The Bologna connection

1 The Bologna connection Lo codi, the birth of elaborate Romance and its Latin background Johannes Kabatek (University of Zurich) 1. Introduction Gonzalo de Berceo and the “Roman paladino” In the first half of the 13th century Gonzalo de Berceo, the first Castilian poet known by name, composed the Vida de Santo Domingo de Silos, a collection of poems written in “cuaderna vía”, a typical form of clerical poetry. It opens with the following wellknown verses: (1) Quiero fer una prosa en román paladino en qual suele el pueblo fablar con so vecino, ca non so tan letrado por fer otro latino, bien valdrá, como creo, un vaso de bon vino. (Gonzalo de Berceo, Vida de Santo Domingo de Silos, ed. Dutton) ‘I want to write a prose in simple Romance in the way people speak to their neighbours because I am not educated enough as to write in a different Latin. I believe, however, it will be worth a glass of wine’ At first glance (and following Berceo’s captatio benevolentiae), to write in Romance instead of Latin seems to be an indicator of ignorance, and indeed the emergence of vernacular writing has sometimes been attributed to the fact that such writers lacked competence in Latin. But it does not take a very sophisticated eye to perceive the educated and well-trained background at underlies Berceo’s writing. His verses follow the specific rhetorical principles that must be mentioned in a prologue: the intentio and the utilitas of what follows. Moreover, we know that Berceo’s poetry is built on Latin prose texts that tell the same stories; the poems are, thus, lyric transformations of previously existing texts in prose that circulated around the monasteries of northern Spain in the 13th century (Timmons and Boenig 2007). However, they are not only testimonies of a general clerical context; they also show a close connection to legal knowledge and traditions in the Iberian Peninsula. This much is widely known 2 (Bermejo 1969) and has been noted in many editions of Berceo’s work. Yet as we showed in Kabatek (2005), he not only alludes to legal concepts from time to time in his work, but also plays with the contrast between Castilian local legal norms and the “new” legal concepts that had arrived on the Iberian Peninsula due to the renaissance of Roman law from the late 12th century onwards. Berceo, far from being ignorant, can be seen as among the most educated people of his time in his region. He had studied legal science at the beginning of the 13th century in Palencia, which was, between 1212 and 1250, the location of the first Spanish university, and where there had previously been an influential law school. It is known that at the end of the 12th century, the Italian Ugulino de Sesso taught Roman law there, and his Summulae on certain concepts of Roman law (Iglesia Ferreirós 1998) testify to the teaching methods there and provide some examples of how the “new” law (in fact, classical Roman law, see below) that had been rediscovered in Italy, namely in Bologna, had been taught in Spain years prior to Berceo’s contact with that school. This brings us to the central point of the present paper, the claim that Berceo is far from being an isolated case, and that vernacular writing in the 13 th century is closely linked to what had been called “The Renaissance of the 12 th century” (Haskins 1927) and the revival of Roman law that spread from the University of Bologna across Europe. Let us call the direct or indirect links to Bologna the “Bologna connection”, an intellectual movement with numerous consequences. But before addressing this central issue, I will reconstruct some elements in the complex puzzle that underlies it, beginning with a short description of a text that I believe is key to the whole story and which thus can be seen as constituting the initial point of momentum of what would lead to the creation of elaborate vernacular text traditions from the 13th century onwards: the Occitan summa Lo codi. 2. Lo codi: how a single text leads us to re-write the history of Romance Around the middle of the 12th century (Kabatek 2005, Glessgen 2023), a text of legal prose of some 70,000 words was created somewhere in the south of France in the Occitan language. “Somewhere” refers to a centre of legal studies, with some speculation as to the exact location. This text is Lo codi, ‘the code’, and itself refers to the Codex Iustinianus or Codex Iuris Civilis, the classical Roman-Byzantine collection of legal norms compiled under the auspices of the Eastern Roman Emperor Justinianus the Great in the 6th century. Lo codi is not a translation of the Justinian code, as is sometimes wrongly supposed (Frank and Hartmann 1997, III, 327), but an original summary of the content of the Justinian work, based mainly on the equally southern French Summa Trecensis, a didactic Latin text that structures and summarises the main 3 ideas of classical Roman law. Lo codi consists of nine books of legal prescriptions and meta-legal norms (the remaining three books of the Justinian code were generally left aside in the medieval tradition) and it is written in an independent, rather Latinism-free language that very consciously separates Latin and Occitan (see section 4). The text is highly coherent and homogeneous, and is full of complex and abstract explanations of legal concepts. It gives the impression that writing such an exhaustive work of prose in the vernacular is not only perfectly possible but also almost “normal”. Yet if we look at the historical background, it becomes apparent that the text is not in itself testimony to a tradition, but rather serves as the moment of the very creation of that tradition. How could such a text have been produced, we might indeed say invented, in an area where at that time the most elaborate written texts in Romance were a number of feudal oaths and other legal documents? Of course, the south of France is the most pioneering region in Romance scripturality (Selig 1995), and besides legal vernacular texts, there is also an oral poetic tradition. But nothing compares to Lo codi. However, Lo codi is not an easy case, “Tout, dans le Codi, pose problème” (‘everything about Lo codi is problematic’, Gouron 2002, 1; Glessgen 2023), from its precise localisation to its exact genesis. The sudden and mysterious appearance of this radically innovative Romance text demands an explanation. I will explore the textual background of the text in more detail in the following section, but here I would like to claim that, all mysteries apart, the text shows obvious evidence for the existence of an independent, elaborate Occitan written language in the second half of the 12th century, and that it is a pioneering work that merits a prominent place in the canon of Romance linguistics. It seems strange that this is not the case: the text has been known since the early 19th century and historians of law have never doubted its importance. I see three main reasons for this lack of attention: first, a profound misunderstanding of the general evolution of Romance in the Middle Ages; second, the myth of early scripturality arising from erroneous and anachronistic dates; third, scant and partly misleading knowledge about the text itself. Let us discuss these three aspects briefly. The first point concerns problematic conceptions about the emergence of written Romance, which is indeed one of the habitual issues in Romance. Traditional manuals dedicate long chapters to this matter (in Tagliavini’s famous introduction (1959) over 50 pages, for example). These manuals say nothing about Lo codi, and they neither mention the Renaissance of the 12th century nor the importance of Bologna. There is a lot of discussion of why and when Romance emerges, with several imprecise anachronisms (e.g. in the case of the Strasbourg oaths), which I will not go into here. We know that Romance appears after