roman custom<: Pliny NH 28,18,6; Tac. Ann. 6,32,2; Suet. Div. lul. 24,2; Gell. 10,23,1 (all of imperial date). 212 category of cultus Romanus could consistently determine re- ligious activity. In other words, whereas religious activity may have been accompanied by the assertion of a Roman cul- tural identity in the little tradition, that mere assertion is not sufficient proof that religious activity was informed by such a criterion. As I suggested above, the search for a link between religious activity and a Roman religious tity iden- is misleading, since the Roman religious culture which the little tradition tried to identify was not a monolithic entity with pre-defined boundaries. As a consequence, native communication, both in the little and the in great tradi- tion, can provide only a relative approximation to the Roman religious system. Scholars often rely on the normative as- sertions of the little tradition in particular when attempting to determine what the Romans' religious identity was, but such reliance is unwarranted. The problematic status of several of these assertions will be the subject of the following pages. Consider patrius ritus. Ritus is akin to mos or lex and denotes what is cus- tomary in religious behaviour. 107 Under the rius pat­ Republic, ritus is used in a normative rather than in a descrip- tive sense and seems to be limited to literature. In De gibus, Cicero construed a religion which adhered to ances- tral religious custom, the patrii ritus. Like Varro, he lieved le- be- that the best customs were those of the most distant past. 108 Livy saw one reason for the decline 107 Cf. Livy 24,3,12 ritus mores legesque; ritusque; DURAND & SCHEID 1994. of the 29,1,24 res mores 108 Leg. 2,27: quoniam antiquitas proxime accedit ad deos. Patrii ritus: ibid. 2,19, 22-3, 27, 40. On Varro, see above, 2.4. 213 publica in the neglect of these traditional religious val- ues. The contamination of ancestral religious custom by foreign rites is one of his hidden leitmotifs: of the acceptance alien! Titus, mores legesque was a phenomenon of the so- cial and political insecurity of the Second Punic War (24,3,12), just as the Bacchanalian affair of 186 led to the abandonment of patrii mores ritusque (39,16,10). 1 ° 9 We can also trace this moralistic discourses that as in the politicized were inextricably linked to the parameters of elite communication. This early stereotype stereotype can be found as the second century, when Cato the Elder idealized the mos maiorum of the past and criticized the decline of moral values among fellow-members of the political elite; 110 or when Plautine comedy, through its mock-heroic inversion of this stereotype, parodied the idolization of the veteres as well as its underlying discourse about moral decline. 111 The Roman elite's emphasis on mos maiorum and patrius ritus forms a sharp contrast to the limited importance which Athenian public oratory attached to the TionrpLOS VOUOQ. To be sure, appeals to the authority of one's ancestors were made in the Athenian lawcourts, and the TtonrpLOi vou-oi could be 109 Cf. Livy 1,20,6: ne quid divini iuris neglegendo patrios ritus peregrinosque adsciscendo turbaretur; 1,31,3, 35,5; 5,52,9; 25,1,7; 29,1,24. 110 E.g. frg. 235 Malcovati 4 ap. Cic. Off. 3,104: maiores nostri esse voluerunt; cf. frgs 18, 58, 144 Malcovati 4 . For the Late Republican elite discoursing about moral decline, see in general LIND 1979, esp. 48-56; HAMPL 1980. 111 E.g. Plaut. Trin. 1028-9 (put in the mouth of a slave); W. S. ANDERSON 1979; SEGAL 1987, 70-98; GRUEN 1990, 141-4. Cf. GIZEWSKI 1989 on the juxtaposition of the conflicting moral value systems of Cato the Elder on the one hand and comedy on the other in second century Roman society. 214 invoked when legitimizing policies in the civic assemblies. Yet, in Classical Athens socio-political authority was tinely reinforced by reference to the Athenian &ffojLOC rather than by calling upon of past authority of the the normativity achievements of a collective or of individuals. 112 This principle precluded the of rou- ancestors, prominent instrumentalization which we find in Rome. There, the glorifica- tion of the mos maiorum was but the corollary of a system of elite competition in which the public display of their an- cestors brought social and political power to members of the aristocracy. 113 Since the individual aristocrat's use of his ancestors made sense only if the authority of ancestral tradition as such remained accepted, elite discourse was in constant need of re-creating the kind of we find normativity which attributed to ancestral moral and religious custom in the little tradition. However, while that need may ex- plain the elite's interest in emphasizing the normativity of ancestral religious tradition, it does not prove that the little tradition's assertion of the authority of the patrius ritus could determine the religious outlook of the great tradition. If that had been the case, the little tradition's complaints about moral and religious decline would have been superfluous. As regards Romanus and Graecus ritus, the distinction between great and little tradition for once does not seem to matter. For it is performed Romano ritu, commonplace that Roman sacrifices were that is with 112 Cf. JOST 1936; OBER 1989, 181-2. 113 ROLOFF 1938; FLOWER 1996, 60-90. one's head veiled, 215 the Graecus ritus entailed sacrificing capita aper- whereas to. 114 Sacrificial reliefs in Rome or Italy and, in the perial period, im- throughout the Empire show Republican magi- strates or the emperor undertaking sacrifices capite velato. These scenes arguably display a paradigmatic haviour religious be- as regarded Roman sacrifice in the civic domain. 115 Yet, this is not to say that one could classify Roman reli- simply according to the category of sacrificing Romano gion ritu, since there were a number of >Roman< deities, such Honos or Saturnus, to as whom sacrifice was explicitly made with one's head unveiled. 116 On the other hand, sacrifices like the one described the Elder Cato by or the sacrificial sequence in the acts of the Arval Brothers followed a distinct grammar of prescribed actions and linguistic utterances which allowed Dionysius of Halicarnassus, writing in the Augustan period, to a detailed undertake comparison of the differences between Greek and Roman sacrificial practices. 117 However, a descriptive parison is not the same as a normative account of Roman re- ligion which employs Romanus ritus in an explicitly rical com- catego- sense. To Varro, the meaning of Romanus ritus was not restricted to particular sacrificial or ritual practices. He 114 Romanus ritus: Varro LL 5,130; FREIER 1963. Cf. Cato Orig. frg. 18 Chassignet: incincti ritu Sabino, id est togae parte caput velati, preferred to the mss. variant ritu Gabino by A. RUMPF, Kleiner Pauly 1 (1975), 1190-1. 115 By way of contrast, military commanders sacrificed capi­ te aperto. I owe this information to Dr Valerie Huet who is about to publish a collection of sacrificial reliefs in Roman Italy. For a preliminary visual illustration, see TURCAN 1988, 2, nos 66 and 68. 116 Cf. LATTE I960, 214 4 . 117 Dion. Hal. Ant. Rom. 7,72,15-8. Cf. Cato Agric. 134, analysed by CANCIK 1994, 389-94; PHILLIPS 1997. See further above, 2.7.2. 216 wrote normative theology in the little tradition when ing the link- well-being of Rome to the performance of all reli- gious ritual Romano ritu, as if an undiluted system of Roman religion true could be reconstructed from antiquarian re- search. 11 8 The phrase sacra publics populi Romani is epigraphically attested in Late Republican and Augustan Rome. ence to the With refer- duties of Vestals and Rex sacrorum, it is em- ployed to circumscribe religion's civic organization in Rome in the purely administrative context of Late Republican city administration. 119 And two colleges of the accompaniment for religious publica in the course of their These Roman cult musicians rituals public to sacra functionaries defined themselves through city's religious as organized by the civic authorities rather than through specifically Roman deities technical refer self-representation. the involvement in the maintenance of the infrastructure providing meaning underlies or rites. 120 This the legal distinction between the two distinct domains of the sacra publica of the Roman People as a whole (as represented by the city's authorities) 118 RD frgs 49-50 Cardauns: nostro ritu sunt facienda civi libentius quam Graeco castu. (50) Et religiones et castus id possunt, ut ex periculo eripiant nostro. For nostro ritu, cf. Varro Log. Catus frg. 35 Mueller ap. Nonius 494 L. Castus is often, albeit not exclusively, used in conjunction with Greek cults: FUGIER 1963, 24ff. 119 Tabula Heracleensis (RS no. 24), line 62: sacrorum publicorum p(opuli) R(omani) caussa-, cf. Livy 34,1,3. 120 OIL 6,2193 dis manibus collegio symphoniacorum qui s(acris) p(ublicis) p (raesto) s(unt); ILLRP 185: ... c[ollegiei] teib(icinum) Rom(anorum) qui s(acris) p(ublicis) p (raesto) s(unt). See further below, 4.1.2. 217 the sacra privata of families or individuals. 121 By way and of contrast, the phrase terminology, Romana sacra, introducing ethnic is used exclusively in the little tradition of elite discourse to provide a normative account of the gious customs of the populus Romanus as a whole. 122 This evaluative use contrasts with that of appear reli- Graeca sacra, which to have signified one specific cult at Rome, that of Ceres. 123 Later writers use sacra Romana in the same alized way when referring to a collective body of Roman rites. Normally, a specification of what these rituals is not gener- is >Roman< about offered. It is no doubt due to this imprecision that sacra Romana, unlike the concrete and evaluative non- phrase sacra populi Romani, is only infrequently found in literature before the fourth century CE. 124 The evaluative meaning of Romana sacra forms in Livy's a stark contrast to the merely descriptive use of sa- cra publica or sacra populi Romani in the epigraphic or in work the record distinction between civic and private cults. As regards Romanus ritus, it is hard to assess to what extent its conscious creation as a normative category in the little tradition caused a revaluation of religious behaviour in the 121 E.g. Cic. Har. Resp. 14: sacra religionesque et privatas et publicas; Leg. 2,20, 2,22; Livy 1,20,6, 5,52,4; Asconius 21 C; Dion. Hal. Ant. Rom. 2,65,1; WISSOWA 1912, 380-408; DUMEZIL 1970, 553-75; CANCIK 1994, 377-81; below, 4.1.1, 4.2.1 and 4.2.2. 122 Cf. Livy 1,31,3 (in opposition to the sacra Albana which had been abandoned); 5,40,10; 5,50,3; 7,20,7 (all referring to the sacra housed in Caere, the sacrarium po­ puli Romani, during the Gallic invasion). For once, Livy is more specific when narrating the Sibylline Books' prescription of human sacrifice after the defeat at Cannae: minime Romano sacro (22,57,6). 123 Cic. Leg. 2,21; Paulus 86 L. 124 E.g. Varro ap. Cell. 3,2,8; Pliny NH 28,39,5. For fourth century instances, see below, 3.3. 218 tradition. A similar problem applies to Graecus ritus great and to peregrina sacra. Graecus ritus first appears in the Cato Elder. It is normally assumed that the fragment concer- ned belongs to the oration in which Cato, in his censorship, justified the decision to deprive L. Veturius of at his horse the recognitio (which amounted to the latter f s expulsion from the equestrian ordo). Apart from charging Veturius with moral debauchery, Cato apparently ligence denounced him behaviour. 125 In this religious irrespons- rhetorical context, Cato must have regarded Graecus ritus as an integral part of Roman culture, rather than emphasizing any foreignness. John Scheid has suggested that the concept of Graecus was a ritus secondary classification of the origins of some ele- ments of Roman religion. According to Scheid, it was ted inven- when Rome created her cultural consciousness at the be- ginning of the second century. »typically Roman Graecus ritus signified a way of honouring the gods«; it was an at- tempt to explain with reference to Rome as an open city its neg- in matters of family cult; he highlighted the sanc- tity of religious ritual as a foil for Veturius 1 ible for why religion was a synthesis of Roman, Hellenic and foreign traditions. As Scheid has again reminded us, the category of Graecus ritus was applied to some of those elements of religion that, in religious practice, were consultation hybrids: the Roman of the >Greek< libri Sibyllini, the Roman cel- 125 orat. frg. 77 Malcovati 4 : Graeco ritu fiebantur Saturna­ lia. Cf. frg. 72: quod tu, quod in te fuit, sacra stata sollempnia capite sancta deseruisti. For further discussion of the latter fragment, see below, 4.1.4. For the Saturnalia, see LATTE 1960, 254-5. Cf. RAWSON 1991, 80-101 who discusses what she sees as a renewed elite interest in ancestral religion during the mid second century. 219 ebration of the >Greek< lectisternia and supplicationes, or of the Ludi saeculares, whose programme juxtaposed Roman and Greek religious elements. 126 The category of peregrina sacra displays a similar tern. In pat- the Augustan period, it was defined as comprising those cults quae aut evocatis dis in oppugnandis urbibus Romam sunt coacta, aut quae ob quasdam religiones per pacem sunt petita, ut ex Phrygia Matris magnae, ex Graecia Cereris, <ex> Epidauro Aesculapi, quae coluntur eorum more a quibus sunt accepta. The Mater magna had her Phrygian clergy brought from the East, the cult personnel of Aesculapius on the Isola Tiberina included Greek priests, and that of Ceres consisted of Greek priestesses from Velia and Naples. Rituals formed in Greek. 127 However, were made by the cult sacrifices ' PCDIKXLCDV voy,ouc. 128 But the The of Ceres was a recent addition to the deity's ancient Roman cult (documented by the existence of a Cereris). to urban praetor on behalf of the state dvd Ttdv ETOQ and X<XT<X TOTJQ >Greek< per- the Mater magna, the former Cybele, received a Roman cult name; and goddess were even the flamen first priestess was granted Roman citizenship, whereas later religious functionaries were the daughters of Roman citizens. 129 These instances of peregrina sacra do not document any particular foreignness, but de- 126 SCHEID 1998. I am grateful to the to see this article prior to its openness of the city of Rome, see 127 Mater magna: Dion. Hal. Ant. Rom. and Ceres: Cic. Balb. 55; Dion. Val. Max. 1,1,1; Pliny NH 35,154; LE BONNIEC 1958, 379-400. author for allowing me publication. On the above, 3.1.1. 2,19,4-5. Aesculapius Hal. Ant. Rom. 6,17,2; Festus 86 and 268 L; 128 Dion. Hal. Ant. Rom. 2,19,4-5, 52,19. See further above, 3.1.2. 129 ILLRP 61; JLS 3343; cf. Cic. Balb. 55. 220 monstrate that these cults were acknowledged hybrids characteristic of a synthetic religious culture. Although I am in general agreement with Scheid, I one ought to think date the cultural process through which such conscious classifications were first introduced to the third rather than to the second century, as he does; and unlike Scheid, I believe that the little tradition involved in that process was redefining the boundaries of religion at Rome rather than simply affirming its openness. 130 Above all, is it important to stress that when we encounter a phraseology which provides us with religion it categories for pigeonholing Roman is through the little tradition of elite writ- ing. Unfortunately, it is hard to assess to what extent such classifications in the little tradition determined the religious system at Rome. However, whereas to many may be a matter scholars of convenience to apply Graecus ritus or Romanus ritus when categorizing the entirety of Roman gion, 131 I would it reli- suggest that it is problematic to invest these categories with more significance than they originally had. They were artificial classifications of the religious data in the little tradition, which could be superimposed to provide exegesis and theological speculation. This tendency is most clearly expressed in the acts of the Augustan and the Severan Secular Games of 17 BCE and 204 CE respectively. In the acts of the Augustan Ludi, carefully composed under the auspices of the emperor and the XVviri , Roman sacrifi- 130 See above, 3.1.1. 131 E.g. KIRSOPP LAKE 1937; GAGE 1960, passim; LATTE 1960, 242-63; DUMEZIL 1970, 567-70; GRUEN 1992, 228-9; CANCIK 1994, 351-2. 221 cial terminology (imnolare) is juxtaposed with the prescription to conduct the sacrifice of the victim Achivo ritu. The Severan acts adopt this juxtaposition, but link Roman terminology (immolatio) to a classification which is linguist- ically over-determined (Graecus Achivus ritus). represent a self-conscious These acts theological re-creation of the tradition of the Secular Games. 132 For instance, the X(V)viri were said gious to perform reli- rites in accord with the Graecus Titus. 133 As regards their function in the civic religious system, however, description is misleading: technically speaking, that the X(V)viri were in charge of the Sibylline Books and therefore involved only in those cults and rituals that resulted the from college's interpretation of the Sibylline verses. 134 In the imperial period, the XVviri's supervision of the cult of Mater magna in Italy and Rome is attested. Since many other cults were transferred to Rome from Magna Graecia which were not supervised by that priestly college, it is unlikely that quindecimviral authority over cults resulted from the cate- gory of Graecus ritus. Rather, the college's link cult with the of Mater magna resulted from the fact that the goddess 132 PIGHI 1965, 113-4, 90-1 and 155-6, iv, lines 4ff., cited by SCHEID 1998. On the inventiveness of the >theology< of these acts, see CANCIK 1996, 100-3, 109-10. 133 varro LL 7,88: et nos dicimus XVviros Graeco ritu sacra, non Romano facere; Livy 1,7,3, 25,12,10-3; PIGHI 1965, 125. 134 Cf. Cic. Leg. 2,20 alterum [sc. genus publicorum sacerdotum] quod interpretetur fatidicorum et vatium ecfa­ ta incognita quorum senatus populusque asciverit; DIELS 1890, passim; RADKE 1959, esp. 217-8; Id. 1980, 119-20. 222 had been brought to Rome after consultation of libri the Sibyllini. 135 The religious rituals which the X(V)viri recommended after consultation of these Greek books often employed Roman rather than Greek ritual means of procuration. Moreover, from the third century >Greek< supplicationes were not only by the X(V)viri proposed after consulting the Sibylline Books but also by the pontiffs or the haruspices. Yet, ultimately the Roman Senate was in charge of ordering supplica- tions to take place. 136 As one would expect from a synthetic religion, the organized system of religious reality con- stantly blurred the categorical distinctions that the little tradition made. Therefore, Graecus ritus can as a native be understood exegetical category which explained religious behaviour in Republican Rome. As a distinct descriptive ca- tegory of Roman religion, it is unhelpful. 3.3 Excursus: >Romana religio< The function of the following excursus, which makes no claim to comprehensiveness, is to suggest that the peculiar tionship of the great and the little tradition in Late Re- publican and, for that matter, Imperial Rome can understood rela- when be better it is compared with the period of conflict between paganism and Christianity in the later third and 135 Cf. BENDLIN 1997, 52 for the imperial supervision of the cult of Mater magna by the XVviri. 136 Ex decreto pontificum: Livy 27,4,15; 37,4; 39,22,4. Ex haruspicum response: Livy 32,1,14; 41,13,3. Cf. LATTE 1960, 245-6; DUMEZIL 1970, 569. 223 fourth centuries CE. For it is toward the end of that period that direct and immediate influence of normative catego- a ries of Roman religion on the pagan great tradition observed. A discussion can be the phrase Romana religio will of help to demonstrate that point. That phrase has often been thought to represent a Latin equivalent to the modern notion of >Roman religion<; as we have seen, that assumption rests on spurious grounds. 137 Judging from the available evidence, the exact phrase Romana religio does not occur prior to very the end of the second century CE; and the qualification of religio by the addition of an adjective connoting ethnic affiliations is a new development. 138 This is one reason why it is methodically misleading to take religio Romana as a descriptive denomination for the Roman religious system. 139 It is first used by the Christian Tertullian in the allegedly legal phrase crimen laesae Romanae religionis, which is built on and imitates the legal charge of crimen laesae maiestatis. Tertullian's legal phraseology insinuates that Roman authorities the charged Christians on a systematic basis t for their neglect of Roman religious cult activities: 137 See above, 1.1. 138 Contrast the various uses of religio in e.g. Cic. Rep. 1,23: perturbari exercitum nostrum religione et metu; Livy 42,3,1-11: obstringere religione populum Romanum (>religious scruples<). Religio, as cultus deorum, is not limited to the Romans: e.g. Cic. ND 2, 8; Div. 1,93 and 96, whereas communis religio (Cic. Div. 2,28) is once employed when justifying cult practice as politically expedient. For religio sepulcrorum and religio privata denoting private cult at the tomb, see Cic. Leg. 2,55-8; below, 4.2.1. 139 E.g. KOEP 1962, 46: »In jener Zeit, da das Christentum dem romischen Reich und damit der offiziellen religio Romana begeget, sind religio ... und ritirs engstens miteinander verbunden, wenn nicht nahezu identisch: die religio Romana auSert sich in ihren Riten, in ihren Zeremonien«; WLOSOK 1970. For further critique, see FEIL 1986, 77-82. 224 Omnis 1sta confessio illorum, qua se deos negant esse quaque non alium deum respondent praeter unum, cui nos mancipamur, satis idonea est ad depellendum crimen laesae maxime Romanae religionis. Si enim non sunt dei pro certo f nee religio pro certo est; si religio non est, quia nee dei, pro certo, nee nos pro certo rei sumus laesae religionis. 140 If this claim were true, a strong case could be made for a systematic Roman religious policy in the Roman Empire, which inflicted capital punishment on Christians for disobedience to a clearly defined system of Roman religious activity. 141 The problem with such a claim is that there legal no body of regulations to that effect, just as a compre- hensive persecution of Christians did not mid existed occur until the third century. Prior to that period, the suppression of Christians was motivated by socio-political which the level of local communities. occurred mainly at considerations Roman officials, when persecuting Christians, would merely respond to local popular pressure, which was channelled into short-lived pogroms and thus provided an outlet for communal pagan unrest. 142 Therefore, no supposed Roman religious po- licy can be reconstructed on the basis of mark. His Tertullian's pseudo-legalistic phraseology must rather be ex- plained by the striving for rhetorical effect. 143 What scholars re- since some Mommsen have unwittingly accepted as an ex- pression of legal authenticity, is a contextual phrase in- 140 Apol. 24,1. 141 Cf. MOMMSEN 1907, 394-416, at 395: »... eine strengere [Auffassung der maiestas populi Romani ] ..., welche auch die Verletzung der dii populi Romani auffasste als Beleidigung der herrschenden Nation und die Anwendung der Capitalstrafe also auch hier erforderte.« 142 Cf. STE CROIX 1974; MOLTHAGEN 1991, esp. 42-3, 73-5; NIPPEL 1995, 107-8. Cf. PHILLIPS 1991b, 268-9, for the absence of exact legislation that could have dealt with any kind of »unsanctioned religious activity*. 143 Cf. RIVES 1995, 243-4. 225 vented in order to juxtapose the Romana religio and the vera religio veri del of the Christians, and to turn back on pa- gans the charge of atheism by construing crimen the dichotomy of laesae Romanae religionis and the verum crimen verae irreligiositatis: in Rome, any god could be worshipped except the one true god. 144 Further on in the same chapter, Tertullian defines Roma­ na religio as the worship of the deities of the city of Rome (Apol. 24,8). Whereas other Empire peoples throughout the Roman were permitted to worship their own gods, that right was denied to Christians, who were instead forced to worship the gods of the Romans. 145 Above, Tertullian portrayed as the exponent of Rome paganism capable of incorporating any national deity except the Christian god. Now, Tertullian expanded that argument through the metaphor of ethnic affiliation: the various pagan nationes throughout the Roman Em- pire followed their ethnic beliefs and worshipped their tive gods. Why then should the Christians not be permitted to follow their belief and worship their god? This comes close in ethnic the Christian apologists in general: ethnic terms for pagans such as gentes or prior argument to defining Christianity in terms of ethnicity (Apol. 24,7-9). Moreover, it accords with the use of phraseology na- nationes were employed to the non-ethnic term paganus, and Christian authors used the denomination Christiana gens for themselves. 146 144 Apol. 24,1-2; cf. ibid 24,10. 145 Apol. 24,9: sed nos soli arcemur a religionis proprietate. laedimus Romanos nee Romani habemur, quia non Romanorum deum colimus. 146 E.g. Arnob. Adv. Nat. 1,1; FREDOUILLE 1986. 226 Tertullian's phrase Romana religio must be understood as a conscious invention in the context of the definition of religious affiliations in terms of ethnic affiliations. Once Christianity had been defined in ethnic terms in opposition to pagan national religions, an alleged Roman ethnicity came the prime target of the apologists. For in their tele- ological scheme, it was Rome, the worldly empires - fifth and replaced by the last of the and its national religion - which while resisting the Christians' missionary zeal be be- would ultimately civitas del. 147 Yet, ethnicity, as we have seen, is an unsuitable category for defining pagan re- ligious affiliations. The notion of Romana religio therefore fails to provide a common ground that would have been shared by Roman pagans and Christians in the second century. The phrase Romana religio reappears in of Cyprian. 148 Summoned before the martyr-acts the proconsul Paternus at recognos- Carthage on 30 August 257, Cyprian was ordered to cere Romanas colunt, debere caerimonias, Romanas for qui non Romanam religionem caerimonias recognoscere. When Cyprian refused to comply, he was sent into exile. At a second summoning in 258, Cyprian was again compelled to recog­ noscere Romanas caerimonias; at his renewed refusal, he sentenced to death. 149 The evaluation of the phrase Romanas caerimonias recognoscere scholars. was has proved difficult to modern It was Heberlein's thorough philological analysis which established the phrase's exact meaning in the 147 Tert. Nat. 2,17,18-19; cf. Min. Pel. Oct. Div. Inst. 7,15-7. 148 Acta Cypr. 1,1. 149 For the chronology of the events veniently SCHWARTE 1989. of context 25,12; 257/58, see Lact. con- 227 of the Latin literature of the third century. Recognoscere Romanas caerimonias means »to perform Roman ties^ Therefore, not expected to observances activi- according to the martyr-acts Cyprian was observe the entire system of religious as it was practiced by Romans (Romana religio), but merely forced to perform a ritual proconsul. ritual Whereas activity before the religio entailed a more systematic con- cept of cult practice, caerimonia denoted the behavioural element of ritual performance. 150 In the past, scholars have doubted whether acts the martyr- of Cyprian represent the authentic wording of the pro- consular proceedings. Accepting the received view that Roman religion ought to be defined as a ritualistic religio, ritus or caerimoniae being between Roman religion with homonyms, they have taken issue with the idea that an official tiation affair, Roman differen- (as which they translated religio) and the mere performance of cult (caerimoniae or ritus) would have been possible. The same scholars have succumbed to the view that the Roman state actively promoted the acknowledgement of its religion by the inhabitants of the Empire and therefore could not issue a phrasing which in effect undermined such an acknowledgement. 151 However, the 150 HEBERLEIN 1988. Incorrect translations still persist, invalidating modern interpretations of the Christian persecutions under Valerian. Cf. most recently RIVES 1995, 252: »to acknowledge Roman rites«. On the distinction between religio and caerimoniae, cf. ROLOFF 1953. 151 E.g. KOEP 1962, 51-2; M. SORDI, ANRN 2,23,1 (1979), 369-70. Agreeing that a differentiation between religio and caerimoniae should not be attributed to the martyracts, SCHWARTE 1989, 123-7 tries to defend the text against the charge of inauthenticity by following a textual variant and reading qui Romanam religionem colunt. 228 performance of caerimoniae, rather than religio, is exactly what Roman the observance officials expected from Christians on a number of similar occasions. 152 the of As far as of religio and caerimoniae is concerned, the dichotomy text of the martyr-acts is unproblematic. This is not to say that the martyr-acts should as an official be read document of undisputable authenticity. For despite their stylistic matter-of-factness and seeming tual accuracy, and despite the fact that the name acta pro- consularia is sometimes attached to them, they very represent a Christian clearly rather than an official Roman docu- ment. Their style is informed by the and fac- internal organization titulature of the church and coloured by a partisan de- scription of the martyr Cyprian. Their tone suggests that they refer to the official questioning of Cyprian before the proconsul, but rephrase the material to hand for homiletic purposes and for a Christian audience, probably of Carthage after the and caerimonias jRomanas church events of 257/58. 153 How does this observation influence the evaluation of the religio the phrases recognoscere? Romana Heberlein pointed out that the latter phrase, in the sense of »to perform caerimoniae«, is the earliest linguistic paradigm functiones extant example recognoscere. 154 earlier evidence may simply be lost, it is thus 152 E.g. Acta Scillit. 3; cf. WLOSOK 1970, 45-6. of the Although certainly 153 Cf. Acta Cypr. 1,2: Cyprianirs episcopus dixit; 2,(l),i: Cyprianus sanctus martyr electus a deo; A. A. R. BASTIAENSEN (ed.), Atti e passioni del martiri 2 (1990), esp. xxviii-xxxvi, 202-4, noting that other martyr-acts are also very likely to have been Christian redactions of the official proceedings, reshaping the text for the purposes of a Christian audience. 154 HEBERLEIN 1988, 97. 229 worth pointing out that the phrases Romana religio and Romanae caerimoniae are, to our knowledge, first used by Chris- tians. Whereas Cicero could employ omnis populi Romani reli­ gio in the little tradition, phrases such as Romana religio received prominent treatment from Christians like Tertullian or the authors of the martyr-acts. The Christian persecutions of 257/58 were response an immediate to the emperors' litterae which had been distribu- ted to provincial governors. 155 There can be no doubt that both Valerian's orations in the Roman Senate and the resulting litterae were directed specifically at the Christian clergy. 156 However, they were most probably phrased so as to demand the recognition of the pagan gods in general terms. A sacrificial offering to any god would do, although this in some local instances entail a sacrifice pro salute could imperatoris or might involve worship of the emperor. 157 At a parallel hearing genius of the before the prefect of Egypt, which resulted from the same libellae that led to the summoning of Cyprian in Carthage, Dionysios, the Alexandria, was told Christian god. 158 were the use- As we have seen above, worshipping the gods as such was not an unusual deities of to worship the gods who preserve the Empire and who were worthy of respect rather than less bishop requirement. As these not specified to the sacrificer, their choice 155 These letters originated from imperial orations to the Roman Senate. They were orations rather than rescripta, since the latter were not issued to the Roman Senate: NORR 1981. Contra MILLAR 1992, 277, 569-71. 156 SCHWARTE 1989, 109-19. 157 Cf. MILLAR 1973. 158 Euseb. Hist. eccl. 7,11,7-9; cf. Rufin. Hist. eccl. 7,11,6-7. Cf. Acta Scillit. 3; Orig. Exhort. Mart. 7,1, 40, 17,32. 230 must have been dictated with a view to regional preferences, the availability of specific gods in a local pantheon, or the pressure exerted by individual Roman officials. 159 This pagan background lends strong support to the thesis that the phrasing of the acta Cypriani must be It inauthentic. is therefore very likely that Romana religio and Romanae caerimoniae were introduced by the produced the martyr-acts of Christian redactor Cyprian. As has already been observed in the case of Tertullian's use of Romana the who religio, phraseology of the Acta Cypriani construes the category of Roman ethnicity with regard to religion. 160 This is a consciously Christian discourse, which defines its own religion in opposition to Rome and her religious system. It is in this process of Christian self-definition that writers developed Christian concepts for >their< and for >our< reli- gion - Romana religio or Romanae caerimoniae on the one hand and religio nostra or religio Christiana on the other - well before Constantine. Yet, I do not know of a case in which pagan a would have used a phrase such as, let us say, religio Isiaca with a similar meaning. 161 The contrast with a pagan environment whose self-con- sciousness was based on religious traditionalism exemplified by obedience to the mos maiorum, rather than on ethnic definitions, is worth emphasizing. For ethnic peculiarities ap- 159 Cf. SELINGER 1994, 77-140. See above, 2.7.2. 160 For further instances in the Acta Cypr., see 3,(1),4: diis Romanis, 3, (2),4; 3,(l),5: Romana mente. Cf. the Passio Crispinae of 304 CE, 1,4: sacra deorum Romanorum; 2,4: cole religionem Romanam; Lact. Div. Inst. 1,20: religiones Romanorum. 161 Religio nostra: e.g. Min. Pel. 29,2. Religio Christiana: e.g. Arnob. Adv. Nat. 1,3, 3,1. 231 pear to have been marginal, as long as innovations could incorporated into a traditional framework. Political, mili- tary and social prosperity still rested, so it could be gued, be ar- on scrupulous observance of the religion of one's an- cestors. 162 For instance, the Roman equites who made dedications to the combined dii patrii and dii Mauri the in second half of the third century stressed the reconcil- iation of their ancestral and of the the hospites oppression local gods following of local revolts, rather than dichotomizing between different ethnicities. 163 In the context of the synthetic religious systems of ancient paganism, ethnic sity diver- could be overcome by conceptual assimilation. However, although the Latin Christian apologists might wish to define their religion through the issue of ethnicity, when ing Christianity against the charge of disobedience to the traditional deities, they religio vera or defend- religio employed non-ethnic terms like Dei, as well as their opposites, religio(nes) deorum or falsa (e) and vana(e) religio (nes) , rather more frequently. 164 This is not to deny the possibility that the pre-Con- stantinian Christian apologists stood for a little tradition which was not only predominantly Christian literary elite, but whose representative literature was of the mainly written for a Christian elite audience. 165 However, I am not 162 Cf. most recently STROBEL 1993, 324-40. 163 OIL 8,8435, 21486; FENTRESS 1978. 164 Religio vera: e.g. Min. Fel. 38,7; Tert. Apol. 24,2, 35. Religio Dei: e.g. the Latin translation in Iren. Adv. Haer. 1,16,3. Cf. FEIL 1986, 50-82, esp. 80-1. 165 Cf. MACMULLEN 1984, 20-1, 131 14 . CAMERON 1991, 22-3 allows for a wider dissemination of Christian literature at least among upper-class pagans. 232 concerned with the interrelation of two little traditions, Christian and pagan, in the first and second Nor do I wish centuries CE. to determine the impact of Christianity on pagan perceptions of their religion in the great tradition during that period. Rather, I would tentatively suggest that a trend can be established concerning the pagan great tradition's response to Christianity in the third and fourth centuries. In a third century inscription, the dating be, to Christianity. 166 Other known phrase restitutor sacrorum refer to the the reference may as the first editor suggested, to Decius' victory over his predecessor Philip, who was accused tached 251, Decius is celebrated as restitutor sacrorum et emperor libertatis by the res publica Cosanorum. The well to defender of pagan religious of being at- instances of the emperor Julian as tradition in the fourth century. 167 Whereas in these instances it was unnecessary to specify which sacra were meant - they were no doubt the traditional pagan ones -, the phrasing of an inscription which was set up between 361 and 363 in the province of Numidia is noteworthy for its explicitness. It praised the emperor Ju- lian for the restoration of freedom and of practices: restitutori libertatis et Romanarum num. 168 I say >noteworthy< because this is a which illustrates how Roman rare religious religioinstance the phrase Romanae religiones, pre- 166 AE 1973,235. For Philipp's alleged philo-christianism, cf. Euseb. Hist. eccl. 6,29,1. 167 Rec. de Constantine 27 (1892) 255 = ILS 8946, alluding to Julian's oppression of Christian religion. Cf. AE 1907,191: recreator sacrorum et exstinctor superstitionis. 168 MEFRA 14 (1894), 77-8 no. 130. Both OIL 8,4326 and ILS 752 print Romanae religionis, but on epigraphic grounds the plural is clearly preferable. 233 viously used by Christian writers, was adapted by a pagan great tradition in the fourth century. I would suggest that this adaptation of Christian terminology must be seen as a pagan adjustment to munication about societal religion, which by the latter half of the fourth century had not only been shaped by but was also creasingly dominated by in religious opposition to Christianity employed an ethnic terminology which the Christian writers had process in- Christians. 169 In this climate of marked Christianization, pagans who defined their behaviour com- provided. This may be described as the creation of a non-Christian religious consciousness which did not simply imply the reassertion of traditional pagan values; rather, Romana had come religio very close to becoming a metaphor for pagan reli- gious practices as such. Similarly, the evaluative re-use of the phrase sacra Romana in fourth century must be seen pagan literature as the pagan recollection of a lost heritage rather than as a faithful reproduction of the precise sacred terminology of an earlier age. 170 Arguably, both the time when the great and the little pagan traditions displayed an distinct »Romanness« with regard to religious behaviour seems to have been the fourth century CE. Comparing the Late Republic and the fourth century CE entails problems of me- thodology. Yet, I believe that the juxtaposition of these 169 The influence of upper-class Christians in society in the early fourth century is often under-estimated. BARNES 1995 plausibly suggests that the Christian aristocracy occupied a majority of public positions in the city of Rome as early as the 320s. 170 For sacra Romana, see SHA Hadr. 22,10; SHA Did. lirl. 7,10; Nonius 834 L; Serv. Aen. 2,116, 8,698. Cf. above, 3.2.2. 234 two periods is illustrative. The analysis in terms of longue duree shows how socio-political changes began to affect the religious self-definition of the pagan great tradition in the fourth century. The peculiar parameters of these changes by implication re-inforce the thesis that similar forms of self-definition did not apply in the great tradition in Late Republican or Early Imperial Rome. 3.4 Roman religion and Late Republican Italy Returning to the second and first centuries BCE, during that period the scope of what could be meant >Roman religion< was by the notion increasingly widening. First of all, >Roman religion< would denote the local religious system the city of Rome. of of But should the phrase also be employed when comprising the religious behaviour of Roman citizens in Roman coloniae or viritim in Italy and abroad following as well as, the Social War, the religions of the entire Ital- ian peninsula and, since 49, of Transpadane Gaul? As pointed out above, with such geographical and cultural heterogeneity Roman citizenship proves an insufficient category termining cultural and religious when de- behaviour. What further aggravates the problem: even with reference to the city of Rome the identity of Roman religion has been shown to defy a categorical classification. 171 After the foregoing discus- sion it would be hard to maintain that Republican Roman lonization or the municipalization co- of Italy in the early first century simply represented the imposition of a defined 171 see above, 2.1, 3.1 and 3.2. 235 system of Roman religion on the provincial or Italian landscape. Rather, the Roman, Italian interaction and of religious structures provincial -, which were by themselves highly variable, will be the subject of the following pages. 