On Aristotle's World

Created on 7 December 2013 at 21.30 hours page 311  RS T  T E ’S W RD T E   KUKK E  . ntroduction S  opens his monumental commentary on ristotle’s On the Heavens, the only full commentary extant from antiquity, with the words: ‘lexander [of phrodisias] says that the subject matter [skopos] of ristotle’s treatise On the Heavens has to do with the world [peri kosmou].’ The statement is surprising for a couple of reasons. irst and most importantly, the term kosmos does not appear anywhere in the first chapter of ristotle’s work, which to the late antique commentator will have been the first place to look when it came to discerning ristotle’s intentions. n De caelo .  ristotle speaks of ‘the all’ (to pan), of everything (ta panta), and of that which is complete and in that sense perfect (to teleion) in the third chapter the uppermost region (to anōtaton) and the outermost heaven (ho eschatos ouranos) are added to the mix, signalling that the immediate discussion has to do with the celestial region. The word kosmos occurs for the first time only in the eighth chapter it is evoked in conjunction with the universe’s uniqueness (. –) and its ungenerated nature (. , a). verall, despite the tendency on the part of modern commentators to side with lexander and to call ristotle’s On the Heavens his cosmology, the term kosmos does © Taneli Kukkonen   thank audiences at the University of Victoria, the University of ritish olumbia, the University of Western ntario, the University of Dayton, and ew York University for comments and feedback, as well as an anonymous reviewer subsequently revealed to be Peter damson. The research for this article was supported by European Research ouncil project o. , SST.  In De caelo . – eiberg cf. .  ff. ll references to the reek commentators on ristotle are to the ommentaria in ristotelem raeca (editors’ names are given at first occurrence). or clarity’s sake, in this article kosmos is always translated as ‘world’, while to pan is rendered as ‘the ll’, to holon as ‘the whole’, and ouranos variably as ‘heaven’ or ‘the heavens’ according to context. This leaves ‘the universe’ as a neutral descriptor for miscellaneous purposes. n this article  have sometimes made use of existing English translations from the ncient ommentators on ristotle series if so, the technical terminology has been rendered uniform where necessary without separate notice. Created on 7 December 2013 at 21.30 hours page 312  Taneli Kukkonen not appear to have any special significance in ristotle’s exposition in this particular treatise. or does the term kosmos figure in the summary of natural philosophy given at the outset of ristotle’s Meteorology, a fact to which Simplicius draws the reader’s attention (In De caelo . –.  . –). This matters, because the Meteorology’s opening account of the parts of natural philosophy—a list Simplicius supplements with zoology, for reasons that will be made clear—could be, and often was, taken as a guide to what ristotle’s intentions were in framing the study of nature the way he did. So for ristotle to have tackled the kosmos in On the Heavens, one would expect to see the term appear there, rather than the mere mention of the orderly character of the astral motions (phoran diakekosmēmenōn astrōn). ut such is not the case. or again is there any evidence that the original reek title of ristotle’s On the Heavens would have included the term kosmos. The compound atin title De caelo et mundo, which became commonplace in medieval atin scholasticism and which thence found its way into conventional listings of ristotle’s works, evidently was either an rabic innovation or else based on an earlier Syriac model. t is worth noting in this connection how in the rabic translations of ristotle ʿālam, or ‘world’, is substituted a few times where the reek original has ouranos, which means that the ear Eastern translators very probably introduced the two-part title in pleonastic imitation of the way ouranos and kosmos are equated in De caelo . , b–. Yet a further reason to regard Simplicius’ report of lexander as odd is that a separate work, bearing the very title On the World (Peri kosmou), did circulate under ristotle’s name: lexander of ph r indeed anywhere else: see . onitz, Index Aristotelicus, nd edn. (erlin, ), s.v. kosmos, and see sect.  below.  See e.g. lymp. In Meteor. . –.  Stüve.  ndeed, the title Peri ouranou is of later provenance: see D. . llan, ‘n the anuscripts of the De caelo of ristotle’, Classical Quarterly,  (), – P. oraux, Aristote: Du ciel (Paris, ),  D. . llan, Aristotelis De coelo libri quattuor (xford, ), p. iii.  See bn al-adīm, Kitāb al-fihrist, ed. . lügel,  vols. (eipzig, ), –  further . Endress, Die arabischen Übersetzungen von Aristoteles’ Schrift De caelo [Übersetzungen] (diss. Ph.D., rankfurt a.., ), –, –. (lexander’s lost commentary is, however, according to bn al-adīm’s testimony, supposed to have been simply on ristotle’s treatise ‘n the eavens’ [Fī al-samāʾ].)  