the Council of Tours in 813, when vernacular languages were recognised as different from Latin. We also know that Romance appears in Latin contexts, in quotations and in testimonies of the spoken language, that it appears in religious contexts, and that it may reflect the act of reading aloud to an audience that was not proficient in Latin (Wunderli 1965, Koch 1993). There is also 4 much discussion about the emergence of what we might call “sporadic scripturality” and its relationship with the different reforms of Latin from Charlemagne, through Cluny, to Burgos (see Wright 1982). But can we explain the appearance of a text like Lo codi simply as a slow continuation of that early sporadic emergence of the vernacular? Is the evolution of written Romance from the 9th to the 12th century nothing but a slow advance of Romance into the domains of Latin? I think that the idea of such a slow advance is a myth and that nothing fundamental changes until the mid 12th century: Latin (and, in some regions, Arabic) remains the written language in a clear diglossia, with Romance only appearing from time to time, sometimes indirectly in Latinised forms, sometimes directly in fragments of text that might make it into entire short texts, as we will see in the next section. But the path from texts like the Strasbourg oaths to one like Lo codi is not a continuous evolution. The sudden appearance of Lo codi is an event at least as radical as the emergence of written Romance some centuries before, and the explanation of the former, as we will see, is in some ways comparable to that of the latter. The second reason as to why the importance of Lo codi has been overlooked is the general image we have of written Romance in the 12th century. Why should Lo codi be understood as a surprising text if we have so much other literature here, from the Occitan Trobadors via the Chanson de Roland, Chrestien de Troyes and Marie de France to the Cantar de mio Cid? Since Rayonuard (who, incidentally, was the first Romance scholar to highlight the importance of Lo codi) there has been a long tradition of an absence of historical criticism, in which supposed and estimated dates were accepted as real ones, with little critical attention to alternative interpretations. Let us take the example of the Castilian Cantar de mio Cid. Traditionally, and indeed still frequently, it is regarded as the first extensive text written in Castilian: the traditional history of Spanish as set out by Ramón Menéndez Pidal (who was a genius but who also defended a series of ideas that are problematic in our actual view) dated the Cantar, an epic work about one of the mythological heroes of the Reconquista, to 1140. There were many reasons for this: the reconstruction of the historical Cid, his appearance in chronicles from the 13th century, the “archaism” of the language of the text… However, there exists only one incomplete manuscript, probably written at the beginning of the 13th century and preserved in a copy from the beginning of the 14 th century. This is why the text appears as marked in “red” in Rodríguez Molina and Octavio de Toledo’s (2017) fantastic Cordemáforo, a catalogue of documents contained in the Spanish Royal Academy corpus CORDE that distinguishes between “green”, “yellow” and “red” documents, the latter considered problematic due to an unclear relationship between the original date of composition of the text and the date of the manuscript. Similar problems of textual transmission arise when we look at the exhaustive documentation of medieval Occitan and French literature: generally speaking, there is little from before the second 5 half of the 12th century, a moment when the influence of the “Bologna connection” is already at its height. The third reason is the marginality of studies on Lo codi: most writing about the text were written by historians of Roman law (who in fact call themselves “Romanists”) and not by Romance linguists. Derrer’s (1974) important edition of the Occitan Ms. A was a university dissertation written in German, and my 2005 book on Lo codi and the Bolognese Renaissance also had the problem of being written in German: Germanica non leguntur. However, recent years have seen new interest in the text; not only have there been several publications within the area of Romance linguistics, but also several research projects, and thus we can be sure that Lo codi will finally find its way into the canon of Romance linguistics1. Sometimes things move quite slowly, but if something genuinely important is left aside, the passing of time has a tendency to cure such omissions, however long it takes. 3. Lo codi and its models 3.1 Two different worlds Before comparing the text of Lo codi to its models, I would like to underline the contrast between this text and the production of early “sporadic” Romance through a brief comparison of both. Let us take the Spanish Glosas emilianenses (considered to be one of the first short written vernacular texts, if not the first, in the Iberian Peninsula) as a contrast; we could equally look at other Romance texts from the first centuries, so this is simply an illustrative example. The Glosas emilianenses (García Turza and Muro Munilla 2006) consist mostly of vernacular word translations as found in a Latin text that seems to have been used for the teaching of Latin. At one point, on fol. 72 of the Codex Aemilianensis 60, there is more than just a short translation, with a marginal gloss comprising a brief and more or less independent text in Romance with a strong Latin influence. The following two images provide a comparison of this marginal gloss (below in the right margin) within ―― 1 There have been several recent publications by Frédéric Duval (2018a, 2018b) specifically on the French versions of Lo codi and on translations of the Justinian Codex. Very recently, Maxime Paratte published the results of a graduation thesis on the Latin translation of Lo codi (Paratte 2023) and Martin Glessgen (2023) offers a new overview on the state of the art with some linguistic analyses of the Occitan Ms. A. Duval, Glessgen and myself are also engaged in new editorial projects. In a project within the field of Digital Humanities supported by the University Foundation of Zurich, we are currently updating the partial, parallel edition from the early 2000s, introducing new texts and new tools for analysis (see www.locodi.ch). Following the publication of the ms. A by Derrer in 1974, Zurich (in collaboration with Paris), again emerged as a centre for the study of Lo codi and its traditions. 6 the Latin context on the left, and a page (one of almost 200!) from one of the Spanish manuscripts of Lo codi on the right: Fig. 1: Glosas emilianenses (Codex aemilianensis 6, fol. 72, left), Lo codi (ms. K, BN Madrid 10816, f 3r, right). There are worlds between these two examples, and a comparison of the gloss (the most extensive one in the manuscript) with a short extract from Lo codi is enlightening: “sporadic” Romance: emilianenses Cono aiutorio de o nuestro dueno, dueno christo, dueno salbatore, qual dueno get ena honore. e qual duenno tienet ela mandatione cono patre cono spiritu sancto enos sieculos delos sieculos facanos deus omnipotens tal serbitio fere ke denante ela sua face gaudioso segamus. amen [With the help of our lord Lord Christ, Lord Savior, Lord who is in honor, Lord that has command with the Father, with the Holy Spirit Glosas elaborate Romance: Lo codi De todas las cosas del mundo que son mas nobles & mayores & meiores. & mas onrradas. & que se mas deuen fazer. & guardar & onrrar son las cosas que a Dios pertenecen. . & por esto deuemos primera mient dezir de la fe & de la trinidat las quales dos cosas pertenescen mas a dios que otras cosas nombrada mient fe & trinidat. Et deuen seer guardadas & tenudas a todos los omnes del mundo. segunt que fueron ordenados en quatro conceios . que fueron fechos en quatro logares nombrados […] [Of all the thing in the world those that are most noble and biggest and best and most honourable and that must be done and preserved and honoured most are the things that pertain to God. And this is why we will in first place have to talk about faith and trinity which are the things that most pertain to God 7 for ever and ever. God namely faith and trinity. And they must be preserved Omnipotent, make us do and observed by all the people of the world as they such a service that before His were commanded in four councils that took place in face joyful we are. Amen.] four named places.] Tab. 1: Glosas emilianenses vs. Lo codi The short marginal Romance text (the “glosa”) inserted in a Latin manuscript shows Romance (diphthongs, articles, prepositional cases) as well as Latin elements (conservation of Latin consonants, Latin case markers, Latin spelling). It consists, after some introduction formulae, of one short sentence with subordination and is in fact a short prayer. The text of Lo codi here is merely a brief extract from the beginning of the first chapter, and announces what is going to be dealt with in the chapter. It is written in a rather coherent Romance and differs in many aspects from the Glosas example. The following table summarises some of the main contrasts: “sporadic” Romance: Glosas emilianenses Romance within the Latin context short text (44 words) with strong presence of Latin elements, no stable spelling connection to oral traditions “by-product” of Latin scripturality sporadic appearance of the vernacular without tradition consequence of Carolingian reform elaborate Romance: Lo codi - independent Romance context large texts with a low presence of Latin elements (almost 70,000 words) conscious separation of languages within the text product of an “independent” Romance scripturality beginning of a textual tradition consequence of the “Bolognese Renaissance” Tab. 2: Glosas emilianenses vs. Lo codi: some major differences There is of course a temporal and a local gap between both texts, and it is perhaps justifiable to claim that this is like comparing apples and pears. It was traditionally assumed that the Glosas were produced in the 10th century in the Rioja, but since Wright’s criticism (1982, 1997), there are good arguments to consider them as texts of the late 11th century. Lo codi is thought to have been written in the second half of the 12th century, between 1149 and 1170 (Pfister 1978, 286; Prawer 1962, Glessgen 2023). The Occitan manuscript A (Sorbonne 632) stems from the time immediately after this (Derrer 1974; Frank/Hartmann 1997, III, 327). Yet the comparison of both texts is just an exemplary comparison of two Romance texts from two different periods, and it shows that there are not only quantitative differences, but that there is also a qualitative gap that can hardly be accounted for by the notion of a slow, progressive evolution without any sudden qualitative, formative event. This event, indeed, can and must be 8 explained, and it is, in our view, the Bolognese Renaissance of Roman law and its spread through Europe catalysed by the Catholic Church (see section 3.3). Even if the scope and characteristics of the text seem revolutionary for that time, Lo codi does not appear out of the blue. It is based largely on two models: the Occitan writing tradition of the 12th century and the Latin Summae of Roman law written around the middle of the 12th century. Occitan is known to be the first Romance language in which vernacular writing not only appears sporadically, as in the aforementioned example, but where textual and the practice of habitually writing in the vernacular emerge. The context of this emergence is also a legal one, but not one of abstract texts but of a concrete application: the Occitan charters, documents of buying and selling, testaments, and feudal oaths. All these documents have in common that they contain a referential section with descriptions of goods or territories, plus a performative section that describes what will happen to these things. The charters are intended to be read out aloud to the participating parties, and the vernacular is introduced for two reasons: the referents themselves (specific places and individuals named with vernacular toponyms and vernacular anthroponyms), and the need to be understood by illiterate participants. This is where the notion of “discursive zone” (Kabatek 2015) becomes relevant; in this case, the shift from one language to another does not take place in a general way but begins by means of entering the text in certain specific areas. Even if the formulaic parts in the opening remain in Latin, the vernacular penetrates more and more into charters until an independent Occitan writing tradition emerges. It seems as if the charters – at least in the first period, i.e. at the beginning of the 12th century – represent a gradual flowing of the vernacular elements into the Latin texts “from below”, beginning with the free sections and then spreading to the more formal sections, while Lo codi is an expression of an elaborate form of writing in Romance “from above”. This leads to the main critical arguments against a direct relationship between the charters and Lo codi. The latter is a text whose content and textual form has few parallels with the charters. There are, however, elements that can be related to the existing vernacular tradition of Occitan or partly Occitan legal documents and vernacular feudal oaths (see next section), already richly documented in the classic collection by Brunel (1926, see also Selig 1995). Another argument against a relationship between Lo codi and the charters is the geographical origin of both. According to the prevailing research, Lo codi stems from a region where the tradition of Occitan documents is not particularly strong (Kabatek 2005, 124). However, in a recent contribution, Glessgen argues against the traditional localisation of the text in the region of St. Gilles, and introduces, after an examination of several manuscripts, a difference between the Ms A and the original text; he then locates the original production to the Valentinois in around 1149 and the Ms A to the Rouergue, a region with a rich production of charters, in around 1163 (Glessgen 2023, 150). 9 3.2 Occitan legal documents as a source Lo codi is rather an abstract text on general legal principles. As to the choice of language, we can only speculate. The most reasonable explanation can probably be found in a new division of domains after the reception of Roman law, when a separation takes place between the new, “scientific” treatment of law on the one hand, and on the other its application to, and connection with, the domains of traditional local law. Lo codi is not a scientific text about the new law, but a compendium that summarises this law for less educated people or for legal practitioners. By using the vernacular, the author (or authors) expresses precisely this separation of two systems of knowledge. As noted above, Occitan had already made its way into the domain of legal application during the 12th century. The majority of Lo codi cannot itself build on vernacular textual traditions, but the important breakthrough of writing in the vernacular had already been made in the charters. There are, however, passages within Lo codi where a link to the existing written tradition in Occitan is established. The text not only summarises the content of the Codex Iuris Civilis, it also introduces examples to illustrate this abstract content. If we look at specific points in the text, we can identify a certain degree of similarity between some passages in Lo codi and the feudal oaths (Selig 1995, 209ff). For example, in the second book, the way that different participants in a trial have to swear in court is described: (2) XXVII. §1. «[…] li