3.4.7 Intervention and laissez-faire In one form or the other, Roman intervention in the life Italian communities is copiously of attested. By the latter half of the second century, public works such as road-building or water-supplies, documented from the late fourth century, had transformed the geography of many parts of peninsular Italy. These public works were under the responsibility of Roman magistrates, normally the censors, and the literary sources (usually historiography) already in the third make clear that century these magistrates undertook road-building in agro peregrine. However, the legal basis on which road-building was conducted status are frustratingly in unclear; areas without Roman and even the extent to which the creation of an Italian road-system might have followed a grand strategy of imposing Roman control on Italy is hard to assess. Moreover, it is noteworthy that such scale public works are most frequently attested in Latium and Campania but less so in other parts period under consideration, of Italy. In the from the early second century onwards, the main Roman roads served the needs of of large- the city Rome: supported by road-building on a much grander scale than previously, they supplied the links between the polis and metro- its more immediate hinterland. In the second and particularly in the first century, their most important eco- 236 of nomic function was to facilitate the transport Rome to ties commodi- order to satisfy the demands of the city in population. 1 7 2 magistra- By way of contrast, the intervention of Roman tes in public building outside Rome appears to have been restricted to those communities with Roman legal status. For Flaccus in 170, the year of his censorship, was not limited communities, to Rome but extended to other Roman the construction of Potentia Sinuessa. and vius' use of public building censorial ities, but met with when money involving for Pisaurum and Fundi and Capitolia for fortifications as road-building, water-supply as well Yet, Fulvius Q. by instance, the building programme supervised It is worth pointing out that Fulwhich activity criticism for earmarked been had was welcomed by these communfrom his fellow-censor. 173 Fulvius interfered in an allied community by reat Lacinia moving the marble tiles from the temple of luno Croton to embellish his temple of Fortuna Equestris at Rome, the senate forced him to return them and commanded an exhad piatory sacrifice to the goddess. The censor encroached on an allied city, but also violated the rights of luno. By virtue of his command, Fulvius populus only not Romanus; the represented therefore, expiating his sacrilegious ac- tion became the responsibility of the Roman People as a whole. 174 172 Road-building: WISEMAN 1987, 144 (chronology), 152-3 (second century building activities). Distribution in Italy: LOMAS 1997, 26-9. Economic function: MORLEY 1996, esp. 173, 177-80. 173 Livy 41,27,11-3: cum magna gratia colonorum. 174 Livy 42,3,1-11: [Fulvium] ... obstringere religione pulum Romanum; cf. Val. Max. 1,1,20. po- 237 This is not to deny that Rome assumed the right croach upon the of 186 extended the legal the Roman Senate's ruling concerning the worship- pers of Bacchus to the Italian allies. The extent diate en- juridical affairs of local communities in Italy. The SC de Bacchanalibus of force to and direct of imme- Roman intervention in the affairs of the Italian communities has often been over-stated: for whereas of the Senate's provisions in Rome and in fora et violators conciliabula were instantly prosecuted, the distribution of the SC throughout Italy did not entail similar Roman juridical action in the allied communities. Italians who wished to perform the rites were advised to seek the permission of the Roman Senate through the urban years, it seems, the Bacchanalian instrumentalized to interfere praetor. affair with In the ensuing was occasionally some frequency in the juridical proceedings of the allies. For instance, quaestiones were conducted by Roman magistrates in Southern Italy in 184 and 181. 175 Yet, by the standards of the second encroachment of century this sort was exceptional rather than dis- playing a consistent or systematic Roman strategy of ference with Italian affairs. Furthermore, the extent to which the Roman Senate's provision of 186 directly the juridical affected and religious life of the allied communities is hard to quantify. Similarly, the extent regulations inter- concerning the to which Roman expulsion of undesirables from Rome and Italy in the second and first centuries were re- 175 ILLRP 511, 2-9: folderatei, Latini, socii; Livy 39,14,7: non Romae modo sed per omina fora et conciliabula conquiri . .. edici praeterea in urbe Roma et per totam Italian edicta mitti. Cf. Livy 39,41,6-7; 40,19,9-10; GRUEN 1990, 36-45; below, 4.2.3. 238 garded as technically binding in the Italian communities is difficult to assess. Nor is quantification possible gards re- the Roman Senate's religious regulations which occur- red with increased frequency in the triumviral and periods, such Augustan as the stipulation to erect statues of Divus lulius in Rome and Italy. 176 In have as these cases, action must depended largely on the Italian elites' initiative. As will be shown below, it is in the course of the first century that these elites would become increasingly more willing to comply with the centre's demands. A change in Roman attitudes toward Italy is perceivable in the period following the enfranchisement of the peninsula in the early first century. By the imperial period, ager Italicus or solim Italicum had acquired complementarity in status with ager Romanus. Both were distinguished from solum provinciale in pontifical law with regard to their superior religious status (Gaius Inst. 2,3-9). I Roman believe this classification of Italian territory originated in the final years of the Republic. Between 82 and the that first 78, Sulla was Roman to extend the pomerium at Rome. Only those who increased the size of Roman territory were entitled to do so; and Sulla met this requirement by moving the boundary between Italy and Cisalpine Gaul. Referring to Sulla's ac- tion, the younger Seneca reports a dispute as to whether the addition of Italian or of provincial soil entitled to extending the pomerium a Roman at Rome. Allegedly, it was mos apud antiques which held that the extension of the 176 ILS 73, 73a; AE 1982,149; ALFOLDY 1991, 305. pomerium 239 was warranted only if ager Italicus had been added to Roman territory. Seneca's authority for this assertion is an antiquarian, whom he had recently heard lecturing on that very topic. The contemporary debate which Seneca represents was occasioned by the extension of the pomeriim under Claudius in 49/50 CE. Claudius, however, like lulius Caesar before him, added provincial rather than Italian soil to Roman Therefore, when affirming the supposed legality extension, territory. of Seneca's antiquarian, without further specifica- tion, implicitly denied the legality of Caesar's and to a Republican discussion of the modus operand! of extend- the pomerium: that discussion started when the need for an exegesis first arose. The work De auspiciis by rius Clau- Yet, Seneca's mos apud antiques is likely to go back dius'. ing Sulla's Messalla Rufus M. Vale- (cos 53, augur 82/1) was used in Gel- lius* discussion of the Roman pomerium; and there are points of contact between the arguments presented in Seneca and Gellius. It is thus possible that Seneca's antiquarian re- ferred back to a controversy to contributed in in the which Messalla Rufus had aftermath of Caesar's extension of the pomerium in 45/4. 177 It is noteworthy that the antiquarian debate made the extension of the pomerium conditional on the acquisition of ager Italicus. That debate must have been 177 Sen. Dial. 10,13,8: Sullam ultimum Romanorum protulisse imperium, quod numquam provincial! sed Italico agro adquisito proferre moris apud antiquos fuit; Gell. 13,14,1-4; GRIFFIN 1962, 107-11. Caesar: Gell. 13,14,4; Cass. Dio 43,50,1. Claudius: OIL 6,1231; Tac. Ann. 12,23,5. Nothing warrants the view of RAWSON 1985, 93 that Messalla, who survived into the Twenties, wrote his work before the Civil War between Caesar and Pompey. 240 inspired by a new Roman consciousness as concerns the status of Italian territory in the aftermath of the enfranchisement of peninsular Italy: 178 ager Italicus could now be as belonging to the regarded populus Romanus, whose territory was expanded by extending the size of Italy. By way of contrast, both the antiquarian discussion and the testimony referred of Gaius to above confirm that with regard to the status of provincial territory an uneasiness prevailed at Rome long into the imperial period. Yet, the Roman revaluation of Italian territory first century did that the the not entail that Roman and Italian reli- gious practices had become identical. The Augustan tion in sacra formula- of municipia ought to be observed in accordance with local custom reflects Late Republican Roman, rather than merely religion; Augustan, attitudes towards municipal as regards the involvement of the pontifical col- lege at Rome, the past tense is used: municipalia sacra vocantur quae ab initio habuerunt ante civitatem Romanam accept am. Quae observare eos voluerunt pontifices, et eo more facer e quo adsuessent antiquiYet, the preservation of local religious custom, as well as of local sacred law, implied that municipal religion differed from the religious system of the city of Rome, as was the case with the rules concerning the inheritance of family sacra in Arpinum observed by the Elder Cato, or burying dead the within the physical boundaries of Italian towns, which 178 Cf. ROPKE 1990, 35. 179 Festus 146 L = Ateius Capito frg. suppl. 69 Strzelecki. 241 would be impossible at Rome. 180 The Roman acknowledgement of municipal religion no doubt reflected the compatibility of the respective religious cultures at Rome and in the Italian munidpia. I shall return to the reasons for that compatibility below. But it is also noteworthy that the relevant ca- tegory on the basis of which religious behaviour was authorized is antiquitas. The mos maiorum was, above, an as we have seen extremely flexible category which did not neces- sarily entail the definition of a binding core of religious behaviour. In the case of municipalia sacra, the Augustan author's specification concerning the involvement of the Roman pontifices is significant. However, pontifical college, through the law, de facto encroached the extent application to which the of pontifical upon local Italian tradition is hard to assess. It is routinely held that, following the enfrachisement of the peninsula, pontifical law and pontifical authority applied to Italy as a whole. 181 But given the de iure preservation of local religious structures, view would entail a such clash between Roman and local sacred law, which is difficult to prove. Furthermore, we simply not know a do to what extent the pontifical college at Rome was consulted during the Late Republic when the repair of tombs 180 Cato Orig. frg. 2,31 Chassignet = 61 P: si quis mortuus est Arpinatis, eius heredem sacra non secuntur. As the discussion of the inheritance of family sacra at Rome in Cic. Leg. 2,48-53 (referring to pontifical law) suggests, this local custom ought to be taken as belonging to sacred law rather than to the law of persons. For the survival of local religious custom, see conveniently CRAWFORD 1996, 426-30. 181 For this thesis, based on a legalistic understanding of the diffusion of Roman religion, see WISSOWA 1912, 408. 242 or the transferral of corpses, which fell under the province of pontifical law, occurred in Italy. For the imperial pe- riod, the consultation of the college on mainly attested in the these matters city of Rome itself but rarely in other Italian communities. Pliny, when writing to the ror, is empe- as the official head of the pontifical college, in or- der to inquire about the applicability of pontifical law provincial tombs, assumes that consultation on this matter by the pontifical college is customary in the city rather than in to Italy of Rome as a whole. Trajan's reply suggests that Pliny follow the established provincial modus operand!. His practical reasoning - durum est iniungere necessitate™ provincialibus pontificum adeundorum - may also have applied to Italians. 182 The comparatively scarce evidence for Ital- ian consultation of the Roman pontifices under the Empire may thus have been the result of individual or local initiative: an inscription which recorded local religious behav- iour as approved by the pontifical college, whose official head was the emperor, could be used to improve a community's or an individual's social standing in a local or an Italian context. Yet, given the evidence it would be wrong to that such imply accumulation of >symbolic capital< occurred on a regular basis in the first century BCE. It is noteworthy that Ulpian, in burial of the a discussion of the dead and the transferral of corpses, recom- mended that imperial legislation, that is the imperial re- 182 Ep. 10,68-9. Rome: e.g. ILS 1792, 7947, 8226-9, 8282, 8380, 8382-3, 8386, 8292; AE 1909,92, 1926,48. Italy: ILS 8110 (Beneventum), 8381 (Tarracina), 8390 (Sabine territory). Cf. MILLAR 1992, 359-61. 243 scripta, supersede municipal law. 183 It is likely that these rescripta, responding to individual claims and queries, were formulated on an ad hoc basis. This is what seems to have been the case with the regulation that sacred groves were in the legal possession of the populus Romanus by as represented authorities of the capital rather than of a colonia the or municipium, even if these groves were situated within the latter's boundaries. In Ulpian, this particular regulation follows the imperial legati's responsibility for sacred places in the provinces as laid down in the emperors' mandata; and it is very likely that by issuing such a regulation emperors tried the to protect sacred land against the economic interests of private possessores. 184 As regards Italian re- ligion, in 22 CE the Roman Senate established that all ceremonies/ temples and divine images in Italian towns were un- der Roman jurisdiction and command. This comes very close to the view that in (provincial!) solo dominium populi Romani est vel Caesaris, held by plerique according to Gaius (Inst. 2,7). Yet, on closer examination the Senate's stipulation turns out to be another ad hoc regulation constructed in the capital: since the equites Romani vowed a statue of the on behalf well-being of Julia to Fortuna Equestris, and since (to everyone's embarrassment) no temple of that goddess could be found at Rome, the jurists declared Italian sanctuaries to be under Roman jurisdiction, as a result of which 183 Ulpian Dig. 47,12,3,3-5. 184 Front. 48 Thulin (= 56 Lachmann): in Italia autem densitas possessorum multum improbe facit, et lucos sacros occupant quorum solum indubitate p. R. est, etiam si in finibus coloniarum aut municipiorum. Cf. MILLAR 1992, 314-5, for mandata. See further below, 4.1.1. 244 the statue could properly be dedicated in the temple of Fortuna Equestris at Antium. 185 By way of contrast, it is unlikely that tionist imperial such interven- views concerning religion would have pre- vailed in the Late Republic. The Tabula Alcantarensis, documenting the deditio of an unknown Spanish town to the impe- rator L. Caesius in 104, establishes the self-imposed limits of Roman encroachment upon non-Roman communities; provided that the populus senatusque Romanus authorized the of that decision Roman official, the community in question would be free to live according to its customary laws and habits: [L. Caesius] ... imperav[it ut omnes] captivos equos equas quas cepisent [traderent. Haec] omnia dederunt. Deinde eos ... [liberos] esse iussit. Agros et aedificia leges cete[ra omnia] quae sua fuissent pridie quam se dedid[issent quae turn] extarent eis redid!t dim? populus [senatusque] Roomanus vellet.^ 86 In contrast to the imperial readiness to impose Roman in an Italian religious context, the scarcity of active Roman encroachment upon provincial or Italian religion ing rules dur- the second and first centuries is remarkable. In parti- cular, following the Social War Roman displayed a Republican officials striking indifference to the regulation of the 185 Tac. Ann. 3,71,1: repertum est ... cunctas ... caerimonias Italicis in oppidis templaque et numinum effigies iuris atque imperil Romani esse. Was Ateius Capito, human! divinique iuris sciens and a loyal servant of Tiberius, involved? Note his presence in Ann. 3,70,1-3. A Roman temple to Fortuna Equestris had been dedicated by Fulvius in 173 (Livy 42,10,5). It still existed in 92 BCE (Obseq. 53) and under Augustus (Vitr. 3,3,2). A dedication to Fortuna was an apt choice of indirect divinisation, for the iconographic representation of female members of the imperial family resembled the divine iconography of goddesses like Fortuna: ZANKER 1987, 236-7. 186 For a text of this important inscription, see J. S. RICHARDSON, Hispaniae (Cambridge 1986), 199-201. Cf. CRAWFORD 1989, 97. 245 religious habits of what had become a the significant part of populus Romanus. As will be shown in what follows, this Roman indifference to religious matters noteworthy if the even more compared with the imposition of a Romanizing municipal or colonial administrative fected becomes juridical and structure, administrative which af- autonomy of the Italian peninsula and of provincial communities with colo- nial or Latin status. 3.4.2 Colon! ae and muni dpi a The municipalization of Roman Italy in the first century was achieved on the basis of an equilibrium between the new communities' relative autonomy and their adoption of Roman ministrative and juridical structures. ad- To judge from the surviving fragments of the municipal charters of the first century, a significant number of institutions, practices and constitutional elements were common to the municipia as a whole. It is therefore likely that established a Roman model charter minimal expectations and thus imposed a minimal administrative framework on these new Roman communities. the other hand, notwithstanding the temptation construct an >ideal charter< from the fragmentary such a charter to On re- evidence, may not have existed in the first century. For instance, the surviving parts of the Lex Tarentina, dating to the forties of the first century, relate to exclusively the municipium of Tarentum, thus suggesting that the mu- nicipal charters were solely relevant communities, to their respective which adapted a Roman matrix law to their par- 246 ticular needs. 187 The notion of Romanization in this context entailed the adoption of Roman administrative thinking. Yet, this process of acculturation was not causally determined by enfranchisement or by direct documented by the Roman intervention. This is Lex Osca tabulae Bantinae, a charter of the Oscan community of Bantia, which presumably dates to the Nineties (RS no. 13). This document was but written in Oscan, used Latin script. Moreover, though composed before the enfranchisement of the peninsula, this statute ized< in was >Roman- character in that it employed a Roman administra- tive nomenclature and Roman institutional practices. Never- theless, the statute outlined the organization of the administrative life of the Oscan community of Bantia. The Lex Osca is a forerunner of the municipal tions of the period after 90. While marking an important step towards administrative unification, this exemplifies the complex that there statute is illustrative To be of the was no straightforward connection between (political) enfranchisement and (administrative) ization. also process of Italian adaptation and choice. Furthermore, the Lex Osca fact constitu- municipal- sure, the municipia were Roman communities through the conferral of Roman citizenship and the municipal law(s) of the 80s; the municipales were Roman citizens by affiliation to the census list and to a tribus at Rome. This status remained unaffected by the individual municipal charter, which was non-political in so far as it was concerned with the administrative demands of the communal life of particular municipium. It one was individuals who constituted 187 Lex Tarentina: RS no. 15, lines 1, 7-8, 11, 18-20, 26-9, 43-4, passim. 247 the municipium. These no doubt used Roman matrix laws when providing the respective municipia with their constitutions. By way of contrast, Roman authorization, while resting with the popular assembly at Rome, must have been a pure formalicapital's ty. This constitutional process again reveals the relative indifference to its new citizens. 188 affect How did these municipal charters the newly enfranchised? Due to the charters' frag- of life with caution, be augmented by a more is answer. to difficult information they contain may profitably, if the Therefore, question this state, mented religious the comprehensive closely and related much group of statutes from the imperial pe- riod, the Flavian municipal laws of individual Spanish muni­ cipia (LFlav) issued under Domitian. Arguably, this strategy is not unproblematic: firstly, the LFlav imposes Imperial but does not necessarily reflect Late Republi- regulations, can conventions. Secondly, the comprehensiveness the of LFlav is due to the fact that the Roman imperial lawyers who drew up the charters intended to provide the Spanish towns with a Roman ideal of municipal charter that life than rather with a be implemented instantly or that would could truly represent the actual administrative structure of these new municipia. 189 Nevertheless, it will the LFlav preserves become well. By way that some rules and expectations which ap- plied in the context of Late Republican city as clear administration of contrast, the colonial charter of the Caesarian colonia of Urso in Baetica (LUrs), preserved in 188 See conveniently LINTOTT 1993, 132-45; BISPHAM esp. 52-64, 256-63. Cf. CRAWFORD 1996, 421-4. Cf. FEAR 1996, 131-69. a 1994, 248 form which is contemporary with the LFlav, preserves the Late Republican religious regulations in a Roman colony. Like the municipal charters, it was >given< by an individual founder and formally passed in Rome. Yet, already the Roman sources commented on the difference between a Roman colonia a municipium: the colonial constitution was supposed to and follow the capital's administrative organization, whereas municipales had the privilege to issue their own local statutes and use their particular legal systems. 190 The distinction between colonial dependence on the capital and municipal autonomy ought to be taken cum ils. The grano sa- LUrs requires the scribae responsible for the fi- nancial records of the colonia to take an oath »in a contio, openly, before the light of day, on a market day, <facing> the forum, by luppiter and the Dei Penates«. The phrasing of this provision is directly adopted from Roman Republican political life: there, as we have seen, magistrates swear obedience to had to Roman statutes by luppiter and the Dei Penates. 191 When this Roman oath reappears in the LFlav, it is significantly enlarged: ... in contione per lovem et divom Augustum et divom Claudium et divom Vespasianwn Augustum et divom Titum Augustum et genium imperatoris Caesaris Domitiani Augu­ st i deosque Penates. 192 The text of the LFlav represents a Roman municipal vision of paradigmatic life. In that context, the phrasing of the 190 E.g. Gell. 16,13,6-9: Municipes ergo sunt cives Romani ex municipiis legibus suis et suo iure utentes . .. populi Romani istae colon!ae quasi effigies parvae Simulacraque esse quaedam videntur. 191 LUrs (RS no. 25) ch. 81, lines 17-9: in contione palam luci nundinis in forum <verso> ius iurandum . .. per lo­ vem deosque Penates. See above, 3.1.3. 192 LIrn ch. 26, IIIB, lines 40-3. 249 oath by luppiter, the Di Penates and various perors (including the Genius of divinized Domitian) formulates straightforward imperial expectations concerning tion of Roman the adop- religious practices in those municipia that were covered by this municipal statute. One can only late as to em- specu- how the capital's expectations would have been realized in a remote town in Roman Spain. However, since the LFlav establishes a framework of implicit imperial tions expecta- put together in the Roman capital, and since the oath by luppiter, the Di Penates, the living emperor and his vinized predecessors was presumably used di- in the city of Rome, the different phrasing of the LUrs is noteworthy. This colonial statute, being republished around the time when the LFlav was composed by Roman Republican version lawyers, apparently of the of the oath, rather than complying with imperial expectations. This instance suggests a degree uses significant colonial autonomy with respect to administrative matters towards the end of the first century CE. The Republican oath by luppiter and the Di Penates as was, I suggested above, tailored to the religious infrastruc- ture of the city of Rome. While oaths were no doubt employed in the administrative life of the Italian municipia ing the municipalization of the first century BCE, it is unclear which deities were invoked Yet, that it is most likely in these choices communities. were, as in Rome, constrained by the local pantheon to hand, just as municipal constitutions were the new geared to local needs. It is possible that the new municipia would wish phrasing follow- to implement a which closely followed Roman usage; but unlike the 250 imperial authors of the LFlav, Republican Rome did not gest its Italian municipales a particular Roman oath by to luppiter and the should be Penates, incorporated or in demand that these charter deities the pantheons of these commun- ities. To put it in a slightly different way, the founding of a Roman colon!a in a new environment not surpri- singly formulated matters of political and religious istration in terms which closely administrative structure admin- resembled the capital, whereas the constitution of a new municipium, Roman sug- with a pre- already in place, would be inclined to use the model provided by Rome much more flexibly. This distinction between coloniae and municipia became operative with respect to the constitution of the local pantheon. The LUrs prescribed the institution of the cults and games in honour of the Capitoline triad and of Venus Gene- trix. 193 The choice of the Capitoline triad is explicable by the statute's desire to transfer a central religious element from Rome to the provincial periphery. Venus Genetrix, on the other hand, was a deity whose Roman cult voured by had been fa- lulius Caesar, the founder and first patronus of the colonia. 194 It is worth noting that the colonial adop- tion of that goddess did not entail that her cult attained a disproportionate importance in the new colony. For instance, Venus Genetrix was not included in the scribes' oath. At the same time, the LUrs specified the prerogative of the colo- nial ordo decurionum to re-establish 193 LUrs (RS no. 25) chs 70-1. 194 WEINSTOCK 1971, 84-7. the calendar of the 251 colonia whenever new Ilviri entered office. The ordo was thus able to revise the colonial choice of public dies festi and sacra every year, even though in practice class the decurial was likely to endorse already existing regulations on LUrs a routine basis. 195 But it is worth stressing that the imposed only minimal religious obligations upon the new co- lonists. As foundation regarded the pantheon of colonia, the its charter entailed only very limited requirements; the actual text implies that any further development was at the discretion of the local ordo decurionum.^ 96 By contrast, our knowledge is limited immediate concerns Italian municipia of the first It is clear that responses to the political Roman- ization of the peninsula differed from community to ity. the impact of enfranchisement and municipalization on the pantheons in the new century. as commun- Moreover, the changes to the municipal religious land- scape largely seem to have been self-regulated, depending on the respective communities' traditions and resources as well as on the eagerness of individual members of the local eli- tes to publicize overtly Romanizing strategies. A preliminary pattern can be reconstructed from the Late Republican dedications that individual magistrates in the Italian muni­ cipia made to a number of dedications different deities. 197 Municipal to luppiter optimus maximus at first sight seem 195 Ch. 64. This is the implication of quicumque SCHEID 1992, 130 12 ; ROPKE 1995a, 535. (line 9): 196 Cf. ch. 70, lines 8-9: ludi scaenici for the Capitoline triad and yet unspecified del deaeque; ch. 72, 11. 32-4 quae sacra ... ei deo deaeue cuius ea aedes erit facta fuerint. 197 For an updated corpus of the inscriptions by municipal magistrates, including dedications to deities, see the catalogue compiled by BISPHAM 1994. 252 to display the Romanization of the Italian landscape at most its extreme. In Vitruvius' normative account of a paradig- matic Roman town, the Capitoline prominent site of triad occupies the most the urban centre. To be sure, Capitolia became an important feature of many Roman colon!ae as a visual means of asserting their Roman environment. The reference status in a non-Roman to the Capitoline triad in the LUrs, mentioned above, can easily be supplemented by ar- chaeological evidence concerning the construction or renovation of colonial Capitolia in second and first century Ita- ly. 198 Generally, if not invariably, communities acquiring Capitolia however, in the the Italian second and first centuries appear to have been Roman colon!ae rather than the new municipia. Moreover, the extent to which the to a deity dedication of the Capitoline triad by the magistrate of a municipium is tantamount to the municipal acceptance of Roman the concept and organization of the cult is impossible to assess. 199 This is not to deny that municipal seriously choices were indebted to the example of Rome. For instance, at Canusium, a community previously exposed to Hellenic culture and language, individual magistrates chose to tions dedica- to deities that must have been borrowed from Rome. 200 Dedications to Apollo and which make was Victoria in Marsian territory, previously not exposed to any Hellenizing influ- 198 Vitr. 1,7,1. Capitolia in lic: TODD 1985, 56-62. At of the Capitolium has not 199 luppiter optimus maximus: CIL 1 2 ,3164 (Petelia). (Bantia). coloniae under the Late RepubUrso itself, the actual site been identified. e.g. OIL 10,1572 (Puteoli), Minerva: Arch. Class. 1969, 15 200 Mars: CIL I 2 ,3182; Vesta: CIL I 2 ,3183; Vortumnus: CIL I 2 ,3184; Concordia: CIL 10,5159. Hellenization: Hor. Sat. 1,10,30. 253 degree ences, also betrays a high of the from borrowing Roman pantheon. 2 ° 1 This picture becomes more Italianized when one focuses on dedications from a wider social range in the Italian context of rural pagi and vici. There, outside the direct im- pact of the urbanized centres of central Italy, the phic epigra- record points to a diminishing influence of Romanizing strategies on the choices made by worshippers, who addressed native, or rather >nativized< divinities from a Italian By background. 202 traditional way of contrast, the continuing Primi- prominence of Hercules Victor in Tibur or of Fortuna genia in Praeneste betrays the local loyalty to deities with a long-standing tradition. The wide variety of choices made in an Italian context - on the one hand the the Roman from borrowing pantheon at Canusium or in Marsian territory and on the other hand the worship of deities with a strong local identity - shows that these choices were not dictated by active an religious policy. Rather, they must have been Roman constrained by the degree to which religious infrastructures were firmly established and by the Romanizing local elites. These choices illustrate adaptation of the Roman pantheon supra-local context of when ambitions of the only partial transferred to the peninsular Italy, and document the limits of the universalization of >Roman religion<. While 201 Mars i : CRAWFORD 1981, 158. For similar processes of local acculturation, cf. e.g. OIL I 2 .3167: Minerva Victrix in Tarentum; AE 1990,303: Victoria in Trea. 202 LETTA 1992, 117-24, who draws attention to dedications made to luppiter, Hercules, Feronia (OIL 9,4321, 3602) or Vacuna (OIL 9,4636, 4751-2; AE 1979,199). Cf. the dedications to local divinities such as Mefitis in Potentia (OIL I 2 ,3163a) or Gemina in Gales (AE 1989,176). 254 provides a unifying category for the de- citizenship Roman scription of first century Italy, >Roman religion< fails do so. Rather, to local adaptation of the Roman pantheon en- tails that the Roman gods and goddesses themselves altered their identities, thus elucidating the process of the additive extension of an open system< outlined above. 203 One final example serves to illustrate that of transferral of complete the Roman religious system to a colonial or municipal context was not at issue. The LUrs appointment a colonial specifies the pontifices and augures as well as their privileges: the exemption from military service and mirnera, the attribution of military campaigns and privileged seating arrangements at games and gladiatorial shows. These priestly privileges were closely modelled on respective those the priests at Rome. 204 Pontiffs, augurs, as well as flamines, are also attested in municipal contexts. is of unlikely that municipal Yet, it priestly offices of the first century closely resembled the organization of priesthoods at Rome. Rather, whereas Roman nomenclature was imposed on cal lo- religious roles, the nature of these roles was probably left unchanged. The fact that Roman nomenclature can be found in coloniae and municipia is not surprising, given the character and composition of colonial and municipal consti- tutions. In the case of the latter, however, priesthoods and religious administration existed prior to municipalization. 203 See above, 2.7.2. 204 LUrs (RS no. 25) chs 66-8; SCHEID 1984, 258-9. It is questionable whether the exemption of Roman priests from military service dates back to a statute of the early fourth century, as Plut. Camill. 41,7, Marcell. 3,4 and App. BC 2,627 maintain; cf. FLACH 1994, 270-1. 255 As seen, have we the municipal administration in general adapted Roman matrix laws to its particular local needs. It is therefore highly plausible that the organization of munipriesthoods as well rephrased the local status quo in cipal a new Roman format. Indirect from comes of thesis this the LUrs. For I would suggest that, whereas the specifications concerning the of privileges and pontiffs that colonial statute follow a Roman model, this in augurs confirmation the Romanizing tendency does not entail that functions of these priests would necessarily have resembled those of pontiffs and augurs at Rome. To be sure, as far as the organization was and its coloniae were or municipia. provided with (like their counterparts at the funds: cult public there are several similarities between Rome concerned, priests of a Rome) Colonial magistrates and number of appari tores, who were paid public from LUrs mentions a haruspex and a tibicen, respec- tively responsible for divination and the musical accompaniment of sacrifices (ch. 62) . In Urso, these appari tores were apparently free citizens of the colonia. In analogy city of to Rome, public slaves, including those who performed the actual sacrifice, were also made available. 205 In tion, the LUrs prescribes addi- the appointment of magistri ad fana templa delubra, who have a further the the responsibility for organization of the circus games, of sacrifices and the setting up of pulvinaria (LUrs ch. 128) . particular institution does However, as this not directly resemble a Roman 205 Lirn chs 19-20. For appari tores at Rome, see the Lex Cornelia (RS no. 14) col. I, lines 1-6 and the Tabula Heracleensis (RS no. 24) lines 80-1. 256 organizational model, the status of these magistri is not entirely clear. For the overall administrative and financial responsibility for the maintenance of temples in the city of Rome, in Roman colon! ae and in municipia rested with the aediles. 206 On the other hand, these magistri to are unlikely have been religious functionaries of lower social stand- ing. The procedure outlined in the LUrs - magistris creandis (ch. 128, line 19) - implies that the election of persons of considerable social status was concerned. suggests comparison with The nomenclature the office of magister municipi, involved in various civic and religious duties at a munici- pal level and drawing on members of the local upper classes; or with the office of curator tempi! , which implied personal and financial liability for the upkeeping of shrines by in- dividuals of high social status performing a munus . 2 ° 7 Another close non-Roman parallel is provided by the gistri fanorum of individual deities in Republican Capua before its transformation under Caesar in 59. As Martin Frederiksen has shown, these boards of magistri, persons of social standing and sufficient wealth to pay a surma honora­ ria and contribute to public expenses from their own funds, were responsible for the maintenance and Capua administration of and its territory. Supervised by the central pagus of 206 E.g. LIrn ch. 19. Cf. Cic. Fam. 13,11,1, discussing the aedilician administration of sacra ... et sarta tecta aedium sacrarum locorumque communium in the mimicipiinn of Arpinum. There, the maintenance of temples, supervised by the local board of aediles, was financed from the rents that the community received from property let to tenants in Cisalpine Gaul. See further below, 4.1. 207 Magister municipi: PACI 1980. Curator tempi!: RIVES 1995, 36-7. For a curator tempi! of the temple of luppiter Dolichenus at Rome, see CIL 6,30758 = ILS 4316. 257 Capua, these boards of magistri, freedmen, financed comprising free-born the building and repair of civic sacred and secular buildings, either on their own or in tion with their colleagues. necessitated by the collabora- It is worth noting that this form of organizing the maintenance was and of Capua's sanctuaries exceptional lack of a decurional class in the local community. 208 A similar rationale must for the introduction of magistri ad fana templa de- account lubra in the Roman colony of Urso. At Rome, members aristocracy were expected to of the contribute to the building, rebuilding, refurbishing and maintenance of the city's sanctuaries. The existence of such an ethos of could not be taken elite behaviour for granted in a recently established colonia, where a new colonial decurional elite still had be to formed. In the absence of the deeply rooted aristocratic euergetism which informed the political life of the capital, the office of magister would therefore offer a means of promotion to the new colonists and at the same time guarantee the emergence of individual munificence at a colonial level. Moving on to another striking feature, it is illuminating to see how the existence of magistri in colonial and municipial constitutions entails a functional differentiation of reli- gion's administration in these communities. A similar juxtaposition of de iure aedilician authority and diffusion of the de facto administrative responsibilities characterized the system of civic religion at Rome. 209 208 CIL 10,3918, 10,3924; 10,4620; FREDERIKSEN 1984, and 286-7. 209 see further below, 4.1.2. 264-84 258 In contrast to the detailed administrative system of organizational, and financial regulations provided, the LUrs gives only minimal specifications concerning the actual ligious duties re- the two priestly colleges of pontifices of and augurs in the colonia of Urso: pontific(es) augures cra sa­ publica ... facient (ch. 66). In the capital, the civic system of XVviri, religious Vllviri, functionaries fetiales - pontifices, augurs, and Vestal virgins (to name but the most important functional groupings) - was characterized by the specialization of ritual as well as other obligations of religious administration. 210 The increase in of members number of these priestly colleges can be seen as a re- sponse to the demands of representing civic metropolis the that outgrew the city-state. Turning to the LUrs, confines a religion of the noteworthy in a nuclear discrepancy becomes evident between the degree of specialization at Rome and in the colonial statute. In particular, there is a dis- proportionality regarding the sophisticated responsibilities of the pontifical college at Rome as laid down by Roman sacred law and the obligation of sacra facere in the case of their colonial colleagues. As we have seen above, Pliny when inquiring into the applicability of pontifical law to pro- vincial communities sent his query to the head of the pontifical college at Rome. It was the capital where a certain 210 Cf. WISSOWA 1912, 479-566. BEARD 1990, 19-43 and RUPKE 1996a, 252-5 stress the diffusion of religious authority as a result of the division of responsibilties and the lack of internal hierarchization; cf. above, 2.4.2. On >specialization<, see ROPKE 1996a, 255: »[Specialization is a useful term ... it was only the membership within the colleges that defined the religious function: political position and social prestige ... did not suffice to establish the special religious and functional competence*. 