This follows the example set by Plato, Tim.   cf. also De caelo . , a. Created on 7 December 2013 at 21.30 hours page 313 On Aristotle’s World  rodisias was probably familiar with it and may have used it as an (uncredited) aide in constructing his own treatise On the Principles of the All, now lost in the original reek but extant in an rabic translation. ny desire on lexander’s part to frame On the Heavens as a work on the kosmos thus seems to put the two works in direct contention. ut the most puzzling aspect in all of this is that Simplicius sees fit to open with a citation from lexander’s lost commentary (for that is certainly what we must take the reference to be) in the first place. otwithstanding lexander’s high standing among the commentators in Simplicius’ eyes, the former is brought up in this instance only in order to be refuted. ollowing a detailed rundown of the previous positions that had been taken with respect to the scope of ristotle’s study, Simplicius concludes that On the Heavens, far from harbouring cosmological ambitions, instead functions merely as an account of the simple bodies—first aether or first body, then the four sublunary elements. The suggestion seems plausible enough on the face of it. Yet precisely because of this, it is unclear why lexander’s conflicting opinion would assume such prominence at the outset. The detail is all the more puzzling since Simplicius’ problematization of the skopos of On the Heavens was not the norm even in his own time. rom among Simplicius’ fellow trainees under mmonius, ohn Philoponus in his introduction to ristotle’s natural philosophy casually refers to On the Heavens as the treatise in which correlates to eternal things (idia tois aidiois) are discussed, while lympiodorus, commenting on ristotle’s insis See lexander of phrodisias, On the Cosmos [Mabādīʾ], ed. . enequand (eiden, ),  and – regarding lexander’s supposed use of De mundo, . Kupreeva sounds a cautionary note in her review of enequand, in Ancient Philosophy,  (), – at –.  See the discussion in . Rescigno, Alessandro di Afrodisia: Commentario al De caelo di Aristotele. Frammenti del primo libro (msterdam, ), –.  . altussen, Philosophy and Exegesis in Simplicius: The Methodology of a Commentator [Philosophy and Exegesis] (ondon, ), –, raises the question of why lexander is so prominent in this passage, but does not give an answer beyond the trust Simplicius regularly puts in lexander when it comes to setting the scene for any given investigation—which in this case is no explanation at all, given that Simplicius’ attitude towards lexander in the present context is expressly critical.  In De caelo . –, . –, . –. , .  ff., . –.  Simplicius’ basic position is accepted explicitly, although on the basis of aegerian arguments, by . P. os, On the Elements: Aristotle’s Early Cosmology (ssen, ) it is adopted silently by e.g. . Solmsen, Aristotle’s System of the Physical World (thaca, Y, ), .  Philop. In Phys. . – Vitelli. The contrast is with On Coming-to-Be and Created on 7 December 2013 at 21.30 hours page 314  Taneli Kukkonen tence on the overall immutability of even the sublunary order in the Meteorology (. , a ff.), happily takes on lexander’s description of De caelo as concerning the kosmos (In Meteor. . –  ayduck).  do not have a ready answer as to why lexander would have claimed that ristotle’s On the Heavens takes the world as its subject matter. The claim is puzzling from a Peripatetic standpoint, for reasons that will become clear presently. ut  think a thing or two can be said about why Simplicius held the statement up to scrutiny and why he rejected its implications. The discussion neatly highlights certain developments that pertain to the demands made of an account (logos) concerning the kosmos in later reek philosophy, at the same time that it reveals the pressures faced by Peripatetic philosophy in meeting these expectations. . The (well-)hidden unity of On the Heavens We may begin from an observation that is largely uncontroversial from a modern standpoint but would surely have offended Simplicius (and perhaps lexander). This is that the overall impression given by On the Heavens is one of a range of loosely related discussions collected under one banner—‘something of a rag-bag’, as R. . ankinson puts it. s is the case with many of the ristotelian works presented to us as integral treatises, On the Heavens appears to have originated as a set of separate studies strung together due to a certain family resemblance. Tensions between certain aspects of the treatise—above all, the question of whether the heavens should be regarded as animate or not, and the related issue of whether a separate immaterial mover is responsible for the celestial rotations— have moreover led some contemporary scholars to believe that De caelo is a patchwork effort, with elements stemming from different points in ristotle’s career, and that the treatise’s internal consistency and coherence are questionable at best. Passing-Away, which according to Philoponus treats of things proper to generable and perishable substances still, Philoponus’ characterization of De caelo’s remit would be difficult to square with that given by Simplicius.  