actor deu jurar primeirament en aital guisa: ‹aquel demandament que eu fasz en aquest plaig, eu lo cui faire per mun dreig e per mun radon, ed aco que li altre parz me demandara, que eu sabrei que sia uers, eu no li en demandarei garent ni proua, ni non i demandarai alongament per mal engeing, si non aquel que me sera obs›». If we look at the following extract of a feudal oath, there are some parallelisms: (3) «Haus tu Raimuns, fils Aialmus! Heu Ponz, fils de Garsia, del castel de Fos et de Eiras et d’Aix not decebrai, ni nols ti tolrai, ni om ni femena per mon conseil, las forzas qui i sunt ni que azenant faias i serant; et s’om era o femena quels ti tolges, ab aquels fin ni societad non auria, fors pels castels a recobrar, et con recobrati los auria, heuls te rendria sens logre et sens engan. Aquels castels c’aici sunt escrit, euls te rendrai con m’en comoras o comonroe m’en faras sens engan. (Serment prêté à Raimon de Saint-Gilles, marquis de Provence, par Pons, fils de Garsie, de tenir fidèlement les châteaux de Fos, Hyères et Aix; Brunel 1926, 13s.). The way of swearing in the first person future tense is common to both texts as well as the choice for some formulae (Codi: engeign; Oath: engan; Codi: per son cosseil, Oath: per mon conseil). The example of swearing seems to have evoked the already established tradition of the feudal oaths, which were originally oral and subsequently documented in writing. As in the feudal oath (e.g. ni om ni femena), the references in 10 Lo codi (garent ni prova) are sometimes doubled. This general characteristic of legal language originally stems, as is the case here, from the need for exactness in the reference for reasons of legal precision, but then becomes independent as a stylistic element, one which is also used to imitate legal or administrative language or to mark more formal discourses. The main differences between the two texts here are obvious: a feudal oath is an act of application in a concrete case, whereas in Lo codi it is about swearing in general and the insertion of a specific swearing formula is an exception in a text about general principles. There are further parallelisms. A recurrent technique in legal documents is an almost pedantic form of textual deixis in order to leave no doubt as to the correct referent. In the Occitan charters we find passages such as the following: (4) «[…] de tot aquest do d’aquesta terra sobre escriuta e del bosc e del deime sobre escriut» (Brunel, Nr. 65; Toulousain 1151) […about this whole donation of this aforementioned land and of the forest and the aforementioned tithe.] This finds an echo in Lo codi with reference to the abstract content of the text: (5) «[…] tot aco que es escriut en aquest paragrafo d’aquelz procuradors es uers, zo es d’aquelz omes qui uolunt esser auctors» (Codi A II. 6. 9). [Everything that is written in this paragraph about those procurators i true, i.e. of those men who want to be plaintiffs]. There are also some rare passages of exemplification through direct speech in Lo codi that reflect the style of the feudal oaths: (6) I. §3. «[…] si eu dic: ‹tu me deus.X.sol.›, e tu dides a me que eu t’en aia feita couenenza que non los te demandaria, ed eu te respon: ‹uers es, mas tu me fedist poissas couenenzas que tu los me darias e que eu los te pogues demandar›». (7) III. §11. «[…] si eu te couen: ‹aitals om te faia aital causa›». (8) VI. §18. «[…] aco que uos en fedesz ses mon mandament non ual». [What you would do about it without my command would not apply.] (9) X. §3. «Peire sias mos heres, e si tu non eras mos heres, ma moiller sia mos heres». [Peter you shall be my heir, and if you were not my heir, my wife shall be my heir]. All this is only marginal within the broader context of the book; the main contribution of the Occitan written tradition is the fact that the vernacular language can be used for writing legal texts. However, most of the elaboration to be found in Lo codi is new and derives from Latin rather than Occitan models. 3.3 The Latin models of Lo codi We ought to bear in mind that the content of Lo codi is basically that of the Justinian Codex Iuris Civilis, and that the main difference between the original Latin model and 11 the Occitan Summa can be found in the ordering, the systematization and the textual representation of this content. Yet the textual filiation is far more complex than that of a direct adaption of the original Justinian text into an easily readable, didactically modified vernacular version. In the Justinian corpus there are already different kinds of texts; on the one hand there is the Codex Iuris Civilis, the main collection of laws, and on the other, the Digesta or Pandectes, composed after the Codex and containing a collection of the main legal writings of about 40 classical authors. This is the most complex part of the corpus, and it coexists with texts that are more easily accessible, such as the Institutiones, a rather short compendium about the foundations of law that served for initial studies in the field. But when Roman law is rediscovered or newly addressed in Bologna and other places from the second half of the 11th century onwards, not only are the classic texts copied and used for teaching, there are also several newly created didactic forms that emerge which are used mainly for the teaching of laws and legal principles. The most common form is the gloss, originally a side-comment in legal manuscripts that comments on the content of a legal text. The gloss gradually extends in use during the time of the Bolognese Glossators from the 12th century onwards. The glosses partly evolve into independent texts that summarise the content of the Codex, the so-called Summae codicis, didactic texts that convert the content of the Code into clearly structured schemes of (indirect) questions and answers. This didactic scheme reflects a medieval tradition that combines insights from dialectics and rhetorics with legal knowledge. There exist shorter Summulae and extensive Summae that comment on the whole Codex. Several legal scholars, it is supposed, are the authors of such extensive Summae codicis. We cannot go further into the question of the contrast between the classic texts of Roman law and their systematic elaboration in the Summae, but it should be noted that this elaboration is one of the major achievements of medieval jurisprudence and that its mark on continental law has continued to the present. The classic Justinian texts2 were collections of statements and opinions of various legal experts that reproduce a whole process of discussion; the Summae, by contrast, extract the essence of the results here and present them in an ordered and didactic system. Lo codi then transforms this didactic summary into an even more explicit and less abstract compendium in Occitan. In the case of Lo codi, legal historians have identified the so-called first version of the Summa Trecensis as the main source of the text. The Summa Trecensis received its name due to a manuscript found in the Library of Troyes in France (Codex Trecensis ―― 2 An online version of Samuel Scott’s translation of the Codex Iuris Civilis, which gives an impression of the configuration of the Justinian Code, is available at https://droitromain.univ-grenoblealpes.fr/Anglica/CJ2_Scott.htm. An online version of extracts from the Latin Digests can be found at http://www.thelatinlibrary.com/justinian/digest50.shtml 12 1317). It was formerly ascribed to the famous but somewhat mysterious Irnereus, the supposed founder of the University of Bologna. There is, however, no evidence for this adscription, and the Summa was probably written by a legal expert named Geraudus in the south of France in the 1130s (Gouron 1986). There is a second, later version that stems from around 1170 and which can hardly be said to have served as a model for Lo codi, as shown in the following