259 local knowledge about matters of sacred law existed, and where pontifical law could be expected to apply. 211 By trast, con- it is unclear whether the pontifical law of the city of Rome would have applied in colonial or municipal communities, or whether Pliny would have been able equally to receive an authoritative answer from the colonial or municipal pontifices. This instance illustrates that despite the colonial and municipal willingness to adopt Roman administrative models the religion of the city of Rome to a significant extent remained a local religion, which did not easily >travel<. Only its administrative epitome was imposed on the colonial charters or transferred to the municipal tions. 211 Cf. above, 3.4.1. constitu- 4 RELIGION IN LATE AND SOCIETY REPUBLICAN dvdpCDKOL OLXiau xevau Ydp TlOU TtoXlQ ou6e aToat ROME eCJTiy, ou6 3 dXX' OUX dyopat dv6pov >Augustus< ap. Cass. Dio 56,5,3, ing to >Nicias< ap. Thuc. 7,77,7 allud­ The previous chapter illustrated the difficulty of adequate­ ly outlining the system of beliefs, deities and we cults that unwittingly tend to see as a stable entity called >Roman religion<. Whereas the result may seem largely negative, it suggests that progress cannot be made by applying tradition­ al models of socio-political chapter will therefore revaluate city of Rome's religious or cultural identity. This the organization of the system beyond these models. In­ stead, it will suggest more complex parameters that consti­ tuted the differentiated local religion of Rome. 4.1 Organizing local religion at Rome The ancient city-state was responsible for providing a reli­ gious administrative infrastructure for its citizens, re­ garding the financing and maintenance of civic temples, fes­ tivals, rituals and sacrifices. At Rome, this civic respon­ sibility was retrojected to the >foundation< of state reli 261 gion by Numa; 1 and this civic obligation persisted until the later fourth and early fifth centuries CE, when the Chris- tianization of the agents of organized religion resulted in the demise of paganism in its civic form. 2 4.1.1 >Sacrum< and >publicum< According to Roman sacred law, through its dedication a tem­ ple (like any other object) became sacer and thus the pro­ perty of the deity in question; it no longer belonged to the populus Romanus or to private individuals. 3 Yet, the provi­ sion and maintenance of property than that was sacrum, rather being left to the divine realm itself, operated in the framework of publicum as managed by the Following civic authorities. pontifical decision, a dedication in loco publico which had not been authorized by the city-state was not sa­ crum; and tombs could not be placed in loco publico, as they were religiosa, belonging to the realm of private religion. 4 The civic administration of public religion entailed that in terms of the Roman law of property sacrum and publicum were closely related. As concerns building activities in the city of Rome, the Tabula Heracleensis, presumably dating 1 2 3 4 to 45, Livy 1,20,5: quibus hostiis f quibus diebus, ad quae templa sacra fierent, atque undo in eas sumptus pecunia erogaretur. Cf. Fest. 284 L = Ateius Capito suppl. frg. 70 Strzelecki: publica sacra quae publico sumpto pro populo fiunt. METZLER 1981. See BARNES 1995 on the Christianization of the Roman elite in the early 320s; and SALZMAN 1993 for the Christianization of elites in response to imperial pressure. CANCIK 1995 discusses the ensuing transforma­ tion of paganism. E.g. Trebatius Testa ap. Macrob. Sat. 3,3,2. Cf. Aelius Gallus ap. Festus 348-50 L; Gaius Inst. 2,4-8. Dedications: Cic. Don. 127-37; LINDERSKI 1986, 2249 407 , 2272-9; TATUM 1993. Tombs: Cic. Leg. 2,58; cf. Paulus Dig. 47,12,4. 262 specified aedile's the responsibility for repairing that part of a road which ran beside aedis sacra seive aedificium of publicum seive locus publicus, whereas the repair adjacent to private activity was left to those who owned property that property. Similarly, parts transport building to relating at sacred sites as well as to civic building would be permitted by the city's authorities during those times of the day when public transport was from banned otherwise Rome. 5 It is such juxtaposition of the publicum which has led scholars to assume the conceptual embeddedness of religion in the public realm. these scholars and sacrum of realms To be sure, are prepared to accept that in the Late Re­ public Roman legal and administrative conceptualized texts three kinds of domains, sacred, public, and private. »But it is also clear that throughout the Republic the domain of the sacred and the domain of the public were very close to each other and that the essential boundary did not lie between the divine and the human.« 6 As outlined above, the problem with such a modern view is that its understanding of the native Roman conceptualization of religion depends on the definition of Roman religion as an external and largely public affair. The now dichotomy of disreputable >sacred< and >secular< has thus been replaced by the dichotomy of >public< and >private<; and the view that the sacred domain belonged to the publicum is supposed5 6 RS 24, lines 29-31, 56-61. Note lines 62-5, whose phras­ ing is echoed by Livy 34,1,3, outlining that the ban on vehicles in the city of Rome did not apply to the Vestal virgins and the Rex sacrorum, when acting sacrorum publicorum p(opuli) R(omani) caussa, or to generals dur­ ing their triumph. CRAWFORD 1989, 94. For further documentation of this view, see above, 2.4.2. 263 ly reinforced by the testimony of legal and administrative texts or the interpretations of the elite administrators of religion, priests and magistrates. The danger of cir­ Roman cular thinking is apparent, since these Roman textual genres address only the external aspect of religious behaviour, and underly­ therefore are a priori incapable of falsifying the modern proposition concerning the external character of ing have Roman religion. Below, I will demonstrate that we will to go beyond these dichotomies, if we wish to reassess the religious behaviour of the agents of ever, on Roman How­ religion. internal grounds alone the close interrelation of not sacrum and publicum as postulated in scholarship is an unproblematic proposition. con­ For the Roman administrative and legal texts could ceptualize sacrum and publicum as different domains, if (and if) only the distinction mattered. Still with the adminis­ tration of property, financial funds that were private not were divided into pecunia publica sacra religiosa. The muni­ cipal Lex Tarentina, when dealing with the embezzlement of these funds, treats them as belonging to the municipium, without further discriminating the legal terms of ownership. Yet, the Lex Tarentina seems to present an altogether sim­ plified version of a Roman model, expedient as municipal administration regards but not a truthful conceptualiza­ tion of the interrelation of sacrum and publicum. At tum, the Rome, Taren- embezzlement of money was covered by civil law. A fourfold fine applied, modelled on legislation on at the peculatus yet a local magistrate exacted the penalty, which had to be paid to the municipiirni. By contrast, at Rome pecu- 264 latus, the misuse of public money, and sacrilegium, the mis­ appropriation of sacred money or objects, were distinct fences, even of­ though both were part of public criminal law, and although sacrilegium appears to have been subsumed under peculatus. And whereas at Tarentum pecunia religiosa, presumably relating money to tombs and the cult of the Di Manes, belonged to the same category as public and sacred money, it was a res religiosa and thus private rather than public ac­ cording to Roman sacred law. At Rome, the pontifical college had the right to fine persons desecrating graves, but this particular offence could be prosecuted by zen. 7 In the any Roman citi­ absence of an organizational structure which allowed for temple jurisdiction and policing, it was a prac­ tical concern for the protection of the sacred domain prevailed, when Ulpian (Dig. 48,13,1), referring to the Lex lulia de peculatu, suggested that in terms of sacrilegium which criminal law was subsumed under the offence of peculatus. In the case of loss, damage or embezzlement of temple property, which was divini iuris, capital punishment could thus be inflicted on temple personnel (Pliny NH 34,38) . The Roman anxiety over preserving the public and boundary between private domains is responsible for the frequent juxtaposition of sacrum and publicum in legal or administra­ tive contexts. At the very beginning of his De architecture, 7 Lex Tarentina (RS no. 15), lines 1-6, with the editors' commentary. Rome: [Cic.j Ad Her. 1,12,22 (peculatus); Cic. Invent. 1,11, 2,55 (sacrilegium). For the nature of pecunia religiosa, cf. the definition of religiosum in Aelius Callus ap. Festus 348-50 L; Gaius Inst. 2,6, and the discussion of GNOLI 1979, 71-132. Desecration of graves: FIRA ch. 15; Ulpian Dig. 47,12,3; BEHRENDS 1978. See further below, 4.2. 265 Vitruvius divided building activity into public and private, and public buildings into fortifications, fana and sanctuar­ ies, and civic buildings like baths, theatres this was a contextual choice: what Vitruvius, To (1,3,1). porticos and public mattered in this context was the distinction between buildings, whereas the further sub-division of private and of books first the public buildings mirrored the structure of five his work on architecture. Similarly, the distinc­ tion between public and private was of fundamental importan­ ce in relation to the status of land. A document such as the Lex agraria of 111 or the Republican practice illustrate land the concern over private rights of Roman the usufruct and property. In such contexts, gether of sacrum and surveying of grouping to­ is of minor significance, publicum since their situational juxtaposition entails only an unsys­ tematic conceptualization on the these texts. part of the authors of Rather, treating sacred land as if it consti­ tuted a category similar to that of ager publicus was an expedient protection of the sacrum against private misappro­ priation. 8 However, the contextual relationship of sacred and civic domains was redefined as soon as the internal administration of the sacred domain became an issue. As regarded the polic­ ing of temple regulations, a sanctuary's maintenance was organized with a view to the involvement of the civic autho­ rities. The Lex templi of the temple of luppiter at Furfo in Sabine 8 territory, laid down by the sanctuary's two dedica- Lex agraria: RS no. 2. Cf. CIL I 2 ,402-3: censuere sacrom aut poublicom ese with CRAWFORD 1989, 95. aut 266 tors in 58, 9 placed the sale of the tions received (venditio) as donations (lines dedica­ well as the usufruct of any resulting income (locatio) at the discretion aedile and of the local 8-9). Furthermore, the prosecution of embez­ zlement or theft of sacred property was under the control of the aedile and the vicus Furfensis. Since they fixed the resulting fine, penal money was presumably paid to the local community (14-6) . The adaptation of Roman legal and sacred terminology suggests that this statute drew on the model of Roman sacred and pontifical law. If so, the boundary that the Lex Furfo- nensis draws between sacred and civic realms copies Roman conceptions. For the statute's intention concerning venditio and locatio was to enhance the temple's prosperity in that any income resulting from the aedile's transactions had to be spent on the temple itself. 10 This was achieved by a pro­ cess whose complexity underlines the distinction between sacrum and publicum. The sanctuary's property was a res sac­ ra. Since the temple's prosperity demanded that this proper­ ty was made available for the aedile's financial transaction and usufruct, and since in that process the aedile had to be protected against the charge of sacrilegiim, the property was transferred into a res profana: (7-8) Sei quod ad earn aedem donum datum, donatum dedicatumque erit, utei liceat oeti, venum dare; ubi venum datum erit, id profanum esto ... (11-2) Quae pequnia ad eas re data erit, profana esto, quod dolo malo non erit factum. 9 10 ILLRP 508 with DOLL 1972, 288-93. Lines 10-1: Quae pequnia recepta erit, ea pequnia emere conducere locare dare, quo id templum melius honestius seit. For the economic importance of such temple proper­ ty, see below, 4.1.2. 267 The validity, and the frequency, of the transfer of property from the sacred domain to the domain of human usufruct, im­ plying their conceptual differentiation, is attested by Trebatius Testa's paraphrase of the meaning of profanum: profanum .. . quod ex religioso vel sacro in hominum usum proprietatemque conversum est ... proprie profanatum quod ex sacro promiscuum humanis actibus commodatum est 11 In a further step, however, the Lex Furfonensis any ruled that acquisition resulting from the aedile's transactions on behalf of the temple of luppiter took the character of a res sacra, thus moving back from the human domain to the domain of the sacred: (12-4) Quod emptum erit aere aut argento ea pequnia, quae pequnia ad id templum data erit, quod emptum erit, eis rebus eadem lex esto quasei sei dedication sit. Through these detailed specifications the temple of luppiter protected its property, which had again become a res against misappropriation. Although the temple's rights were safeguarded by the civic authorities, the Lex theless sacra, managed to shield sacra never­ the temple of luppiter against civic encroachment. For its classification of temple proper­ ty as a res sacra prevented the domain of the publicum from having unauthorized usufruct of the temple's resources. The interaction of the two separable domains of publicum and sacrum also informs the content of the sacrae. Consider so-called the inscription from Cignoli in Macerata, datable to 6 CE: M(arco) Lepido L(ucio) Arrunti(o) co(n)s(ulibus) d(ecreto) d(ecurionum) posit(us) Qui intra stercus 11 Leges Ap. Macrob. Sat. 3,3,4. Cf. DOLL 1972, 284. 268 fuderit multae a(sses) IIII d(abit) 12 Prima facie, it is inviting to link the tion of cleanliness within intended preserva­ the boundaries of this sacred place to wide-ranging theological concerns about sacred pol­ lution. At the same time, practical considerations ing general hygienic and administrative aspects must have played an important role, since and financial prevailed, as Leges sanctuary's functioning such practical considera­ this and parallel texts do not openly address the theological issue of these a income would have been compromised by pollu­ tion and obstruction. Arguably, tions concern­ sacred purity. 13 Nor do sacrae deal with particular religious or theo­ logical concerns such as the exclusion of individuals on grounds of gender or ethnic origin. 14 As was the case with the Italian Leges Lex tempii from Furfo, the sacrae refer to the authority of magistrates for protection and policing; and the local civic authorities exacted the fine resulting 12 13 14 from the overstepping of the PACI 1987, for text and commentary. For parallel texts, see e.g. ILLRP 485: neiquis intra terminos propius urbem ustrinam fecisse velit neive stercus cadaver iniecisse velit. Stercus longe aufer ne malum habeas; ibid. 504-8. Cf. NEMETH 1994, who argues that practical, rather than religious considerations, prevailed in the case of similar temple regulations in the Hellenic world. Contrast the Lex sacra on the fifth century Corcelle altar, which appears to have limited access to the sanc­ tuary by certain groups of women on religious grounds: MOREAU 1988 who, at 319-20, gives further regulations, thought to date back to the archaic period, concerning the exclusion of certain categories of women - paelices, virgins or remarried women - from specific cults of fe­ male deities. 269 statutes' prescriptions. 15 In the case of the sacred groves addressed in ILLRP 504-7, the Leges sacrae propose a twofold fine, consisting of an offering form of of expiation, the an ox to be sacrificed to luppiter, and a multa in cash. The receiver of the multa must community taking protecting the sacred have been the local grove. The public domain invariably employed such fines for civic and religious com­ munal activities alike. The LUrs ruled that fines exacted in connection with the raising of vectigalia should be used for the colony's sacra (ch. 65). According to the municipal Lex Tarentina, a magistrate could spend part of the fine result­ ing from deceitful maintenance of public property on building public and games (RS 15, lines 27, 32-8) . The Roman aedi- les employed fines as a means of financing the construction and decoration of the city's temples. 16 However, it is noteworthy that the twofold specification a fine in these Leges sacrae positively established a distinction between the sacrum and the publicum. If fence of an of­ was committed within the boundaries of sacred groves, a sacrifice to luppiter as the instance dealing with offen­ ces against the divine realm applied. But only if the offen­ ce was 15 Civic authority: e.g. PACI 1987, line 2: decreto decurionum; ILLRP 485, line 1: L. Sentius C. f. praetor de sen(atus) sententia. Civic involvement: e.g. ILLRP 505, lines 8-10 and 506, lines 11-2: dicator[ei] exactio est[od]; ibid. 508, lines 14-5: aedilis multatio esto. 16 committed wittingly and with evil intent, it became E.g. Pliny NH 33,19 (Concordia in 304); Livy 10,23,11-13 (games and golden paterae dedicated to "Ceres); 10,31,9 (Venus in 295); 10,33,9 (Victoria in 294); Pliny NH 18,286 (Flora in 241 or 238); Livy 24,16,19 (Libertas in the 230s); 33,42,10 (Faunus in 194). 270 an issue of human law as well/ for which complementary a fine in cash was levied: Sequis advorsum ead violasit, lovei bovid piaclum dato; seiquis scies violasit dolo malo, et lovei bovid piaclum dato et a(sses) CCC mo[ltai simtoj. 17 distinct The twofold fine, addressing two the ceptualized con­ instances, domains of sacrum and publicum as two dis­ tinct entities, to which different kinds legal of under­ standing applied. 18 The major Roman priesthoods had the usufruct of, as well as owned, landed or urban property. In exceptional cases of crisis, this property could be appropriated by the monetary the state. Unfortunately, it is unclear whether property or value in cash was later restored to the priests. equivalent of Roman civil law accounted for the possibility bequeathing testators to priests and temple personnel. 19 By legacies contrast, only a few respected deities could be nominated as whereas it is not (Ulpian legacies heirs or recipients of infrequently assumed frg. But 22,6). that temples in did Italy, unlike the sanctuaries of Greece and Asia minor, not own land, this was not universally true. The office of vilicus Dianae, the bailiff of (the temple of) Diana Tifatina at Capua, or that of saltuarius Virtutis, attested temple 18 19 a of Virtus near modern Ferrara, suggest that at least some temples 17 at in Roman Italy possessed and administered ILLRP 506, col. 2,2-10 = ILLRP 505, col. 1,10-col. 2,7. Cf. Tac. Ann. 1,73,4: deorum iniuriae dis curae; Cic. Leg. 2,22: periurii poena divina exitium, humana dedeCUS, 2,43-4; WISSOWA 1912, 388. Property: e.g. Sic. Flaccus 162 L; Festus 204 L. Appro­ priation: Oros. 5,18,27; App. Mithr. 22,84, referring to the years 91-88. Cf. BODEI GIGLIONI 1977, 33-45; ROPKE 1995b, 281-2. Legacies: Scaevola Dig. 33,1,20. 271 landed property of their own, which in legal terms would have been sacred land belonging to the respective deity, but whose estates were maintained as private land by an organi­ zational structure resembling Villae rusticae. 20 This observation raises a final question with respect to the Roman conception of temple property. Wars of During the Civil the Late Republic and the triumviral period, Roman generals not only pillaged private money, for which a number of Roman and Italian sanctuaries served but also used the as depositories, 21 property of the temples of Roman Italy itself as a means of financing their military campaigns. the In self-representation of the Late Republican condottieri, the appropriation of sacred money was portrayed as a tem­ porary borrowing of divine property, which would subsequent­ ly be restored to its rightful owners. By contrast, Caesar presented his opponents negligent as sacrilegious and deliberately of the boundary between the respective domains of human and divine law: Tota Italia dilectus habentur, arma imperantur, pecuniae a municipiis exiguntur, e fanis tolluntur, omnia divina humanaque iura permiscentur. 22 20 21 22 Vilicus: CIL 10,8217 = ILS 3523. Saltuarius: OIL 5,2383 = ILS 3524. Cf. BODEI GlGLIONI 1977, 41-2; CARLSEN 1994. Sulla donated land to Diana Tifatina in gratitude for his victory over the consul C. Norbanus in 83: Veil. Pat. 2,25,4; DE FRANCISCIS 1966. Cf. Cic. ND 3,49 and Livy 24,3,4-6 for landed property belonging to Greek sanctuaries and to the temple of luno Lacinia at Croton respectively. BROMBERG 1939-40. Caes. BCiv 1,6,8. Despoiling of temple property: e.g. Pliny NH 33,16 and Diod. 38-9, frg. 14 (Marius and Sul­ la); Suet. Icil. 54,3 (Caesar); Caes. BCiv 3,33,1; Cass. Dio 48,12,4: xai xpTpaTa dnavTaxodev xai ex T&V ispcov fiOpotaav; App. BCiv 5,24,97, 5,27,106 (Octavian and Marc Antony). Rhetoric of restitutio: Caes. BCiv 2,21,3; Au­ gustus RG 24; Suet. Aug. 18; Cass. Dio 51,16,3. For the 272 Furthermore, the restitutio named the deity as the owner had of the sacred property that misappropriated. distinction between the public domain, to which sources In law of property/ these Roman generals made a the of terms been re­ sacred had temporarily been transferred, and the domain of the deity, to which these resources belonged. 23 Moving to a more nomenclature observation, general mirrored official Roman conceptual distinction between this sacrum and publicum. The estate managed by the vilicus nae above mentioned bore the Dia- name of p(raedia) D(ianae) T(ifatinae) (OIL 10,3828): both the office and the land do­ cumented the ownership of the goddess. Similarly, the phrase aedituus Dianae Plancianae (AE 1971,31) referred to the tem­ ple in terms of an office obliged to one particular warden deity, rather than suggesting that the office merely repres­ ented a functionary appointed by the Roman state, whose sponsibility it was to supervise public property. At the same time, however, these offices, as well as the sphere their activity, were which of placed in the framework of civic au­ thority. As a consequence, we find ourselves in a in re­ situation any attempt to define the relationship of the two domains of sacrum and publicum through the received dichoto­ mies of >sacred< and >secular< or of >public< and is a >private< priori compromised. 24 One way of avoiding the trap of these dichotomies is to apprehend how this relationship ful- 23 24 religious propaganda of the generals of the first centu­ ry in general, see JAL 1961; Id. 1962. E.g. U. LAFFI, Athenaeum 49 (1971), 45: e(go) v(olo) vos c(urare) ... utei Lusias ... resrtituat deo fa[num e]t in eo inscribeatur Imp. Caesar Deivei f. Augustus re[stituit]. Cf. e.g. SCHEID 1984, 259. 273 fils the criteria of the interaction of >open can systems<. We fully assess the relationship of these two domains only when applying a model that allows us to understand entities which are them as closely interrelated, yet which at the same time preserve their internal autonomy marked by bound­ aries. 25 As we shall see, the conceptualization of the realm of religion as an autonomous area characterized the organ­ ization of religion at Rome. 4.1.2 The diffusion of religious administrative authority * The name aedilis was etymologically linked to aedes, and the aedium sacrarum procuratio was only one of the numerous pects as­ of the aedilician cura urbis under the Republic. Yet, contrasting the minute administrative apparatus available to the aediles and their vast administrative obligations in the Late Republican capital, the practical efficiency actions of their must remain doubtful. 26 In order to meet the admin­ istrative demands of the city of Rome, there existed a clear territorial division of aedilician responsibilities later forties the of the first century. By contrast, Cicero in 70 seems to imply that his duties ferentiated in as future aedile, dif­ into the domains of organizing the Ludi, of up- keeping the sacrae aedes, and of administering secular busi­ ness, covers the entire city. In this context, however, cero intends to magnify the office of aedile before the Roman People, rather than detailing 25 26 Ci­ the functions of For the methodological framework, see above, 2.7.4. one Cf. ROBINSON 1992, 59-82; KUNKEL & WITTMANN 1995, 487-8; NIPPEL 1995, 17. Etymology: Varro LL 5,81; Paulus Festus 12 L; Dion. Hal. Ant. Rom. 6,90,3. 274 particular aedile. Therefore, it is unwarranted to believe on the basis of this passage only that the territorial divi­ sion in the competence of aediles postdates the year 70. 27 The aediles' procuratio sacrarum aedium involved the administrative organization of the restoration of civic tem­ ples/ shrines and altars. This did not necessarily entail the use of public money. The aedilician involvement in the refurbishment of the mosaic floor in the temple of Asclepius on the Isola Tiberina in the early second century is an ex­ ample of the interaction of two autonomous domains. For aediles 1 contribution was restricted to the locatio, which was entirely paid for from donations and fees that the ple had the tem­ received and administered as its own (sacred) pro­ perty. 28 The aedilician procuratio extended, it seems, to the regular annual supervision of the physical state of temples, of aeditui, and of the property of civic temples. The aedi- tuus of Varro's De re guests rustica is unable to welcome the he had invited to celebrate the Feriae Sementivae in the temple of Tellus because he has been summoned to see the aedile. 29 In general, aeditui were subject to the directions of Roman magistrates (Livy 30,17,6). At the same time, 27 28 29 how- Cic. Verr. 2,5,36. Forties: Tabula Heracleensis (RS no. 24), lines 24-31. Cf. Varro RR 1,2,2: (aedilis) cuius procuratio huius templi est. The dramatic date of RR book 1 is 45-37 rather than 59-56, as is traditionally assumed; see FLACH 1996, 7-15, 38-9. CIL I 2 .800 = ILLRP 39: ... aed(iles) d(e) stipe Aesculapi faciundum locavere, eidem pr(aetores) probaverej LATTE I960, 277. Cf. CIL 1 2 ,807; CIL 6,36807, for aedil­ ician involvement; ILLRP 186 and 191, for stipes spent on small-scale sacred building. RR 1,2,2. Cf. the aediles' cura annua mentioned in [Asconius] 251 St. 275 ever, the office of aedituus serves to illustrate the preca­ rious interaction of the domains of sacrum and publicum. As we have seen, these aeditui documented their the obligation to temple and to the deity rather than to civic authority. They commemorated the social fact that head they stood at the of a civic temple with a distinct legal status and so­ cio-economic function, supported by its own administrative apparatus and internal hierarchy. Men like Varro's aedituus, who was freeborn and in a position to manumit slaves (RR 1,69,2), were no doubt eager to enhance their social posi­ tion by seizing the administrative responsibilities that the aediles delegated. This process contributed to the decen­ tralization and diffusion of civic authority in matters of religious administration. The providers of the basic means of daily cult routine in the city constituted one particular group over which pub­ lic control could have been exerted in the final years of the Late Republic. For the senatorial legislation de giis colle- of the sixties and fifties also extended to those col­ legia without whom the smooth running of the would sacra publica have been impossible. In the Augustan period, the re­ strictive reorganization of collegiate life in Rome entailed that a collegium symphoniacorum would resort to in public that its raison d'etre was its participation in the city's sacra publica, and that this contingent advertising raison d'etre was upon the authority of the Senate and that of Au­ gustus. 30 The prescriptive phrasing that shines through this 30 OIL 6,2193: dis manibus collegio symphoniacorum qui sacris publicis praesunt ... quibus senatus c(oire) c(onvocari) c(ogi) permisit e lege Julia ex auctoritate 276 act of collegiate self-representation reflects the izing control over central­ collegia and suggests a regulative ap­ proach towards the organization of civic religion stan Rome. By contrast, in Augu­ Late Republican evidence for the monitoring of these collegiate providers of cult services is lacking - or seriously fragmented. 31 In any case, systematic legislation concerning collegia did not exist prior Caesarian to the and Augustan Leges luliae de collegiis. The Impe­ rial attitude towards collegia, which informs the respective positions of the LFlav and of Trajan and the Pliny, rested on principle that the public usefulness of a collegium had to be proved to the civic authorities before its knowledgement could legal ac­ take place. Without that civic permis­ sion, a collegiate organization would incur the legal charge of illegally assembling (coetus situation causa). While this legal presumably provoked the Augustan symphoniaci into asserting their utility as the providers of services for the realm of civic religion ludorum causa, and affirm the centrality of caused them to the domain of civic religion, a comparable form of centralization does not seem to have been achieved, nor indeed intended, by the Late Republican Roman Senate. 32 31 32 Aug(usti) ludorum causa. SC de collegiis of 64: Asconius 7 C. Leges luliae de collegiis: Suet. lul. 42,3; Joseph. Ant. Iird. 14,215; Aug. 32,1. Cf. LINDERSKI 1995, 217-23. On collegia, see further below, 4.2.3. ILLRP 185, referring to a college teib(icinum) Rom(anorum) qui s(acris) p(ublicis) p(raesto) s(unt) e[x ?s(enatus) c(onsulto)], could be a Republican example. But the restoration is uncertain, and dating rests on speculation. LFlav ch. 74; Pliny Ep. 10,33-4. Cf. Gaius Dig. 3,4,1. Since LUrs ch. 106: coetum conventum coniu[rationem ..., dealing with illegal gatherings, is fragmentary, we can­ not rule out the possibility that this chapter of the 277 re­ The diffusion of civic administrative and financial sponsibilities also take the form of public spending would on the remuneration of private contractors, who provided for those things quae ad sacra resque divinas opus erunt (LUrs ch. 69). Civic sacred buildings subject to censorial locatio would be let to private entrepreneurs by auction for a li­ mainte­ mited period of five years. The person aquiring the nance (tuitio) of a sacred building for a set amount of mo­ ney, paid out of the public undertake obliged himself to its upkeeping and repair. 33 The tuitio of private entrepreneurs included the objects, treasury, for catered regular maintenance of sacred painting the cult statue of luppiter Capitolinus, organized the feeding of the sacred geese on the Capitoline hill, rendered the horses for the annual Ludi circenses, or provided the trombonists who served to announ­ ce the comitia centuriata. 34 In addition, private maintenan­ ce covered the nonregular refurbishment of sacred statues that had been damaged. The inefficiency of this practice illuminated by is a famous incident from the years leading up to Cicero's consulship. In 65, lightning struck the Capito­ line hill, liquefying bronze tablets inscribed with laws and damaging 33 34 honorific statues, a statue group of the she-wolf Caesarian charter already included collegia. In general, cf. BEHRENDS 1981, 174-8; below, 4.2.3. E.g. Cic. Verr. 1,129-134; Festus 428, 254 L; Paulus Festus 429 L; Tertull. Apol. 13,5; KUNKEL & WITTMANN 1995, 446-61. The sacred buildings on offer were dis­ played in the tabulae censoriae: Cell. 2,10,1 (censoris libri). The tabulae censoriae listing the private sacellirni that formed part of Clodius 1 house on the Palatine (Cic. Har. resp. 30) presumably were censorial records of private property; cf. LENAGHAN ad loc. . Capitoline hill: Cic. Sext. Rose. 56; Pliny NH 10,51; Plut. Quaest. Rom. 98. Ludi: Livy 24,18,10; Paulus Fe­ stus 8 L. Trombonists: Varro LL 6,92. 278 with the twins as well as a luppiter on a column. After con­ sultation with the haruspices, the consuls of that year gotiated the respective locationes. In the event, work on the statue of luppiter could not commence prior to presumably to the ne­ heavy 63, due demands which deterred potential contractors, since as a result of the recommendation of the haruspices the locatio entailed replacement by a larger sta­ tue and its subsequent relocation. 35 In effect, the Republican system zation transferred of religious organi­ public control over routine administra­ tion to private individuals. The >state< entered into an economic contract with an equal partner: this legal arrange­ ment limited the control of the public domain. The absence of systematic state control and its corollary, the decen­ tralization of religious authority, are striking features of the Roman Republic, which the advocates of the civic reli­ gion model do not suffiently take into account. Partly as response to the inefficiency of the Republican system, and partly with the intention of created the a office increasing control, Augustus of curatores aedium sacrarum et operum locorumque publicorum. The Augustan reorganization of the maintenance of sacred and civic buildings meant that private contractors 35 would still be involved, but that civic super- Cic. Div. 2,46; Obsequens 61; Cass. Dio 37,9,2, 34,3-4. Cicero preferred to portray the ensuing delay of two years as an act of divine providence, since the statue's eventual replacement in December 63 coincided with his exposure of the Catilinarian conspiracy: Cat. 3,20-21; De consulatu suo frg. 10 FLP, 33-65. 279 vision, though nominally still a senatorial prerogative, and financing became subject to the emperor's direct control. 36 It is commonplace that the of administrative organization the major religious festivals of the city of Rome relied on public money. According to the Fasti dating rum, to the Tiberian Antiates ministro- period, between 380,000 and 760,000 HS of public money were spent on the Romani, Ludi the Ludi plebei and the Ludi Apollini respectively. However, to very considerable extent these public funds were sup­ a the plemented by the magistrate organizing the games, since actual costs could amount to up to 3,000,000 HS. 37 On such pub­ an occasion, popular sentiment expected the display of lic munificence from the magistrate, who exploited the op­ portunity for self-representation provided by the office and festival. 38 On a smaller scale, the LUrs expected Ilviri and aediles to spend no less than 2000 HS of their own money Ludi on for the Capitoline triad and Venus as well as for gla­ diatorial shows as summae honorariae, although public funds could be accessed for supplementing the expenses if the mon­ ey was spent on civic sacrifices (chs 70-1). The Lex Taren- tina legally obliged municipal magistrates to spend on lic building and games, but accounted for the possibility that a magistrate would wish to fulfil his 36 37 38 pub­ summa honoraria Suet. Aug. 36-7; ECK 1992; KOLB 1993. Private contrac­ tors: OIL 6,31338a; 15,7241. See e.g. OIL 6,8478, 9078; 10,529; 11,3860, for the combination of public and the princep-s* private funds. Fasti Antiates ministrorum: Inscr. It. 13,2,26 p. 208-9; Cass. Dio 54,17,4; ROPKE 1995b, 274-6. E.g. Cic. Har. Resp. 22-9, on Clodius organizing the Ludi Megalenses as aedile in 56; WISEMAN 1974, 159-169; Id. 1985a, 36. Aedilician munificence did not necessari­ ly lead to political success: GRUEN 1992, 188-95. 280 by using half of the money received from fines exacted on illegitimate handling of public property. 39 position of Such a juxta­ the public treasury's contribution and private munificence, of civic funds and individual sumnae honora- riae, served to diffuse financial responsibilities. In the city of Rome, it was civic munificence that the began emperor's control of to undermine the Republican system of decentralized financial and administrative author­ ity in the organization of civic religion. The of temple organization building and repair illuminates this change. For instance, throughout the Late Republic individual aristo­ crats could be expected to repair religious shrines built by their ancestors, an expectation still applying in the years after Actium and either continued or revived under Tibe­ rius. 40 Cicero in 55-54 was commissioned to rebuild the tem­ ple of Tellus, whose magmentarium had been incorporated in the adjacent house of Cicero's neighbours Clodius and Appius Claudius. The refurbished shrine contained a Cicero statue of Q. and, very likely, references to the consular himself and his achievements. 41 Such practice continued in the year immediately after Actium. Munatius Plancus restored the tem­ ple of Saturn ex manib(iis), as the inscription, carrying the restorer's name, reported (OIL 6,316 = ILS 41). The tem­ ple of Apollo on the border of the Campus Martius, 39 40 41 restored RS 15, lines 27 and 35-8. Cf. ILLRP 648 (Pompeii): ex ea pequnia quod eos (sc. Ilviros) e lege in ludos aut in monumento consumere oportuit; LOMAS 1997, 32-6. E.g. Cic. Verr. 2,4,79-80; Cass. Dio 53,2,4; Tac. Ann. 3,72,1-2. Cic. Quint. Fr. 3,1,14; cf. Har. Resp. 31: aedes Telluris est curationis meae with E. COURTNEY, CR (1960), 198-200. 281 by C. was hence known as that of Apollo Sosianus, Sosius, and the temple of Diana on the Aventine, restored by L. Cornificius after his triumph in 33, was subsequently referred to as that of Diana Cornificia. 42 By contrast, with the rise of autocracy restoration the of temples, just as public building in the city of Rome in general, increasingly became the prerogative of the princeps. When referring to storation of eighty-two Augustus 1 use of Republican senatus) auctoritate in temples city the his of formality constitutional re­ Rome, (ex misrepresented the facts of realpoli- tik: now the princeps could even afford to restore the Capitoline theatre and the theatre of Pompey sine ulla inscrip- tione mei. 43 4.1.3 Decentralizing financial responsibilities It has often been assumed that under the Republic victorious generals fulfilled their vows to the gods by building tem­ ex manubiis on return from their military campaigns. 44 ples view The that construction these of their generals themselves financed the temples could be taken as entailing that with respect to the particular category of Roman votive shrines the Senate's influence crude form, such was marginalized. 43 44 this a view is untenable, since it underesti­ mates the Roman Senate's participation in 42 In the construction Apollo Sosianus: Pliny NH 13,53; 36,28. Diana Cornificia: Suet. Aug. 29,5; OIL 6,29844,2; OIL 6,4395 = ILS 1732; ECK 1984, 139-40. Augustus JRG 19,1-2, 20,1, 20,4; Suet. Aug. 30; Cass. Dio 53,2,4; ECK 1984, esp. 139-142. The Republican pattern of financial decentralization persisted in Italy and in the provinces; cf. RIVES 1995, 28-39. E.g. SHATZMAN 1975, 90-1, passim; PIETILA-CASTREN 26-7, passim. 1987, 282 of temples vowed before or during military campaigns. The Senate was not only responsible for the allocation of to the new sanctuary, but also provided the administrative framework relating to the foundation and dedication temple space concerned. 45 Furthermore, of the only very few Republican temples are known to have been financed from the general's manubiae, whereas a significant number of temples vowed by a general were financed from public funds. In most cases, how­ ever, the actual means of financing remain obscure. 46 At the same time, the generals possessed carte blanche as to the use to which they would put their booty. In effect, part the manubial money was spent on the decoration of the tem­ ples vowed while on campaign. Rather than pression in of the financing finding its ex­ of the sacred site itself, the self-representation of a successful commander in the temple took the form of displaying spoils, of art objects or honor- 45 46 See ORLIN 1997, 139-87, drawing attention to the inter­ nal delegation of senatorial responsibility to the re­ spective committees of duumviri aedi locandae and duum­ viri aedi dedicandae, subject to the Senate's authority. Cf. the Senate's appointment of Illviri for the mainte­ nance of temples, also operating under the Senate's su­ pervision: Livy 24,47,15-6. ORLIN 1997, 117-35, 199-202, drawing up a list whose most prominent feature is the frequency of the entry >not recorded<. The temples securely attested as having been financed from manubial money are (1) Fors Fortuna, financed by the consul Sp. Carvilius Maximus in 293 (Livy 10,46,14); (2) Mars Invictus, by D. lunius Brutus, consul in 138 (Val. Max. 8,14,2); (3) Honos and Virtus, by Marius de manubiis Cimbris et Teuton<ibus> (ILS 59). ORLIN 1997, 130-1, 194 adds three further temples, name­ ly (4) Fons ex Corsica (Cic. ND 3,52); (5) Felicitas ex TOU 'Ipepixou TioXeuou (Cass. Dio 22, frg. 76,2); (6) Hercules Victor (OIL I 2 ,626 = ILLRP 122). Unfortunately, the evidence for (4) and (5) is inconclusive, while (6), the triumphal inscription, positively fails to mention financing from manubial money. 