ote that kosmologia appears to be a neologism. Simplicius, On Aristotle’s n the eavens . –, trans. R. . ankinson (ondon and thaca, Y, ), .  See W. K. . uthrie, Aristotle: n the eavens (ambridge, ass., ),  Created on 7 December 2013 at 21.30 hours page 315 On Aristotle’s World  ll this would have been utterly unacceptable to the late antique commentator. ot only was the ‘lesser harmony’ of ristotle with himself an unquestioned postulate in Simplicius’ act of commenting on his works—across ristotle’s various treatises as well as, of course, within—and not only was all this merely preliminary to demonstrating the ‘greater harmony’ between Plato and ristotle. What was also inadmissible was the very notion that a work deriving from one of the revered philosophers would have been composed with anything less than perfect authorial control. The assumption was that not only maximal internal coherence but also singularity of intent underlay the writings of the sages: in the notion of skopos, unity of aim matches perfectly the choice of topic, so that nothing is rendered hostage to fortune and no loose ends are left dangling. ccordingly, Simplicius, when introducing the question of what ristotle’s On the Heavens might be about, baldly posits that ‘each treatise demands a single subject [skopos] dealing with one thing, in relation to which it weaves together its individual parts’ (In De caelo . –, trans. ankinson). The implication is that this is what ristotle would have aspired to as well, and attained. rmed with this assumption, Simplicius can develop his own preferred interpretation, according to which On the Heavens treats the simple bodies, these being the ‘parts of the ll’ (merē tou pantos) which it is appropriate to examine immediately after the Physics’ study of the principles of nature is complete. Simplicius in fact intimates that an appeal to the simple bodies would have formed the second half of lexander’s original characterization. ccording to this second formulation of lexander’s position, ristotle’s pp. xv–xxv . Elders, Aristotle’s Cosmology: A Commentary on the De caelo (ssen, ), posits that De caelo exhibits signs of at least four layers of additions and (imperfectly executed) editorial harmonization, which seems excessive.  or the former point see altussen, Philosophy and Exegesis,  n.  for the latter see e.g. the useful set of translations in R. Sorabji, The Philosophy of the Commentators – ,  vols. (thaca, Y, ), i. –, and chs.  and  in R. Sorabji (ed.), Aristotle Transformed: The Ancient Commentators and their Influence (ondon, ).  adopt the terms ‘lesser harmony’ and ‘greater harmony’ from Rob Wisnovsky.  ompare Proclus on Plato’s Parmenides, In Parm.  ousin for other examples in Proclus see . ansfeld, Prolegomena: Questions to be Settled before the Study of an Author or a Text (eiden, ), – for an anonymous Platonist account on how best to identify the skopos of a work see . . Westerink, Anonymous Prolegomena to Platonic Philosophy, X, nd edn. (Dilton arsh, ), –.  In De caelo . –. Created on 7 December 2013 at 21.30 hours page 316  Taneli Kukkonen intention was to treat the world and the simple bodies (In De caelo . – . –). ut if this is so, then we should find all the more curious Simplicius’ decision to pick the more controversial first half of lexander’s formulation for inspection and refutation, rather than focusing on that aspect of it which was common ground. Simplicius’ move effectively creates a controversy where one could easily have been suppressed. n the event, Simplicius seems vaguely apologetic about the whole business, insisting that if lexander had only stuck to talk about the simple bodies and not brought the world into it, then there would have been no cause for quarrel (In De caelo . –). t would also have been all right if lexander had spoken of the world in the sense of that term covering all the simple bodies in the world (kath’ hoson peri tōn haplōn pantōn tōn en tōi kosmōi, In De caelo . ). Such an interpretation can be teased out of amblichus, if we are to believe Simplicius, and in essence it is what a number of later Peripatetics ended up saying (see below). las, lexander can only be taken as having meant that ristotle treated of the world (as world) first, the simple bodies second. t is the conjunctive kai that Simplicius objects to. . The Platonic kosmos The puzzle can be put in yet more pointed terms. Why should an account of the simple bodies alone not constitute a cosmology n other words, why would lexander’s second characterization (‘the simple bodies’) not be equivalent to the first (‘the world’), given that the two end up being one and the same, extensionally speaking Why is an elemental physics not equivalent to a cosmology ll that the sensible world contains, after all, is composed of the simple bodies, meaning that in at least one sense, when one has accounted for the elements, nothing has been left out when it comes to describing the physical world. This seems to be the implication, innocent on the face of it, of the way that lexander handles ristotle’s introduction to the Meteorology. ccording to lexander, On the Heavens treats the five simple bodies, since these are the elements of the kosmos (tauta gar stoicheia tou kosmou, In Meteor. . –. ). quinas hits upon much the same formulation when he attempts to reconstruct an acceptable interpretation of lexander on the basis Created on 7 December 2013 at 21.30 hours page 317 On Aristotle’s World  of what he reads in Simplicius’ account (this is based on William of oerbeke’s translation of Simplicius): On the Heavens looks at the simple bodies, but it does so through the filter of seeing these as the most elemental constituent parts of the universe, on the basis of which a fuller scientific account of the more complex embodied substances can be constructed. To answer the question why this will not do, a brief look at the general connotations of the word kosmos may prove helpful. s has been noted many times, the reek term originally appears to have meant any beautiful and purposefully arranged whole. Thus, omer famously uses kosmos to describe a sleeping regiment whose arms are so arranged that they are ready for battle at a moment’s notice (Il. . ) by contrast, a chaotic retreat is called ‘acosmic’ (. ). The assembly of clothes, jewellery, scents, and sandals worn by the goddess era likewise constitutes a kosmos (. ), indicating that the aesthetic connotations of the word were foregrounded early on. n the philosophical tradition Pythagoras is said to have been the first to call ‘that which contains the whole’ the kosmos on account of the order which is in it (ek tēs en autōi taxeōs, ët. Plac. . . ). f we are to believe the doxographers, the Pythagorean philosopher Philolaus, moreover, explicated the unity of the kosmos in terms of all of its different parts exactly mirroring one another, with a common centre and origin at the middle, which is to say the sun (  DK). Philolaus appears to have been engaged in an attempt to apply mathematical reasoning to his peculiar brand of cosmological speculation, something for which ristotle chides him—so it appears, at any rate, although the criticism is anonymous—in On the Heavens (. , a–b). n the opposite side, a late  ‘de simplicibus corporibus determinatur in hoc libro secundum quod sunt partes universi constitutivae’ (quinas, In De caelo, prooemium).  . R. Wright, Cosmology in Antiquity (ondon, ), – W. Kranz, ‘Kosmos’, Archiv für Begriffsgeschichte,  (), – and –.  Text and translation in . ansfeld and D. T. Runia, Aëtiana: The Method and Intellectual Context of a Doxographer, ii/ (eiden, ). or further materials see e.g. . Kerchensteiner, Kosmos: Quellenkritische Untersuchungen zu den Vorsokratikern (unich, ) for an interpretation critical of the notion that the Presocratic kosmos would designate anything like the world see . inkelberg, ‘The istory of the reek Kosmos’, Harvard Studies in Classical Philology,  (), –.  See D. W. raham (trans. and ed.), The Texts of Early Greek Philosophy: The Complete Fragments and Selected Testimonies of the Major Presocratics,  vols. (am- Created on 7 December 2013 at 21.30 hours page 318  Taneli Kukkonen materialist monist of the likes of Diogenes of pollonia could be said to ‘construct the world’ (kosmopoiei) in terms of air condensing here while becoming rarefied there (  DK), this leading to infinite kosmoi forming in a boundless void ( ): and the principle tying all this together, we are told by ristotle and others, is that everything is in some sense composed of the same stuff, given that otherwise things within each world would be unable to affect one another. While such lines of speculation can prove hard to grasp with any precision, they serve to illustrate how even in early reek philosophy, questions regarding the kosmos addressed not only the building blocks of the physical universe, so to speak (what the universe is made out of), but also what would tie them together and make them a single whole. n other words, the question was how to advance from the observation of all things (ta panta) to an understanding of the ll (to pan). ut if this is the aim, then it is immediately clear how ristotle’s work On the Heavens fails to accomplish much of anything.  