scheme (Kabatek 2005, 131), with a supposed date of conclusion of Lo codi in 1149 (Glessgen 2023, 133). First version of the Summa Trecensis (ca. 1130) Lo codi (ca 1149) Second Version of the Summa Trecensis (around 1170) Fig. 2: Filiation Lo codi - Summa Trecensis In order to illustrate the relationship between both texts, let us take an example: Summa Trecensis (ed. Fitting; 6) De fide ac religione dictum est. ecclesia quidem mater fidei ac religionis constituta est, ideoque statim de ea adnectit. set et res necessarie sunt tam ecclesiis quam etiam ceteris locis uenerabilibus. nec non et sine loco pia actio constituta est … Lo codi (ed. Derrer) Translation Lo codi II. §1. de fe e de Trinitat auem dit; aras digam de las gleisas que son maires de fe e de religion. mas cum zo es causa que las causas de·l mun, si cum sunt terras e uignas e maisos ed altras causas mundanas, sunt obs a gleisas e ad hospitals e ad altres locs honorables, bes es que nos digam de las causas de las gleisas e de·ls autres locs uenerables. About fauth and Trinity we have spoken; let us now speak about the churches that are masters of faith and religion. But as the worldly things, such as land and vineyeards and houses and other worldly things are necessary for the churches and for hospitals and other venerable places, it is good that we speak about the things of the churches and the other venerable places. Tab. 3: Summa Trecensis and Lo codi, example of comparison Even if the content is practically the same, Lo codi is not merely a translation of the original text, and introduces a number of differences. The impersonal passive of the 13 Latin text (dictum est ‘it is said’) is transformed into a personal auem dit ‘we have said’; and the following topic is explicitly introduced (aras digam de las gleisas ‘let’s now talk about the churches’). Then, whereas the Latin text only mentions ‘things’ that pertain to the church, the Occitan text explains what the “causas del mun”, ‘worldly things’, are: “si cum sunt terras e uignas e maisos ed altras causas mundanas” – ‘such as land and vineyards and houses and other worldly things.’ In general, Lo codi personalises the impersonal style of its source and introduces examples and explicit explanations and paraphrases. We cannot judge whether this personification occurs because the passive was not felt to be adequate in the vernacular or if the text was consciously transformed so as to adopt a more accessible style. The text in Tab. 3, from the first book (in both the Summa and Lo codi – the structure is the same), refers to the Catholic Church, and I already mentioned that the church is the main European catalyser for the spread of Roman law: in 1140, the Decretum Gratiani, a collection of legal prescriptions for the Catholic Church, unifies the Canon law within the context of the Bolognese Renaissance of Roman law, and many of the legal principles, especially those referring to the organisation and processes of the legal system, were adopted from the Justinian civil code. The “new” ecclesiastic law is dealt with alongside the Roman Civil law, and “both laws” are studied together and are introduced into the European centres of the Catholic Church, generally by Bishops or other Iuris Periti, legal experts with links to Bologna or to other places where the new law is being taught (see next section). In general, the comparison of the Summa Trecensis with Lo codi makes the latter appear to be more explicit; it uses examples, it personalises the style, and it introduces deictic elements, as we can see in the following examples Summa Trecensis (III, VI; Ausg. Fitting 1894) Lo codi (III, XII; Ausg. Derrer 1974) De iurisdictione et ubi iudicia peragenda sunt. Apud eum qui iurisdictioni preest iudicia expedienda sunt aici ditz en cal log deu om metre altre em plaig Possessionis causa ibi agitanda est, ubi possessio turbata est. Relicta ibi petenda sunt, ubi hereditas relicta est, uel ubi maior pars hereditatis, uel ubi est res ipsa que relicta est. pois que nos auem dit de∙ls iutgues ed en cal mesura deuunt li iutgues donar iudicias, ara digam davant cals personas deuunt esser faig li plaig. Li plaig deuunt esser faig dauant aquellas personas que ant iurisdiccion, zo es poestat si es plaiz de possessios, dauant aquill iutgue deu esser fait lo plaitz en cui poder es la tenedons. si eu li deman auer o autra causa que eu diga que alcuns om me donet a sa mort, aqui deu esser lo plaiz tenduz un es la heretaz de∙l defunc o la maier parz de la heretat, o aqui on es la causa qui es demandada. 14 Tab. 4: Summa Trecensis and Lo codi, personification of impersonal forms Almost systematically, Latin impersonal gerundive constructions are replaced by personal ones: iudicia peragenda sunt iudicia expedienda sunt causa ibi agitanda est petenda sunt deu om metre altre em plaig li plaig deuunt esser faig dauant aquill iutgue deu esser fait lo plaitz deu esser lo plaiz tenduz Tab. 5: Summa Trecensis and Lo codi, substitution of gerundive forms Definitions of legal terms are generally short and rarely repeated (e.g. iurisdictio est potestas iuris dicendi (III, 6, 1)) in the Trecensis, whereas Lo codi not only offers short definitions of terms but tends to make them more explicit by paraphrases or repetitions. In many places terms are paraphrased in short subordinate clauses, for example in the following passage at the beginning of the third book: (10) «[…] poiss que nos auem dig de sobre d’aquellas causas que sunt obs e∙l iudizis, si cum es de edendo, zo es de manifestar ad altre per cal radon el lo uol metre em plaig, ed de in ius vocando, zo es de clamar em plaig omen, e poiss que nos auem dig de-ls arbitres, zo es d’aquellas personas que receubon fermanzas de plaig ni non ant autra jurisdicion, zo es altra poestat, aras digam….». In the Trecensis, by contrast, we find a general prologue at the corresponding point, without repetition of all the previously treated points: (11) «Intendunt quidem principes aequitatem constituere et eam in preceptione redactam tam ab omnibus uolentibus quam etiam inuitis obseruari et lites singulorum publica auctoritate dirimi ac exsecutioni mandari. eaque ratione premissis iuditiorum preparatoriis inde non inmerito eorum tractatus apponit». These examples suffice to give an impression as to the main differences between the two texts. Lo codi offers a double transformation of its Latin source, a translation from Latin into consciously separated vernacular writing, and a further step of didactification of a Latin that is already quite didactic in comparison to the original sources of the Codex Iuris Civilis (for more details, see Kabatek 2005, 130-161; cf. also Paratte 2023, 180). 4. Lo codi, “overt” and “covert” Latinisation After indicating, in the previous section, the main background of Lo codi (the Occitan vernacular, the Occitan written traditions and the Latin Summa Trecensis), I will now 15 turn briefly to the language of Lo codi, limiting my observations to two areas: the lexicon and the techniques of clause linkage. Given that Lo codi was based on a Latin model, we must bear in mind that the Occitan language at the time of composition did not have an exhaustive inventory of legal terminology or many other technical terms. Thus, once the decision was taken (or the order given) to write an Occitan text, the Latin model must have obliged the translator or translators (we know nothing about the exact history of the process of translation) to create something new. Within these conditions of asymmetry between a language in which everything required already existed and a vernacular where many of these means were lacking, the easiest way to innovate would have consisted of the mass adoption of Latin elements. Indeed, we can observe that several technical terms appear in Lo codi in their Latin form without modification, including the following: (12) comodatum, contumelia, cruciatum, edictum, escriptum, integrum, inventarium, iudicatum, iudicium, legatum, litum, mancipacium, pecculium, qualiter lis contestatur, restitutum But in these cases, there is a clear awareness of the separation of two different languages, and Latin terms are sometimes explicitly marked as such: (13) ara digam de prestanzas que / fai uns om ad autre ses auer per amor. aquist prestanza es appellada per latin “como datum”. primeirament deu esser esgardat, que es comodatum [let’s now talk about the loans that one man gives to another out of love, without payment. This loan is called in Latin ‘como datum’. In first place, it must be explained what comodatum is] Hence Latin terms are quoted as elements pertaining to another language. Moreover, terms are frequently introduced with clear definitions, as in the following example: (14) II. 6. 