283 ific statues and paintings, including self-portraits, which commemorated the exploits of their dedicators. 47 The Senate's financial and administrative in founding temples participation vowed by individual generals could be interpreted as the state's encroachment upon individual ligious re­ choices and the subordination of personal interests to a system of state religion. That view would be a variant of the modern model of civic religion, which perceives indi­ vidual religious activity in Rome as fundamentally embedded in the civic domain. However, as we have seen, public admin­ istrative and financial involvement as such did not a priori constitute a case of civic encroachment. By contrast, it is noteworthy that the commanders' choices concerning the divi­ ne addressees of their vows were unconstrained by inter­ ference from the Senate. On the contrary, by erecting sacred shrines for deities chosen by individuals the Senate promot­ ed individual religious predilections to the status of munally supported cults. 48 Moreover, the modern concept of >the state<, which informs the notion of a unified ial body com­ senator­ administering >state religion< against the inter­ ests of individual aristocrats, is problematic. The Roman Senate must not be judged on the basis of a modernizing con­ ceptualization of organized state control over society, but could only be as unified as the interests of its members individual would permit. And whereas the civic model construes a development from senatorial authority in religious matters 47 48 Cf. H0LSCHER 1978, 340-6, 344-5; WlSEMAN 1994, 99-100; ORLIN 1997, 132, 135-9. For the generals' authority over manubial money, see SHATZMAN 1972. For the wide range of deities chosen see PIETILA-CASTREN 1987. by the generals, 284 to individualization and fragmentation, the evidence able avail­ suggests that from the fourth century onwards the mem­ bers of a competitive Roman political elite with were concerned their self-representation rather than with the uphold­ ing of a consensus among the upper classes. 49 The idea of senatorial unity, if such a unity existed, has to be located in an even earlier period, where evaluation of the data is based on speculation rather than on sound historical prin­ ciples . The temple of Hercules Musarum in circo Flaminio eluci­ dates the interplay of senatorial participation and individ­ ual independence from senatorial authority. M. Fulvius Nobilior added a portico to an already existing temple of Hercu­ les and, introducing a statuary group of the nine Muses of Ambracia, rededicated the Fulvius' building at sanctuary to Hercules Musarum. the sanctuary seems to belong to the period of his censorship (179-73), and was financed ex pecunia censoria. At the same time, Fulvius' is an instance to turned into a Museion display permanently the booty seized by Fulvius during and after his 49 activity of the aristocratic self-aggrandizement of the second century: the temple was serving building siege of the town of Ambracia in E.g. HOLSCHER 1978, 1980, 1982, for the artistic repres­ entation of individuals from the fourth century onwards. Cf. ILLRP 310; Cic. Cato mai. 61; Livy 8,8,17, 9,1,2; Pliny NH 7,139-40; WlSEMAN 1985b, 3-4; MlLLAR 1989, 148-9, for the third century rhetoric of competition. FLOWER 1996 discusses the use made of ancestors in this process. On the »Selbstverstandnis und Selbstdarstellung der Nobilitat«, see HOLKESKAMP 1987, 204-40, overrating the political elite's homogeneity. See above, 3.1.4. 285 189-87. 50 in addition, Fulvius had Fasti painted on the walls of the sanctuary, which not only included a tradition­ al dedicatory inscription and a list of Roman also introduced Ennian but a selection of the dies natales of temples in the city of Rome compiled by the poet these consuls, Fasti Ennius. No doubt had their chronological culmination in Fulvius f s consulship and in his rededication of the temple of Hercules Musarum: the wall-painting in the sanctuary must be seen as a display of individual ambition in the framework of the politicized culture of the second century. 51 Fulvius 1 temple of Hercules Musarum illustrates the form which the interplay of public authority could take. and individual The generals' priorities lay with self-repres­ entation through temple decoration and not with nancing, independence because temple the financial and administrative involve­ ment of a fragmented Senate provided a welcome framework support, fi­ of rather than impinged upon, individual aristocratic interests. The notion held by some scholars that individual magis­ trates mediated between gods and men as proxies of the Roman state is also problematic. Through the holding of auspicium, magistrates 50 51 and generals took the auspices as representa- RUPKE 1995a, 332-41. Temple: Eumen. Paneg. 9,7,3; Serv. auct. Aen. 1,8. Booty: Livy 39,5,14; Pliny NH 35,66. Games: Livy 39,22,1-2; GRUEN 1992, 195-6. Macrob. Sat. 1,12,6; RUPKE 1995a, 341-68. Invention of consular Fasti: RUPKE 1995c, 199-202. For Ennius as the likely author of Fulvius' Fasti, see SKUTSCH on Enn. Ann., pp. 144-6, 313-4. Ennius and Fulvius: Enn. Ann. 268-86 Skutsch; L. Aelius Stilo fr. 51 Funaioli ap. Gell. 12,4,5; Cic. Tusc. 1,3. Ennius' Fasti became a source of lunius Gracchanus' commentary De Fastis: frgs 1-2 Funaioli ap. Varro LL 6,33-4; cf. Macrob. Sat. 1,12,6. On the genre of Libri fastorum, commentaries about Fasti<, see ROPKE 1994, 125-31. 286 tives of the People and the Senate. Nevertheless, we not be thinking should in terms of an abstract category of reli­ gious charisma: auspicium showed in the magistrate's or general's the taking of individual auspicia publica-, and polit­ ical or military success, linked to scrupulous ritual per­ formance, enhanced the political authority of the respective magistrate or general. The laudatio funebris for L. Caeci- lius Metellus or the votive inscription of L. Mummius stress this relationship between religious obligation and individ­ ual authority: auspicio suo maximas res geri, duct(u) auspicio imperioque eius. 52 Whereas the Roman Senate accepted the financial and administrative responsibility for constructing votive temples, the actual dedicatio was generally, if not invariably, performed by the relatives. And respective commander or his it seems that this special relationship be­ tween the family of the person that had undertaken the vow that the and the deity addressed played a decisive role. Moreover, it was the individual's achievement votive inscription over the temple doors publicized. For example, M. Aemilius Lepidus, Fulvius' colleague in 179, as censor dedicated the temple of the Lares marini in campo, vowed by L. Aemilius Regillus eleven years earlier and ap­ parently financed from public money. Nevertheless, the voti­ ve inscription supra valvas templi publicized Regillus' mil­ itary 52 53 achievements. 53 Apparently, the association of a Pliny NH 7,139-40, discussed above, 3.1.4; ILLRP 122. Cf. RCPKE 1990, 44-5; KUNKEL & WlTTMANN 1995, 28-37. Livy 40,52,4-7: a copy was placed in aede lovis in Capitolio supra valvas. Cf. OIL I 2 ,626 = ILLRP 122, the vo­ tive inscription of the temple of Hercules Victor near the Tiber bank, dedicated by L. Mummius in c. 142. That text does not corroborate the view that Mummius erected 287 temple foundation with the self-representation of a particu­ lar individual preceded the invalidates Late Republic. Such evidence the view that this association of votive temple and individual was a sign of the erosion of state authority in the wake of increasing individualization. 54 Instead, from the early second century onwards these sacred shrines were linked to individuals, rather state, thus than to an abstract seemingly disregarding financial responsibili­ ties. For instance, the temple of Hercules in vowed and Roman Foro Boario, dedicated by P. Cornelius Scipio Aemilianus, was known as the aedes Aemiliana. 55 Whereas that temple was pre­ sumably constructed from public money, the temple and Virtus, its Honos referred to as the aedes Mariana, was financed ex manubiis by the homo novus Marius. 56 That has of latter temple own revealing history: Marius' choice of Honos and Virtus clearly imitated the undertaking of M. Claudius Mar- cellus (cos 222, 215, 214, 210, 208), who restored Q. Fabius Cunctator's temple to Honos and added the cult of Virtus. In Marcellus' new dual temple exactly the same pair of deities as in Marius' aedes was worshipped. The earlier temple was duly dedicated by Marcellus' son, the consul of 196, outside the Porta Capena in 205. The tomb of the elder Marcellus, situated ad Honoris et Virtutis, provided a visual link tween 54 55 56 the be­ two deities in question and that particular ari- the temple from manubial money, as e.g. ZIOLKOWSKI 1988, 309 or ORLIN 1997, 193-4 assume. But the temple could have been a private aedicula: LATTE 1960, 219-20; PIETRILA-CASTREN 1987, 143. For this view, see e.g. ORLIN 1997, 192-8. PIETILA-CASTREN 1987, 134-8; ZIOLKOWSKI 1988, 311-4. JLS 59; Cic. Sest. 116; Plane. 78; Div. 1,59; Vitr. 3,2,5; Val. Max. 1,7,5; D. PALOMBI, LTUR 3 (1996), 33-5. 288 stocratic family. 57 Elizabeth Rawson gested that through his has tentatively sug­ choice Marius insinuated that he wanted to be regarded as a successor to Marcellus in terms of his military success and political prestige, appropriate­ ly expressed by the qualities of virtus and honos. 53 More likely, by choosing Honos and Virtus Marius made a more neral ge­ political claim. The two deities in question belonged to those deified abstractions that embodied on a divine pla­ ne the competitive value which was system individual excellence instrumentalized by Rome's aristocratic elite. 59 The abstract constituents civic of worship (though of of that value system received course instituted by individual initiative) from the late fourth and the third century on­ wards. 60 By making his choice the homo novus Marius demanded recognition as a member of that very elite. Furthermore, the construction of his temple from manubial money may document some opposition from the political establishment as well as defiance on the part of Marius. The displacement of state authority onto individual the level of responsibility in matters of sacred building was the result of the notion that the erection or restoration of public monumenta would enhance 57 an individual's honos and 58 Monumenta: Asconius 12 C. For the deities, see WISSOWA 1912, 149-50, and for the temple D. PALOMBI, LTUR 3 (1996), 31-3. RAWSON 1991, 158-63, at 161-2. 6° Cf. HOLSCHER 1978, 348-50; HOLKESKAMP 1987, 238-40. 59 Cf. HOLKESKAMP 1987, 208-21 with further literature. That is not to say that the shared communication about this value system created a homogenous elite: see above, 2.5, 3.1.4 and passim. 289 This decentralization of the state's responsibi­ memoria. 61 lities in the final century of the Late Republic, the corol­ lary of the euergetism of a competitive aristocratic as well as the elite, subsequent reversal of that process under autocratic rule is well documented. A famous example is Pom­ pey's theatre-temple in the Venus Victrix Campus dedicated Martius, to in 55, whose architectural design included a statue of Pompey which carried a globe and was surrounded by thus fourteen nationes, portraying Pompey's conquest of these regions and the promotion of the imperium populi Romani to the margins of the orbis terrarum. The theatre was known as the theatrum Pompei. 62 Similarly, the temple of Venus Genetrix, the ancestral goddess of the Aeneadae lulii, was the centre-piece of the architectural design of Caesar's Forum lulium. 63 On a less conspicuous plane, there existed many more instances of a deity or temple being linked to the building activities of individuals. 64 61 62 63 64 Paulus Festus 123 L: monumentum est ... quicquid ob memoriam alicuius factum est, ut fana f portions. Cf. Cic. Verr. 2,4,69; WISEMAN I985a, 20-1. Tac. Ann. 3,72,2; Pliny NH 36,41; Suet. Nero 46,1; SHATZMAN 1975, 392; COARELLI 1996, 360-81. In Pompey's triumph of 61, statues of the OLXOTJUEVTI and the conquer­ ed regions had been displayed: Cass. Dio 37,21,2; Pliny NH 7,98, Plut. Pomp. 45,2; App. Mithr. 116,568; WEINSTOCK 1971, 35-59. WEINSTOCK 1971, 80-90; ULRICH 1994, 117-55. E.g. Porticus Metelli, constructed by Metellus Macedonicus (praetor 148) and renamed as Porticus Octaviae fol­ lowing its Augustan restoration: Cic. Verr. 4,126; Veil. 1,11,3; PIETILA-CASTREN 1987, 129-34. Diana Planciana, presumably relating to the aedile of 54: PANCIERA 1970-71, 125-34; SHATZMAN 1975, 388; BEARD 1994a, 737; L. CHIOFFI, LTUR 2 (1995), 15. Delubrum Domitii, erected by Domitius Ahenobarbus in 49: Pliny NH 36,26. Hercules Pompeianus in the aedes Pompei Magni, the restored tem­ ple of Hercules Invictus ad Circum Maximum: Pliny NH 36,26; Vitr. Arch. 3,3,5; ZlOLKOWSKI 1988, 313; F. COARELLI, LTUR 3 (1996), 20-1. Hercules Sullanus: D. PALOMBI, LTUR 3 (1996), 21-2. 290 It would be disproportionate to the realities of financ­ ing temples and cults in the city of Rome their to believe that maintenance ever relied solely on the civic domain or on elite euergetism. Wishing to set an example to low-citizens, Augustus while Roman temples. 65 By contrast, implying that the Roman custom of collecting stipes was wide-spread, accepted it Mater fel­ used the stipes collected among the city population to restore Cicero, his for the cult of the magna, yet wanted to have such practice banned in the case of other cults, because he not only believed lection the col­ of money to be reflective of superstition, but also thought that it was economically detrimental. 66 More impor­ tant than such occasional collecting of stipes was the regu­ lar income a temple received. The statute of the temple at Furfo made detailed provisions regarding the its financial property With the to which should be put, thereby, as we have seen, guarding itself against private priation. use and civic misappro­ same protectionist intention, the LUrs ruled that a stips given to a temple had to be used for that temple and its cult only (ch. 72). The case of the temple of Asclepius on the Isola Tiberina mentioned above shows that such stipes, comprising votive offerings and fees, could be considerable. of the The flourishing healing cults of Asclepius or Minerva Medica in the second and first centuries also meant that a large number of worshippers brought massive financial income, which remained 65 66 CIL 6,456-7; 6,30974; Suet. Aug. 57,1, 91,2. Cic. Leg. 2,22 2,40; Ovid Fasti 4,350-2; Apol. 13,6. In general, see DESNIER 1987. cf. Tertull. 291 the property of these temples. 67 A temple like the shrine of Clitumnus near Spoletum prospered mainly due to its urban clientele, wealthy as the existence of sacella, inscriptions and material dedications suggests. Presumably it was the temple's function as an oracular shrine that attracted urban worshippers. 68 >Traditional< Roman temples could also rely on income through private votive dedications: togae praetextae et undulatae literally covered the golden cult statue of Fortuna Virgo in the Forum Boarium, while on the occasion of someone's birth, his reaching adolescence or his death luno Lucina, luventas and Libitina could expect to receive stipes from relatives. 69 The great Italo-Roman shrines such as the temple of luppiter optimus maximus Tibur, the sanctuaries at Praeneste and Nemi were renowned for their wealth in land, material publicae or dedications charged and individuals money. 70 for Moreover, aedes the use of the temple's infrastructure. Worshippers had to pay for what was on of­ fer, both in terms of temple personnel and the paraphernalia required for daily cult practice, which ranged from the sup­ ply 67 68 69 70 of sacrificial animals to the provision of warm water. On the prominence of these healing cults, see GUARDUCCI 1971; Roma medioreppublicana 138-48. For the restoration of the temple of Asclepius in the mid-first century, see D. DEGRASSI, Athenaeum 65 (1987), 521-7. Pliny Ep. 8,8,5-7; BEARD 1991, 39-44; NORTH 1995, 136-7. Rome herself never possessed a cultic site where priests used sortition to divine the gods' will, but the prac­ tice itself was well known: e.g. Plaut. Cas.; Cic. Inv. 1,101, Div. 1,34, 2,85-8; Varro LL 7.48. Calpurnius Piso frg. 14 P ap. Dion. Hal. 4,15,5. Fortuna Virgo: Ovid Fasti 5,670; Pliny NH 8,194, 197; Arnob. Adv. Nat. 2,67; BOMER on Ovid Fasti 6,569; GLADIGOW 1994, 13-4. E.g. App. BCiv 5,24,97, 5,27,106; BODEI 51-4; BLAGG 1986, 214-8. GlGLIONI 1977, 292 In the words of their Christian critics, the pagan gods were venales - >marketable< and >venal<. 71 This decentralized system of financial economic entities semi-detached from, rather embedded ually in, administra­ created a >market< of small religio- responsibilities tive and concept­ than civic system. 72 The prosperity of the these entities was contingent upon the attraction of a par­ ticular cult or divinity to worshippers, and depended on the continuous economic support of the temple's wealthy urban clientele, whose contributions sustained the sanctuary. the later second and For first centuries, this dependence is documented by the demise of those extra-urban sanctuaries of framework Italy that failed to achieve integration into the of a local pagus or to come under the patronage of the urban once elites, urbanization and migration transformed the Italian landscape. 73 But the same kind of dependence applied in the city of Rome. Following Augustus' dedication temple of of luppiter Tonans in the vicinity of luppiter opti- mus maximus, the old luppiter, appearing to the princeps a dream, in complained that competition with the new cult de­ creased his own revenues; as a result, the imperial age the patron­ of luppiter optimus maximus immediately re-intensified. In his temple at Pompeii, the old god was less fortunate, since the cult statue of luppiter optimus maximus was repla- 71 72 73 Tertull. Apol. 13,6: exigitis mercedem pro solo tempi!, pro aditu sacri; non licet deos gratis nosse, venales sunt. For a list of fees, see e.g. OIL 6,820. For the notion of the market model of religion, see BERGER 1965; GLADIGOW 1990, 239-41. E.g. LA REGINA 1976, 223-9; LETTA 1992, 110-24; NORTH 1995; COARELLI 1996, 328-33; CRAWFORD 1996, 427, 430-3; E. CURTI & al., JRS 86 (1996), 179. 293 ced by a cult statue of luppiter Tonans and subsequently deposited in the temple favissae. 74 The market model allows us competition between to conceptualize both the different religious choices, cults and gods and the disappearance of some of these choices as natu­ ral processes in a about self-regulating system. Elite the demise of traditional gods and their cults in the Late Republic must be seen in this context of the system's constant optimization religious in times of changing fash­ ions. 75 Cicero's complaints about the desuetude the laments into which auspicia privata apparently had fallen by the mid-first century should not be taken as a decline symptom of the supposed of religious practice. The appearance of more expe­ dient methods of divination meant that augury was in danger of losing out to the competitive services offered by profes­ sional haruspices, astrologers or harioli. 76 By contrast, the model of civic religion, based on the gious centralization, notion of reli­ is incapable of accounting for these decentralized processes of choice, fashion and competition between individual Roman deities or their cults and temples. 4.1.4 The limits of state interference As regards the suppression of cults, the Republican Senate's occasional encroachment upon this market 74 75 76 Suet. Aug. 91,2; MARTIN 1988, 255. of cult alterna- E.g. Varro RD frg. 2a Cardauns; LL 6,19; NORTH 1976, 11-2. Disappearance of auspicia privata: Cic. Div. 1,27-8, 2, 73-4; ND 2,9; DUMEZIL 1970, 618-20; RAWSON 1991, 152. Greater expediency of the new forms of divination: ROPKE 1995a, 577-8. Haruspices: e.g. Plaut. Amph. 1132-3; Cato Agr. 5,4; ILLRP 186. >Prophets<: WlSEMAN 1994, 49-67. 294 unsystematic and followed the constraints of Ta- was tives gespolitik rather than religious the under time, policies. 77 at civic is to not same prescribed particular cults. 78 deny that under certain circumstances the Senate demanded collective worship. Under the Republic, occasions which on the rituals and festivals or necessitated the participation of Roman citizens in This the there was no official religious Republic calendaric document which either positively attendance At the the active participation of Romans or Italians was required appear to have been restricted to supplicationes. At stake was the communal pax restoration of the deorum through the procuration of portents, once it had been established that these portents related to the populus Romanus as a whole rather than to a particular individual. 79 According to Mommsen, such state encroachment upon Roman citizens was foreign to the formalism of old Republican ligion, where magistrates and priests conducted rituals and sacrifices on behalf of their fellow-Romans. In marked trast on supplicationes, 79 conducted Graeco ritu or the precedent of Greek ritual tradition and in­ volving the populace, 77 78 con­ to these religious traditions, from the third century onwards public moulded re­ would herald the Hellenization of See above, 2.7.1. Cf. RUPKE 1995a, esp. 283-6, 366-8, rebutting the view that the Late Republican Fasti Antiates maiores repres­ ent a religious document reflective of the traditional Roman state religion. E.g. Livy 7,28,7-8: non tribus tantum supplicatum ire placuit sed finitimos etiam populos, ordoque Us quo quisque die supplicarent statutus; 21,62,6-9: supplicatio ... universe populo; 27,4,15; 32,1,14; 34,55,4: edicturn est ut omnes qui ex una familia essent supplicarent pariter; 40,19,5: per totam Italian. Cf. LUTERBACHER 1904, 29-34. 295 Roman religion. 80 However, as we have seen above/ the modern classification of supplicationes as Greek rites is problem­ perfor­ atic, because such expiatory rituals were sometimes suggestion the at med of the X(V)viri, but could also be proposed by the pontiffs or the haruspices. 81 Moreover, wide communal participation was also rituals Roman stipulated in traditional of procuration such as the sacrum novendiale or the ver sacrum, the offering of the produce of one entire Rom. spring to luppiter. 82 Dionysius of Halicarnassus (Ant. 1,6,1) or Strabo (5,250) saw the supposed Greekness of these rituals as proof of their thesis that Rome was a Hellenic community, but it is unlikely that their contempora­ Roman ries or Romans of the later third and second centuries would have reached a similar conclusion. Rather, the state's ex­ and ceptional encroachment upon its own citizens Italians can be explained the by the same proportionality in religious thinking that we encountered rituals upon above. these Since responded to portents addressing the entire populus Romanus, the Roman Senate resorted to increasing the number of possible participants: the more people involved, the more successful the procuration would be. 83 With the rise of autocracy religious civic interference in the life of Roman citizens changed. The Roman Senate, rather than responding to prodigies that concerned the popu- 80 E.g. MOMMSEN 1899, 568; WARDE FOWLER 1911, 226-7; Wisso- 81 WA 1912, 399-400; LATTE 1960, 242-63; DUMEZIL 1970, 567-71, at 569: »Rome frees herself in this way from the rigid confines of formalism*. Cf. NORTH 1989, 604-5; GWYN MORGAN 1990, 32-3. See above, 3.2.2. S3 22,10,1-6; LATTE I960, 124-5; RADKE 1980, 110-6. cf. above, 2.7.2. 82 Livy 21,62,6: prope tota civitas opera ta firit; 296 re­ lus Romanus, involved the population of Rome in rituals lated to the emperor and his family. The triumviral edict of 42 BCE, which ruled that all citizens celebrate the birthday of Divus lulius or otherwise be accursed and, if of senator­ ial status, fined (Cass. Dio 47,18,5) is an example of this for development. The specifications concerning the funerals the of members imperial family, namely the Parentalia for Lucius and Caius Caesar or the honours for Germanicus, ruled the closing of temples and positively encouraged public par­ ticipation at the funeral ceremonies. 84 Yet, the success the of state's interference with the religious life of the in­ habitants of the city is pation another matter. preoccu­ Livy's with the frequency of attendance at collective reli­ participating would have varied. On the same principle, Au­ gustus stressed the exceptional Roman those of gious rituals implies that in reality the numbers frequency with which all citizens had made supplications to the gods on behalf of his health. 85 It is noteworthy, however, that senatorial encroachment concerning religious limited to calling upon the elite's or imperial activity very often was participation. As we have seen above, the (alleged or actual) neglect of a family sacrifice by an equestrian was instrumentalized by Cato the Elder when expelling L. Veturius from the equestrian Although 84 85 it order. seems unlikely (given, as we have noticed, the Parentalia: JLS 139-40; SCHEID 1993b. Germanicus: Tabula Siarensis frg. b col. I (RS no. 37), lines 1-14, model­ led on the Parentalia for L. and C. Caesar; Tabula Hebana (= RS no. 37), lines 54-62. Cf. ARCE 1988; LEBEK 1989; Id. 1993. Livy 3,63,5; 5,23,2-3; 10,23,2; 27,51,8-9; gustus RG 9,2. 45,2,7. Au­ 297 Roman distinct legal status of sacra privata in the that law) latter's failure to perform a family rite could be used among the censorial charges which led to Veturius' expulsion of illustration further doubt no from the ordo, Cato this used that for incident a equestrian's moral debauch­ ery. 86 As regards the civic realm's encroachment upon mem­ of the elite when the Roman People, rather than an in­ bers dividual's religious behaviour, were concerned, mention has already been made of the civic authorities ruling in respon­ se to that portents the Roman upper class matrons should to from their own funds contribute stipes, Regina. luno cuncti The consuls (Livy magistratus triumviral 42 of edict of be offered to 169 demanded sacrifice from Disobedience 43,13,8). resulted to the fines for senators, in whereas the rest of the population merely incurred curses. imperial funerals in specifications The concerning the Tabula Hebana contented themselves with monitoring the equites Romani, even though the absence could be excused by of individual knights ill health or the obligation towards domestica sacra. Q7 The fact that civic control would often be limited to the Roman elite betrays a lack of interest in systematically imposing 86 87 elite expectations on the urban population at Cato frg. 72 Malcovati 4 , cited above, 3.2.2. The wording of the text (... quod in te fuit ...) may suggest that Veturius 1 neglect of the sacra privata was his own busi­ ness, since in terms of sacred law Veturius himself, rather than the Roman People, would be held accountable by the gods: sacra ... capite (sc. Veturii) srancta. For this common legal distinction, see above, 4.1.1. RS no. 37, lines 55-6: ... gui ordini[s equestris erunt ...J ... gui eor(um) officio fungi volent et per valetudinem perq(ue) domestic[a sacra officio fungi poterunt; Cf. the Rome frg. (b) (= RS no. 37), line 1. 298 large. 88 Moreover, a look at the organization of large scale models religious events exposes the severe limits of modern which the claim of religious behaviour at centralization genres of Augustus, from the and the code instruc­ receiving college of XVviri announced the and sacrifices and Ludi in a contio programme the which applied in a city of nearly communication 1,000,000 inhabitants. Consequential on tions document 17 of Rome. The acts of the Ludi saeculares ritual the published of behaviour at the rituals in an album for further dissemination: sacrificium saeculare ludosque ... de ea re quae more exsemploque maiorum in contione palam ediximus ... item in albo posuimus uti si qui a contione afuissent aut non satis intellexissent ... cognoscerent quid quemquam eorum quoque die facere oporteret. 89 Contio and album are the instruments through which religious information was transmitted in nouncement an oral society. legal at the Ludi this occasion, of the city saeculares could not be taken for granted. Rather, the functionaries of civic nizing suggests and religious information of this kind was nor­ mally passed on orally. 90 Yet, the attendance population an­ of feriae by the Rex sacrorum to a Late Republi­ can urban populace on the Nones of each month also that The religion orga­ the XVviri, resorted to competitive means of attracting potential attendants. The singularity of the occasion was stressed: quod tali spectaculo nemo intererit, iterum neque ultra quam semel ulli mor[talium eos spec- tare licet ludos] (lines 54, 56). The freeborn population was encouraged to come in great number (10, 21). Trombonists 88 89 90 Cf. above, 3.2. PIGHI 1965, 110,24-8; CANCIK 1996, 108-9. Varro LL 6,28; Macrob. Sat. 1,15,12-3; RuPKE 1995a, 212-4. Oral culture: HARRIS 1989, 159-284; SCHEID 1990. 299 were employed for advertising purposes (86-8). Fumigants for purification were freely distributed in front of the temples of luppiter optimus maximus and luppiter Tonans on the Capitoline hill and that of Apollo on the Palatine (29-33). Cer­ tain religious and legal restrictions applying to the at­ tendance of rites were temporarily many as freeborn success that so same time, priestly the in instrumental!zing marketing strate­ gies caused the partial failure to cater for the demands the as possible would be able to attend propter religionem (52-7, 110-4). At the organizers' abolished, of city population: due to the need for the fumigants dis­ tributed in the city, the XVviri had to stipulate that noone or be permitted to collect the fumigants more than once to send their wives (64-6). 4.2 Public and private As we have seen, the use of the market alternatives model religious of entails competition between the religious ser­ vices offered by the Roman state and those offered by indi­ vidual temples and cults. In addition, the foregoing discus­ sion has made clear that the consumers' demands would not necessarily be controllable by temples or by the civic au­ thorities. At the same time, the individual consumer's needs were not catered for exclusively by the religious services which these public competitors had to offer. They provided the infrastructure for what Roman terminology would call the publica sacra, but did not organize the private sacra of the inhabitants of the city of Rome. Roman sacred law distingu- 300 ished the domains of publicum and privatum: publica, loca when being dedicated to the gods by magistrates on behalf of the populus Romanus, became loca sacra. Loca privata, by contrast, when being dedicated by private loca came religiosa. individuals, Privata religio, the private cult of the the Di manes at the tomb, while never attaining of be­ status public cult, was nonetheless protected by the state: any legal impingement was subject to pontifical Under jurisdiction. 91 the Republic, public encroachment on religious behav­ iour remained limited to the domain of publicum. Even the imperial Lex Hebana, when monitoring the public behaviour of the Roman equites, excluded the domain of domestica sacra, i.e. the privata sacra, from its encroachment. If individual religious obligations and public duty were in conflict, dividual in­ obligations prevailed; in the LUrs, the magistrate was excused from conducting the trial arising out of a multa on behalf of his community if the following circumstances applied: Si Ilvir praef(ectus)ve, qui earn rem colonis petet, non aderit ob earn rem quot ei morbus sonticus vadimonium iudicium sacrificium funus f ami Hare feriaeve de<n>icales erunt quo minus adesse possit ... 92 The domains of private law (vadimonium and iudicium) and private religion of thus were protected against the civic do­ main. Here, we can see how a Late Republican administrative document conceptualized the differentiation of the domain of privatum from the publicum and defined it as comprising both 91 92 Cic. Ear. resp. 9, 14, 30, 32; Gaius Inst. 2,5-6; Ulpian Dig. 1,8,9 pr.; Marcianus Dig. 1,8,6,3. Privata religio: Cic. Leg. 2,58; KASER 1978; CRAWFORD 1989, 95. Pontifi­ cal jurisdiction: Aelius Callus ap. Festus 348-50 L; OIL 10,8259; BEHRENDS 1978. Ch. 95, lines 19-24. 301 the secular realm and the realm of the private religion of the Roman family (sacrificium funus fami Hare feriaeve denicales) . 4.2.1 Privata religio The discussion of the preservation of sacra privata most is the substantial part of Cicero's constitutio religionum in book two of De legibus (2,22, 47-58). Cicero's obvious con­ cern about the preservation of family cults finds its paral­ lel in his appeal to religious sentiment in the case of Clodius' transition to the plebs by way of adoption in 59. Clodius was sui iuris, that is a pater in his own right, and thus responsible for the family sacra. Since the process adrogatio would of in such a case entail that the adoptee ab­ andoned his sacra familiae in favour of those of the adopt­ er, this special form of adoption could not take place with­ out the approval of the People. This complex legal procedure documents the Roman anxiety about preserving the cults of the families of both the adoptee and the adopter. Cicero insinuates that Clodius' transition to the plebs implied the demise of the family sacra, but it is unclear whether Cice­ ro's narrative actually corresponded with the facts. 93 In De legibus, Cicero concurred with the law that position of pontifical the maintenance of sacra familiae, comprising the preservation of its feriae and caerimoniae, must be linked to, rather than dissociated from, the acceptance of inherit­ ance. 93 His insistence on linking heirdom to the duties of Cic. Dom. 34-9, 77; AdAtt. 2,7,2, 2,12,1; VERNACCHIA 1959; CLASSEN 1985, 235-8. Adrogatio: Gaius Inst. 1,99; Cell. 5,19,8. 302 burial and commemorating the deceased accorded with the views expressed by imperial jurists. Similarly, popular sen­ timent associated the preservation of family sacra with the acceptance of inheritance - and vice versa. 9 * Cicero's emphasis on the hereditary principle of succes­ sion shows that to his audience the familiae preservation of was a matter of hereditary affiliation (not neces­ sarily entailing that heirs were relatives) rather agnatic sacra than of family lineages. These family cults operated in the realm of small social units rather than of large gentes that would organize and maintain the cult. This observation curs with the fact that in Roman society the >nuclear fami­ ly^ consisting of established as father, mother and children, the also be expressed imperial jurists who expected heirs or close family members to preserve tombs or family sacra. Only of can the norm, at least among the upper classes. The principle of hereditary affiliation was by con­ if neither these two groups were willing to enter into the inherit­ ance, or if they were non-existent, then sors intestate succes­ or cognate kins fulfilled the family obligations. 95 In order to ensure the ritual commemoration at the their death, testators tomb after made specific regulations in their wills relating to the maintenance of tombs and cult; once an heir entered into the inheritance, he was expected these to obey regulations. Collegia performed similar, if much more limited, services for their members, ranging from small 94 95 Cic. Leg. 2,47-53; Ulpian Dig. 11,7,3-5. SHAW 1984, 125-7. E.g. Ulpian Dig. 11,7,12,4. SALLER & SHAW 1984, 125-6. Cf. HOPKINS Cf. SALLER 1983, fi- & 205-6; 303 nancial contributions in form of funeraticia to the full maintenance of tombs and religious ceremonies for those left legacies who and bequests or named the collegium as their heir. 96 The Roman anxiety about posthumous commemoration the context of in the religious duties of families and heirs (privata religio), which these practices express, cannot be integrated into a model of homology between religion and the state. 4.2.2 Roman religion, families and individuals The legal distinction between public and private cult vities was complemented by the differentiation of feriae publicae and feriae privatae. The civic acti­ former were subject to authority: they were announced by the Rex sacrorum on the Nones of each month. The latter were in the lity of families and responsibi­ individuals. Roman law divided them into feriae fami 11 arum and feriae singulorum: Sunt praeterea feriae propriae familiarum, ut familiae Claudiae vel Aemiliae seu luliae sive Cornell ae et siquas ferias proprias quaeque familia ex usu domesticae celebritatis observat. Sunt singulorum, uti natalium fulgurumque susceptiones, item funerum atque expiationum. 97 First, feriae familiarum. The history was of cults seen by some scholars as a process in which the rise of the archaic city-state undermined the gentes and gradually social functions of took over their cultic responsibili­ ties. This view is unacceptable. To be sure, in the family some cases civic realm preserved family rites by transforming them 96 DE VISSCHER 1966, 189-207; ANDREAU 1977; AUSBOTTEL 1982, 97 Macrob. Sat. 1,16,7-8; cf. Festus 282,14-16 L. 59-71; CHAMPLIN 1991, 155-182; SALLER 1994, 155-180. 304 into public cults, when the families concerned became ex­ tinct. Thus the worship of Hercules at the ara maxima by the Potitii and Pinarii passed into the hands of the state in 312. Or the civic authorities could reward a family for services to the community by administering cultic obliga­ tions on behalf of the family. Thus the family sacrifice the Octavii to Mars of in Velitrae became a civic sacrifice which accorded the Octavii a place of honour. 98 But only its it was with the insight that the rise of the gentes and their socio-religious institutions must be linked to the rise of the city-state that the independence of their religious tra­ ditions from the alleged encroachment of the state could be assessed more precisely. 99 The ambiguity of the literary sources gious about the reli­ role of the archaic gentes is illustrated by an inci­ dent taking place during the Gallic invasion. On Livy's count, C. Fabius Dorso, despite the siege manages to traverse to the Quirinal to perform genti ac­ of the Arx, a sacrifice Fabiae on a fixed day in obligation to the family sa­ cra. In contrast, Cassius Hemina, less inclined to follow family tradition, homogenizes the narrative by claiming that this sacrifice was performed on behalf of the People by Fa­ bius Dorso in his capacity as pontifex (maximus). It is per­ ilous to hypothesize about Cassius Hemina's motives: it could be that his narrative connected the liberation of Rome from 98 99 the Gauls with the exemplary pietas shown by a civic Hercules: Livy 1,7,12; Verg. Aen. 8,268ff.. Octavii: Suet. Aug. i. Cf. A. v. BLUMENTHAL RhM 90 (1941), 317-322. Cf. CORNELL 1995, 81-6. 305 priest to the gods. 100 The historicity of this event is no doubt disputable. However, its conceptualization in the Late Republican and triumviral period demonstrates to what extent sacrificia privata constituted a realm in their own right. If we accept Livy's account, Fabius Dorso temporarily disso­ ciated himself from warfare and civic obligations to his obligations in the private context of family religion. As the passage from the LUrs mentioned the fulfil above demonstrated, obligation to perform a private sacrifice was safeguar­ ded by the city. 101 Yet, the nuclear character of the upper class Roman mily fa­ and the hereditary principle of succession entail that the received concept of gentilician religion becomes proble­ matic when applied to the second and first centuries. over, as regards the feriae familiarum mentioned by Macro- bius, one should not necessarily be gentilician religious thinking means nuclear household. elite Upper-class of were families blended the idea of nuclearity with the deliberate ex­ tension of family relations. As a result, the class terms of increasing a network of family alliances that complemented the thus in continuity. In Late Republican Rome, divorce and serial marriage among the political one More­ family Roman upper- was a fluid structure of familial dislocation 100 Family sacrifice: Livy 5,46,2-3, 5,52; cf. Val. Max. 1,1,11. Cassius Hemina frg. 19 P ap. App. Gall. frg. 6; cf. Plut. Cam. 21,3; Cass. Dio 7,25,2. On Cassius Hemi­ na, see RAWSON 1991, 246-257, at 256. 101 Cf. RUPKE 1995a, 501-506, at 502. For the synonymity of familia and gensr in such contexts, see SALLER 1994, 78-9. 306 through divorce, just as the immediate underwent household constant restructuring through remarriage. 102 This is not to say that the category of gentilician re­ ligion becomes obsolete. In some form, gentilician religious traditions (as opposed to traditions preserved in the nu­ clear family) continued. For instance, in century the late second the genteiles luliei dedicated an altar to Vediovis pater in Bovillae. 