comparison with Plato’s Timaeus will prove particularly unflattering. To cite Simplicius, ristotle clearly does not explain the world in this treatise as Plato did in the Timaeus, where he treated both of the principles of natural objects, matter and form, motion and time, and of the general composition of the world [koinēn sustasin tou kosmou], and gave a particular account both of the heavenly bodies and of those below the moon, in the latter case occupying himself both with atmospheric phenomena and with the minerals, plants, and animals on the earth up to and including the composition of man and of his parts. ere, however, very little is said about the world as a whole [tou kosmou pantos], and only such things as it has in common with the heaven, i.e. that it is eternal, limited in size, and single, and that it has these features because the heaven is eternal, limited in size, and single. (In De caelo . –, trans. ankinson) bridge, ), i.  ff. Simpl. In De caelo . –.  connects ristotle’s exposition here with his lost treatise on Pythagoreanism.    DK = GC . , b– the term kosmos is explicitly evoked in the famous parallel passage   DK.  D. W. raham, Explaining the Cosmos (Princeton, ), calls the earliest attempts at philosophical speculation instances of the enerating Substance Theory, as opposed to material monism, in order to emphasize how they represent attempts to provide ‘Theories of Everything’ in both the elemental and the generative senses of the word. n ristotle’s terms, raham’s hypothesis would mean that, contrary to ristotle’s exposition (Metaph. Α , a–), already the earliest ilesians would have been investigating the moving cause as well as the material one. Created on 7 December 2013 at 21.30 hours page 319 On Aristotle’s World  The underlying note of protest sounded by Simplicius is that the world as a whole does not figure as an object of investigation in ristotle’s account: but this can be taken in a number of ways. The first and more mundane understanding is that ristotle does not truly take into account all kinds of physical beings. t most, Simplicius ventures, he examines what the all-encompassing celestial body must be like and then extends the analysis of features that pertain to it to fit the rest. (In De caelo . –) This is indeed broadly consistent with ristotle’s practice in the work On the Heavens and with his third and last definition of ‘heaven’ as everything encompassed by the outermost sphere. Even assuming such a lowly goal, Simplicius says, ristotle falls well short of Plato, who in a single elegant treatise—the Timaeus, a towering achievement in the eyes of the late antique commentators—manages to recount the entire composition of the world, from the heavens through the meteorological phenomena and from the constitution of the mineral world all the way to plants, animals, and human beings. Drawing an analogy that must have seemed obvious to the late Platonist observer, Simplicius suggests that those who would wish to inspect ristotle’s theory of the world should therefore turn to all of ristotle’s works on nature put together, because all in all they cover roughly the same ground the Timaeus does (see In De caelo . –. ). This, it now turns out, is why Simplicius earlier saw fit to include the zoological treatises in an account that otherwise was based on the Meteorology: the purpose is to make the analogy with the Timaeus more complete. Simplicius even evokes as a witness the ugustan Peripatetic icolaus of Damascus, who—so Simplicius says—had produced a work entitled On the All in which he proceeded to write on everything in the world species by species (peri pantōn tōn en tōi kosmōi kat’ eidē: In De caelo . –). owever, this only serves to uncover a more fundamental flaw in the ristotelian approach to the kosmos. n account—any account—that proceeds as icolaus is said to have done, merely counting off species one by one, by definition fails to treat the world  De caelo . , b– see similarly lex. phr. In Metaph. . – ayduck (commenting on Metaph. Β , b–) and cf. verroes’ gloss on ristotle, Metaph. Λ , a–, which points to a similar elision between heaven in the singular and the world as all physical reality: In Metaph. Λ, comm.  = bū l-Walīd bn Rushd, Tafsīr mā baʿd al-t.abīʿa [Tafsīr], ed. . ouyges,  vols. (eirut, – ), iii. . –. Created on 7 December 2013 at 21.30 hours page 320  Taneli Kukkonen as an entity in its own right, as the world.  catalogue of natural kinds, in other words, whether on the level of genera or species, is not going to be enough to determine what the overall shape of the sensible world is and what accounts for its unity—not for the Platonist, at any rate. lexander’s claim to have found in ristotle’s On the Heavens a treatment of the world as a whole (peri tou pantos kosmou) is what really irks Simplicius: this is evident from his desire to restate lexander’s position several times and from his need to reiterate his disapproval as well (see In De caelo .  . –). The comparison, once again, is with the Timaeus. t will have seemed sufficiently obvious to Simplicius for him to leave the matter implicit, given how forcefully wholeness and unity figure in Proclus’ famous commentary on Plato’s work. n discussing the Demiurge’s third gift to the world, which is its perfect unicity, Proclus explains that the ll is properly (kuriōs) said to be a whole: this is because ‘the ll is wholly a whole, seeing as it is a whole made out of wholes’. This sets up the way in which the Platonic mode of exposition in describing the world is superior to all others. Proclus claims that Timaeus (as well as the Timaeus) proceeds methodically from wholes to parts and that, moreover, this top-down model of explanation replicates the direction in which the kosmos itself unfolds. Such a mereological understanding of the act of kosmopoiēsis may appear odd, but it accords well with the top-down, deductive, and demonstrative method that Proclus claims for the Platonic school. The world, and the very nature of nature itself, will naturally appear as unitary objects of attention when the adopted viewpoint is that of the supernal creative principles and their necessary outcomes and the metaphysics of participation. ne is still not doing theology as such—that would be confusing the remits of the Timaeus and the Parmenides, and Proclus’ focus is still on nature, phusis, even if in a roundabout manner—but assuredly, the lar In Tim. i. . –.  Diehl puts the matter succinctly: Plato, according to Proclus, speaks both about encosmic things and the world in its entirety (peri enkosmiōn dialeksetai pragmatōn kai peri kosmou tou sumpantos). The immediate contrast is with any strictly theological account that would take the higher principles as its primary object of study.  to men pan holon holikōs estin, hōs holon ex holōn, In Tim. ii. . – cf. Plato, Tim.  .  See D. altzly’s notes to his translation volume: Proclus: Commentary on Plato’s Timaeus, iii. Book , Part : Proclus on the World’s Body (ambridge, ),  ff.  See arije artijn’s comments against ernould’s theologizing interpretation: . artijn, Proclus on Nature (eiden, ), –. Created on 7 December 2013 at 21.30 hours page 321 On Aristotle’s World  ger degree of unicity thought to prevail in transcendent reality is allowed to guide the Platonist in the direction of viewing physical reality, too, as a unity whose very parts can be shown to be grounded in greater wholes. When Simplicius cites the failure of On the Heavens to address the principles of natural objects (tas archas tōn phusikōn), this again is in comparison with Plato, who had elegantly folded a discussion of matter, form, motion, and time into the kosmopoiēsis described by Timaeus. These would be topics covered by the treatises of the Physics, of course: but then, the Platonist follow-up would be to say that the immanent explanations offered by ristotle still amount only to ancillary causes (sunaitia) and never show how worldly events rely on transcendent principles. Proclus’ judgement is characteristically harsh. ccording to Proclus, ristotle, for all that he drew out his discussion of nature over several treatises (itself a misguided attempt at outshining Plato), only ever really attended to lowly material explanations and rarely gained so much as a formal understanding of things, to say nothing of the higher causes (In Tim. i. . –. ). Whereas ristotle along with the Presocratic phusikoi speaks at best about the material and (immanent) formal causes, Plato invokes the higher principles of productive, paradigmatic, and final causation (In Tim. i. . –), thus invoking a comprehensive list of origins or archai. n Plato’s own authority, the Timaeus was read as a treatise that addresses the nature of the ll (peri phuseōs tou pantos, Tim.  ). This kind of phusiologia appears to have been framed from at least the days of tticus as an exercise in working out the operations of the divine within sensible reality. Such an interpretation effectively fused the providentialist and physicalist perspectives when it came to establishing the purpose of the dialogue. ll of this is in evidence in Proclus’ Timaeus commentary, in a form, moreover, that neatly highlights the significance of the terminology of kosmopoiēsis  See Proclus, In Tim. i. . –. . Steel, ‘Why Should We Prefer Plato’s Timaeus to ristotle’s Physics Proclus’ ritique of ristotle’s ausal Explanation of the Physical World’, in R. W. Sharples and . Sheppard (eds.), Ancient Approaches to Plato’s Timaeus [Ancient Approaches] (ondon, ), –.  In Tim. i. . – see i. . –. , where the instrumental, i.e. the immanent moving cause, is added to the list of sunaitia.  See . Siorvanes, ‘Perceptions of the Timaeus: Thematization and Truth in the Exegetical Tradition’, in Sharples and Sheppard (eds.), Ancient Approaches, – at – and esp. nn. –.  