1. «[…] aquist an num ‹procurador›: zo es aquel om qui per lo meu mandament aministra lo meu negoci». [these are called ‘procurator’, i.e. the man who manages my affairs on my behalf] Frequently, the Latin terms are explained and paraphrased by a sentence that begins with “zo es”, ‘this means’, and it is clear that the reference is Latin. The main way of proceeding, however, is to adapt the Latin terms to the vernacular or to create new vernacular terms. In example 14, the Latin word procurator is adapted slightly to Occitan as procurador, with Occitan declension (procuradors in the nominative case). Later in the text the term is further integrated and appears with the Occitan ending -aire as procuraire. If we take the text as a whole, we observe that the 16 direct adoption of Latinisms is rather infrequent, and that the vast majority of terms are in fact recreated in Occitan3. And there is a third way of creating terms in Occitan on the basis of Latin models: the projection of the Roman content on a different, existing form; that is, the creation of “calques”. Different forms of agreement are called couenenz (<Lat. convenientia) in Lo codi, a term that stems from the feudal tradition. This is a way of linking the “new” law to the traditional one. We can summarize these techniques schematically (see Fig. 3): there can be overt Latinisms, effects of positive (=positively attested) Latinisation, such as Lat. procurator > Occ. procurator. Then there can be “negative” Latinisation, i.e. Latinisation as a means of avoiding overt Latinisms, to varying degrees: a) with minimal adaption (procurator>procurador) b) with further adaption (procurator>procuraire) c) with substitution of form (couenenz). In the case of the substitution of the form, there can be a projection of new content on an existing form (a calque) or the creation of a new form within the possibilities of the target language. Fig. 3: Types of Latinisation If we turn now to another aspect of the text, combining clauses, we can observe the following: clause-linkage techniques are well-known indicators of differences in textual traditions (Raible 2001, Kabatek/Obrist/Vincis 2010). Lo codi is a complex text ―― 3 The following Latinisms are adaptions in Occitan: adreitament, alongament, comandament, cossentiment, defendement, demandament, deuediment comenzament, difiniment, entendament, mandament, pagament, requerrement, retenement, sagrament, saluament, tardament, testament; fidanza, uenganza, fermanza, desebranza, deseuranza, usanza, doptanza, credenza, uedienza, audienza, temenza, couenenza, presenza, paruenza; accion, adoptions, cumpradons, distinccions, eleccion, escepcion, excusation, gestion, iurisdiccion, mencion, messions, peregrinacions, prescripcion, promessions, question, radons, stipulacion, suspicion, tenedons, transaccions, transactions, uendedons; amesuradament, comunalment, cotidianament, cuminalment, dreitament, eissament, engalment, espressament, generalment, leument, malament, nominadament, primeirament, propriament, sebradament, simplament, singularment, solament, tacitament, uniuersalment, utilment. 17 that requires argumentative structures that are not easy to find either in the spoken vernacular or in the written tradition of the charters. It must reproduce the whole range of subordinating patterns that can be found in the Latin model. Thus, we find not only frequent additional junctors (for the term, see Raible 2001) such as eissament, aregers, atrestals and atressi, but also a great many examples of the conditional junctor si (‘if’), so characteristic of legal texts. The argumentative passages also require causal, consecutive and concessive junctors. Concessives are, due to their cognitive complexity, generally considered to presuppose a rather high degree of linguistic elaboration. Concessive conjunctions in the Ms A of Lo codi are mainly ancara (366 times), but also ia sia zo que (12 times), ben que (three times), ancara que, mas ancara si ben, each of which triggers the subjunctive mood. The Latin concessive conjunctions etsi, etiamsi, quamvis, etc. have not found successors in the Romance languages, but are replaced in Gallo-Romance by complex innovations of the type iasia que, etc. It seems unlikely that Lo codi is a place of creation here, since most of the forms can be found in other early Old Occitan texts. It is nevertheless important to note that a relatively frequent number of concessive constructions can be found here, and that these do not occur sporadically, but belong to the fixed, established inventory of the text. For more examples and the whole range of subordination techniques, see Kabatek 2005; in our present context, it is important to highlight that the more complex the techniques of conjunction are, the more likely will they be accounted for by the technique of substitution, according to Fig. 3, because there is no oral vernacular continuity of the Latin forms. 5. Lo codi and the Romance languages Even if Lo codi seems to be a rather isolated text in its origin, I tried to show in Kabatek (2005) that, contrary to the claim that the text did not have much impact (see Selig 1995), the existing manuscripts of the Occitan version, its translations into other languages, and its traces in other texts, are all evidence of a considerable presence. According to common medieval estimates (Glessgen 2023, 128), this presence should probably be multiplied in light of the lost and currently unidentified manuscripts that were produced. At the moment, we have four complete Occitan manuscripts of the texts and several fragments. The text was translated into Latin by a translator whose name indicates Tuscan origins (Ricardus Pisanus, see Fitting 1906, Paratte 2023). There is single manuscript with a Franco-Provençal translation (Royer and Thomas 1929), and there are several French translations. On the Iberian Peninsula, we have two manuscripts of a complete Castilian translation as well as texts where the presence of Lo codi is obvious, such as the Catalan Costums de Tortosa (Massip i Fonollosa 1996). An interesting Spanish example, one that links Lo codi to the Toledo tradition of vernacular 18 texts based on Roman law established by Alfonso the Wise in the second half of the 13th century, is a manuscript of the Fuero Real, one of the most significant Castilian legal texts of that time, with marginal glosses that quote parallel texts from Lo codi (Kabatek 2007). We cannot be sure if Lo codi was known at the Toledo court, where the first extensive vernacular elaboration based on Roman law was created and in which the identification of Alfonso the Wise with the emperor Justinianus might be supposed: Both Justinianus and Alfonso had in mind the creation of an universally valid legal system without limits in time and space. The Italian Iuris Periti present at the Toledo court, such as the famous Jacobo de Junta (also known as Jacobo de las Leyes ‘James of the laws’, see Roudil 1985), may have known the text, even if they also had direct access to the Latin sources. The two manuscripts of the Castilian translation are from the 14th century, but copies of the text might have circulated there even in Alfonsine’s time. In any case, it can be supposed that further places in which the impact of Lo codi left traces will continue to be identified. 