103 An annual gentilician sacrifice appar­ ently took place in the sacellum of Diana in Caeliculo. though we do not know which gens was involved, Cicero im­ plies that the significance of the realm of Al­ the this sacrifice transcended nuclear family. As part of the Senate's action against the collegiate structures associated with the cult of Isis Capitolina, L. Calpurnius Piso's destruction of the shrine in 58, to which Cicero refers, may have responded to collegiate activities in the area. Cicero Yet, creates pathos among his audience, the Roman Senate, by pointing out how the consul impinged on the gentilician religious tradi­ tions of those present. 104 However, while the relationship between various Roman gentes and their ancestral deities has been collected, 105 and while this material has been used to construct a picture of gentilician religion in the Late Re­ public, the actual commitment of individual aristocrats to a 102 Cf. BRADLEY 1991, 130-9, 156-76; 1994, 219-23; above, 3.1.4. 103 ILLRP 270; WEINSTOCK 1971, 5-12. CORBIER 1991; SALLER 104 Ear. Resp. 32: L. Pisonem quis nescit ... maximum et sanctissimum Dianae sacellum in Caeliculo sustulisse? Adsunt vicini eius loci; multi sunt etiam in hoc ordine qui sacrificia gentilicia illo ipso in sacello stato loco anniversaria factitarint. For the shrine, see D. PALOMBI, LTUR 2 (1995), 13-4. Destruction of collegiate structures in 58: below, 4.1.3. 105 E.g. WISSOWA 1912, 33, 404, passim; WEINSTOCK 1971, 4-5. 307 great gentilician tradition is obscure. By contrast, from the generation of Marius and Sulla onwards, the reverses coins those show of deities with whom individual aristocrats has process parallel wished to be associated. 106 Above, a been observed in relation to temple foundations: individuals would exploit the opportunities of a market of cultural and needs. the Given nature of Roman family relations, it is therefore not surprising that on drawing own their religious choices, but would accommodate them to the political elite, while gentilician lineages and their religious tradi­ immediate tions, was nevertheless mainly concerned with the family or with individual self-aggrandizement. of The institutionalized occasions cult, the feriae also singulorum, individual private on nuclear focussed groups, rather than on a wider public. To events of commemoration be ritual sure, such as birthdays or funerals and their anniversaries involved the participation a of large number of people. While a person's dies natalis was a matter of individual concern, its celebration, dependent on his social status, embraced family members, friends and clients. Individuals expressed the bond of pietas towards friends superiors by commemorating their birthday. Horace, for in­ stance, observed the birthday of Maecenas with on the or a sacrifice Ides of March (Carm. 4,11). The dies natalis of the dominus constituted feriae for his slave familia. 107 Despite the involvement of varying social groups, however, it is 106 E.g. CRAWFORD, RRC 1,502-511, passim. Marius and Sulla: LUCE 1968. Individual aristocrats and >their< deities: JAL 1961; Id. 1962. 107 Tib. 2,2,5-6; Hor. C. 3,17,14-16; Sen. Ep. 110,1; NH 2,16; SCHILLING 1978; ARGETSINGER 1992. Plin. 308 noteworthy that an individual's birthday had the status of feriae privatae. It is not before the early imperial period that the civic authorities undermined this legal distinction between the domains of publicum and privatum. In 8 BCE, the Roman Senate decided to celebrate the birthday as a of Augustus public event, at which Ludi circenses were performed. Hence, the birthdays of feriae, the principes constituted public and no longer feriae that had to be observed by one particular familia, by friends or clients only. 108 Birthdays constituted only one of the many alized occasions on which Romans would be able to celebrate festivals relating to the family. 109 duals created their Macrobius of what In addition, indivi­ own ferialia as a reflection of their life cycle, which transcend the legal in institution­ definition preserved might count as feriae privatae. For instance, Cicero celebrated the anniversary of the Nones December in commemoration of his suppression of the Catili- narian conspiracy in 63. As the flame flaring up during rites for have the the Bona Dea in the consul's house on that night had been interpreted as a sign of may of celebrated the divine approval, Cicero occasion with a sacrifice to the goddess. 11 ° 108 Cass. Dio 55,6,6, passim; Fast, fratr. Arv. on September 23 = Inscr. It. 13,2, p. 35. 109 Cf. NICOLAI 1968; HARMON 1978, for an impressive list of festive events. 110 M. lunius Brutus ap. Cic. Ad Brut. 1,17,1: non omnibus horis iactamus Idus Martias similiter atque ille Nonas Decewbris suas in ore habet; Ad Att. 1,19,6. Divine ap­ proval: Plut. Cic. 20,1-2; Cass. Dio 37,35,4. Cf. MOREAU 1982, 15-9. 309 Normally, however, individuals would tend to synchronize the by private ritual celebrations with the dates provided city's Fasti. The bachelor Horace celebrated the anniversary of his rescue from a falling tree on the day of the Matronalia (Martiis Kalendis, C. 3,8). The feriae and caerimoniae days: of of the funerary cult centered on a limited number on the Parental!a in February, the dies violae in March, the in May, and on the birthday of the deceased. Testa­ Rosalia tors would choose from and combine these ferialia therefore became dates. Non-public an integral part of the private foundations of the funerary cult (e.g. OIL 5,4489). The tak­ ing of the toga virilis took place on the day of the Liberalia (17 March). Cicero, writing February 50, required from information his concerning 20 on province the day on which the Liberalia would fall: Quinto togam puram Liberalibus cogitabam dare (mandavit enim pater). was necessitated by Cicero's inquiry a lack of information about the exact how­ date of intercalation in Rome. His pragmatic solution, shows ever, that he did not feel a particular religious obligation to follow the Fasti of the city of Rome, but rely wished to synchronize date and event: ea sic observabo quasi intercalation non sit. The public calendar welcome temporal, ted. 111 Private a private ceremony opera­ events with a religious significance could be structured by the administrative the provided but otherwise indeterminative, framework within which the performance of a which me­ and temporal patterns city provided. However, the mere synchronization 111 Att. 6,1,12; ROPKE 1995a, 292-5. Cf. Cic. Att. 5,21,14 (13 February 50) : cum scies Romae intercalation sit neenet velim ad me scribas certum quo die mysteria futura sint. 310 of events does not prove the homology of private and religion. Choosing public days that the public Fasti earmarked as days of rest (feriae) was a matter of convenience. It served to structure one's own life cycle and helped to it with the social, economic or religious obligations of friends and relatives. The content of gious these reli­ not contingent the physical city. Despite the synchronization of pri­ vate ritual and civic calendar, it is in the private events remained dissociable from a civic context and, as the Ciceronian inquiry demonstrates, was upon co-ordinate differentiation of private cult these and areas that civic religion should be located. 4.2.3 Collegiate associations and publica sacra The foregoing discussion has employed Late Republican definitions to legal demonstrate that Romans would have concept­ ualized sacrum and publicum as well as publicum and privatum as interrelated, yet at the same time distinct, domains. Conversely, this paragraph will suggest that a predominantly legalistic approach has its limitations. Returning to the religious behaviour in the public domain, it will question the criteria that define >public< religion at Rome. Publica sacra quae publico surnptu pro populo fiunt, quaeque pro montibus pagis curis sacellis; at privata quae pro singulis hominibus familiis gentibus fiunt. 112 Comparing the definition of popularia sacra by the Augustan jurist Antistius Labeo, Wissowa identified Festus' sacra pro montibus as the Septimontium, the sacra pro pagis as the 112 Festus 284 L = Ateius Capito suppl. frg. 70 Cf. Macrob. Sat. 1,16,4-7. Strzelecki. 311 Paganalia, the sacra pro curiis as the Fornacalia f and those pro sacellis as the Compitalia held in the vici: Popular!a sacra sunt, ut ait Labeo t quae omnes cives faciuntt nee certis f ami His adtributa sunt: Fornacalia, Parilia, Laralia, Porca praecidanea. 113 Labeo and his anonymous contemporary used in Festus 1 defini­ tion of publica sacra defined civic religion as the religion of the Roman People in their entirety. To both of them, the sacra performed on behalf of montes t curiae, pagi and vici were part of the publica sacra. Both authors perceived these entities as mere territorial subdivisions of the civic com­ as a whole, and their religious festivals as expres­ sions of civic religion. This view was accepted by Mommsen and Wissowa; and the legal categorization of these festivals munity public religious events of the city-state is taken as an implicit justification of their exclusion from the analysis as of private religious behaviour in modern studies. 114 Yet, it is important to note that perspective, Augustan Labeo represents an which is very likely to postdate the territorial reorganization of the city's vici by the princeps. As has been mentioned above, the reorganization of the city's administrative and Augustus entailed a degree of existence cannot be religious imperial retrojected infrastructure under encroachment whose into Republican Rome. By contrast, Varro, referring to Late Republican Rome, distin­ guishes montes and pagi from the People as a whole: Septimontium and Paganalia constitute feriae for the inhabitants 113 Festus 298 L; WISSOWA 1912, 398-9. Antistius Labeo: Tac. Ann. 3,75; Pompon. Dig. 1,2,2,47; KUNKEL 1967, 32-4. 114 MOMMSEN 1887, 3,112-26; WISSOWA 1912, 380-404, passim. Exclusion: e.g. BARKER 1994, 1-4; below, 4.2.4. 312 of monies and pagi respectively, but not for the populus in its entirety. 115 Moreover, Varro does not describe mere ter­ ritorial sub-divisions of the city as a whole, but tions associa­ of people in a territorial context, montani and paga- ni, who have their own festivals. acter becomes clear from which, if not genuinely the Their associative char­ Commentariolum petitionis, Republican, uses Late Republican sources. It implies that the organizational structure, hier­ archy and leadership of these territorial associations re­ sembled those of traditional Roman collegia. 116 Mommsen his successors were of course aware of the nature of these associations as corporate bodies. Yet, their these entities as mere Roman People, biased by sources and territorial the perception of sub-divisions of the Augustan perspective of the they privileged, prevented them from fully grasping the true nature of these collegiate organizations. In the Late Republic, the origin of collegiate struc­ tures stricto sensu was open to speculation. The creation of the most respected professional invariably dated back to the publican collegia of artisans was regal period. 117 In mid-Re­ Rome, collegia seem to have been subject to censo­ rial supervision. However, a censorial prescription concern­ ing luxury regulations, issued in 220 in relation to the 115 LL 6,24: dies septimontium nominatus ab his septem montibus in quis sita urbs est; feriae non populi sed montanarum modo, ut Paganalibus qui sunt alicuius pagi. 116 [Q. Cic.] Comment. Pet. 30: deinde habeto rationem urbis totius f collegiorum montium pagorum vicinitatum; ex his principes ad amicitiam tuam si adiunxeris t per eos reliquam multitudinem facile tenebis. Cf. Cic. Dom. 74; FLAMBARD 1981, 149-54. On the authorship of the Commentariolum petitionis, see DAVID & al. 1973. 117 BEHRENDS 1981, 154-67; GABBA 1984. 313 collegium of fullers, had to be referred to the popular as­ sembly for ratification. 118 In citizens the second century, could enter, and even found, professional collegia without prior permission from the authorities. The ment Roman from censorial supervision develop­ to free organization may suggest that the original intention of founding professional collegia, namely the provision of an institutionalized asso­ ciation for members of one longer upheld. particular profession, was no At any rate, from the Early Imperial period onwards, even if the old name was preserved, membership of such collegia could comprise artisans from different profes­ sions, tradesmen and freedmen as well as slaves - the latter depending on their masters' prior permission. Already under the Late Republic, non-professional collegia were composed of members of similarly diverse backgrounds. 119 The deeply rooted elite suspicion of any collegiate structure as a potential source of conspiracy and armed vio­ lence resulted in the senatorial legislation which banned certain collegia and prohibited 64. 120 However, the Compitalian Games in the fluidity of political life in Late Re­ publican Rome referred to above guaranteed that senatorial consensus about the suppression of collegia was short-lived. Clodius re-instated the Games on 1 January 58, and a few days later succeeded with his bill on the re-activation of collegiate associations in the city. His bill was understood 118 Pliny NH 35,197: ... dedere ad populum ferendam. 119 BEHRENDS 1981, 168-74. Slaves: Marcian Dig. 47,22,3,2. KNEISSL 1994 shows that collegia in the imperial period, while retaining a name that suggests an organization of one profession or trade group, were in fact no longer mere >Berufsgenossenschaften<. 1 20 LINDERSKI 1995, 165-217; NlPPEL 1995, 72-3. 314 to be part of a political deal between the leading members of the political establishment, which by everyone concerned. Even entailed compromises Cicero acquiesced. As a con­ sequence, Clodius was able to pass his Lex Clodia de colle- giis without serious opposition from his opponents. 121 As we have seen above, systematic legislation, which exerted con­ trol over collegiate activity or which found new collegia made the right to contingent upon senatorial or imperial permission, had its origin in the Augustan Leges luliae de collegiis and informed later imperial practice. 122 The relative autonomy of professional collegia Republican Rome extended to in Late the collegiate structures of montani, pagan!, the members of vici or curiae, and to those collegia devoted to the cult of autonomy one deity. This relative is central to an understanding of their social and religious role in the first century. Most ciations collegiate resembled the socio-political structure of the ci­ ty-state in that they were organized ad exemplum rei cae. These and elected concilia (Cic. of the Plane. officers whose nomenclature, tenure of office and internal hierarchy faintly resembled trature publi- collegiate structures had their own assemblies, conventus and conventicula et quasi 36-7), asso­ Republic; the magis- they had internal constitutions (leges, alba) which ruled the purpose of their existence; they enjoyed benefactions (stipes) and possessed common pro­ perty and financial means (pecunia) ; they set up inscrip­ tions and public monuments, the latter subject 121 Cf. TATUM 1990. 122 Cf. above, 4.1.2. to the au- 315 thorization of the aediles; and they organized their own games, employed their own priests, sacerdotes or flamines (ILLRP 698) and performed their own sacra. 123 The jurists of the imperial period specified the legal elements constitu­ tive of such collegiate associations: these comprised the existence of a magisterial office, a causa, res communes and an area communis. 124 It is noteworthy that the SC de Bacchanalibus of 186 BCE already embraced this legal understanding, since meetings banned that had not received prior permission, prohibited the offices of magister and sacerdos, ruled out ence it of the exist­ a common purse (pecunia communis), and limited the number of male members to two so that meetings would fall foul of the rule of »tres faciunt collegium*. 125 The Baccha­ nalian affair of 186 is often seen as an attempt by the po­ litical elite to reinforce state control over religion and suppress the individualization of religious life in Rome and Italy in the early second century. 126 While moral concerns and the threat to elite control over political stability must have played an important part in the suppression of the 123 Gaius Dig. 3,4,1,1; BEHRENDS 1981, 171-2; FLAMBARD 1981; PURCELL 1994, 673-6. 124 Gaius Dig. 3,4,1; BEHRENDS 1981, 170-3. 125 ILLRP 511, lines 10-22. BAUMAN 1990, 343 thinks that the Senate limited attendance to five people so that the substituted wills and false seals which were allegedly one of the charges in the Bacchanalian affair could not be arranged; for the mancipatory will five witnesses, the testator and the libripens were required. This hypo­ thesis is problematic because the five witnesses had to be male; cf. CHAMPLIN 1991, 75-6. The relevance of the restriction on attendance must lie in the ratio of two male and three female members so that a collegiate asso­ ciation could not be formed. 126 E.g. GALLINI 1970, 81-90; NORTH 1979, 92-7; SCHEID 1981, 158-9; FLAMBARD 1981, 162; BEARD I994a, 761-2; NIPPEL 1996, 29. Contra GRUEN 1990, 55-78. 316 of Bacchus, the instrumentalization of the Bac­ worshippers chanalian affair in the elite's internal is struggles at as noteworthy: the Postumii and Aebutii, instrumental least in the suppression of the affair, rose to the in fame years after 186, whereas the Sempronii Rutili and Atinii, previously involved in the worship disappeared political modest from the of Bacchus, political scene. 127 However that may be, it is worth pointing out that the SC of 186, even if its cult, ultimate intention was the destruction of a religious had resort to impinging on the organizational structure to pseudo-collegiate of a crimes association: it attacked alleged which fell under the rubric of criminal law, but me­ rely discouraged, rather than overtly prohibited, Roman cit­ rite izens, Latins and allies from performing the religious in question. 128 The worship of a collegiate organization could focus one particular deity. At Beneventum, a collegium tibicinum of the Mater magna from around 100 is attested, with men and slaves on serving freed- as magistri.^ 29 Isis was not only worshipped by members of the senatorial elite or by the guild of pastophori, established under Sulla, 130 but also by the plebs urbana, freedmen and slaves. The goddess 1 promi- 127 M0rai concerns and popular hysteria: FORSYTHE 1994, 385-96. Threat to political stability: BAUMAN 1990, 347-8; GRUEN 1990. Rise and fall of families: ROUSSELLE 1989. 128 Discouragement: ILLRP 511, 1-9. Charges of criminal law: Livy 39,14,8, 16,3, 18,4 with BAUMAN 1990, 335-7, 342-3. 129 AE 1925, 117 = K. ScHiLLiNGER, Untersuchungen zur Entwicklung des Magna Mater-Kultes im Westen des romischen Kaiserreiches (1979), nos. 233-6. 130 The Caecilii Metelli: ILLRP 159. Sulla: GWYN GRIFFITHS 1975, 342-5. Pompey: HAYNE 1992. Pastophori: Apul. Met. 11,30. 317 the nence among these groups should presumably be linked to structure on the Capitoline hill, which the Sen­ collegiate ate repeatedly tried to destroy in the fifties of the Similarly, the religious activities of montani, century. 131 pagani or the members of vici surpassed the first of organization the their interest in Fornacalia, Parilia, Laralia, or Compitalia mentioned above. The Magistri Herculani (or Magistri Herculis), elected by a decree of the pagus, celebrated games, possibly for the eponymous deity. 132 JLLRP tions a magister 703 organizing games in a wooden theatre (as Mommsen restored the text) for Hercules magnus. Three stri, two from men­ magi- pagi and one vicus Sulpicius, dedicated an altar which showed Maia and Mercurius (ILLRP 702) . Four dif­ ferent collegia commemorate donations to Fors Fortuna (ILLRP 96-9). A territorial association, the pagus lanicolensis, instructed its magister to have porticum cellam culinam aram built de pagi sententia (ILLRP montani montis Oppi, through their rebuilt the 699, 700). Another, the magistri and flamines, sanctuary of luppiter Fagutalis from their own funds. 133 It must be assumed that such financial investments were the result of thorough deliberation and close investi- 131 59: Cic. Ad Att. 2,17,2 with WlSSOWA 1912, 358 4 . 58: Tert. Apol. 6; Ad Nat. 1,10. 53: Cass. Dio 40,47,4. 50: Val. Max. 1,3,3. 48: Cass. Dio 42,26,6 with LATTE 1960, 286. Collegiate association: COARELLI 1984; ALFOLDI 1985, 52-74. For Republican sacerdotes Isid(i)s Capitolin(a)e, possibly to be linked to that collegium, see OIL I 2 ,986; ILLRP 159. 132 OIL I 2 ,984 = ILLRP 701, dated to 58 by Mommsen. The ex­ act content of this inscription is obscure. The magistri may have been magistri vici, since the pagi appear to have elected only one magister as head of their organi­ zation. 133 ILLRP 698: sacellum claudend(um) et coaequandum et arbores serundas. On the deity's identity, Degrassi ad Joe. compares Varro LL 5,49. 318 of the relationship between the local deity in ques­ gation tion and the association concerned. No doubt these sanctuar­ ies provided more than just convenient pagan! the and meeting places for montani. Arguably, the collegiate focus on such shrines created a local religious identification. The Compitalian Games in the Republican period may serve to illustrate this pattern of decentralized religious affil­ iations. The Compitalia were organized by the magistri vicorum as heads of the collegia compitalicia, focusing on the worship of the local Lares in the various quarters' sacella. Lares and their festival attracted the slaves and the Both the poor but also members of the professional collegia before the Augustan reform of the Compitalia, some of the rich. 134 When conceptualizing the Republican Compitalia, should not a we be thinking of their Augustan successor. Scenes of Compitalician worship in Augustan and later Imperial display and., relative uniformity: art vicomagistri, their heads veiled, pour libations to the sound of tibiae, as victimarii and sacrificial animals stand by. These idealizations re­ flect the cult's Augustan transformation into a uniform rit­ ual which focused, through the worship of the Lares Augusti at Rome's cross-roads, on the emperor it himself. 135 However, would be misleading to assume that such uniformity would 134 FLAMBARD 1977; Id. 1981, 156-8; BOMER & HERZ 1981, 32-56; LINDERSKI 1995, 165-203. On the traditional clientele of the Lares, see above, 3.1.3. 135 Illustrations: VON HESBERG 1978, 914-9; TURCAN 1988, 2, 61. Cf. BARKER 1994, 118-33; LACEY 1996, 184-6. The in­ dex to OIL 6 lists twenty-one shrines of the Lares Augu­ sti attested by inscriptions. For their continuity into the fourth century CE, cf. the evidence presented in N. LEWIS & M. REINHOLD, Roman civilization: select rea­ dings3 (1990), 2,514-30. 319 acter was such that the Republican Senate, rather than alternative resort to thinking in terms of reforming the cult, had the char­ Its necessarily have prevailed in the Late Republic. to between outright prohibition in 64 and re­ luctant acceptance. One should therefore make allowances for the perception of these Lares as context of deities distinct in the a vicus' assertion of its identity, and for ri­ valry and competition between different vici and their magistri . The appeal of a Plautine character to the Lares viales for support, or the various topographical epitheta the Lares acquired in local worship, suggest a close identification of divinity and place and point to an essentially personalized conception of the Lares' divine nature. 136 religious This is not to say that local identification in the context of vici, monies, pagi or other collegia was a straightforward process. The freedman Clesipus Geganius put up an inscription (litteris magnis et pulcherrimis according to Mommsen) recording that he held the office of magister in the colleges of Luperci and Capitolini. 137 Aulus Castricius, in the Augustan period, managed to four collegia: We was do not that position in magister collegiorum Lupercorum et Mercurialium et Capitolinorum niensium. 138 he hold know et paganorum Aventi- how these two men perceived their various religious affiliations; yet, the realm of col­ legiate associations, fragmented as it appears to us, sug- 136 Plaut. Merc. 865; G. WlSSOWA, ML 2,2 (1894-97), 1868-97, at 1885-7. Personalization of divinities: below, 4.3.2. 137 OIL I 2 ,1004 = ILLRP 696. On Clesipus Geganius, see Pliny NH 34,11. 138 ILS 2676. Cf. Livy 2,27; 5,50. The Capitolini and Mercu­ rial i could dare to eject a Roman eques from their mem­ bers; Cic. Q. Fr. 2,5,2. 320 gests religious identities which cannot easily be subsumed under the category of civic religion. The Augustan jurists mentioned at the outset of this paragraph suggested that the religious activities of montes, pagi, vici and curiae were mere territorial reflections of the Roman People as that publica of the a whole. The collegiate forms of worship, which created their own forms of invalidate sacra legalistic semi-publicity, dichotomy. serve to In Late Republican Rome, there were various forms of possible religious identi­ fications in the context of public religious observance. Cicero conceptualized the life of a civic community as taking place in public spaces (Off. 1,53) : Multa enim sunt civibus inter se communia, forum fana portions viae, leges iura indicia suffragia, consuetudines praeterea et familiaritates multisque cum multis res rationesque contractae. The material record of the Late Republic period seems the Augustan to support this holistic portrait of society. Building activities focused on civic ples, and basilicas, porticos, projects, fora, curiae and macella. The subse­ quent creation of semi-public and private spaces, the lae tem­ scho- of collegia, private chapels of deities like Mithras or Isis, and the prominence of baths in Empire, cities all over the seemed to herald a partial detachment from the pre­ vious commitment to civic society. 139 Accordingly, the advo­ cates of the civic change model view religious evolution a from the embeddedness of religion in the traditional city-state to the dislocation, and privatisation, gious as choices in an 139 E.g. ZANKER 1994. 140 See above, 2.7.1. of reli­ increasingly complex environment. 140 321 With a view to the public nature of Republican society, behaviour the of collegiate associations has been understood as an imitation of the elite. The quasi-political active ex­ perience in the collegia would have compensated for what was impossible in the popular assemblies. The display of social status through inscriptions, dedications and games, a privi­ lege normally reserved to the aristocracy, would have coun­ teracted the socio-political discrimination, providing inte­ gration and acceptance of the fabric of civic society. 141 While such a functionalist analysis has its attractions, the category of >integration can only depict the unintended consequence, rather than the raison d'etre, of behaviour. The such category of integration ultimately fails to explain events such as the political violence of years social the final of the Republic or the crowd behaviour at the funeral of Clodius, when popular social power temporarily subverted the rituals of elite behaviour and threatened the fabric of society. 142 On the functionalist view of the civic model, religion was a deeply integrated affair of a closed society, in which the religious behaviour of Rome's collegiate ciations reflected the religion of the city-state, since everybody worshipped the same deities as the integrationalism can asso­ conceptualize city. 143 Such competing choices only when they become manifest in deviant religious behaviour. It cannot explain Augustus 1 motives for changing the of the Compitalian cult; his reorganization of the cult structure must have reflected the wish to control 141 E.g. FLAMBARD 1981, 165-6. 142 NIPPEL 1995, 75-84; SUMI 1997. 143 E.g. BEARD I994a, 748-9. structure cult be- 322 haviour which would otherwise have remained uncontrollable. A more plausible suggestion is that the collegia of organizations such as freedmen and slaves (as we have seen, the division between these organizations collegia • and the professional was highly permeable) through imitating the socio­ political elite's behaviour in fact impinged on that elite's privileges. These associations unashamedly took to the ditional means of status display and adopted traditional forms of religious communication, while preserving a as stability. 144 subverting Mention has the the same already foundations of the of Roman of in first century. 145 Accordingly, I would religious such status groups as the embodiment of their relative autonomy in what rather social freedmen suggest that we ought to conceptualize the public behaviour time been made above of the socio-economic independence achieved by course at social openness and independence which to the elite appeared the tra­ remained an integrated society than as a further sign of the gradual disintegration of civic religion in Late Republican Rome. 146 Given the relative autonomy of in Rome in collegiate associations the first century, it is surprising to find no traces of buildings, scholae or templa, owned by profession­ al collegia or other collegiate bodies in the Republican material record in Rome, the colony of Cosa or on Delos. The earliest buildings which are undisputably recognized as scholae of professional collegia in the city of Rome date to 144 Cf. PURCELL 1994, 671-2. For freedmen imitating the artistic values of the elite, see ZANKER 1975. 145 See above, 3.1.1. 146 For the phrase >relative autonomy< and the underlying notion of an >open system<, see above, 2.7.4. 323 the early Augustan period: the scholae of the tibicines, the fabri tignarii and the scribae librarii et praecones of the curulian aediles. 147 The collegiate associations of the Late Republic may not have possessed the scholae of their Imper­ ial successors, which visually embody the the detachment public sphere; nevertheless, the professional collegia, the organizations devoted to one deity, the montani, or from pagan! the inhabitants of Rome's quarters are not simply illus­ trative of Cicero's closed society, and their religious haviour be­ is not merely reflective of civic religion. Rather, this behaviour serves to undermine the legalistic categories of public and private. Yet, by questioning the dichotomy of public and private, I have already run ahead of my argument. 4.2.4 Roman religion: public or private? As has been elaborated above, Wissowa understood Roman state religion as the formalistic, yet true religiosity. reflection from Roman In his account, private religion, the religion of the populace, became increasingly emotionalized moved of such true religious behaviour. and re­ The People's alienation from its traditional religious practices led to the decline of state religion. By contrast, succeeding gene­ rations of scholars made the alleged decline of religion in the Late Republic contingent upon this religion, formalism of Roman which was unsuited for providing an emotional ex­ perience. This Schleiermacherian perspective was complement­ ed by the view that the privata sacra of families and viduals, indi­ as described in the second part of Festus' defini- 147 BOLLMANN 1993. 324 tion of public and private religion, served to provide quintessential religious this experience. Even now, such a view appeals to students of Roman religion and ancient historians alike. 148 By contrast, recent approaches stress the public fundamentally nature of Roman religion, which was the corollary of the quintessentially public nature of Roman social life der un­ the Republic. This focus on the public aspect of social behaviour has resulted in the revision of the received para­ digm that the dichotomy of >public< and >private< of behaviour and internal experience, ritual should be resolved in favour of the latter. Now, as a corollary of the modern sus­ picion of man's internal world, the pects of religion city penetrated private as­ disappear in a society characterized by public religious behaviour; and if the allegedly the religious practices in distinction between >public< and >private<, it is due to the all-embracing normativity of the former that the distinction no longer matters: »It might seem a possibility that ... private cults would have afforded a separate religious world within which the individual Roman might have found the personal experience of superhuman beings, the sense of community and of his place in it, which the remoteness of the of­ ficial cult denied him, but which he needed to make sense of the world ... [H]owever, it is not so easy to believe in this deep but unattested religious life Almost all the evidence we have suggests that in Rome in particular religious life focused on the public cults, on the relationship between the city and the city's gods 148 See above, 2.2, 2.3. Cf. e.g. DORRIE 1978a, 248: »In altesten Zeiten diirfte die private Ausiibung des Kultischen das eigentliche Feld gewesen sein, auf dem sich Religion und Religiositat manifestierten«; MANTELL 1979, 217: »the sacra privata ... were the Roman's real reli­ gion, the cults that really mattered*, quoted with ap­ proval by LACEY 1996, 173. 325 and goddesses; the citizen participated through his identification with the city and its interest ...«. 149 To be sure, such a view has an initial plausibility. For the fabric of Late Republican urban society was held together by the interacting interests of the freedmen elite, the patronage by the resulting profusion of strong system of interrelated differences. 150 mutual little group space for an interests which transcended To most of its inhabitants the city would which pro­ total retreat or absolute privacy. undesirable goal. Rome, with its finite number of sions, and commitments, created a Indeed, the fabric of mutual dependence would make solitude of dependence, of Rome, notoriously congested and overcrowded, vide bonds or amicitia. Roman society was an >integrated so- ciety< in the qualified sense that status plebs, and slaves, who shared accommodation, work and the means of recreation, and who were linked the urban Moreover, deities complete the religion at and ritual occa­ were employed by all strata of society, would easily suggest a uniformity of behaviour and a congruity of belief systems to the modern beholder. 151 In the preceding discussion I have Roman society tried show the legal dis- of these spheres does not a priori entail that public and private religious activity needed to be ually how defined the difference of public and private spheres of religious activity. To be sure, tinctiveness to different. concept­ At the same time, however, the modern em­ phasis on the public and social nature of Roman religion has 149 NORTH 1989, 605-6. Cf. above, 1.3, 1.4, 2.3. 150 E.g. PURCELL 1994, 668. 151 Cf. above, 2.7.3. 326 led to the rash conclusion that these two spheres of activi­ ty were homologous. In what follows, I will suggest that it is the underlying dichotomy of public and private which com­ promises both old and more recent models, and which must be overcome. Consider the sphere of domestic cult. Generally, if invariably, the deities of household shrines were not dif­ ferent from those receiving worship in material not evidence for civic temples. The domestic cult seems to suggest that cult objects and paintings were placed in the representative areas of Late Republican and class Imperial Roman upper- houses (in atria and peristyles) rather than in those parts (vestibules, might early bedrooms or accentuated rooms) which have provided retreat and privacy. Atria and peristy­ les served to display the house owner's socio-political sta­ tus to the Roman public at appears large. The logical to be that religious practice, while not lacking an affection for the deities who presided over home, conclusion was one's physical not considered to be a private affair that would have been conducted in spiritual retreat. 152 Yet, the mate­ rial evidence for the domestic cult in atrium-peristyle hou­ ses in Pompeii and Herculaneum is more complex. For a signi­ ficant quantity of religious objects or paintings can be found in kitchen as well: »The large majority of the Campanian evidence is found in the kitchens (27.4%), atria-alae (25.6%), peristylia (18.4%) and viridaria (20.2%) ... Only a few shrines are documented in the vestibules, bedrooms, accentuated 152 E.g. BEARD 1994a, 732; FEENEY 1998, WISEMAN 1987, Household and esp. 88-95. 6. Representation: 393-413; WALLACE-HADRILL 1994, 60, passim. upper-class family: SALLER 1994, 74-101, 327 rooms (oeci, tablina, triclinia), corridors, latrines, and stables. Cult-rooms are found especially in the viridaria, sometimes in the atria-alae and peristylia, or underground ... in virtually all kinds of rooms paint­ ings can be found. The majority is located in the kitch­ en (39). Large amounts of niches are found in the kitch­ ens, atria-alae, peristylia, and viridaria: 37, 39, 26, and 29 respectively. With one possible exception (in the tablinium) aediculae and pseudo-aediculae are found in the peristylia, viridaria, and atria-alae only.« 153 A study of paintings counts in the Lararia of Pompeian houses seventy paintings of Genius, Lares and/or Di Penates in the kitchens and service-areas, twenty two in representa­ tive rooms of smaller houses, and twenty four in representa­ tive rooms of medium-sized houses. Paintings in kitchens and other work areas depicted the Lares and the Genius, those in atria and peristyles the Penates. Owners of upper-class hou­ ses preferred the more expensive niches or aediculae rather paintings. 154 The logic behind these statistics is ap­ than parent: the exclusive appearance of expensive aediculae in representative rooms on the one hand, and the generally high proportion of cult installations in kitchens plus the number of comparatively cheap paintings in that area on the other, point to a divide between representative rooms as the sphere of upper-class inhabitants and the kitchen as the realm of their slave familia. 155 This evidence poses a problem to the view that domestic cult illustrates the public and social nature of Roman reli­ gion, or that it documents the disappearance of the notion of individual religiosity. For in order to include the evi- 153 BARKER 1994, 39-40. The atrium-peristyle type resembles Late Republican upper-class houses: BARKER 1994, 27; GEORGE 1997, 303-n. 154 FROHLICH 1991, 28-9, 178-9; Id. 1995, 205. 155 Cf. above, 3.1.3. 328 dence of religious installations in kitchens, that area of religious activity, too, has to be labelled >public<. Of course, in some sense a kitchen was a public sphere; yet, at the same time its public status was very different from that of the atrium in which the house-owner would receive clients and visitors. This example illustrates the terminological dilemma behind the traditional dichotomy of >public< and >private<. One way of solving this problem may be to look at those rooms of upper-class houses which, following the convention­ al dichotomy, fall under the category of non-representation­ al or private rooms: vestibules, cubicula and accentuated rooms. To be sure, such areas as cubicula or triclinia could be used for representative functions as well. Consider cubi­ cula: the Republican and suggest that Early Imperial With reference to location it was which stresses locate prevailed, and permanent of the year. This the temporal and functional dif­ ferences in room use, can explain why cannot receiving common to use different parts of the house on different occasions at different times approach, for these rooms it has therefore been suggested that a flexible use of space that sources cubicula were sometimes used for sleeping. On other occasions, they provided the visitors. literary the material record furniture in these so-called non- representational or private areas of the house: the flexibi­ lity of use required constant relocation. 156 156 Cf. NEVETT 1997, 288-92, 294-8. Sleeping: Cic. Deiot. 129; Quint. 4,2,72; Plin. Ep. 5,6,21. Public function: Cic. Scaur. 26,4; Apul. Met. 1,23. 329 In conclusion, it seems preferable to replace the itional dichotomy of >public< and >private< areas of the Roman elite household by the more adequate be trad­ distinction, to taken cum grano sails, between rooms designated for per­ manent representative functions and those which were put to a more flexible use. This explanation is compatible with the material evidence for domestic cult. House-owners would not be interested in installing expensive niches in areas whose and aediculae interior was open to relocation, but would prefer permanent locations such as the atria, peristylia viridaria for cult installations. The slave familia, on the other hand, would wish to install permanent cult objects those or in areas which had a permanent service function: the cu­ ll na and other work areas. Moreover, this conclusion suggests that the modern clas­ sification of >public< and >private< religious areas of social activity does not do justice to the Roman eviden­ ce, which does not divide space into areas of public sentation and repre­ on the one hand and private retreat on the other. For even those spaces commonly associated with privacy, na­ mely bedrooms, appear to have had shifting functions, public or private, at different times. The striking absence of rooms from the Roman household which permanently to >belonged< one individual household member correlates with the lack of a literary conception of personalized space in the elite house. 