19 Fig. 4: Ms 10816 F 1r (Biblioteca Nacional de España), Castilian translation of Lo codi 6. Beyond legal texts: Roman law and Romance writing Having seen the genesis, background and impact of Lo codi in different areas of Europe, let us return to our initial statement about Gonzalo de Berceo and consider Lo codi in a broader context. As we have seen, the text is a particular manifestation of the Renaissance of Roman law in the 12th century: a text that is part of what we have called the “Bologna connection”, the spread of “new” ideas from the University of Bologna – or from satellite centres of the study of Roman law in southern France and other places – throughout Europe. Lo codi opens a tradition, probably a rather isolated one at the beginning but then increasingly continuous and spreading, of vernacular writing on Roman law. A significant blossoming of this new tradition can be seen in the Castilian 20 codification in the second half of the 13th century, with texts such as the Alfonsine Espéculo, the aforementioned Fuero Real and, finally, the Siete Partidas, the most exhaustive medieval vernacular legal work inspired by Roman law. But we believe that there is even more than this: if we look at where Romance vernacular writing emerges, we can see a close connection to the legal traditions, as in the case of Gonzalo de Berceo, who had studied legal science in Palencia and was familiar with the principles of Roman law, and I believe this is not just coincidence. If we take a closer look at early elaborate vernacular production, we can find several cases where the “Bologna connection” played a role. Is there an explanation for this? Without entering into excessive speculation, we can state the following: - Bologna emerges as a new intellectual centre in the 12 th century. The study of Roman law, especially the Digest, is limited to a small elite that comes to Bologna from different European regions. - Bologna not only receives students from the whole continent, but those who have studied there go on to occupy important positions in places across the continent. But what licenses the intellectual authority of Roman law? The classical Roman idea expressed in Justinian’s code that the law is for ‘all peoples’, “cunctos populos”, and the abstract and general character of many of the legal prescriptions, associates Roman law with universality and superiority with respect to local legal traditions. There must, however, be a powerful institution that supports this superiority, and this institution is the Catholic Church and the adoption of Roman law for the Canonical Process. The Canonical law is the catalyser and the carrier substance of the new ideas. - having returned to the peripheries, those scholars coming from Bologna or related centres (such as in southern France) were aware of the gap between the local traditions and the intellectual milieu of Bologna, of which they now formed a part. - their evident double identity, embracing both local relations and a Bolognese background, led them to contribute to the construction of an openly visible “diglossification” between the academic Latin of their studies and the newly constructed vernacular writing. A clear example of this is the Catalan Vidal de Canyelles, who studied in Bologna in the 1220s and then became Bishop of Huesca and author of the Compilatio Maior, also known as Vidal Mayor, an extensive compendium based on Roman law written in Aragonese (Del Arco 1951). - vernacular writing was not limited to legal texts but spilled over into other textual traditions, where it took up elements from orality as well as from the scant written vernacular production that existed at the time. 21 7. Conclusions It has become the fashion in the Humanities to begin papers with the rather boring and stereotypical enumeration of what is going to be presented in the text, followed by the initial hypotheses. I will not enter here into the discussion of whether we always need initial hypotheses in the Humanities, but I would like to end with four hypotheses that are in fact my conclusions, and which go from weak to strong. Hypothesis 1: The renaissance of Roman law, a movement that occurred essentially within Latin contexts, had an impact on the creation of elaborate romance texts. As we have seen, this is not only a hypothesis, but it is supported by evidence if we look at the specific case of Lo codi. I would therefore propose a rejection of the idea of a continuity in the evolution of vernacular writing from orality and texts of “conceptional immediacy” towards text of “conceptual distance” (Koch and Oesterreicher 2012). As is well known in Romance linguistics, the ingenuous idea of continuity and the slow emergence of vernacular writing in Latin contexts (model 1 in Fig. 5) is erroneous, and Romance does not in fact emerge “by itself”. Fig. 5: The ingenuous idea of continuous penetration of Romance into scripturality Rather, the emergence of Romance writing is a consequence of the Latin spelling reform in Carolingian times (Wright 1982) and the rising discrepancy between the spoken vernacular and the new reading style of written Latin. This discrepancy leads, directly or indirectly and at different points of time and space, to the emergence of “sporadic” texts such as the Strasbourg oaths or the Glosas emilianeneses (Model 2 in Fig. 6): 22 Fig. 6: Impact of the Carolingian reform Thus, Hypothesis 1 extends this view to the historical impact of a second intellectual movement, the “Bolognese Renaissance” (Model 3 in Fig. 7): Fig. 7: Impact of the Carolingian reforms and the Bolognese Renaissance Hypothesis 2 is that the Occitan summa Lo codi is a key text, maybe the key text, in the link between the renaissance of Roman law and Romance textuality, in that it suddenly presents a new textual and linguistic model for a clearly structured, hierarchically organised and exhaustive Romance legal prose with few elements of overt Latinity. The spread of this model to other Romance areas via translation puts this model at the disposal of writers in other regions and is the basis for further elaboration such as that found in the Catalan Costums de Tortosa or in the Castilian Codi as well as in the prealfonsine and alfonsine legal prose. Hypothesis 3 goes beyond the mere description of facts and interprets elaborate Romance production as a consequence of the intellectual gap between the “new legal science” and the former legal traditions, as well as the emergence of a third space of transmission of “scientific” Bolognese legal knowledge to the world of practical legal applications. The same people who were proficient in dealing with the complex Latin prose taught at Bologna produced texts in Romance. This also takes into account the fact that only the most educated had the ability to elaborate on a new written vernacular language. Finally, Hypothesis 4, the most adventurous one, goes beyond legal texts to claim that there is a phenomenon that 23 originally appears to be more a playful or creative side-product alongside the principal goal of activity: that of the production of literary Romance texts. The strongest way of formulating this hypothesis is as follows: wherever we find the influence of Roman law in the Romania, there is the high probability that literary phenomena will emerge. This can also be formulated in the opposite direction: wherever Romance poetry or other literary texts appear in the 12th or 13th century, it makes sense to ask whether there is any direct or indirect link in the background to Bologna; this includes also indirect effects such as the writing down of previously orally transmitted forms of literature. Of course, many scholars will reject this last hypothesis, on the grounds that my view has a legal bias and that I see a background involving Roman law even when there is no real evidence of this. However, the claim might be understood rather as a research programme than a final conclusion; be that as it may, I am sure that uncovering of the Bologna connection will continue to be relevant in more and more cases in the future. References Bermejo Cabrero, José Luis. 1969. El mundo jurídico en Berceo. Revista de la Universidad de Madrid 70–71. 