157 The material and the literary record thus emphasize a profound cultural difference between and the modern concept 157 GEORGE 1997, 300-1. Roman the Roman of individualism. When drawing on 330 these insights, studies of religion sometimes make the im­ plicit assumption that the quintessentially public nature of Roman houses prevented a conceptualization of privacy in a domestic environment. Before accepting that may assumption, it be useful to ask whether these scholars wish to suggest a mere cultural difference private spaces, in conceptualizing public or whether they intend to postulate a more profound distinctiveness of the Romans, which shows undesirability and in the of some sense of privacy. 158 A cultural dif­ ference is no doubt apparent. Yet, postulating a profounder distinctiveness would be erroneous. For the literary evidence relating to upper-class Romans clearly suggests that there was an interest sense in creating a of personal privacy. Pliny's letters about his villas illustrate the concern for separating rooms and areas of his houses by means of doors, windows with shutters or curtains in order to create a temporary sense of seclusion from other members of his household. Although Pliny's villas were no doubt permanently shared domestic spaces, individuals to achieve a degree of personal privacy even in the context of spheres which could be put to communal use. 159 the tried upper class Moreover, villa in the countryside, notwithstanding its economic function, provided a place for private retreat. Its artificial landscape of Hellenizing culture counter­ balanced the political culture of the city of Rome. Thus the spread of the villegiatura from the early second century onwards exemplifies the differentiation of 158 Cf. above, 1.4, 2.3. 159 E.g. Pliny Ep. 2,17; LEFEVRE 317-8. 1987; the GEORGE spheres 1997, of esp. 331 upper class activity, and the creation of non-political do­ mains by the political elite. 160 Such private have been more difficult to achieve retreat must in the constricted spaces of urban houses. However, the temporal and functional distinctions between the use made of triclinia enabled or peristyles demonstrate same principle, were employed for private activities, once the flow of visitors had ebbed away. These examples that the primary function of public representa­ tion in the rooms of the elite their cubicula inhabitants to use these areas for reading or rest­ ing, when guests were not received. On the atria or secondary function for household did not exclude temporary retreat in mutual respect for personal privacy. While religious studies use the material record to prove their assertion about the penetration of the into public sphere the physical areas of private life, they tend to over­ look these functional differences in room use, which led the to flexible employment of different areas of the household for differentiated activities at different times. A more flexible approach, which takes into account these functional differences, should help us to overcome the dichotomy of >public< and >private<, whose static nature tempts to think in scholars terms of incompatible dualisms, and to favour unduly either the private or the public aspect of social and religious behaviour. The very fluidity of in the ancient material these categories and literary record implies that 160 Cicero called his villa at Puteoli the >Academia<: Tullius Laurea FLP pp. 182-3 ap. Pliny NH 31,6. Cf. FLAIG 1993; K. SCHNEIDER 1995, 73-104. Hellenizing architec­ ture: NEUDECKER 1988; R. FORTSCH, Gnomon 64 (1992), 520-34. 332 Roman conceptions representation of in the household ranged we public the domus frequentata to private retreat in the sanctum pertugiurn. 161 In inversion view, from of the orthodox ought to think in terms which resolve the dualism of public and private. Concerns for privacy, including those for the private reflection on cult, the importance of domestic rather than being excluded by the Romans' representa­ tional behaviour, penetrated into the physical areas of pub­ lic representation. To be sure, societal behaviour in general was both lic pub­ and private. Consider, for instance, the motives of the testator in Roman society: »The Roman will was indeed an expression of deepest emo­ tion, particularly of affection in the form of concern for the future happiness or security of family and friends. But it was also a solemn evaluation of the sur­ rounding world, ... and it was an insurance that the individual would be remembered by others both in life and in death.« 162 This insight into the fundamentally dialectic nature of cial behaviour should also serve to resolve the discussion between two fundamentally opposed conceptualizations of ligion at Rome sphere. The position, while doing justice to the realm of public rituals conducted by magistrates or priests, all re­ either as a private religion or as a civic religious system whose focus lay in the public latter so­ must perceive forms of religious behaviour as essentially public, and must treat such public religious behaviour as being deter­ mined by the civic realm. We do not know how exactly, and to 161 Domus frequentata: Sen. Ep. 21,6; cf. Cic. Att. 1,18,1, 2,22,3. Sanctum perfugium: Cic. Catil. 4,2; Vatin. 22; Dom. 109; cf. Gaius Dig. 2,4,18; Paulus Dig. 50,17,103, cited by GEORGE 1997, 300-11, at 300 1 . 162 CHAMPLIN 1991, 27-8. 333 what extent, Romans used the privacy of their households for their religion. Yet, as I have suggested above, it is metho­ to exclude on a priori grounds the unwarranted dologically from realm of private religious emotions and belief systems our analysis. 163 Rather, we should try to re-investigate these individual motivations and beliefs, they as instru- mentalized the religious infrastructure of the city of Rome. 4.2.5 Public places, private concerns The public and social nature of dedications has become com­ monplace: dedications served as permanent display of indivi­ dual motives and attitudes to a larger social environment. In apparent imitation of the publicity of civic on occasion publicized their pri­ individuals processions, sacrificial vate sacrifices. Proceeding through the city to the ary, they advertised sanctu­ bystanders their motives for the to sacrifice by means of a titulus carried by an attendant. 164 Such overt publicity would no doubt offend Schleiermacherian sensibilities. Yet, it is difficult to see how this display of motives is more than a distinct of public cultural behaviour, and how it can preclude the existence of religious motives that deviate from assuming civic ideology. By that such deviation is excluded, scholars a priori exclude from their frame searching characteristic for motives of analysis the possibility which find their realization in the realm of public display, but which may nevertheless carry 163 Cf. above, 1.4, 1.6. of a 164 VEYNE I983b. Cf. SCHEID 1985, 12-5; BEARD I994a, 732-3: »Thus what might have been 'merely 1 private devotion became part of public, city life.« 334 of dimension internal meaning. Thus these scholars unwitt­ ingly assume that the public assertion of religious behav­ could be direct proof of the public nature of interior iour motives and thoughts. 165 The actual motives of the individuals raised who in­ scriptions or made dedications to the gods are often unknown to us. For instance, we can only speculate about the reason­ ing of lulia Sporis, the wife of the aedituus of Diana Planmentioned above, who made a dedication to Silvanus ex ciana visu - or about the situation which led to the god's appari­ tion in the first place (AE 1971,31). Personal economic terests at lie the centre in­ of dedications to Hercules in origin Tibur, where entrepreneurs of Italian and Roman re- payed the deity for supporting their financial transactions. The language of these dedications resembles the world of business, trade and profit. 166 The applying our problem hermeneutic of own preconceptions to these dedications is in danger of resulting in circular thinking; arguably, however, we should invert the view that private commitment ought forms of religious to be seen as part of the religion of the city. On the contrary, the penetration of public temples and 165 Cf. above, 2.7.3, 2.7.4. 166 E.g. ILLRP 136: M. P. Vertuleius C. f. quod re sua d[if]eidens asper afleicta parens timens heic vovit, voto hoc solut[o de]cuma facta poloucta leibereis lube<n>/tes donu[m] danunt Hercolei maxswne mereto. Semol te orant se voti crebro condemnes; ibid 149: Sancte de decuma victor tibei Lucius Munius donim moribus antiqueis pro usura hoc dare sese visum animo suo perfecit, tua pace rogans te cogendei dissolvendei tu ut facilia faxseis t perficias decuman? ut fad at verae rationis, proque hoc atque alieis donis des digna merenti; BODEI GIGLIONI 1977, 51-4. 335 places with individual religious concerns affected all areas of the city of Rome. The Capitoline hill, the religious and political centre of the city, was not excluded from the diffusion of personal religious aims. On January 1, the consuls upon entering of­ fice conducted a sacrifice to obtain divine favour and made vows on behalf of the People on the Capitoline hill. 167 Scipio Africanus the Elder is said to have regularly visited the temple of the Capitoline triad at night, »as if he to discuss political issues with luppiter optimus maxi- mus«. 168 On a different level, C. Crispinius Hilarus, a man were Ro­ citizen from the plebs Faesulana, conducted a sacrifice to luppiter on the Capitoline Hill, at which he was accompa­ nied by eight children, including two daughters, twenty sev­ en grandsons, eighteen great-grandsons and eight granddaugh­ ters. His motivation for this private sacrifice is Augustus included unclear. Crispinius Hilarus in the acta diurna in support of his policies on procreation. 169 Yet, there to be no seems direct connection between this private sacrifice and civic religion. On their visit to the Capitoline temple, Crispinius larus and Hi­ his family may have met worshippers behaving not unlike those depicted in Seneca's De superstitione. 167 SCHEID 1985, 130-1; ORLIN 1997, passim. In the 168 Gell. 6,1,6: in Capitolium ventitare ... atque ibi solum diu demorari quasi consultantem de re publica cum love with AYMARD 1953; GLADIGOW 1994, 16-7. LIND 1992, 14, following Gellius and Livy 26,19,3-9, accuses Scipio of religious hypocrisy. On Scipio's famously close relation with the Capitoline luppiter, add Val. Max. 8,15,1; App. Pun. 104, 109; Cass. Dio 16,57,39; WALBANK 1985, 120-37; RAWSON 1991, 88-9. 169 Plin. NH 7,60. 336 sanctuary of the Capitoline triad, luppiter is looked after by a nomenclator, a horae nuntius, a lector, 170 and an unc- tor. Coiffeurs and maids serve luno and Minerva. Individuals ask the deities for support and advice on legal quarrels. An aged actor performs for them. Some women sit in the sanctua­ ry and believe they are loved by luppiter: (fr. 36) In Capitolium perveni: pudebit publicatae dementiae, quod sibi vanus furor adtribuit officii. Alius nomina deo subicit, alius horas lovi nuntiat, alius lietor est, alius unctor qui vano motu bracchiorum imitatur unguentem. Sunt qui lunoni ac Minervae capillos disponant: longe a templo non tantum a simulacro stantes digitos movent ornantium modo. Sunt quae speculum teneant. Sunt qui ad vadimonia sua deos advocent. Sunt qui libellos offerant et illos causam suam doceant. Doctus archimimuSt senex iam decrepitus, quotidie in Capitolio mimum agebat, quasi dii libenter spectarent quern homines desi erant. Omne illic artificum genus operantium diis immortalium desidet. (frg. 37) hi tamen ... etiamsi supervacuum usum, non turpem nee infamem deo promittunt. Sedent quaedam in Capitolio quae se ab love amari putant nee lunonis quidem, si credere poetis veils, iracundissimae respecter terrentur. 171 It is not entirely clear whether these services were perfor­ med by cult personnel on a daily routine basis or by indivi­ duals as part of their worship. The services offered to lup­ piter correspond to those performed by slaves on an institu­ tionalized basis in an upper-class over, the daily maintenance Roman More­ household. of cult statues was a common phenomenon in the ancient world. Yet, the wording the of text suggests that at least some of these services were ren­ dered by individual visitors to the sanctuary, who put themselves temporarily in the subordinate position of 170 The mss offer lector, gested Iirtor which Cf. H. FUNKE, JAC 17 comparing Tert. Idol. serv- lictor or luctor. H. Georges sug­ Dombart-Kalb accept, Linker litor. (1974), 149, who accepts lictor, 18. 171 Sen. De superstitione frgs 36b-7 Haase ap. August. 6,10 (p. 268,24-269,10 D-K). Cf. Epist. 95,47-50. CD 337 their obeying ants masters, thus contradicting the social status they may have had outside the temple precinct. 172 identify Wissowa suggested that we should as sequence the lovis, performed in the form of a epulum lectisternium twice a year, on 13. 173 ritual this September 13 and November However, Seneca does not describe a ritual conducted not on a regular basis, but criticizes a religious practice by any temporal frame. In marked antithesis, he re­ limited Isis strains his judgement on the ceremonies in the cult of (however superstitious and they appeared to embarrassing him), since those rites were conducted only once a year, from October 31 until November 3. It was the absence of such a confined temporal scheme which offended Seneca in the case of practices in the temple of the Capitoline triad. 174 the To Latte, this incident illustrated superstition by new religious creeds which arrived from the East tainted and subsequently replaced Roman popular an authentic, though weakened, religion. 175 Versnel's suggestion to explain this in­ cident by comparison with a Republican dedication put up by Paulla Toutia M. f. et consuplicatrices has the advantage of avoiding such traditional evolutionist stereotypes of au­ thenticity and decline. He offers a potential context for the scene on the Capitoline Hill which firmly links the pas­ sage back to Roman religious practice. 176 172 E.g. qui vano motu bracchiorum imitatur unguentem; longe a templo non tantum a simulacro stantes digitos movent ornantium rnodo. Cf. GLADIGOW 1994, 24. 173 WISSOWA 1912, 423 3 . 174 Sen. De superstit. 36a: huic tamen ... furori [the rites in the cult of Isis] certum tempus est. Tolerabile est semel anno insanire. Cf. LAUSBERG 1989, 1891. 175 LATTE i960, 327-31, at 328. 176 VERSNEL 1981, 30-1, at 30 18 , comparing ILLRP 301. 338 Yet, Versnel does not clarify whether his suggestion entails that we should link the incident described by Seneca to a supplicatio organized by a group of female worshippers. If this so, would be misleading. For although we lack the mo­ most basic information concerning the social status and of tives these worshippers, this puzzling passage does not appear to portray a formal ritual supplication by a cohesive group. Rather, this passage seems to describe an incoherent assemblage of religious practices through which individuals, in imitation of social stereotypes, those of temple servant and deity, slave and master, beloved and lover, transcended the norms temporarily and boundaries of their traditional social bonding in Roman society, as found outside the sanc­ tuary's confines. The context in which this passage is placed by Augustine supports such an interpretation. For the Seneca's portrait of practices on the Capitoline Hill as preserved by Augus­ tine is immediately followed by Augustine's own remark, men­ tioned above, that Seneca took the liberty of systematically compromising the theologia civilis. 177 Augustine's delight at Seneca's deconstruction of ritual practice becomes under­ standable once we realize that the separation between the rites of oriental cults and traditional Roman religion, rou­ tinely made in modern scholarship, is decidedly by portrayal of such religious practices in De super- Seneca's stitione. He introduces the nature blurred of ritual behaviour section on the superstitious by censuring the behaviour of 177 Cf. above, 1.1 and 2.4.1. Cf. August. CD 6,10 267,9-12 D-K): civilem ... et urbanam theologian. (p. 339 Yet, Isis. 178 the in worshippers Mater of cults cult Bellona and he immediately proceeds to the discussion of superstition in the cult of the this magna f Capitoline triad, linking back to the preceding superstitious practices by the category of furor. Significantly, all of these practices are listed under one heading, de ritibus. 179 Seneca refuses to make a substantive discrimination, but suggests that emo­ and cognitive responses in the cults of Mater Magna, tional Bellona and Isis on the one hand and of the Capitoline triad on the other were indistinguishable, and equally repulsive. Furthermore, by connecting the practices on the Capito­ line Hill to the religious behaviour of individuals in cults of so-called oriental deities, Seneca implies that the reli­ gious behaviour described in the context of the worship of the Capitoline triad was that of individual well, rather than that of worshippers as organized groups or religious functionaries. To Seneca, this passage described an integral part of daily religious routine taking place in of the the temple Capitoline triad. It took a form which was unaccep­ table to the philosophical critic, yet which apparently mained inoffensive here we are rather than re­ to the civic authorities. The fact that concerned with normal cult routine, ritus, with exceptional religious activity that could be marginalized, calls into question the very foundations of the models of >Staatskultus< and >civic religion<. 178 Sen. De superstit. 34-5 Haase (ap. August. 267,32-268,23 D-K); MAZZOLI 1984, 989-91. CD 6,10 p. 179 August. CD 6,10 (p. 267,31-269,10). Furor as an expres­ sion of superstition: MAZZOLI 967-8; LAUSBERG 1989, 1894-5. Cf. 1.1. 340 4.3 Polytheism position We are now in a ultimately that problem to the identify underlies methodological many current models of religion Roman religion. Neither Wissowa's concept of Roman as >Staatskultus< nor its recent replacement by the concept of >civic religion< manages to question the notion of >religion< itself. Though presenting defini­ competing the substance or the ideological content of reli­ of tions various ask gion at Rome, these scholars have unanimously failed to whether the very notion of >religion< is capable of compris­ ing variety the constellations of different cults and of deities in a polytheistic context. The problem is concept of >religion< the that is closely adjusted to monotheistic religious choices such as Jewish religion, Christian reli­ gion or Muslim religion. By talking about these >religions<, we imply their that identity can cum grano sails be re­ constructed by means of a religious dogma, their theology or orthodoxy. When speaking about >Roman religion<, many schol­ ars fail to realize that they do not deal with entity, cohesive and a that the >identity< of a pagan reli­ gious system cannot be reconstructed on the basis ciples similarly of prin­ which are geared to monotheistic religions. Scholars do not sufficiently take into account that the notion of >religion< is incongruous with the nature of the polytheist­ ic system at Rome which they intend to describe. 180 This incongruity results in the reductionist assumption that the complexity of the religious data at Rome could find complete 180 Cf. AHN 1993; GLADIGOW I997a, 103-5, for the problem of applying the monothetic concept of >religion< to a poly­ theistic context. 341 representation in monothetic account of the identity of a this >religion<, be it >Staatskultus< or >civic religion<. the to tribute Many scholars of Roman religion pay passing phenomenon of polytheism. Yet, their assessment of that phenomenon's importance for super­ remains studies their ficial at best. More often, they display an apparent uneasi­ ness when facing a plurality of potentially conflicting cults and deities, that undermines their monothetic of tions concep­ Republican religion. As long as this plurality of cults and deities under the Republic can be marginalized, the apparent extension of religious choices under the Empire must be seen as dissolution. 181 The comprehensive monogra­ phic treatment of individual deities and the their cults over last decades has not remedied this dilemma. These stud­ ies rarely conceptualize the internal logic of a polytheist­ ic system: how are different deities perceived in relation to another; one what parameters constrain a worshipper's place, his obligations and choices? Instead, >polytheism< is treated as though it were a mere appendix to monothetic con­ arbitrary ceptions of ancient religion - an assemblage of different deities in a certain place. 182 The history of these modern the notion difficulties. of polytheism Considering young. Philon of Alexandria antiquity of the Graeco-Roman paganism, the term >polytheism< is the relatively introduced noXudeia and 6o£a TioXudsoc in the early first century CE in order tualize illustrates to concep­ worship of more than one deity. At its concep- 181 Cf. above, 2.7.1. 182 por notable exceptions, see MORA 1995; ROPKE I995a. 342 tion, >polytheism< was a polemical term, author writing in the of by an Jewish diaspora of Alexandria, who defined belief in the one true god ex background introduced negative before the a hostile pagan environment. 183 A systematic pagan conceptualization of the dichotomy of polytheism and monotheism is not found prior to the second century CE, when pagan philosophers, confronted with the Christian attacks on polytheism, tried to justify the pagan choice of worshiping more than one deity. 184 After the official demise of ism, the pagan gods became aesthetic objects as part of a Late Antique cultural tradition. It was in the Renaissance rediscovered rediscovering Roman poetry like Ovid. pagan­ and the this vein that Graeco-Roman gods, when mythology through authors However, the notion of >polytheism< did not be­ come an issue in its own right before the seventeenth centu­ ry, when the contemporary interest in primitive polytheistic societies unearthed ancient polytheism as comparative research. The enlightenment a discipline of tradition of the eighteenth century could regard polytheism as an alternative model to contemporary monotheism, exemplified in the tyranny of French Catholicism. At the same time, the debate origins of society human principle religion. Evolutionist models either postulated a primeval monotheism which subsequently degenerated into lytheism; the addressed the problem of whether mono­ theism or polytheism was the first organisational of on or developed a evolution that started from po­ teleological model of religious fetishism or animism, passed 183 Philo 1,41; 1,609 with SANDELIN 1991. Contrast the pagan use of noXudeoc, denoting that which is the property of many deities; e.g. Aesch. Suppl. 424: itoXudeoc e6pa. 184 MOMIGLIANO 1987, 313-28. Cf. BURKERT I985b. 343 through the stage of advanced polytheism, and finally culmi­ nated in the Christian belief in one deity. 185 Polytheism as a transitional stage in the religion: we as evolution have seen, this evolutionist paradigm was twentieth century. 186 Due to this paradigm, the and evaluation of failed cults, the polytheism, to plurality of it sufficient scholarly attention. attract choices of polytheism. For do not matter, as long as these choices are controlled by a civic religious para­ new is not surprising that the model of >civic reli­ gion < has also excluded the phenomenon individual and deities Given the many points of contact between old and digms, nine­ the accepted by various studies of Roman religion in teenth of behaviour. As identity determining individual a matter of fact, the civic model has also been applied to monotheistic societies such as Re­ naissance Florence. 187 4.3.1 »Si deus unus est... « By way of contrast, worship of a number of different deities was conceptualized by Romans as one of the guiding princi­ ples of their religion. 188 Therefore, the lack of an adequa­ te systematic expression of this (internalized) principle is not significant. This does not mean that the number of dei­ ties would not become an issue in Late Republican Rome. Lac- 185 For ample documentation, see MANUEL 1959; SCHMIDT GLADIGOW 1998, 315-6. 186 See above, 2.2. 187 E.g. MOLHO & al. 1991. 1985; 188 E.g. Cic. Red. Sen. 30: ... in ipsis dis irmortalibus non semper eosdem atque alias alios solemus et verierari et precari. 344 tantius, trying to rebut the could be raised against philosophical arguments that the existence of just one deity, cites a passage from Cicero's dialogue Hortensius, with the persona of Q. Hortensius Hortalus (114-50) speaking: Sed fortasse quaerat aliquis a nobis idem illud quod apud Ciceronem quaerit Hortensius si d e u s u n u s e s t , q u a e e s s e b e a t a solitudo q u e a t . Tamquam nos, quia unum dicimuSt desertum ac solitarium esse dicamus. Habet enim ministros quos vocamus nuntios. Et est illud verum quod dixisse in Exhortationibus Senecam supra rettuli, genuisse regni sui ministros deum. 189 According to the most recent editor of the Hortensius, passage this is unique in that pagan philosophy before the early second century CE neither conceptualized monotheism principle of nor conceived any polemic against it. 190 A reas­ sessment of this passage, however, will Lactantius' the show that despite interpretation to the contrary its content does not take issue with monotheism. Cicero's dialogue Hortensius was written in the early summer of 45. Set in presumably the summer of 62, the perso­ na of Hortensius compared the respective worth of rhetoric and philosophy, praising the former and denying the ness of the latter or its constituting parts, physics, log­ ic, and ethics. The fragment must have formed attack useful­ on physics, part of his cosmology and theology. In particular, Hortensius' remark must have been part of the attack on the Platonic and Stoic conceptions of divine providence by means of which one divine principle governs the world. The sugges­ tion that this fragment's discussion of one divine principle 189 Frg> 62 straume-Zimmermann = frg. 47 Div. Inst. 1,7,3-5. 190 STRAUME-ZIMMERMANN 1990, 347-8. Grilli ap. Lact 345 was complemented by a related attack on the innumerable gods of the Epicureans, thus representing Sceptical argumentation in utramque unlikely. If this had been the is partero, 191 case, Lactantius might have referred to such an argument his in defense of monotheism. Furthermore, it is probable that the persona of Cicero responded to Hortensius' attack on the Platonic and Stoic conceptions of divine providence with the traditional argument that the order of the universe necessi­ tates the existence of divine guidance. 192 A Ciceronian ply to Hortensius 1 alleged re­ attack on Epicurean theology, however, is not documented. The defense of Epicurean theolo­ gy by Cicero may have been considered to be inappropriate; its absence from the dialogue would thus be unexceptio­ and nal . However, it is more in accord with the agenda behind the Epi­ attack on philosophy that Hortensius deliberately used curean arguments one divine principle in order to against that demolish the Platonic and Stoic position, thus showing any philosophical consideration of theology rested on ques­ erat demonstrandum. quod - tionable logical premises and therefore was useless There was no need to attack the Epicu­ rean position, namely that there existed many deities, since this was a common sense view shared which did not require by philosophical popular sentiment, proof. An Epicurean would have been in a position to agree with Hortensius' gument from ar­ common sense, since Epicurean epistemology was based on the fact that the true 191 STRAUME-ZIMMERMANN 1990, 347-8. perception 192 prg. 97 Straume-Zimmermann = frg. 83 Grilli 402,17 M; cf. Cic. Fin. 1,63-4, 4,11-2. of the ap. gods Nonius 346 (TipoXfjipLC) was reinforced not only by the images of the gods received in accepted societal Hortensius dreams but also by the persistence of commonly and cultural about them. 193 was finally won over to the cause of philosophy. It is far easier to imagine that Epicurean ideas an alleged proponent of doctrines admitted defeat and became convinced of the usefulness of philosophy than to assume that a Sceptic would have been won over, in particular since the definition of philosophy as a Socratic enterprise, as proposed by Car- neades 1 New Academy, seems to have been reserved for Cicero himself in the closing speech of the dialogue. 194 It has already been seen in the past that the anthropo­ morphizing tone of the fragment in question suggests an Epi­ curean pedigree for Hortensius 1 argument. 195 The gods of the Epicureans, although blessed and imperishable, were concei­ ved as living and anthropomorphic founder beings, not by roughly contemp­ with Cicero's treatises of the forties, portrayed the gods as bodily entities with lifestyles and sentiments our the himself, at least by Epicureans in the first centu­ ry. The third book of Philodemus 1 De dis, orary if own, except like that the gods are indestructible. In that work, Philodemus thought in strikingly anthropomorphic, not to say anthropocentric, terms. For instance, it was not dif­ ficult to identify the language in which the gods would be 193 Cf. LONG & SEDLEY 1987, 1,78-97. 194 Frg. 102 Straume-Zimmermann = frg. 115 gust. De Trin. 14,26. 195 E.g. GRILLI 1962, 85-6. Cf. Id. 1992. Grilli ap. Au­ 347 communicating: it was Greek, the language of philosophy culture. 196 The third book of Philodemus' De parallel or dis also and provides a for Hortensius 1 argument from happiness. Happiness blessedness Hortensius, (eudcuuovLcx, beatitude) , according to cannot be achieved in loneliness and isolation. If there were just one god, he would be lonely and therefore unhappy. Since perfect happiness or blessedness is acteristic of the a char­ divine, there cannot be one single god. The gods of the Epicureans were self-contained material bod­ ies which existed individually (TOI X<XT<X iiepoe) , just as hu­ mans were individuals. According to Epicurean doctrine, hap­ piness could not be achieved in isolation but only through the concept of justice and friendship happiness in in society. Perfect an Epicurean community could only be achieved »when everything will be full of justice and mutual friendship«. 197 This state is perfected in the gods where, habitation of the according to Philodemus 1 portrait in the third book of De dis, justice and mutual friendship among the gods guarantees their eternal eu&aiuovLd. The Epicurean doctrine of communality per se started from premises similar to those of the Peripatetic concept of the zoon politikon which must participate in a community in order to develop and prove its virtuous life, 198 and those of the Chrysippean idea universe as a of the habitation, a true city in the Stoic sense, 196 DIELS 1916 for a preliminary text; cf. Philodemus De pietate lines 124-99 = OBBINK 1996, 114-8 for a summary treatment by the same author. On Epicurean doctrine, see LONG & SEDLEY 1987, 1,139-49, esp. 148-9. 197 Diog. Oen. new fr. 21,1,4-14. Cf. LONG & SEDLEY 1987, 1,125-39. 1 98 E.g. Arist. NE 1097 b 9; 1099 b 4; passim; above, 2.5. 348 common to gods and (wise) men as rational beings governed by reason. 1 " impossi­ In expounding the Epicurean doctrine about the of bility the of existence only one god, the persona of Hortensius in the dialogue may have been attracted by the similarity between popular sentiment about the nature of the and Epicurean epistemology which argued for the tradi­ gods tional anthropomorphism of the divine pantheon from is It tion. percep­ not impossible (though ultimately unprovable) dissenting that Hortensius also voiced his concern that any philosophical speculation on the nature of the gods not only was useless, systems. 200 The of charge but also posed a threat to traditional belief Epicureans atheism took delight in turning the back on other philosophical schools. In that debate, the number of the gods becomes a central issue. The Epicureans asserted »that there exist not only all those [gods] that the Greeks say but many more as well«, since the endless number of immortals matches beings. This assertion is part the number of mortal of the critique of Stoic theology in Philodemus 1 De Pietate. 201 For Philodemus, as for other Epicureans, the that suggestion there might be only one god was an assault on the doc­ trine of the original and true conception of the gods as 1987, SEDLEY & 199 Cf. SCHOFIELD 1991, 57-92; LONG 1,394-401, esp. 399. 200 E.g. Varro RD frg. 8 Cardauns: ... quae facilia intra parietes in schola quam extra in foro ferre possunt aures; above, 2.8.1. 201 HENRICHS 1974, 12-26 for a preliminary text; cf. OBBINK 1996, 552-5. Cic. ND 1,50 is a contemporary parallel whose relation to Philodemus 1 treatise is not entirely clear; the latter may have been used as a source by Ci­ cero. 349 many bodily entities from perception and on their individua­ xcnrd uepoc. Following Epicurus, he could therefore T<X lity the preserving by claim that the Epicureans preserved religion many gods, whereas others like Antisthenes (fr. 39A Deby cleva Caizzi) postulated the existence of many gods only convention (xcnrd vouov) but of only one god in reality (xaT<x like the Stoics, deceived others into thinking or, cpuaiv) ; that there were many gods while in fact assuming that there was only one. 202 The interpretation of the names of the gods played a crucial role in that debate. On the one hand, ety­ mological allegory by metonymy or the de- grammatical construction, or rationalization, of names reduced the divi­ ne pantheon to only one divine principle which manifested itself in various forms. 203 On the other hand, the preserva­ tion of the gods' names by the Epicureans was only a natural consequence of the preservation of the gods' individuality as conceived from perception. 204 The Epicurean conception of the divine was, as seen, anthropomorphic in we have perception. Epicureans therefore not only were inclined to interpret the many gods in anthro­ one pomorphic terms, but insinuated that their reduction to ought principle to follow the same anthropomorphic model. »If there was only one god«, asks Hortensius, »how could achieve he happiness?« This anthropomorphic image was rejected by Lactantius, who regarded this question as monotheism. But an attack on should we assume that a Platonic or Stoic, 202 Stoics: HENRICHS 1974, 20-2. 203 Cf e FEENEY 1991, 8-11, 31-7; LONG 1992; 358-67. 204 E.g. Epic. Ad Herod. 77; De nat. 12-3. OBBINK 1996, 350 thinking in terms of a divine principle rather anthropomorphic god, than of would have understood the attack in a similar way? In other words, does Hortensius attack a losophical an monotheism< in phi­ a polytheistic environment? And would such an alleged philosophical monotheism have provided an alternative presenting to traditional religious belief systems, a monothetic belief system of a metaphysical di­ mension in a society where religion did allegedly not pro­ vide an ethical dimension? 205 The philosophical critique of anthropomorphic depictions of the divine is at least as old as Xenophanes. 206 The Pre- socratics, Platonists, Peripatetics or Stoics postulated one highest deity governing the world, one first principle from which the deities of the traditional pantheon and to which they were were derived subordinate. Modern evolutionist theories have accepted such philosophical views as a confir­ mation of the inevitable passage from polytheism to mono­ theism, and have defined the former as a transition stage in the religious evolution. Yet, there is a crucial difference between the first principle of philosophical speculation and the one god of the Old Testament. Consider the following passage, representative of Stoic thought: 205 Cf. e.g. GRILLI 1962, 85: »Noi sappiamo dello sforzo degli Stoici per una concezione monoteistica propria della filosofia, di fronte a quella politeistica della poesia e della religione statale; sappiamo d'altra parte che gli Epicurei accoglievano la concezione tradizionale.« On the supposed ethical dimension of >philosophical religion<, add e.g. NOCK 1933; GRIFFIN 1989, 36-7: »Philosophy thus played an important role in a society where religion had little metaphysics and less ethics.* For the context and a critique, see further above, 2.3. 206 Cf e FEENEY 1991, 5-56, for a brief survey. 351 et sciendum Stoicos dicere unum esse deum cui nomina variantur pro actibus et of fid is. Unde etiam duplicis sexus nimina esse dicuntur, ut cum in actu sunt mares sint, feminae cum patiendi habent naturam. 207 This text, while postulating the being, existence divine one of allow for the popular notion of male and female can deities. As a result, popular religious thropomorphic an­ and practices conceptions of personalized divinities do not dei­ impair the validity of the philosophical notion of one ty, even though Stoic thought does not conceive of this di­ prin­ vine being as anthropomorphic: postulating one divine ciple which governs the world does not lead to its personal­ Hortensius' rhetorical question - »if there was ization. 208 only one god, how could he achieve happiness?« not does - a Stoic position, since the Stoic concept of the invalidate the divine principle, unlike the Epicurean notion of gods, is not anthropomorphic. However, the contrast between a divine principle and anthropomorphic pantheon did not result in dualistic think­ ing. For instance, in a monotheistic text like the various the Stoic thought Midrash of God must not obfuscate the fact that names one personalized god is addressed. 209 By contrast, or an Platonic did not conceive of one divine being and the many deities of popular polytheism in terms of an exclu­ sive dualism. Yet, an exclusive dualism ought to be expected from a notion of the divine which leads directly into the monotheistic thinking of Jewish or Christian belief systems. The fact that this dualism does not exist in Platonic or 207 SVF 2,1070 = Serv. Aen. 4,638. Cf. Serv. Georg. 2,326. 208 E.g. SVF 2,1057-60. 209 Cf. AMIR 1978. 352 Stoic philosophy suggests a non-monothetic context for these philosophical speculations about the divine. The logical reduction of many possible subjects to one guiding principle resulted in the philosophical postulation which governs the world. In inversion of of one entity anthropomorphism, its attributes were incorruptability, immortality and incorporeality - attributes which necessarily led to assigning divine qualities to this entity. However, unlike truly mono­ theistic options, philosophical speculation did not have substantiate its divine principle; it was already sub­ stantiated as a logical operation, and did not have ceive to to re­ external authorization by being traced back to an om­ nipresent personalized deity. At the same time, the specula­ tive nature of the divine principle meant that gain theological quasi-metaphorical first did not significance either, as it never lost its quality: the notion of »dvdYxr|«, »the principle* or »the unmoved mover«, »voOc« or »ratio«, »6 Se6c« or »deus« did not become a dental it entity personalized transcen­ which would receive worship. Due to the lack of this theological dualism, the existence of a divine prin­ ciple in philosophical thought never meant that the polythe­ istic principle of one god among many gods, personalized and anthropomorphic, would be questioned. This principle did not provide a monothetic construct which could herald the victo­ ry of monotheism. Consider, for instance, Xenophanes' state­ ment, illustrative of that paradox: 210 eCc dsoc, ev TE deoLai xat dvdpanioiai uEYtaTOC; or Valerius reflective of the impact of Orphism: 210 Frg. 23 D-K = 170 Kirk/Raven/Schofield. Soranus' dictum, 353 luppiter omnipotent, regum rerumque deumque progenitor genetrixque, deum deus, unus et cranes. 211 Either text postulates a henotheistic construct which at the same time refers to, and indeed is geared to, its polythe­ istic context. 212 4.3.2 The constitution of belief systems at Rome The monothetic constructs of modern scholarship ing the concern­ organization of a >religion<« facilitate the (mod­ ern) attribution of meaning to the behaviour of the histori­ cal agents. Polythetic organizational principles, by con­ trast, appear to make such an attribution impossible: »A dynamically changing polytheistic system is an ex­ ceedingly problematic place in which to find the ground­ ing for a question like >What were the religious beliefs of Augustus?<. This man ... was participant in and ob­ ject of various new and traditional cults at Rome and throughout the Empire, and an initiate into the myste­ ries of Eleusis since the age of thirty-two. He was ac­ claimed in marble, bronze, papyrus and song as the des­ cendant of Venus and the son of Divus Julius. He was the vice-regent of Jupiter, founder of a new temple of Jupi­ ter the Thunderer, and always carried a sealskin with him as protection against thunderstorms. In which of these contexts is the >core< of belief to be found?« 213 >Belief< is here seen as a monothetic construct which it is difficult to apply in a polytheistic context. This view does not take into account that even within polythetic contexts there are still principles of organization which attempt limit to and make sense of different >belief systems< in rela­ tion to a variety of different deities and religious con­ cepts. Polytheism is not just an agglomeration of many gods, 211 Frg. 2 FLP with Courtney's commentary. 212 Cf. HORNUNG 1990; GLADIGOW 1998, 320. In general, see CANCIK-LINDEMAIER 1979. FRANSOIS 1957 discusses the di­ vine attributes of these philosophical constructs. 213 FEENEY 1998, 13-4. Cf. ibid. 18-9. 354 but possesses an internal structure, even though its logic may escape the modern beholder. research into For instance, ethological behaviour suggests that both individuals and groups are able to conceptualize their interacting with a maximum of up to twenty personalized divine subjects without forfeiting a sense of meaningful interaction. Within these limits, interacting with a number of subjects can be percei­ ved without the feeling that one's the coherence of action or personal identity is compromised; and the attribution of meaning or belief to one's own interactive behaviour is possible even within such a polythetic context. 214 The conceptualization of a polytheistic pantheon follows similar rules: choices allow various differentiated for >belief flexibility in systems< while at the same time maintaining a meaningful sense of personal or identity. The applying communal individual creation of pantheons is a common feature. Ennius, for instance, depicted those twelve deities of the Roman pantheon who, in adaptation of the pantheon traditional of the Twelve Gods in the Hellenic world, were en­ tertained at the lectisternium of 217: luppiter, sta, Minerva, Ve­ Ceres, Diana, Venus, Mars, Mercurius, Neptu- nus, Vulcan and Apollo. 215 Varro in the preface to rusticae luno, replaced Ennius' his Res di urban! by twelve deities qui maxime agricolarim duces stint: luppiter and Tellus, Sol and Luna, Ceres and Liber, Robigus and Flora, Minerva and Venus, 214 Cf. GLADIGOW 1983, 295; Id. 1998, 317; above, 2.7.4 2.8. and 215 Enn. Ann. 240-1 Skutsch ap. Apul. De deo Socr. 2; cf. Livy 22,9,7-10,9. Their bronze statues stood at the foot of the Capitoline hill: Varro RR 1,4: quorum imagines auratae stant. For Hellenic precedents, see LONG 1987, esp. 235-7; E. SIMON, Gnomon 63 (1991), 46-50. 355 Bonus and Lympha Eventus. 216 Vergil, in his invocation of the deities of agriculture, accepted Sol and Luna, Liber and Ceres, and Minerva, but replaced the others by the Fauni, Neptunus, Aristaeus, Pan, Triptolemus and Silvanus. dryads, Vergil added Augustus to this list, celebrating prin- the ceps* numinose power. 217 Vitruvius, in his normative account of sacred the topography of a typical Roman city and its pantheon, listed the Capitoline triad, Mercurius, Serapis, and Isis Apollo, Pater Liber, Hercules, Mars, Venus, Volca- nus, Ceres, and ... ceteri di . However, Vitruvius 1 model has not yet materialized in the archaeological record. 218 It would be misleading to assume that these pantheons provided truthful representations of the religious system of Rome. Instead, they help us to identify individual elements con­ of the overall system which could be put to use in the of private and public religious communication and wor­ text ship. 219 It was essential to possess a local religious know­ ledge which specified how to move between these tions how and to identifica­ use the religious options to hand. This local religious knowledge was acquired in the context of the the religious system of Rome; traditions and customs formed internalized basis of ritual orthopraxy and informed belief systems. Individual choices were constrained by structure and expectations provided the infra­ by the city. Yet, we must not exclude the possibility that they filled in the gap by developing personalized conceptions of their 216 217 218 219 gods which Varro RR 1,4-6 (deos consentis). Verg. Georg. 1,5-23 (twelve gods), 24-42 (Augustus). Vitr. 1,7; BENDLIN 1997, 50, for literature. Communication: above, 3.1.4. 356 went beyond the external framework provided by the civic administration of religion. A passage which serves to illustrate the failure of cur­ rent concepts of Late Republican religion the to conceptualize cognitive and emotional dimension of religion is Ovid's account of the pompa circensis in Rome. The persona poet, sitting of the next to his mistress, rejects identification with most of the divinities carried along in the procession. At last, Venus is praised; and Ovid asks her, and Victo­ ria's, divine support of his erotic undertaking. 220 sed iam pompa venit: linguis animisque favetei tempus adest plausus: aurea pompa venit. prima loco fertur passis Victoria pinnis: hue ades et meus hie fac, dea, vincat amor! plaudite Neptuno, nimium qui creditis undisl nil mihi cum pelago, me mea terra capit. plaude tuo Marti, milesi nos odimus arma, pax iuvat et media pace repertus amor. auguribus Phoebus, Phoebe venantibus adsiti artifices in te verte, Minerva, manusl ruricolae, Cereri teneroque adsurgite Bacchol Pollucem pugiles, Castora placet eques! nos tibi, blanda Venus, puerisque potentibus arcu plaudimus: inceptis adnue, diva, meis daque novae mentem dominae; patiatur amari. adnuit et motu signa secunda dedit. The motivational dimension of Ovid's response to the 45 50 55 statue of Venus is worth noting. By moving, Venus signifies approv­ al: she is ETifixooQ, or exaudiens; at least, this is the poet's interpretation, divining that the moves of the statue are »favourable signs« - an example of private divination in the public context of games organized by the city. 221 Of 220 Ovid Am. 3,2,43-58. 221 Statues in processions: Lucr. 2,624-5: (Mater magna) magnas invecta per urbis munificat tacit a mor tails mirta salute. Exaudiens: VERSNEL 1981, 26-37. Moves of statues used for divination: KRAUSS 1930, 176-9. Cf. P. Yale Inv. 299 (198/99 ^CE) : ... jif)TE 6id XG>ii<xaiac dxaXy.dTcov (sic) with ... Td UTTEp J. REA, avOpCDTtOV TLC El6EV<XL TlpOOTTOLECddOU ZPE 27 (1977), 151-6. Luc. Dea Syr. 36-7 357 the course, poet speaks with his tongue in cheek, for even a dea 59-62). This procession, therefore, evokes in (ibid. maior is him, Venus has to yield to his mistress who, to the persona of the poet a wide range of disparate images of meaning, which are constrained by local religious religious the mobilize knowledge, yet which at the same time poet's bias. This is a far cry from the functionalist in­ interior terpretation of ritual in terms of societal integration and communal meaningfulness. 2 2 2 The Ovidian passage outlines persona the of which functions the the poet assigns to the deities taking part in the pompa circensis. The their of differentiation divine functions reflects the differentiation of cult alternatives. options are compatible. The exclusivity of different These three goddesses; Hippolytus favours Artemis, to worship fails but polythetic illustrates, worship of including its henotheistic predications, in one place and at a certain time another example options stand in relation to their local and temporal contextualization, since the deity, to Aphrodite - in either case with disastrous results for the human. In real life, as the Ovidian one of and theological reflection: Paris chooses between mythology pay level the at religious alternatives is conceptualized deity in another does not place or Rather, the ability to move between exclude at worshipping another time. 223 different deities and interprets divination by the move of statues as priestly imposture. On this tradition, see POULSEN 1945. 222 Cf. LUKES 1975, bias<. esp. 305, for >the mobilization 223 Cf. GLADIGOW 1990, esp. 246-7; Id. 1998, 319-20; 2.7.2. of above, 358 cults characterizes the poet's religious behaviour in their a complex market of cult alternatives at Rome. social structures in complex societies are reflected at the creasing polytheistic respective level of their social pantheons. 224 In­ complexity is represented by an increasing hierarchization number of gods as well as by the subsequent of that fact the Several studies have drawn attention to their respective position. Yet, the differentiation of a cir­ pantheon does not necessarily resemble socio-political cumstances: Graeco-Roman paganism, for example, never linked the complexity of its polytheistic system to a discussion of pluralism or the issue of multinationality in the political Roman Empire. 225 Models which conceptualize the relationship between different deities pantheon the patriarchic normally sociomorphic. The be structured by means of gender differences: can gender are of the nature. supreme society's At the same time, theological specula­ tion questions the gender of a particular scrupulousness the reflects god divinity. 226 The over ritual performance includes the formula sive mas sive femina or si divus si diva. 227 The formula must not be taken as representing an instance of the Romans' supposed inability to conceptualize superhuman beings about whose very substance they were unclear. On the contrary, the 224 Cf. LEMERT 1974; VERNANT 1990, 296-7; Id. 1998, 318-20. 225 MOMIGLIANO 1987, 315-21. 101-19; GLADIGOW 226 E.g Laevius frg. 26 FLP and possibly Licinius frg. 7 FLP, discussing the bisexuality of Venus. 1983, Calvus 227 ILLRP 291-3; Cato Agric. 139; Macrob. Sat. 3,9,7 with APPEL 1909, 80; ALVAR 1985; HICKSON 1993, 41-3. VERSNEL 1981, 15-6 and PHILLIPS 1997, 131 8 add non-Roman paral­ lels. On the related case of taking care over addressing a deity by the right name, see Macrob. Sat. 3,9,9; APPEL 1909, 75; VERSNEL 1981, 15; HICKSON 1993, 33-4, 41, 54. 359 differences gender employs application of a formula which prior personalization, or the attempt of personali­ implies zation, of the deity addressed. 228 The famous evocatio of divinity local town Cilician during of a P. Servilius Vatia's capture of the Isaura Vetus in 75 this illustrates attitude: Serveilius C(ai) f(ilius) imperator, / hostibus victeis, Isaura Vetere / capta, captiveis venum dateis, / sei deus seive deast, quoius in / tutela oppidum vetus Isaura / fuit, -vac.- votum solvit. The general's use of the phrase sei deus seive dea addresses the problem which resulted from the votum under­ routinely prior to the evocatio. For to fulfil the votum, which taken is to ensure the deity's appropriate future worship, his her identity has to be sought by the victorious general. Servilius Vatia resorts to the categorization of concerned as personalized, if in the deity unkown, god or goddess. In this particular case, the unknown deity's evocatio result or did not his or her transfer to Rome or in the conceptual integration into the Roman pantheon. Instead, the inscrip­ tion probably belonged to the deity's new shrine or an altar dedicated by Servilius Vatia in Asia minor. 229 The division of labour and the attribution of functions to particular deities of the pantheon reflects the differen­ tiation of social roles in the human world. The Ovidian pas228 See KOCH 1937, 31-2; ULF 1982, 155-9. Koch never instrumentalized this insight, but postulated the »demythicization« of Roman religion instead: above, 2.2. 229 AE 1977,237 no. 816. For P. Servilius Vatia's career, see MRR 2,99 and 3,197; for his pontificate, cf. SZEMLER 1972, 130. BLOMART 1997, 99-102, 107-8 makes the import­ ant distinction between a deity's evocatio and the sepa­ rate issue of his/her - potential though far from requi­ site - transfer to Rome. Contra e.g. BEARD 1994, 743-4. 360 sage is illustrative in this respect too: Victoria is con­ nected with victory, and Neptune with the sea; Mars ents repres­ the military realm; Apollo is in charge of augury, and the Diana is the patroness of hunting; Minerva is linked to Ceres and Bacchus to agriculture; Castor and Pol­ artisans, lux represent the realm of sport activities; goddess of love. Venus the is The list could be extended ad infinitum. The provinces of individual deities comprise all realms of social behaviour. As a consequence, a structuralist approach postulates the distinction between deities in terms of di­ vine faculties and construes a polytheistic system which is intelligible by virtue of its classifying structure. 230 Yet, the view that a restricted range of functions can be assign­ ed to particular deities is problematic. As we have seen in the case of luppiter optimus maximus, the great particular combined a deities large number of functions, as there was no neat division between divine realms but rather lapping in responsibilities and over­ competition between deities, cults and temples. 231 Moreover, scholars do not sufficiently take into account how epitheta defined additional functions as well as condi­ tioned and further differentiated deities in new, and poten­ tially unexpected, ways. At Rome, there was not just one 230 Cf. DUMEZIL 1970, 174-5, 684-91, passim; VERNANT 1990, 101-19; ZAIDMAN & SCHMITT PANTEL 1992, 176-214, esp. 183-6. For critique of Dumezil's >archaic triad< of lup­ piter, Mars and Quirinus, defended by reference to its archetypical Indo-European origins, see BELIER 1991; CORNELL 1995, 77-9. 231 Competition: above, 4.1.3. The parallel with contempora­ ry Hindu society is illustrative in this respect: there, the position of the gods in a polytheistic context is always relational rather than substantive; cf. BASTIN & FLUEGEL 1988. 361 Stator; not simply one Hercules, but Hercules Vic­ luppiter Capitolio or Minerva Medica; Minera but tor or Hercules Musarum; not merely one Minerva in or Tonans luppiter, but luppiter optimus maximus, luppiter functions attached the meaning further enlarged possible attributions of the and choices of worshippers. 232 Most importantly, however, all of classifications the mentioned above could be combined and choices and thus permitted a refinement of interpretations would be continued at will. Furthermore, a personali­ which fur­ zed conception of the divinities of the pantheon would ther complicate the problem of their identity. To the perso­ na Ovid's Amores, as we have seen, Victoria was not the in goddess of the military sphere, but his in success guaranteed erotic endeavour; Venus was not Genetrix, the ancestral The deity of the lulii, but the goddess of love. of study in Late Republican Rome shows that the political polytheism not and public sphere was, as far as religion is concerned, with homologous society in general. The potential of this approach, both in terms of the differentiation of and choices in terms of individual motivational processes, could only be outlined. As I have tried to show, even when the options presented by Rome's public religious sys­ using Ovid's tem worshippers such as the persona of develop consequence, Rome's polytheistic could society. As a the question about the >identity< of Roman re­ ligion had to be abandoned in favour and poems alternatives of religious behaviour within the dif­ ferentiated system of how, religious with what of the inquiry into motives, individual worshippers would 232 See GLADIGOW 1981; Id. 1990, 242. 362 make use of the religious alternatives which were available to them in the city of Rome; and a new methodological frame­ work has been proposed which may be capable of addressing the latter inquiry more comprehensively than other currently available to the student of Roman religion. models CONCLUSION »Why / Sir, we know very little about the Romans.« Boswell's Life of Johnson (ed. R. W. Chapman, 1970), 464 Many modern of models Roman Wissowa's from religion, >Staatskultus< to the new paradigm of >civic religion<, suc­ cumb to the view that the religious history of Late Republi­ can Rome can be written in dualistic terms: individual reli­ gion either reflective of state religion or must be re­ is garded as deviant. The assumptions view are which underlie such a worth recapitulating. Firstly, individual worship under the Republic addressed those deities who were recogni­ zed by the state. By choosing the deities chose, individual which the state religion reflected the official pantheon of >Staatsreligion< or civic religion. Secondly, Roman reli­ gion was ritual observance that had a public and social sig­ nificance, but lacked an motivation. Thirdly, and independent activity. of individual as a result of these two assump­ tions, individual cult activity was cult level homologous with civic For both individuals and the community must concur on the purpose and meaning of ritual performance once any potentially deviant individuality has been excluded from the analysis, and once the meaning of religion has been lo­ cated in orthopraxy. Fourthly, religious choice, once it has 364 been defined in a purely externalist manner, can only become manifestly deviant when it falls outside the official pan­ theon. On such a view, religious evolution velopment became the de­ from embedded religion to individual interests in new cults and new forms of religious organization outside the traditional area of religion. The critique of these models had to start with the re­ assessment of the categories of belief or individual motiva­ tion. For only the positive revaluation of these categories enables us to undermine the assumptions which underlie those models. Yet, my emphasis on the existence of individual lief be­ systems in Roman religion does not entail returning to a Schleiermacherian position. We must not dichotomize vidual spirituality and indi­ public or social religious behav­ iour. We do not have to postulate that the Romans were vent believers or that they had spiritual inclinations. As Dr Johnson noticed, we are hardly capable of their individual religious reconstructing belief systems. However, the a priori exclusion of the category personal of religious belief, of commitment, individual morality or spirituality is equally unwarranted, since this exclusion starts from lematic fer­ prob­ methodological preconceptions. These preconceptions preclude an understanding of a wide range religious phenomena which are of cultural and suggested by the data: the distinction of sacred and civic domains, the differentiation of religious choices within the framework of the city's re­ ligious infrastructure, the conceptualization of privacy, or the personalization of divinities and beliefs. If we exclude categories like belief from our analytical framework, these 365 phenomena represent a stage in the religious evolution which can only be misunderstood as disintegrational. If an interpretative model which we apply includes those categories, these phenomena become unexceptional. As an implicit quence/ the conse­ religious development from the Republic to the Empire regains a sense of continuity which it was denied by earlier models of Roman religion. Many studies of Roman religion are geared lisms of to the dua­ belief versus scepticism or social religious obs­ ervance versus individualization. It is one advantage of the model of differentiation that the actual ratio of and sceptics at Rome, or of those committed to the cause of civic religion and those who were not, becomes The differentiation of should unimportant. interrelated, yet at the same time autonomous, realms of social activity at Rome we believers entails that expect individuals to combine different forms of social and religious behaviour without forfeiting a sense of personal identity. Q. Cornificius included his membership in the college of augurs in an inscription presumably dating from 45. He was an orator, held civic offices, wrote a Stoicizing work De etymis deorum - and composed erotic poetry. 1 Q. Lutatius Catulus, the consul of 102 and adherent of Car- neades' New Academy, wrote poetry that used religious langu­ age in order to invoke a pseudo-religious atmosphere, but that dealt with a homoerotic sujet: Constiteram exorientem Auroram forte salutans cum subito a laeva Roscius exoritur, 1 ILLRP 439. Priesthood: MRR 2,292. Erotic poetry: Ovid rr. 2,436. Cf. MRR 3,76; FLP pp. 225-7; RAWSON 1991, 272-88. 366 pace mihi liceat, caelestes, dicere vestra, mortalis visus pulchrior esse deo. 2 Both >scepticism< traditionally and >individualization< been seen as elements in the process that un­ dermined the embeddedness of religion in public and cal life. have Scholars have politi­ rightly rejected the Durkheimian dichotomy of sacred and secular domains. Its replacement, the notion of embeddedness and lack of differentiation, how­ ever, is equally misleading. Here I endeavoured to develop a more nuanced view, which accounts for the fact that in Roman society religious and civic domains were conceptualized as interrelated, yet autonomous, entities. While the system made use of religious rituals, social and political activity actually moved between two In 133 the political differentiated realms. pontifex maximus Scipio Nasica, dressed capite velato, killed Tiberius Gracchus. By covering his head, Sci­ pio Nasica signified that he was performing the consecratio of Tiberius, whose attempt to seize royal power had made him a homo sacer in the eyes of his opponents. After Tiberius' death, the Senate sent a delegation of Xviri sacris dis to faciun- the temple of Ceres at Henna in Sicily. The embassy was understood as a means of expiation, since Tiberius' rannical behaviour could ty­ be interpreted as a violation of tribunician sacrosanct! tas, with which the goddess was con­ nected. Yet, while these religious rituals and symbols could be instrumentalized to legitimize Scipio Nasica's and the Senate's killing of Tiberius, the action's legal and politi- 2 Frg. 2 FLP; DAHLMANN 1981. Scepticism: Cic. Lucull. 12; STRAUME-ZIMMERMANN 1990, 381. No priesthood is attested. 367 cal legitimation could be questioned by fellow-senators and the populace. 3 My approach was designed to show that and constitution of Roman religion the organization cannot be defined in terms of >Staatskultus< or civic religion. Even the of local religion proves problematic when applied to the personal belief systems of worshippers religious system. concept Religious in a decentralized behaviour, both in its public aspect and its individual agenda, was not determined by political system or the confines of the city of Rome. As noted above, >Roman religion< is an umbrella-term whose precision the im­ means that it is employed by scholars in a number of different ways. The cults, religions and belief systems which constitute what we call >Roman religion< defy a simple classification in terms of the identity of Roman religion. The polytheistic system at Rome generated different meanings to the native agent; it continues to generate different meanings to the modern observer. 3 Consecratio and delegation: App. BCiv 1,2,16; Plut. Ti. Gracch. 19,4; Cic. Verr. 2,4,108 with SPAETH 1990. Prob­ lems of legitimation: Plut. Ti. Gracch. 19,3, passim; NIPPEL 1995, 32, 60-4. BIBLIOGRAPHY AHN, G. (1993), »Monotheismus - Polytheismus. Grenzen und Moglichkeiten einer Klassifikation von Gottesvorstellungen«, in M. Dietrich & O. Loretz (eds), Mesopotamica - Ugaritica - Biblica. FS fur K. Bergerdorf, Neukirchen, 1-24. ALFOLDI, A. (1985) , Caesar in 44 v. Chr. I: Studien zu Caesars Monarchie und ihren Wurzeln, Bonn. ALFOLDY, G. (1991) , »Augustus und die Inschriften: Tradition und Innovation. Die Geburt der imperialen Epigraphik«, Gymnasium 98, 289-324. ALVAR, J. (1985), »Materiaux pour 1'etude de la formule sive deus sive dea«, Numen 32, 236-273. (1994), »Escenografia para un recepcion divina: la introduccion de Cibeles en Roma«, DHA 20, 149-69. AMIR, Y. (1978), »Die Begegnung des biblischen und des philosophischen Monotheismus als Grundthema des jiidischen Hellenismus«, Evangelische Theologie 38, 2-19. AMPOLO, C. A. (1981), »I gruppi etnici in Roma arcaica: posizione del problema e fonti«, in Gli Etrusci e Roma. Festschrift M. Pallottino, Rome, 45-70. (1991), »Commentary«, in MOLHO & al. 1991, 113-9. ANDERSON, B. (1991), Imagined communities: reflections on the origins and spread of nationalism2 , London. ANDERSON, W. S. (1979), »Plautus f >Trinummus<: the absurdity of official morality«, Traditio 35, 333-45. ANDRE, J.-M. (1975), »La philosophie religieuse de Ciceron. Dualisme academique et tripartition Varronienne«, in A. Michel & R. Verdiere (eds), Ciceroniana. Hommages a Kazimierz Kumaniecki (Roma Aeterna 9), Leiden, 11-21. ANDREAU, J. (1977), »Fondations privees et rapports sociaux en Italie romaine (ler-IIIe s. ap. J.-C.)«, Ktema 2, 157-209. APPEL, G. (1909), De Romanorum precationibus (RGW 7,2), GieSen. ARCE, J. (1988), Funus imperatorum: los funerales de los emperadores romanos, Madrid. ARDLEY, G. W. R. (1973), »Cotta and the theologians*, Prtidentia 5, 33-50. ARGETSINGER, K. (1992), »Birthday rituals: friends and patrons in Roman poetry and cult«, Classical Antiquity, 175-93. ATKINS, E. M. (1990), »Domina et regina virtutum: justice and societas in De Officiis«, Phronesis 35, 258-289. ATTRIDGE, H. W. (1978) , »The philosophical critique of religion under the early Empire*, ANRN 2,16,1, 45-78. AUBERT, J.-J. (1994), Business managers in ancient Rome: a social and economic study of institores, Leiden. 369 AUSBUTTEL, F. (1982), Untersuchungen zu den Vereinen im Westen des romischen Reiches, Kallmimz. Ax, w. (1995) , »Disputare in utramque partern. Zum literarischen Plan und zur dialektischen Methode Varros in LL 8-10«, RhM 138, 146-77. AYMARD, J. (1953), »Scipion 1'Africain et les chiens du Capitole*, R&L 31, 111-6. BADIAN, E. (1990), »The consuls, 179-49 BC«, Chiron 20, 371-413. BARKER, J. TH. (1994), Living and working with the gods. Studies of evidence for private religion and its material environment in the city of Ostia (100-500 AD) (Dutch Monographs on Ancient History and Archaeology 12), Amsterdam. BASLEZ, M.-F. (1977), Recherches sur les conditions de penetration et de diffusion des religions orientales a Delos (Ile-Ier s. av. notre ere), Paris. BARGHOP, D. (1994), Forum der Angst. Eine historisch-anthropologische Studie zu Verhaltensmustern von Senatoren im romischen Kaiserreich, Frankfurt/Main & New York. BARNES, T. D. (1995), »Statistics and the conversion of the Roman aristocracy*, JRS 85, 135-47. BASTIN, R. & P. FLUEGEL (1988), »The Hindu pantheon: addendum and critique«, Man n.s. 23,4, 757-9. BAUMAN, R. A. (1990), »The suppression of the Bacchanals: five questions*, Historia 39, 334-48. BEARD, M. (1986), »Cicero and divination: the formation of a Latin discourse*, JRS 76, 33-46. (1987), »A complex of times: no more sheep on Romulus' birthday*, PCPhS 33, 1-15. (1990), »Priesthood in the Roman republic*, in BEARD & NORTH 1990, 19-48. (1991), »Writing and religion: >Ancient Literacy< and the function of the written word in Roman religion*, in J. H. Humphrey (ed.), Literacy in the Roman world (JRA Suppl. 3), Ann Arbor, 35-58. (1993), »Looking (harder) for Roman myth: Dumezil, declamation and the problems of definition*, in GRAF 1993, 44-64. (1994a), »Religion«, in CAH IX 2 , Cambridge, 729-68. (1994b), »The Roman and the foreign: the cult of the Great Mother in Imperial Rome*, in N. Thomas & C. Humphrey (eds), Shamanism, history, and the state, Ann Arbor, 164-90. BEARD, M. & M. H. CRAWFORD (1985), Rome in the late republic: problems and interpretations, London. BEARD, M. & J. A. NORTH (1990) (eds), Pagan priests, London. BEAUJEU, J. (1969), »Les constantes religieuses du scepticisme*, in J. Bibauw (ed.), Hommages a M. Renard Collection Latomus 102), Brussels, 2, 61-73. BEHRENDS, O. (1978), »Grabraum und Grabfrevel im romischen Recht*, Abh. Gottingen 3,113, 85-106. (1981), »Die Rechtsformen des romischen Handwerks*, in H. Jankuhn & al. (eds), Das Handwerk in vor- und fruhgeschichtlicher Zeit, Gottingen, 1, 141-203. 370 BELIER, W. W. (1991), Decayed gods: origin and development of Georges Dumezil's >Ideologie tripartite<, Leiden. BELL, C. (1992), Ritual theory, ritual practice, New York & Oxford. BENDLIN, A. (1995), »Review of GRAF (ed.) 1993«, JRS 85, 265-6. (1997), »Peripheral centres - central peripheries: religious communication in the Roman empire«, in CANCIK & RUPKE 1997, 35-68. BERGER, P. L. (1965), »Ein Marktmodell zur Analyse okumenischer Prozesse«, Internat. Jb. fur Religionssoziologie 1. BERGER, P. L. & TH. LUCKMANN (1966), The social construction of reality, New York. BERTOLINI, F. (1978), »Wilamowitz a Wissowa e Praechter«, Quaderni di Storia 7, 185-210. BLAGG, T. F. C. (1986), »The cult and sanctuary of Diana Nemorensis«, in M. Henig & A. King (eds), Pagan gods and shrines of the Roman Empire, Oxford, 211-9. BLANCK, H. (1966-67), »Zwei Corneliersarkophage«, RM 73/4, 72-7. BLEICKEN, J. (1957), »Kollisionen zwischen sacrum und publicum«, Hermes 85, 446-80. BLOCK, M. (1992), Prey into hunter. The politics of religious experience, Cambridge. BLOMART, A. (1997), »Die evocatio und der Transfer >fremder< Gotter von der Peripherie nach Rom«, in CANCIK & RDPKE 1997, 99-111. BODEI GIGLIONI, G. (1977), »Pecunia fanatica: 1'incidenza economica dei templi laziali«, RSI 89, 33-76. BOMER, F. (1990), »>The slave had no mana at all< - ?«, Gymnasium 97, 3-14. BOMER, F. & P. HERZ (1981) , Untersuchungen fiber die Religion der Sklaven in Griechenland und Rom. I. Die wichtigsten Kulte in Rom und im lateinischen Westen 2 , Wiesbaden. BOLLMANN, B. (1993), Romische Vereinshauser. Untersuchungen zu den scholae der Berufs-, Kult- und Augustalen-Kollegien in Italien, dissertation Heidelberg. BOYANCE, P. (1972), Etudes sur la religion romaine, Rome. BRADEEN, D. W. (1958-59), »Roman citizenship per magistratum« l CJ 54, 221-8. BRADLEY, K. R. (1991), Discovering the Roman family. Studies in Roman social history, New York and Oxford. BREMMER, J. N. (1993), »Three Roman aetiological myths«, in GRAF (ed.) 1993, 158-174. (1994), Greek religion (Greece & Rome New Surveys in the Classics 24), Oxford. (1995), »The family and other centres of religious learning in antiquity«, in J. W. Drijvers & A. A. MacDonald (eds), Centres of learning: learning and location in pre-modern Europe and the Near East, Leiden, 29-38. BROMBERG, B. (1939-40), »Temple banking in Rome«, Economic History Review 10, 128-31. BRUNT, P. A. (1982), »The legal issue in Cicero, Pro Balbo«, CQ 32, 136-47. 371 (1987), Italian manpower (225 B.C.-A.D. 14) 2 , Oxford. (1989), »Philosophy and religion in the Late Republic*, in GRIFFIN & BARNES 1989, 174-98. BUHL, W. F. (1987), Kulturwandel. Fur eine dynamische Kultursoziologie, Darmstadt. BURCKHARDT, L. A. (1990), »The political elite of the Roman republic: comments on recent discussion of the concepts >nobilitas< and >homo novus<« / Historia 39, 77-99. BURKERT, W. (I985a), Greek religion. Archaic and classical, Oxford. (1985b) , »Herodot iiber die Cotter: Polytheismus als historisches Problem«, MH 42, 121-32. (1995), »Greek poleis and civic cults: some further thoughts*, in M. G. Hansen & K. Raaflaub (eds), Studies in the ancient Greek polls (Historia ES 95), Stuttgart, 201-10. BURNYEAT, M. F. (1982), »Gods and heaps«, in M. Nussbaum & M. Schofield (eds), Language and logos: studies in ancient Greek philosophy presented to G. E. L. Owen, Cambridge, 315-38. BURTON, P. J. (1996), »The summoning of the Magna Mater to Rome (205 B.C.)«, Historia 45, 36-63. BUSCH, J. A. (1979), »Cybernetics III. A system-type applicable to human beings«, Cybernetica 22, 89-103. CAMERON, A. (1991), Christianity and the rhetoric of Empire. The development of a Christian discourse, Berkeley & Los Angeles. CANCIK, H. (1973), »R6mischer Religionsunterricht in apostolischer Zeit. Bin pastoralgeschichtlicher Versuch zu Statius, Silve V 3,176-184«, in H. Feld & J. Nolte (eds), Wort Gottes in der Zeit. Festschrift fur K. H. Schelkle, Diisseldorf, 181-97. (1978), »Die republikanische Tragodie«, in E. Lefevre (ed.), Das romische Drama, Darmstadt, 308-47. (1985-86), »Rome as sacred landscape: Varro and the end of Republican religion in Rome«, Visible Religion 4/5, 250-65. (1994), »La religione Romana«, in G. Filoramo (ed.), Storia della religion! 7. Le religion! antiche, Rome, 349-408. (1995), »Occulte adhuc colunt. Repression und Metamorphose der romischen Religion in der Spatantike«, in KIPPENBERG & STROUMSA 1995, 191-201. (1996), »Carmen und sacrificing. Das Saecularlied des Horaz in den Saecularakten des Jahres 17 v. Chr.«, in R. Faber & B. Seidensticker (eds), Worte, Bilder, Tone. Studien zur Antike und Antikerezeption, 99-113. CANCIK, H. & J. ROPKE (eds) (1997), Romische Reichsreligion und Provinzialreligion, Tubingen. CANCIK-LINDEMAIER, H. (1979), »Die Allmacht und das Pantheon«, in P. Eicher (ed.), Gottesvorstellimg und Gesellschaftsentwicklung (Forum Religionswissenschaft 1), Munich, 64-6. CANFORA, L. (1989), Le vie del classicismo, Rome & Bari. CARDAUNS, B. (I960), Varros Logistoricus irber die Gotterverehrung (dissertation Cologne) , Wiirzburg. 372 CATALANO, P. (1974), »La divisions del potere in Roma*, Studi Grosso 6, 667-91. CARLSEN, J. (1994), »CIL X 8217 and the question of temple land in Roman Italy*, in Id. & al. (eds), Landuse in the Roman empire (Analecta Romana Instituti Danici Suppl. 22) , Rome, 9-15. CHAMPLIN, E. (1991), Final judgements. Duty and emotion in Roman wills 200 B.C.-A.D. 250, Berkeley & Los Angeles. CHANTRAINE, H. (1972) , »Zur Entstehung der Freilassung mit Burgerrechtserwerb in Rom«, ANRW 1,2, 59-67. CLASSEN, C. J. (1985), Recht - Rhetorik - Politik. Untersuchungen zu Ciceros rhetorischer Strategie, Darmstadt. COARELLI, F. (1977), »Public building in Rome between the second Punic war and Sulla«, PBSR 45, 1-13. (1984), »Iside Capitolina, Clodio e i mercanti di schiavi«, in N. Bonacasa & al. (eds), Allesandria e il mondo ellenistico-romano. Studi in onore di A. Adriani, Rome, 461-75. (1996), Revixit ars. Arte e ideologia a Roma. Dai modelli ellenistici alia tradizione repubblicana, Rome. COLE, S. G. (1995), »Civic cult and civic identity*, in Sources for the ancient Greek city-state. Acts of the Copenhagen Polls Centre 2 (Historisk-filosofiske Meddelelser 72), Copenhagen, 292-325. COLE, T. (1990), Democritus and the sources of Greek anthropology2 , Atlanta. COLPE, C. (1975), »Die Vereinbarkeit historischer und struktureller Bestimmungen des Synkretismus«, in A. Dietrich (ed.), Synkretismus im syrisch-persischen Kulturbereich (Abh. Akad. Gottingen 3,96), Gottingen, 15-37. CONNOR, W. R. (1988), »>Sacred< and >secular<. Hiera kai hosia and the Athenian Classical concept of the state*, Ancient Society 19, 161-88. CORBIER, M. (1991), »Divorce and adoption as Roman familial strategies*, in B. M. Rawson (ed.), Marriage, divorce and children in ancient Rome, Oxford, 47-78. CORNELL, T. J. (1991), »Rome: the history of an anachronism*, in MOLHO & al. 1991, 53-69. (1995), The beginnings of Rome. Italy and Rome from the Bronze Age to the Punic Mars (c. 1000-264 BC), London & New York. CORNELL, T. J. & K. LOMAS (1995) (eds), Urban society in Roman Italy, London. CRAPANZANO, V. (1992), »Some thoughts on hermeneutics and psychoanalytic anthropology*, in SCHWARTZ & al. 1992, 294-307. CRAWFORD, M. H. (1981), »Italy and Rome*, JRS 71, 153-60. (1989), »Aut sacrom aut poublicom«, in P. Birks (ed.), New perspectives in the Roman law of property. Essays for B. Nicholas, Oxford, 93-8. (1996), »Italy and Rome from Sulla to Augustus*, in CAH X 2 , Cambridge, 414-33. DAHLMANN, H. (1981), »Das Rosciusepigramm des Q. Lutatius Catulus*, Gymnasium 88 (1981), 24-44. 373 DAVID, J. M. & al. (1973), »Le commentariolum petitionis de Q. Ciceron«, ANRN 1,3, 239-77. DE FRANCISCIS, A. (1966), »Note sui praedia Dianae Tifatinae«, RAAN 41, 241-6. DEGRASSI, A. (1962-67), Scritti vari di antichita, Venice & Trieste. DE RUYT, C. (1983), Macellum: marche alimentaire des remains, Louvain. DESNIER, J.-L. (1987), »Stips«, Revue d'histoire des religions 204, 219-30. DETREVILLE, D. j. (1987), Senatus et religio: aspects of the Roman senate's role in the state religion during the Republic, Ph. D. thesis Chapel Hill. DE VISSCHER, F. (1966), »Les fondations privees en droit remain classique«, in Etudes de droit romain public et prive, Milan. DIELS, H. (1890), Sibyllinische Blatter, Berlin. (1916), Philodemos uber die Gotter. Drittes Buch (APAW Philos.-Hist. Kl. 1916, 4 and 6), Berlin [1917]. DIHLE, A. (1988), »Heilig«, RAC 14, 1-63. DOBBELAERE, K. & J. LAUWERS (1974), »Definition of religion - a sociological critique«, Social Compass 20, 535. DOBERT, R. (1973a), Systemtheorie und die Entwicklung religioser Deutungssysteme, Frankfurt/Main. (1973b), »Zur Logik des Ubergangs von archaischen zu hochkulturellen Religionssystemen«, in K. Eder (ed.), Die Entstehung von Klassengesellschaften, 330-363. Frankfurt/Main. DORING, K. (1978) , »Antike Theorien iiber die staatspolitische Notwendigkeit der G6tterfurcht«, Antike und Abendland 24, 43-56. DORRIE, H. (1974) , »Polybios iiber pietas, religio und fides (zu Buch 6, Kap. 56). Griechische Theorie und romisches Selbstverstandnis«, in Melanges de philosophie, de litterature et d r histoire ancienne offerts a P. Boyance (Coll. de 1'Ecole franc, de Rome 22), Rome, 251-72. (1978a), »Das Gute Beispiel - Kalon Hypodeigma«, in Studien zur Religion und Kultur Kleinasiens. FS K. F. Dorner (EPRO 66), Leiden, 1, 246-62. (1978b), » Vestigia summorum virorum«, Grazer Beitrage 7, 207-220. (1980), »Uberlegungen zum Wesen antiker Frommigkeit«, in E. Dassmann & K. S. Frank (eds), Pietas. Festschrift fur B. Kotting, Minister, 3-14. (1986), »Zu Varros Konzeption der theologia tripertita in den Antiquitates rerum divinarum«, in R. Altheim-Stiehl & M. Rosenbach (eds), Beitrage zur altitalischen Geistesgeschichte. Festschrift fur G. Radke, Minister, 76-82. DOLL, R. (1972), »Rechtsprobleme im Bereich des romischen Sakralrechts«, ANRN 1,2, 283-94. DUMEZIL, G. (1970), Archaic Roman religion, Baltimore & London. DURAND, J.-L. & J. SCHEID (1994), »>Rites< et >religion<. Remarques sur certains prejuges des historiens de la religion des Grecs et des Romains«, ASSR, 23-43. 374 DURKHEIM, fi. (1964), The elementary forms of the religious life, London (fr. Les formes elementalres de la vie religieir.se [1912]) . ECK, W. (1984), »Senatorial self-representation« / in F. Millar & E. Segal (eds), Caesar Augustus: seven aspects, Oxford, 129-67. (1992) , »Cura viarirm und cur a operum publicorum als kollegiale Amter im friihen Prinzipat«, Xlio 74, 237-45. EDWARDS, C. (1996), Writing Rome. Textual approaches to the city, Cambridge. ERRINGTON, R. M. (1988), »Aspects of Roman acculturation in the East under the Republic«, in P. Kneissl & V. Losemann (eds), Alte Geschichte und Wissenschaftsgeschichte. Festschrift fur K. Christ, Darmstadt, 140-57. ERSKINE, A. (1994), »The Romans as common benefactors*, Historia 43, 70-87. FASCIANO, D. (1982), »Deos ... esse nemo negat«, in Melanges offerts en hommage a E. Gareau (Cahiers des Etudes anciennes 14), Ottawa, 191-5. FEAR, A. T. (1996), Rome and Baetica. Urbanization in Southern Spain c. 50 BC-AD 150, Oxford. FEARS, J. R. (1975), »The coinage of Q. Cornificius and augural symbolism on Late Republican denarii«, Historia 24, 592-602. (1981), »The cult of Jupiter and Roman imperial ideology«, ANRN 2,17,1, 3-141. FEENEY, D. (1991), The gods in epic. Poets and critics of the Classical tradition, Oxford. (1998), Literature and religion at Rome. Cultures, contexts, and beliefs, Cambridge. FEIL, E. (1986), Religio. Geschichte eines neuzeitlichen Grundbegriffs vom Fruhchristentum bis zur Reformation (Forschungen zur Kirchen- und Dogmengeschichte 36), Gottingen. FENTRESS, E. (1978), »Dii Mauri and dii patrii«, Latomus 37, 507-16. FLACH, D. (1994) , Die Gesetze der friihen romischen Republik. Text und Kommentar (in Zusammenarbeit mit Stefan von der Lahr), Darmstadt. (1996) (ed.), Marcus Terentius Varro: Gesprache iiber die Landwirtschaft, Buch 1, Darmstadt. FLAIG, E. (1993), »Politisierte Lebensfiihrung und asthetische Kultur. Eine semiotische Untersuchung am romischen Adel«, Historische Anthropologie 1, 193-217. FLAMBARD, J.-M. (1977), »Clodius, les colleges, la plebe et les esclaves. Recherches sur la politique populaire au milieu du ler siecle«, M&FRA 89, 115-56. (1981), »Collegia Compitalicia: phenomene associatif, cadres territoriaux et cadres civiques dans le monde romain a 1'epoque republicaine«, Ktema 6, 143-66. FLASCHE, R. (1989) , ^Religiose Entwiirfe und religiose Wirkungen von Religionswissenschaftlern«, in P. Antes & D. Pahnke (eds), Die Religion von Oberschichten. Religion - Profession - Intellektualismus, Marburg, 203-17. 375 FLOWER, H. I. (1996), Ancestor masks and aristocratic power in Roman culture, Oxford. FORSYTHE, G. (1994), The historian L. Calpurnius Pi so Frugi and the Roman annalistic tradition, Lanham. FOWDEN, G. (1993), Empire to commonwealth. Consequences of monotheism in late antiquity, Princeton. FRANQOIS, G. (1957), Le polytheisme et 1'emploi au singular des mots THEOS, DAIMON, Paris. FRAYN, J. M. (1993), Markets and fairs in Roman Italy, Oxford. FREDOUILLE, J.