33–52. Brunel, Clovis. 1926. Les plus anciennes chartes en langue provençale. Recueil des pièces originales antérieures au XIIIe siècle. Paris: Auguste Picard. Del Arco, Ricardo. 1951. El jurisperito Vidal de Canellas, obispo de Huesca. Cuadernos de Historia Jerónimo Zurita 1. 23-112. Derrer, Felix. 1974. Lo Codi. Eine Summa Codicis in provenzalischer Sprache aus dem XII. Jahrhundert. Die provenzalische Fassung der Handschrift A (Sorbonne 632). Vorarbeiten zu einer kritischen Textausgabe [mit Texttranskription]. Zürich: Juris. Dutton, Brian. 1978. Gonzalo de Berceo, Vida de Santo Domingo de Silos. Estudio y edición crítica. London: Tamesis. Duval, Frédéric. 2018a. Lo codi, une somme au Code de Justinien entre latin et vernaculaire. In Françoise Fery-Hue (ed.), Habiller en latin: la traduction de vernaculaire en latin entre Moyen Âge et Renaissance, 139-168. Paris: École Nationale des Chartes. Duval, Frédéric. 2018b. Les avatars romans du Codi. In Michaela Del Savio, Piero Andrea Martina, Graziella Pastore & Matteo Rivoira (eds.), Fay ce que vouldras. Mélanges en l’honneur d’Alessandro Vitale-Brovarone, 229-242. Paris: Classiques Garnier. Fitting, Hermann. 1906. Lo Codi in der lateinischen Übersetzung des Ricardus Pisanus. Halle (Reprint Aalen: Scientia 1968). Frank, Barbara & Jörg Hartmann. 1997. Inventaire systématique des premiers documents des langues romanes. 5 vol., Tübingen: Narr. García Turza, Claudio& Miguel Ángel Muro Munilla. 2006. Introducción a las Glosas Emilianenses. Logroño: Gobierno de la Rioja. Glessgen, Martin Dietrich. 2023.’Tout, dans le Codi, pose problème…’ Le lieu de genèse et la tradition manuscrite de la Somme du Code occitane. Revue de Linguistique Romane 87. 123-160. 24 Gouron, André. 2002. L’auteur du Codi. Tijdschrift voor Rechtsgeschiedenis 70. 1-20. Haskins, Charles Homer. 1927. The Renaissance of the Twelfth Century. Cambridge (Mass.)/London: Harvard University Press. Iglesia Ferreirós, Aquilino. 1998. Rex superiorem non recognoscens. Hugolino de Sesso y el Studium de Palencia. Initium 3. 1–205. Kabatek, Johannes. 2000. Lo Codi» und die okzitanischen Texttraditionen im 12. und 13. Jahrhundert. In: Angelica Rieger (ed.), Okzitanistik, Altokzitanistik und Provenzalistik. Geschichte und Auftrag einer europäischen Philologie, 147–163. Frankfurt am Main: Lang. Kabatek, Johannes. 2005. Die Bolognesische Renaissance und der Ausbau romanischer Sprachen. Juristische Diskurstraditionen und Sprachentwicklung in Südfrankreich und Spanien im 12. und 13. Jahrhundert. Tübingen: Niemeyer. Kabatek, Johannes. 2007. ‘Muyto he boa grosa’: o Renacemento Boloñés, a elaboración das linguas románicas e a emerxencia do galego escrito. In Ana Isabel Boullón Agrelo (ed.), Na nosa lyngoage galega. A emerxencia do galego como lingua escrita na Idade Media, 21-36. Santiago de Compostela: Consello da Cultura Galega. Kabatek, Johannes. 2008). Das Kastilische und der alfonsinische Hof: über Texttraditionen, Sprache und Geschichte. In Gundula Grebner & Johannes Fried (eds.), Kulturtransfer und Hofgesellschaft im Mittelalter. Wissenskultur am sizilianischen und kastilischen Hof im 13. Jahrhundert, 351-366. Berlin: Akademie-Verlag. Kabatek, Johannes. 2015. Wie kann man Diskurstraditionen kategorisieren?. In Esme Winter-Froemel, Araceli López Serena, Álvaro Octavio de Toledo & Barbara Frank-Job, Diskurstraditionelles und Einzelsprachliches im Sprachwandel/ Tradicionalidad discursiva e idiomaticidad en los procesos de cambio lingüístico, 51-65. Tübingen: Narr. Kabatek, Johannes, Philipp Obrist & Valentina Vincis. 2010. Clause-linkage techniques as a symptom of discourse traditions: methodological issues and evidence from Romance languages. In Heidrun Dorgeloh & Anja Wanner (eds.), Syntactic Variation and Genre, 247-275. Berlin/New York: Mouton De Gruyter. Koch, Peter. 1993. Pour une typologie conceptionelle et médiale des plus anciens documents/monuments des langues romanes. In Maria Selig, Barbara Frank & Jörg Hartmann (eds.), Le passage à l’écrit des langues romanes, 39-81. Tübingen: Narr. Koch, Peter & Wulf Oesterreicher. 2012. Language of immediacy – language of distance: orality and literacy from the perspective of language theory and linguistic history. In Claudia Lange, Beatrix Weber & Göran Wolf (eds.), Communicative spaces: variation, contact, change. Papers in honour of Ursula Schaefer, 441–473. Frankfurt am Main: Peter Lang. Massip i Fonollosa, Jesús. 1996. Costums de Tortosa, edició crítica, amb la col·laboració de C. Duarte i A. Massip. Barcelona: Fundació Noguera. Paratte, Maxime. 2023. La traduction latine du Codi de Richard de Pise: entre tradition(s) discursive(s) et romanisation. Revue de Linguistique Romane 87. 161-188. Pfister, Max. 1978. La localisation d’une scripta juridique en ancien occitan: «Lo Codi» manuscrit A (Sorbonne 632). In: Georges Güntert, Marc-René Jung & Kurt Ringger (eds.), Orbis mediaevalis. Mélanges de langue et de littérature médiévales offerts à Reto Raduolf Bezzola à l’occasion de son quatre-vingtième anniversaire, 285–296. Bern: Francke. Prawer, Joshua. 1962. Étude sur le Droit des Assises de Jérusalem: Droit de confiscation et droit d’exhérédation. Revue Historique de Droit Français et Étranger 40. 29–42. 25 Raible, Wolfgang. 2001. Linking clauses. In: Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible (eds.), Language Typology and Language Universals. An International Handbook, I, 590–617. Berlin/New York: De Gruyter. Rodríguez Molina, Javier & Álvaro Octavio de Toledo y Huerta. 2017. La imprescindible distinción ente texto y testimonio: el CORDE y los criterios de fiabilidad lingüística. Scriptum Digital 6. 5-68. Royer, Louis & Antoine Thomas. 1929. La Somme du Code. Texte dauphinois de la región de Grenoble, publié d’après un manuscrit du XIIIe siècle. Notices et extraits des manuscrits de la Bibliothèque Nationale et autre bibliothèques 42. I-XXXII, 1-138. Roudil, Jean. 1985. Jacobo de Junta el de las Leyes. Œuvres I: Summa de los nueve tiempos de los pleitos. Édition et étude d’une variation sur un theme. Paris: Klincksieck. Selig, Maria. 1992. Die Entwicklung der Nominaldeterminanten im Spätlatein: romanischer Sprachwandel und lateinische Schriftlichkeit. Tübingen: Narr. Selig, Maria. 1995. Volkssprachliche Schriftlichkeit im Mittelalter – Die Genese der altokzitanischen Schriftsprache. Freiburg: Universität Freiburg im Brsg. Habilitation thesis. Tagliavini, Carlo. 1959. Le origini delle lingue neolatine. Corso introduttivo di Filologia romanza. 3a ed. Bologna: Pàtron. Timmons, Patricia & Robert Boenig. 2007. The Miracles of the Virgin and Medieval ‘Spin’: Gonzalo de Berceo and his Latin Sources. The Journal of Medieval Latin 17(1). 226237. Wright, Roger. 1982. Late Latin and Early Romance in Spain and Carolingian France. Liverpool: Francis Cairns. Wright, Roger. 1997. Las glosas protohispánicas. Problemas que suscitan las glass emilianenses y silences, In Claudio García Turza, Fabián González Bachiller & José Javier Mangado Martínez (eds), Actas del IV Congreso Internacional de Historia de la Lengua Española: La Rioja, 1-5 de abril de 1997, 965-973. Logroño: Universidad de la Rioja. Wunderli, Peter. 1965. Die ältesten romanischen Texte unter dem Gesichtswinkel von Protokoll und Vorlesen. Vox Romanica 24. 44–63.