-CL. (1986), »Heiden« 7 RAC 13, 1113-49. FREIER, H. (1963), Caput velare, dissertation Tubingen. FRIER, B. W. (1985), The rise of the Roman jurists. Studies in Cicero's Pro Caecina, Princeton. FROHLICH, TH. (1991), Lararien- und Fassadenbilder in den Vesuvstadten. Untersuchungen zur >volkstumlichen< pompejanischen Malerei, Mainz. (1995), »Pompejanische Lararienbilder«, Antike Welt 26,3, 203-10. FUCHS, M. (1982), »Ein Bildnis der Kybele aus Formiae: Zur Kunst urn Pompeius«, in D. Mannsperger & al. (eds), Praestant interna. Festschrift fur U. Hausmann, Tubingen, 72-8. FUGIER, H. (1963), Recherche sur l f expression du sacre dans la langue latine, Paris. GABBA, E. (1984), »The collegia of Numa: problems of method and political ideas«, JRS 74, 81-6. (1991), Dionysius and the history of archaic Rome, Berkeley & Los Angeles. GADAMER, H.-G. (1978), Die Idee des Guten zwischen Platon und Aristoteles, SB Heidelberg 1978,3. GAGE, J. (1960), Apollon romain. Essai sur le culte d'Apollon et le developpement du >ritus Graecus< a Rome des origines a Auguste (BEFAR 182), Paris. GALLINI, C. (1970), Protesta e integrazione nella Roma antica, Bari. GARDNER, J. F. (1993), Being a Roman citizen, London. GARNSEY, P. D. A. (1988), Famine and food supply in the Graeco-Roman world. Responses to risk and crisis, Cambridge. (1998), Cities, peasants and food in Classical antiquity. Essays in social and economic history, Cambridge. GAWLICK, G. & W. GORLER (1994), »Cicero«, in H. Flashar (ed.), Die Philosophie der Antike 4. Die hellenistische Philosophie, 2 vols, Basle, 2, 991-1168. GEERTZ, C. (1973), The interpretation of cultures. Selected essays, New York. (1975), »Common sense as a cultural system«, The Antioch .Review 33, 5-26. GEORGE, M. (1997), »Repopulating the Roman house«, in RAWSON & WEAVER 1997, 299-319. GIZEWSKI, C. (1989), »Mores maiorum, regimen morum, licentia. Zur Koexistenz catonischer und plautinischer Sittlichkeitsvorstellungen«, in W. Dahlheim & al. (eds), Festschrift fur R. Werner zu seinem 65. Geburtstag, Constance, 81-105. 376 GLADIGOW, B. (1981) / »Gottesnamen (Gottesepitheta) I (allgemein)«, RAC 11, 1102-1238. (1983), »Strukturprobleme polytheistischer Religionen« / Saeculum 34, 292-304. (1990) , »Chrestai theois - Orientierungs- und Loyalitatskonflikte in der griechischen Religion« / in Ch. Elsas & H. G. Kippenberg (eds), Loyalitatskonflikte in der Religionsgeschichte, Berlin & New York, 237-51. (1994), »Zur Ikonographie und Pragmatik romischer Kultbilder«, in H. Keller & N. Staubach (eds), Iconologia sacra. Festschrift fur K. Hauck, Berlin, 9-24. (1995), »Struktur der Offentlichkeit und Bekenntnis in polytheistischen Religionen«, in KIPPENBERG & STROUMSA 1995, 17-35. (1997a), »Erwerb religioser Kompetenz. Kult und Offentlichkeit in den klassischen Religionen«, in H. Binder (ed.), Religiose Kormunikation (BAG 7), Trier, 103-118. (1997b), »Historische Orientierungsmuster in komplexen Kulturen: Europaische Religionsgeschichte und historischer Sinn«, in J. Riisen (ed.), Historische SinnbiIdung, Hamburg. (1998), »Polytheismus«, HrwG 4, 314-23. GNOLI, F. (1979), Ricerche sul crimen peculatus, Milan. GOODMAN, M. (1994), Mission and conversion. Proselytizing in the religious history of the Roman empire, Oxford. GRAF, F. (1993) (ed.), Mythos in mythenloser Gesellschaft. Das Paradigma Roms (Colloquium Rauricum 3), Stuttgart and Leipzig. (1993a), »Der Mythos bei den Romern. Forschungs- und Problemgeschichte«, in GRAF 1993, 25-43. GRIFFIN, J. (1985), Latin poets and Roman life. London. GRIFFIN, M. F. (1962), »De brevitate vitae«, JRS 52, 104-13. (1994), »The intellectual developments of the Ciceronian age«, in CAH IX 2 , Cambridge, 689-728. GRIFFIN, M. & J. BARNES (1989) (eds), Philosophia togata. Essays on philosophy and Roman society, Oxford. GRILLI, A. (1962), M. Tulli Ciceronis Hortensius. Edidit commentario instruxit Albertus Grilli, Varese & Milan. (1992), »Democrito nell' Hortensius e altri echi del dialogo«, in Id., Stoicismo, epicureismo e letteratura, Brescia, 291-318. GRUEN, E. S. (1990), Studies in Greek culture and Roman policy (CCS n.s. 7), Leiden. (1992), Culture and national identity in Republican Rome, London. GUARDUCCI, M. (1971), »L'isola Tiberina e la sua tradizione ospitaliera«, RAL 26, 267-81. GUILLAUMONT, F. (1989), »Ciceron et le sacre dans la religion des philosophes«, Bulletin de 1'Association Guillaume Bude, 56-71. GWYN GRIFFITHS, J. (1975), Apuleius of Madauros. The Isis-book, Leiden. GWYN MORGAN, M. (1990), »Politics, religion and the games in Rome, 200-150 B.C.«, Philologus 134, 14-36. 377 HABERMAS, J. (1981), Theorie des kommunikativen Handelns4 , 2 vols, Frankfurt/Main. HAGENDAHL, H. (1967), Augustine and the Latin Classics, Stockholm. HAHN, A. (1986), »Differenzierung, ZivilisationsprozeS, Religion. Aspekte einer Theorie der Moderne« / in F. Neidhardt & al. (eds.), Kultur und Gesellschaft, Opladen, 214-231. HAMPL, F. (1980), »R6mische Politik in republikanischer Zeit und das Problem des >Sittenverfalls<« / in R. Klein (ed.), Das Staatsdenken der Romer3 , Darmstadt, 143-77. HANSON, J. A. (1958), »Plautus as a source-book for Roman religion«, TAPA 90, 48-101. HANSEN, M. H. (1991), The Athenian democracy in the age of Demosthenes, Oxford. (1993), »The polis as a citizen-state«, in Id. (ed.), The ancient Greek city-state (Historisk-filosofiske Meddelelser 67), Copenhagen, 7-29. HARDIN, M. A. (1983), »Applying linguistic models to the decorative arts: a preliminary consideration*, Semiotica 46, 309-22. HARMON, D. P. (1978), »The family festivals at Rome«, ANEW 2,16,2, 1592-1603. HARRIS, W. V. (1979), War and imperialism in Republican Rome, 327-70 B.C., Oxford. (1989), Ancient literacy, Cambridge/Mass. & London. HAYNE, L. (1992), »Isis and Republican politics*, Acta Classica 35, 143-9. HEBERLEIN, F. (1988), »Eine philologische Anmerkung zu Romanas caeremonias recognoscere (Acta Cypriani 1)«, in U. Kindermann & al. (eds.), Festschrift firr P. Klopsch (Gottinger Arbeiten zur Germanistik 492), Gottingen, 83-100. HECKHAUSEN, H. & B. WEINER (1980), »The emergence of a cognitive psychology of motivation*, in P. C. Dodwell (ed.), New horizons in psychology 2 2 , Harmondsworth, 136-57. HEIBGES, U. (1969), »Religion and rhetoric in Cicero's speeches*, Latomus 28, 833-49. HENRICHS, A. (1975), »Two doxographical notes: Democritus and Prodicus on religion*, HSCP 79, 93-123. (1985), »Der Glaube der Hellenen: Religionsgeschichte als Glaubensbekenntnis und Kulturkritik*, in W. M. Calder iii & al. (eds), Wilamowitz nach 50 Jahren, Darmstadt, 263-305. HICKSON, F. V. (1993), Roman prayer language. Livy and the Aeneid of Vergil (Beitr. z. Altertumskunde 30), Stuttgart & Leipzig. HOLKESKAMP, K.-J. (1987), Die Entstehung der Nobilitat. Studien zur sozialen und politischen Geschichte der Romischen Republik im 4. Jhdt. v. Chr., Stuttgart. (1988), »Das Plebiscitum Ogulnium de sacerdotibus, RhM 131, 51-67. HOLSCHER, T. (1978). »Die Anfange romischer Reprasentationskunst*, MDAI(R) 85, 315-57. (1980), »R6mische Siegesdenkmaler der spaten Republik*, in H. A. Cahn & E. Simon (eds.), Tainia. .Roland Hampe zum 70. Geburtstag, Mainz, 1,351-71. 378 (1982) , »Die Bedeutung der Miinzen fur das Verstandnis der politischen Reprasentationskunst der spaten romischen Republik*, in T. Hackens & R. Weiller (eds), Proceedings of the Ninth International Congress of Numismatics, Luxemburg, 269-82. HOFSTEE, W. (1986), »The Interpretation of Religion - Some Remarks on the Work of Clifford Geertz«, in H. G. Hubbeling & H. G. Kippenberg (eds), Zur symbolischen Representation von Religion: Groninger Abhandlungen zu verschiedenen Symboltheorien, Berlin & New York, 70-83. HOHTI, P. (1975), »Aulus Caecina the Volaterran«, in Studies in the Romanization of Etruria (Acta Inst. Rom. Finland. 5), Rome, 409-33. HOPKINS, K. (1983), Death and renewal. Sociological studies in Roman history 2, Cambridge. (1991), »From violence to blessing: symbols and rituals in ancient Rome«, in MOLHO & al. 1991, 479-98. HORNUNG, E. (1990), Der Eine und die Vielen. Agyptische Vorstellungen 4 , Darmstadt. IRMSCHER, J. (1986), »Sulle origini del concetto Romania«, in Popoli e spazio romano tra diritto e profezia: atti del III seminario internazionale di studi storici >Da Roma alia terza Roma<, Rome, 421-9. (1994), »Der Terminus religio und seine antiken Entsprechungen im philologischen und religionsgeschichtlichen Vergleich«, in U. Bianchi (ed.), The notion of religion in comparative research, Rome, 63-73. JACZYNOWSKA, M. (1976), »Die wirtschaftliche Differenzierung der romischen Nobilitat am Ende der Republik« (1962), in H. Schneider (ed.), Zur Sozial- und Wirtschaftsgeschichte der spaten romischen Republik, Darmstadt, 214-36. JAEGER, M. (1997), Livy's written Rome f Ann Arbor/Michigan. JAL, P. (1961), »La propagande religieuse a Rome au cours des guerres civiles de la fin de la republique«, Antiquite Classique 30, 395-414. (1962), »Les dieux et les guerres civiles dans la Rome de la fin de la republique«, REL 40, 170-200. JEHNE, M. (1995) (ed.), Demokratie in Rom? Die Rolle des Volkes in der Politik der romischen Republik (Historia ES 96), Stuttgart. JOCELYN, H. D. (1966), »The Roman nobility and the religion of the Roman state«, Journal of Religious History 4, 89-104. (1982), »Varro's Antiquitates rerum divinarum and religious affairs in the Late Roman Republic*, Bulletin of the John Rylands Library 65, 148-205. JONES, C. P. (1976), »The Plancii of Perge and Diana Planciana«, HSCPh 80, 231-237. JONES, R. A. (1991), »La genese du systeme? The origins of Durkheim's sociology of religion*, in W. M. Calder iii (ed.), The Cambridge ritualists reconsidered (ICS suppl. 2), Atlanta/Ga., 97-121. 379 JOST, K. (1936), Das Beispiel und Vorbild der Vorfahren bei den attischen Rednern und Geschichtsschreibern bis Demosthenes (Rhetorische Studien 19), Paderborn. KASER, M. (1978), »Zum romischen Grabrecht« / SZ 95, 15-92. KAUFMANN, F.-X. (1989), Religion und Modern!tat: Sozialwissenschaftliche Perspektiven, Tubingen. KEHRER, G. (1988), Einfuhrung in die Religionssoziologie t Darmstadt. KIERDORF, W. (1980), Laudatio funebris. Interpretationen und Untersuchungen zur Entwicklung der romischen Leichenrede (Beitr. zur Klass. Philologie 106), Meisenheim am Glan. KIPPENBERG, H. G. & B. LUCHESI (1995) (eds), Lokale Religionsgeschichte, Marburg. KIPPENBERG, H. G. & G. G. STROUMSA (1995) (eds), Secrecy and concealment. Studies in the history of Mediterranean and Near Eastern religions, Leiden. KIRSOPP LAKE, A. (1937), »The supplicatio and Graecus ritus«, in R. P. Casey & al. (eds), Quantulacumque. Studies presented to Kirsopp Lake, London, 243-51. KNEER, G. & A. NASSEHI (1993), Niklas Luhmanns Theorie sozialer Systeme. Eine Einfuhrung, Munich. KNEISSL, P. (1994), »Die fabri ... in den Stadten Italiens und den westlichen Provinzen«, in E fontibus haurire. FS fur H. Chantraine, Paderborn, 133-46. KNEPPE, A. (1988), »Die Gefahrdung der securitas. Angst vor Angehorigen sozialer Randgruppen der romischen Kaiserzeit«, in I. Weiler (ed.), Soziale Randgruppen und AuBenseiter im Altertum, Graz, 165-76. KOCH, C. (1937), Der romische Juppiter, Frankfurt/Main. (1941), »Pietas«, RE 20,1, 1221-32. KOEP, L. (1962) , »Religio und ritus als Problem des friihen Christentums«, JbAC 5, 43-59. KOVES-ZULAUF, TH. (1978), »Plinius der Altere und die romische Religion*, ANRW 2,16,1, 187-288. KOLB, A. (1993), Die kaiserliche Bauverwaltung in der Stadt Rom, Stuttgart. KRAUS, C. S. & A. J. WOODMAN (1997), Latin historians (G&R New Surveys in the Classics 27), Oxford. KRAUSS, F. G. (1930), An interpretation of the omens, portents and prodigies recorded by Livy, Tacitus and Suetonius, Ph. D. Thesis Univ. of Pennsylvania. KUNKEL, W. (1967), Herkunft und soziale Stellung der romischen Juristen 2 , Graz, Vienna & Cologne. KUNKEL, W. & R. WITTMANN (1995), Staatsordnung und Staatspraxis der Romischen Republik. Zweiter Abschnitt: Die Magistratur (HdAW 3,2,2), Munich. LACEY, W. K. (1996), Augustus and the principate. The evolution of the system, Leeds. LA REGINA, A. (1976), »I1 Sannio«, in P. Zanker (ed.), Hellenismus in Mittelitalien, 2 vols, Abh. Gottingen 3,97. LATTE, K. (1960), Romische Religionsgeschichte, Munich. LAUSBERG, M. (1989), »Senecae operum fragmenta: Uberblick und Forschungsbericht«, ANRW 2,36,3, 1879-1961. 380 LEBEK, W. D. (1989), »Die Mainzer Ehrungen fur Germanicus, den alteren Drusus und Domitian (Tab. Siar. Frg. I 26-34; Suet. Claud. 1,3)*, ZPE 78, 45-82. (1993), »Roms Ritter und Roms Pleps in den Senatsbeschliissen fur Germanicus Caesar und Drusus Caesar*, ZPE 95, 81-120. LE BONNIEC, H. (1958), Le culte de Ceres a Rome des origines a la fin de la Republique (Etudes et commentaires 27) , Paris. LEFEVRE, E. (1987), »Plinius-Studien III. Die Villa als geistiger Lebensraum*, Gymnasium 94, 247-62. LEGARE, M. (1980), »Overlapping functional systems: a theory of vertebrate central nervous system function in terms of informal systems analysis*, Behavioral Science 25, 89-106. LEMERT, CH. C. (1974), »Cultural multiplexity and religious polytheism*, Social Compass 21, 241^53. LETTA, C. (1992), »I santuari rurali nell'Italia centro-appenninica: valori religiosi e funzione aggregativa«, MEFRA 104, 109-24. LEVENE, D. S. (1993), Religion in Livy (Mnemosyne Suppl. 127), Leiden. LEWY, H. (1928-29), »Das Sitzen nach dem Gebet«, Philologus 84, 378-80. LIEBERG, G. (1973), »Die theologia tripertita in Forschung und Bezeugung«, ANRW 1,4, 63-115. (1974), »Considerazioni sull'etimologia e sul significato di religio«, RFIC 102, 34. (1982), »Die theologia tripertita als ein Formprinzip antiken Denkens«, RhM N.F. 125, 25-35. LIEBESCHUETZ, J. H. W. G. (1967), »The religious position of Livy's history*, JRS 57, 45-55. (1979), Continuity and change in Roman religion, Oxford. LIND, L. R. (1979), »The tradition of Roman moral conservatism*, in C. Deroux (ed.), Studies in Latin literature and Roman history 7 (Collection Latomus 164), Brussels, 7-58. (1992), »The idea of the Republic and the foundation of Roman morality. Second part*, in C. Deroux (ed.), Studies in Latin literature and Roman history 6 (Collection Latomus 217), Brussels, 5-40. LINDER, M. & J. SCHEID (1993), »Quand croire c'est faire. Le probleme de la croyance dans la Rome ancienne*, Archives de sciences sod ales des religions 81, 47-61. LINDERSKI, J. (1986), »The augural law*, ANRW 2,16,3, 2146-312. (1995), Roman questions. Selected papers (Heidelberger Althistorische Beitrage u. Epigraphische Studien 20), Stuttgart. LINTOTT, A. (1978), »The Capitoline dedications to Jupiter and the Roman People*, ZPE 30, 137-44. (1993), Imperium Romanum. Politics and administration, London. LOMAS, K. (1995), »Urban elites and cultural definition: romanisation in southern Italy*, in CORNELL & LOMAS 1995, 107-20. 381 (1997), »The idea of a city: elite ideology and the evolution of urban form in Italy, 200 BC-AD 100*, in PARKINS 1997, 21-41. LONG, A. A. (1992), »Stoic readings of Homer«, in R. Lamberton & J. J. Keaney (eds), Homer's ancient readers: the hermeneutics of Greek epic's earliest exegetes, Princeton, 41-66. LONG, A. A. & D. N. SEDLEY (1987) (eds), The Hellenistic philosophers, 2 vols, Cambridge. LONG, CH. R. (1987), The Twelve gods of Greece and Rome (EPRO 107), Leiden. LOVE, J. R. (1991), Antiquity and capitalism. Max Weber and the sociological foundations of Roman civilization, London & New York. LUCE, T. J. (1968), »Political propaganda on Roman Republican coins (92-82 B.C.)«, AJA 72, 25-39. (1971), »Design and structure in Livy: 5.32-55«, TAPhA 102, 265-302. (1990), »Livy and Augustus, and the Forum Augustum«, in K. A. Raaflaub & M. Toher (eds), Between Republic and Empire: interpretations of Augustus and his principate, Princeton, 123-38. LUHMANN, N. (1977), Function der Religion, Frankfurt/M. (1980-89), Gesellschaftsstruktur und Semantik. Studien zur Wissenssoziologie der modernen Gesellschaft, 3 vols, Frankfurt/Main. LUKES, S. (1975), »Political ritual and social integration*, Sociology 9, 289-308. (1992), fonile Durkheim. His life and work: a historical and critical study 2 , Harmondsworth. LUTERBACHER, F. (1904), Der Prodigienglaube und Prodigienstil der Romer. Eine historisch-philologische Abhandlung2 , Burgdorf. MACMULLEN, R. (1980), »How many Romans voted?«, Athenaeum 58, 454-7. (1984), Christianizing the Roman Empire, AD 100-400, New Haven & London. MANTELL, I. (1979), Roman household religion, unpubl. Ph. D. thesis Edinburgh. MARTIN, H. G. (1988), »Die Tempelkultbilder«, in Kaiser Augustus und die verlorene Republik, Mainz, 251-63. MAZZOLI, G. (1984), »I1 problemo religioso in Seneca«, RSI 96, 953-1000. MELLOR, R. (1978), »The dedications on the Capitoline hill«, Chiron 8, 319-30. METZLER, D. (1981), »6konomische Aspekte des Religionswandels in der Spatantike: die Enteignung der heidnischen Tempel seit Konstantin«, Hephaistos 3, 27-40. (1985-86), »Anikonische Darstellungen«, Visible religion 4/5, 96-113. MILLAR, F. G. B. (1973), »The imperial cult and the persecutions*, in W. den Boer (ed.), Le culte des souverains dans l f empire romain (Entr. Fondation Hardt 19), Vandoeuvres & Geneva, 143-65. (1983), »Empire and city, Augustus to Julian: obligation, excuses and status*, JRS 73, 76-96. 382 (1984), »The political character of the classical Roman Republic*, JRS 74, 1-19. (1986) , »Politics, persuasion and the People before the Social War (150-90 B.C.)*, JRS 76, 1-11. (1989), »Political power in mid-Republican Rome: curia or comitium?«, JRS 79, 138-50. (1992), The emperor in the Roman world (31 BC-AD 337) 2 , London. (1995), »The last century of the Republic. Whose history?«, JRS 85, 236-43. MOLHO, A. & al. (eds.) (1991), City-states in Classical antiquity and Medieval Italy, Stuttgart. MOLTHAGEN, J. (1991), »Die ersten Konflikte der Christen in der griechisch-romischen Welt«, Historia 40, 42-76. MOMIGLIANO, A. (I984a), Settimo contribute alia storia degli studi classici e del mondo antico, Rome. (1984b), »The theological efforts of the Roman upper classes*, ClPh 79, 199-211. (1987), Ottavo contribute alia storia degli studi classici e del mondo antico, Rome. MOMMSEN, TH. (1887), Romisches Staatsrecht (Handbuch der Romischen Alterthumer 3,1), 3 vols, Leipzig. (1899), Romisches Strafrecht, Leipzig. (1907), Gesammelte Schriften 3: Juristische Schriften 3 (ed. B. Kiibler) , Berlin. MORA, F. (1995), »Theophore Namen als Urkunden lokaler Religion*, in KIPPENBERG & LUCHESI 1995, 101-17. MOREAU, PH. (1982). Clodiana religio. Un proces politique en 61 av. J.-C., Paris. (1988), »A propos de la lex arae de Corcelle*, in Hommages a H. Le Bonniec (Coll. Latomus 201), Brussels, 316-23. MORLEY, N. (1996), Metropolis and hinterland. The city of Rome and the Italian economy, 200 B.C.-A.D. 200, Cambridge. MORRIS, B. (1987), Anthropological studies of religion. An introductory text, Cambridge. MULLER, K. E. (1972/1980), Geschichte der antiken Ethnographie und ethnographischen Theoriebildung, 2 vols, Wiesbaden. MURR, M. (1948) , Pietas in der romischen Literatur bis zur augusteischen Zeit, dissertation Innsbruck. MURRAY, 0. (1990), »Cities of reason*, in Id. & S. R. F. Price (eds), The Greek city form Homer to Alexander, Oxford, 1-25. MUTH, R. (1978), »Vom Wesen romischer religio*, ANRN 2,16,1, 290-354. NEEDHAM, R. (1972), Belief, language and experience, Oxford. NEMETH, G. (1994), »Me6 5 ovdov eyPaXgv. Regulations concerning everyday life in a Greek temenos«, in R. Hagg (ed.), Ancient Greek cult practice from the epigraphical evidence, Stockholm, 59-64. NEUDECKER, R. (1988), Die Skulpturenausstattung romischer Villen in Italien, Mainz. NEVETT, L. (1997) , »Perceptions of domestic space in Roman Italy*, in RAWSON & WEAVER 1997, 281-98. 383 NICOLAI, K. (1963), »Feiertage und Werktage im romischen Leben, besonders in der Zeit der ausgehenden Republik und in der fruhen Kaiserzeit*, Saeculum 14, 194-220. NICOLET, C. (1967), »Arpinum, Aemilius Scaurus et les Tullii Cicerones*, R&L 45, 276-304. (1976), Le metier de citoyen dans la Rome republicaine 2 , Paris. NICOLIS, G. & I. PRIGOGINE (1977), Self-organization in nonequilibrium systems: from dissipative structures to order through fluctuation, New York. NIPPEL, W. (1990), Barbaren und Wilde. Alte Geschichte und Sozialanthropologie, Frankfurt/Main. (1995), Public order in ancient Rome, Cambridge. NOCK, A. D. (1933), Conversion: the old and the new in religion from Alexander the Great to Augustine of Hippo, Oxford. NO"RR, D. (1981) , »Zur Reskriptenpraxis in der hohen Kaiserzeit«, ZSS 98, 1-46. NOETHLICHS, K. L. (1996), Das Judentum und der romische Staat. Minderheitenpolitik im antiken Rom, Darmstadt. NORTH, J. A. (1976), »Conservatism and change in Roman religion«, PBSR 44, 1-12. (1979), »Religious toleration in Republican Rome«, PCPS n.s. 25, 85-103. (1980), »Novelty and choice in Roman religion«, JRS 70, 86-91. (1981), »The development of Roman imperialism*, JRS 71, 1-19. (1986), »Religion and politics, from republic to principate«, JRS 76, 251-8. (1989), »Religion in Republican Rome«, in CAH VII 2 , Cambridge, 573-624. (1992), »The development of religious pluralism*, in J. Lieu & al. (eds), The Jews among pagans and Christians, London & New York, 174-93. (1995), »Religion and rusticity* in CORNELL & LOMAS 1995, 135-50. NOVAK, K. (1986), Schleiermacher und die Fruhromantik. Eine literaturgeschichtliche Studie zum romantischen Religionsverstandnis und Menschenbild am Ende des 18. Jh. in Deutschland, Gottingen. OBBINK, D. (1996), Philodemus >On piety<. Part 7: Critical text with commentary, Oxford. OBER, J. (1989), Mass and elite in democratic Athens. Rhetoric, ideology, and the power of the people, Princeton. (1993), »The polls as society. Aristotle, John Rawls and the Athenian Social Contract*, in M. H. Hansen (ed.), The ancient Greek city-state (Historisk-filosofiske Meddelelser 67), Copenhagen, 129-160. ORLIN, E. M. (1997), Temples, religion and politics in the Roman Republic, Leiden. OUDEMANS, T. C. W. & A. LARDINOIS (1987), Tragic ambiguity, Leiden. 384 PACI, G. (1980a) , »Magister municipi in una nuova iscrizione di Tolentino e supplemento epigrafico tolentinate«, in Settima miscellanea greca e romana, Rome, 479-524. (1980b) , »Possibili tracce di statuti municipal! in alcune iscrizioni d'ltalia concernenti un particolare tipo di munificenza privata«, in C. Castillo (ed.), Novedades de epigrafia juridica Romana, Pamplona, 125-33. (1987), »Lex sacra da S. Vittore di Cignoli« / MGR 12, 115-36. PAIS, E. (1914) , »Le relazioni fra sacerdozi e le magistrature civili nella repubblica romana« / in Id., Ricerche sulla storia e sul diritto publico di Roma l, Rome, 273-335. PANCIERA, S. (1970-71), »Nuovi documenti epigrafici per la topografia di Roman antica«, RPAA 43, 109-34. PARKER, R. (1996), Athenian religion: a history, Oxford. PARKINS, H. M. (1997) (ed.), Roman urbanism. Beyond the consumer city, London & New York. PARSONS, T. (1966), »Religion in a modern pluralistic society«, The Review of Religious Research 7, 125-46. PENNER, H. (1971-72), »The poverty of functionalism«, History of Religions 11, 91-7. PEPIN, J. (1958), Mythe et allegorie, Paris. PHILLIPS, C. ROBERT (1986), »The sociology of religious knowledge*, ANRN 2,16,3, 2677-773. (1991a), »Misconceptualizing classical mythology«, in M. A. Flower & M. Toher (eds), Georgica. Greek studies in honour of G. Cawkwell (BIOS Suppl. 58), London, 143-51. (1991b), »Nullum crimen sine lege: socioreligious sanctions on magic«, in C. A. Faraone & D. Obbink (eds), Magika Hiera. Ancient Greek magic and religion, New York & Oxford, 260-76. (1992), »Roman religion and literary studies of Ovid's Fasti«, Arethusa 25, 55-80. (1997), »Cato the elder«, in M. Kiley & al. (eds), Prayer from Alexander to Constantine. A critical anthology, London and New York, 128-32. PIETILA-CASTREN, L. (1987), Magnificentia publica: the victory monuments of the Roman generals in the era of the Punic Wars, Helsinki. PIGHI, G. B. (1965) (ed.), De ludis saecularibus populi Romani Quiritium libri sex2 , Amsterdam. PISANI SARTORIO, G. & G. QUILICI GIGLI (1987-88), »A proposito della Tomba dei Corneli«, BCAR 92,2, 247-64. PLEKET, H. W. (1981), »Religious history as the history of mentality: the >believer< as servant of the deity in the Greek world«, in H. S. Versnel (ed.), Faith, hope and worship. Aspects of religious mentality in the ancient world, Leiden, 152-92. POSCHL, V. (1983), Literatur und geschichtliche Wirklichkeit, 2 vols, Heidelberg. PORTE, D. (1989), Les donneurs de sacre. Le pretre a Rome, Paris. POULSEN, F. (1945), »Talking, weeping, bleeding sculptures. A chapter of the history of religious fraud*, Acta Archaeologica 16, 178-95. PRICE, S. R. F. (1984), jRitualsr and power. The Roman imperial cult in Asia minor, Cambridge. 385 (1996) , »The place of religion: Rome in the early empire*, in CAH X 2 , Cambridge, 812-47. PRIGOGINE, I. (1976), »Order through fluctuation. Self-organization and social system*, in E. Jantsch & C. H. Waddington (eds), Evolution and consciousness, London, 93-126. PURCELL, N. (1983), »The apparitores: a study in social mobility*, PBSR 51, 125-73. (1994) , »The city of Rome and the plebs urbana in the Late Republic*, in CAH IX 2 , Cambridge, 644-88. (1995), »The Roman villa and the landscape of production«, in CORNELL & LOMAS 1995, 151-79. RADKE, G. (1959), »Vergils Cumaeum carmen*, Gymnasium 66, 217-46. (1980), »Anmerkungen zu den kultischen MaSnahmen in Rom wahrend des zweiten punischen Krieges*, WiiJbb N.F. 6b, 105-21. (1981), »Die Dei penates und Vesta in Rom«, ANEW 2,17,1, 343-73. RAPPAPORT, R. A. (1971), »Ritual, sanctity and cybernetics*, American Anthropologist 73, 59-76. RAUH, N. K. (1986), »Cicero's business friendships: economics and politics in the Late Roman Republic*, Aevum 60, 3-30. (1989), »Auctioneers and the Roman economy*, Historia 38, 451-71. (1993), The sacred bonds of commerce. Religion, economy, and trade society at Hellenistic Roman Delos, 166-87 B. C., Amsterdam. RAWSON, B. & P. WEAVER (eds) (1997), The Roman family in Italy: status, sentiment, space, Oxford. RAWSON, E. (1985), Intellectual life in Late Republican Rome, London. (1991) , .Roman culture and society. Collected papers, Oxford. RENFREW, C. (1994a), »The archaeology of religion*, in RENFREW & ZUBROW 1994, 47-54. (1994b), »Towards a cognitive archaeology*, in RENFREW & ZUBROW 1994, 3-12. RENFREW, C. & E. B. W. ZUBROW (eds) (1994), The ancient mind. Elements of cognitive archaeology, Cambridge. RIVES, J. B. (1995), Religion and authority in Roman Carthage from Augustus to Constantine, Oxford. ROBINSON, O. F. (1992), Ancient Rome. City planning and administration, London and New York. ROHDE, G. (1936), Die Kultsatzungen der romischen Pontifices, Berlin. ROLOFF, H. (1938), Maiores bei Cicero (dissertation Leipzig 1936), Gottingen. (1953), »Caerimonia«, Glotta 32, 101-38. ROMANELLI, P. (1963), »Magna mater e Attis sul Palatino e nelle sue adiacenze*, MAAL 46, 201-20. ROUSSELLE, R. J. (1989), »Persons in Livy's account of the Bacchic persecution*, in C. Deroux (ed.), Studies in Latin literature and Roman history 5 (Collection Latomus 206), Brussels, 55-65. 386 ROYCE, J. R. & A. POWELL (1983) , Theory of personality and individual differences: factors, systems, and processes, Englewood Cliffs. RUDOLPH/ K. (1994), »Inwieweit 1st der Begriff Religion eurozentrisch?«, in U. Bianchi (ed.), The notion of religion in comparative research, Rome, 131-9. RUPKE, J. (1990), Domi militiae. Die religiose Konstruktion des Krieges in Rom, Stuttgart. (1994), »Ovids Kalenderkommentar: Zur Gattung der libri fastorum«, Antike & Abendland 40, 125-36. (1995a), Kalender und Offentlichkeit. Die Geschichte der Representation und religiosen Qualification von Zeit in Rom, Berlin & New York. (1995b), »Was kostet Religion? Quantifizierungsversuche fur die Stadt Rom« / in KIPPENBERG & LUCRESi 1995, 273-87. (1995c), »Fasti: Quellen oder Produkte romischer Geschichtsschreibung?«, Klio 77, 184-202. (1996a), »Charismatics or professionals? Analyzing religious specialists*, Numen 43, 241-62. (1996b), »Innovationsmechanismen kultischer Religionen: Sakralrecht im Rom der Republik«, in H. Cancik & al. (eds), Geschichte - Tradition - Reflexion: FS fur M. Hengel zum 70. Geburtstag, Tubingen, 2,265-85. (1997), »R6mische Religion und >Reichsreligion<: Begriffsgeschichtliche und methodische Bemerkungen«, in CANCIK & ROPKE 1997, 3-23. SABBATUCCI, D. (1988), »Kultur und Religion«, HrwG 1, 43-58. SAHLINS, M. D. (1985), Islands of history, Chicago. SALEM, J. (1994), »Comment traduire religio chez Lucrece?«, EC 62, 3-26. SALLER, R. P. (1994), Patriarchy, property and death in the Roman family, Cambridge. SALLER, R. P. & B. D. SHAW (1984), »Tombstone and Roman family relations in the principate: civilians, soldiers and slaves«, JRS 74, 124-156. SALZMAN, M. R. (1993), »The evidence for the conversion of the Roman Empire to Christianity in book 16 of the Theodosian Code«, Historia 42, 362-78. SANDELIN, K.-G. (1991), »The danger of idolatry according to Philo of Alexandria*, Temenos 27, 109-50. SCHEID, J. (1981), »Le delit religieux dans la Rome tardo-republicaine«, in Le delit religieux dans la cite antique, Rome, 117-71. (1984), »Le pretre et le magistral:. Reflexions sur les sacerdoces et le droit public a la fin de la republique«, in Cl. Nicolet (ed.), Des ordres a Rome, Paris, 243-80. (1985), Religion et piete a Rome, Paris. (1987), »Polytheism impossible; or, the empty gods: reasons behind a void in the history of Roman religion*, History & Anthropology 3, 303-25. (1990), »Rituel et ecriture a Rome«, in A.-M. Blondeau & K. Schipper (eds), Essais sur le rituel 2 (Bibl. de 1'Ecole des Hautes Etudes, Sciences religieuses 95), Louvain, 1-15. 387 (1992), »Myth, cult and reality in Ovid's Fasti*, PCPhS 38, 118-31. (1993a), »Cultes, mythes et politique au debut de 1'empire*, in GRAF 1993, 109-27. (1993b), »Die Parentalien fur die verstorbenen Caesaren als Modell fur den romischen Totenkult«, Klio 75, 188-201. (1998) , »Graeco ritu: a typically Roman way of honouring the gods«, forthcoming in TAPhA. SCHEID, J. & J. SVENBRO (1989-90), Annuaire de 1'Ecole Pratique des Hautes Etudes. Section des Sciences Religieuses 98, 327-33. SCHIAN, R. (1973), Untersuchungen uber das >argumentum e consensu omnium< (Spudasmata 28), Hildesheim. SCHILLING, R. (1978), »Genius«, RAC 10, 52-83. (1979), Rites, cultes, dieux de Rome (6tudes et Commentaires 92), Paris. SCHLESIER, R. (1995), »Arbeiter in Useners Weinberg. Anthropologie und Religionsgeschichte«, in H. Flashar (ed.), Alter'tumswissenschaft in den 20er Jahren. Neue Fragen und Impulse, Stuttgart, 329-380. SCHMIDT, E. (1909), Kultubertragungen, GieSen. SCHMIDT, F. (1985), »Naissance des polytheismes (1624-1757)«, Archives des Sciences Sociales des Religions 59, 77-90. SCHNEIDER, H. J. (1979), »Zum Begriff der Ursache in den Sozialwissenschaften«, in J. Mittelstrass (ed.), Methodenprobleme der Wissenschaften vom gesellschaftlichen Handeln, Frankfurt/Main, 215-22. SCHNEIDER, K. (1995), Villa und Natur. Eine Studie zur romischen Oberschichtkultur im letzten vor- und ersten nachchristlichen Jahrhundert, Munich. SCHOFIELD, M. (1991), The Stoic idea of the city, Cambridge. (1993), »Plato on the economy«, in M. H. Hansen (ed.), The ancient Greek city-state (Historisk-filosofiske Meddelelser 67), Copenhagen, 183-96. (1995), »Cicero's definition of Res Publica«, in J. G. F. Powell (ed.), Cicero the philosopher. Twelve papers, Oxford, 63-83. SCHOTT, R. (1990), »Die Macht des Uberlieferungswissens in schriftlosen Gesellschaften«, Saeculum 41, 273-316. SCHWARTE, K.-H. (1989), »Die Christengesetze Valerians«, in W. Eck (ed.), Religion und Gesellschaft in der romischen Kaiserzeit (Kolner Historische Abhandlungen 35), Cologne, 103-63. SCHWARTZ, TH. & al. (1992) (eds), New directions in psychological anthropology, Cambridge. SEGAL, E. (1987), Roman laughter: the comedy of Plautus 2 , New York & Oxford. SEIWERT, H. (1981) , »Religiose Bedeutung als wissenschaftliche Kategorie«, The annual review of the social science of religion 5, 57-99. (1987) , »Religionen und Religion*, Ztschr. fur Missionswissenschaft und Religionswissenschaft 71, 225-30. SELINGER, R. (1994), Die Religionspolitik des Kaisers Decius. Anatomie einer Christenverfolgung, Frankfurt/Main. 388 SHATZMAN, I. (1972), »The Roman general's authority over booty*, Historia 21, 177-205. (1975), Senatorial wealth and Roman politics (Coll. Latomus 142), Brussels. SHENNAN, S. (1994), »Introduction: archaeological approaches to cultural identity*, in Id. (ed.), Archaeological approaches to cultural identity, London, 1-32. SHERWIN-WHITE, A. N. (1973), The Roman citizenship 2 , Oxford. SMITH, W. C. (1962), The meaning and end of religion, New York. SOLIN, H. (1982), »Appunti sull'onomastica romana a Delo«, in F. Coarelli & al. (eds), Delo e 1'Italia (Opusc. Inst. Rom. Finlandiae 2), Rome, 101-17. SPIRO, M. E. (1966), »Religion - problems of definition and explanation*, in M. Banton (ed.), Anthropological approaches to the study of religion, London, 85-126. SOURVINOU-INWOOD, C. (1988), »Further aspects of polls religion*, AnnArchStorAnt 10, 259-274. (1990), »What is polls religion?*, in O. Murray & S. R. F. Price (eds) , The Greek city from Homer to Alexander, Oxford, 295-322. SPAETH, B. S. (1990), »The goddess Ceres and the death of Tiberius Gracchus*, Historia 39, 182-95. SPRUTE, J. (1983), »Rechts- und Staatsphilosophie bei Cicero*, Phronesis 28, 150-76. STAAL, F. (1989) , .Rules Without Meaning: Ritual, Mantras and the Human Sciences (Toronto Studies in Religion 4), New York. STE CROIX, G. E. M. (1974), »Why were the early Christians persecuted?*, in M. Finley (ed.), Studies in ancient society, London, 210-49. STEINMETZ, F. (1969) , »Staatengriindung - aus Schwache oder naturlichem Geselligkeitdrang?*, in P. Steinmetz (ed.), Politeia und res publica, Wiesbaden, 181-199. STRASBURGER, H. (1982-90), Studien zur Alten Geschichte, 3 vols, Hildesheim. STRAUME-ZIMMERMANN, L. (1976) (ed.), Cicero's Hortensius, Frankfurt/M. & New York. (& O. GIGON, eds) (1990), Cicero. Hortensius, Lucullus, Academici libri, Munich & Zurich. STROBEL, K. (1993), Das Imperium Romanum im 3. Jahrhundert: Modell einer historischen Krise? Zur Frage mentaler Strukturen breiterer Bevolkerungsschichten in der Zeit von Marc Aurel bis zum Ausgang des 3. Jahrhunderts n. Chr. (Historia ES 75), Stuttgart. SUERBAUM, W. (1977), Vom antiken zum fruhmittelalterlichen Staatsbegriff 3 , Minister. SUMI, G. S. (1997), »Power and ritual. The crowd at Clodius' funeral*, Historia 46, 80-102. SZEMLER, G. J. (1971), »£eligio, priesthoods and magistracies in the Roman Republic*, Numen 18, 103-31. (1972), The priests of the Roman Republic (Collection Latomus 127), Brussels. TATUM, W. J. (1990), »Cicero's opposition to the lex Clodia de collegiis*, CQ 40, 187-94. (1993), »The Lex Papiria de dedication!bus*, CPh 88 319-28. 389 TAYLOR, L. R. (1931), »Aniconic worship among the early Romans*, in G. D. Hadzsits (ed.), Classical studies in honor of J. C. Rolfe, Philadelphia, 305-314. (1937), »The opportunities for dramatic performances in the times of Plautus and Terence«, TAPhA 68, 284-304. (1949), Party politics in the age of Caesar, Berkeley & Los Angeles. TODD, M. (1985), »Forum and Capitolium in the early Empire«, in F. Grew & B. Hobley (eds), Roman urban topography in Britain and the Western Empire, London, 56-66. TREGGIARI, S. (1969), Roman freedmen during the Late Republic, Oxford. TURCAN, R. (1983), Numismatique romaine du culte metroaque, Leiden. (1988), Religion romaine (Iconography of religions 17,1), 2 vols, Leiden. ULF, C. (1982), Das romische Lupercalienfest. Ein Modellfall fur Methodenprobleme in der Altertumswissenschaft, Darmstadt. ULRICH, R. B. (1994), The Roman orator and the sacred stage: the Roman templum rostratum (Coll. Latomus 222), Brussels. VAN DER HORST (1994), »Silent prayer in antiquity«, Numen 41, 1-25. VAN DER LEEUW, G. (1939), Virginibus puerisque: a study on the service of children in vjorship, Amsterdam. VAN DOREN, M. (1954-55), »Peregrina sacra. Offizielle Kultubertragungen im alten Rom«, Historia 3, 488-97. VASALY, A. (1993), Representations: images of the world in Ciceronian oratory, Berkeley. VERNACHIA, J. (1959), Ciceroniana 1, 197-213. VERNANT, J.-P. (1990), Myth and society in ancient Greece (tr. Janet Lloyd), New York. VERSNEL, H. S. (1981), »Religious mentality in ancient prayer«, in Id. (ed.), Faith, hope and worship. Aspects of religious mentality in the ancient world, Leiden, 1-64. (1993), Transition and reversal in myth and ritual (Inconsistencies in Greek and Roman religion 2), Leiden. VEYNE, P. (I983a), Les Grecs ont-ils cru a leurs mythes?, Paris. (1983b), »Titulus praelatus: offrande, sollenisation et publicite dans les ex-voto greco-romains«, Revue Archeologique, 281-300. VIRLOUVET, C. (1991), »La plebe frumentaire a 1'epoque d'Auguste«, in A. Giovannini (ed.), Nourrir la plebe, Basle, 43-62. VON BERTALANFFY, L. (1951), »Zu einer allgemeinen Systemlehre«, Biologia Generalis 19, 114-29. VON HESBERG, H. (1978), »Archaologische Denkmaler zum romischen Kaiserkult«, ANRN 2,16,2, 911-95. WAGENVOORT, H. (1980), Pietas. Selected studies in Roman religion, Leiden. 390 WALBANK, F. w. (1985), Selected papers: studies in Greek and Roman history and historiography, Cambridge. WALLACE-HADRILL, A. (1988), »Time for Augustus: Ovid, Augustus and the Fasti« 7 in M. Whitby & al. (eds), Homo Viator. Classical essays for John Bramble, Bristol, 221-30. (1994), Houses and society in Pompeii and Herculaneum, Princeton. WARDE FOWLER, W. (1911), The religious experience of the Roman people, Oxford. WARDMAN, A. (1982), Religion and statecraft among the Romans, London. WARNKE, C. (1986), »Antike Religion und antike Gesellschaft: Wissenschaftshistorische Bemerkungen zu Fustel de Coulanges 1 >La cite antique<«, Klio 68, 287-304. WATSON, A. (1968), The law of property, Oxford. WEILER, I. (1974), »Von >Wesen<, >Geist< und >Eigenart< der Volker der Alten Welt. Eine Anthologie altertumswissenschaftlicher Typisierungskunst«, Innsbrucker Beitrage zur Kulturwissenschaft 18, 243-91. WEINREICH, O. (1921), »Zu Tibull l,l.ll-24«, Hermes 56, 337. WEINSTOCK, ST. (1966), »The posthumous honours of Germanicus«, in Melanges Andre Piganiol, Paris, 2,891-8. (1971), Divus lulius, Oxford. WELKER, M. (1992), »Einfache oder multiple doppelte Kontingenz? Minimalbedingungen der Beschreibung von Religion und emergenten Strukturen sozialer Systeme«, in W. Krawietz & M. Welker (eds), Kritik der Theorie sozialer Systeme. Auseinandersetzungen mit Luhmanns Hauptwerk, Frankfurt/Main, 355-70. WHITE, G. M. (1992), »Ethnopsychology«, in SCHWARTZ & al. 1992, 21-46. WHITTAKER, C. R. (1995), »Do theories of the ancient city matter?«, in CORNELL & LOMAS 1995, 9-26. WIEACKER, F. (1988), Romische Rechtsgeschichte. Quell enkunde, Rechtsbildung, Jurisprudenz und Rechtsliteratur. Erster Abschnitt: Einleitung. Quellenkunde. Fruhzeit und Republik, Munich. WIEDEMANN, TH. E. J. (1985) , »The regularity of manumission at Rome«, CQ 35, 162-75. WILLKE, H. (1978), »Zum Problem der Integration komplexer Sozialsysteme: ein theoretisches Konzept«, Kolner Ztschr. fur Soziologie und Sozialpsychologie 30, 228-52. WINTERLING, A. (1993), »>Arme< und >Reiche<. Die Struktur der griechischen Polisgesellschaften in Aristoteles 1 >Politik<«, Saeculum 44, 179-205. WISEMAN, T. P. (1971), New men in the Roman senate 139, B.C.-A.D. 14, Oxford. (1974), Cinna the poet and other essays, Leicester. (1979), Clio's cosmetics. Three studies in Greco-Roman literature, Leicester. (I985a), Catullus and his world. A reappraisal, Cambridge. (1985b), »Competition and co-operation*, in Id. (ed.), Roman political life, 90 B.C.-A.D. 69 (Exeter Studies in History 7), Exeter, 3-19. 391 (1987), Roman studies. Literary and historical, Liverpool. (1994), Historiography and imagination. Eight essays on Roman culture (Exeter Studies in History 33), Exeter. WISSOWA, G. (1912), Religion und Kultus der Romer2 , Munich. WLOSOK, A. (1970), »R6mischer Religions- und Gottesbegriff in heidnischer und christlicher Zeit«, A&A 16, 39-53. WOLFF, H. (1986), »Die Entwicklung der Veteranenprivilegien vom Beginn des 1. Jh. v. Chr. bis auf Konstantin d. Gr.«, in Id. & W. Eck (eds), Heer und Integrationspolitik. Die romischen Militardiplome als historische Quelle, Cologne & Vienna, 44-115. WOOD, N. (1988), Cicero's social and political thought, Berkeley. WOODMAN, A. J. (1988), Rhetoric in Classical historiography. Four studies, London & Sydney. WOOLF, G. D. (1990), »World-systems analysis and the Roman Empire«, JRA 3, 44-58. (1992), »Imperialism, empire and the integration of the Roman economy«, World Archaeology 23, 283-293. WORTHMAN, C. M. (1992), »Cupid and Psyche: investigative syncretism in biological and psychosocial anthropology«, in SCHWARTZ & al. 1992, 150-78. YAKOBSON, A. (1992), »Petitio et largitio: popular participation in the centuriate assembly of the Late Republic«, JRS 82, 32-52. ZAIDMAN, L. B. & P. SCHMITT PANTEL (1992), Religion in the ancient Greek city (tr. P. Cartledge), Cambridge. ZANKER, P. (1975), »Grabreliefs romischer Freigelassener«, Jdl 90, 267-315. (1987), Augustus und die Macht der Bilder, Munich. (1994), »Veranderungen im offentlichen Raum der italischen Stadte der Kaiserzeit«, in L'Italie d r Auguste a Diocletien (Coll. de l'6cole Francaise de Rome 198), Rome, 259-84. ZIOLKOWSKI, A. (1988), »Mummius' temple of Hercules Victor and the round temple on the Tiber«, Phoenix 42, 309-33.