Created on 7 December 2013 at 21.30 hours page 322  Taneli Kukkonen to the eoplatonic philosopher. or Proclus, the Platonic phusiologia when taken comprehensively is a study of the nature of the ll at the same time, the nature of the universe just is nature as such. n itself this is something immanent, but a full account of it will lead to a consideration of two transcendent causes, the Demiurge and the goddess Rhea or ecate. ccordingly, the Timaeus offers a theory of the whole stretching from the origin all the way to the ultimate end (ex archēs eis telos, In Tim. i. . ). y this Proclus presumably means to say that not only the prohodos but also the epistrophē is accounted for. nd, in line with this, we discover that the very names ‘world’ and ‘heaven’ carry a concealed import, referring as they do to the different perspectives from which one may approach what is essentially a single outcome (i.e. the sensible universe). Proclus puts it to the reader that the Demiurge’s creation is called the kosmos on account of its place in the order of procession, the way it receives the gifts of the higher realm it receives the name ouranos for the way it reverts back to its source and a third, ineffable name is evoked by Plato in honour of its remaining (monē) in the ather (In Tim. i. . –. ). Timaeus’ supposed evocation of all three names shows how his aim is to account for all three of the visible universe’s relations to what lies beyond it and what grounds it. ll of this positions the world at a very precise juncture within the order of reality, which in turn leads to a very peculiar definition for the kosmos as a whole. or the school of thens, what is constitutive of the kosmos is its mixed ontological stature, one that combines change and immutability: as Proclus puts it concisely, ‘the very being of the kosmos connotes becoming’. Simplicius confirms the same preoccupation in a lengthy meditation on elissus’ views on change and being (In Phys. . –.  Diels) and in  See . artijn, ‘Theology, aturally: Proclus on Science of ature as Theology and the ristotelian Principle of Metabasis’, in . Perkams and R.-. Piccione (eds.), Proklos: Methode, Seelenlehre, Metaphysik (eiden, ), –.  artijn, Proclus on Nature, –.  f. In Tim. i. . –. . The difference between the Timaeus and the Parmenides, meanwhile, is that the former relates things in the visible world to the Demiurge while the latter relates all things equally to the ood: Proclus, In Parm. – (cf. also In Tim. i. . –. ) and the physical nature of the investigation is underlined by the fact that the immanent form or logos and the material subject or hupokeimenon are also given consideration (In Tim. i. . –).  Proclus further specifies that the everlastingness of the kosmos is granted through infinite temporality: hē ousia tou kosmou genesin echei kai to aidion autēs kata tēn apeirian esti tēn chronikēn (In Tim. i. . –. ). Created on 7 December 2013 at 21.30 hours page 323 On Aristotle’s World  a comment on ristotle’s On the Heavens, where he says that ‘the very being of the kosmos lies in some things remaining for ever and throughout, while the generated and corruptible things change into one another’ (In De caelo . –). Simplicius draws from his definition the conclusion that sublunary existents should not properly speaking be called parts of the kosmos but instead its products, as it were (hoion apotelesmata, In De caelo . ). inimally, this reinforces the view that the Platonic kosmos has a being separate from its constituent parts, at least the sublunary ones. nstructive, meanwhile, as regards the perceived gap in ambition between Plato and ristotle is a passage in the Timaeus commentary where Proclus reprimands Theophrastus for the latter’s reluctance to pursue explanations past the postulation of a mover of the outermost sphere. Theophrastus, along with ‘all the Peripatetics’, had simply stopped there, denying that this level of reality would admit of explanation any longer. y comparison, Plato had recognized how even the movers of the ll, ‘whether these be called souls or intellects’, being something participated, ‘have an order far removed from that of the principles’ (pollostēn echei taxin apo tōn archōn), such principles being those more exalted things which are truly divine. Xenarchus of Seleucia, a Peripatetic of the first century , went even further according to the emperor ulian: Xenarchus berated both ristotle and Theophrastus for having bothered with incorporeal or intelligible substance (ousian asōmaton noēton) in the first place, seeing as the whole notion is vapid and unhelpful in the light of the fact that the celestial rotations can be explained solely by referring to natural principles. Though the targets differ (Proclus takes aim at Theophrastus, ulian at Xenarchus), we may notice that the overall point made by both hostile Platonic witnesses is essentially the same. Symptomatic of the Peripatetic style of investigation is to stop far short of the true principles of being and those perspectives that would lead one to a catholic understanding of reality.  See in this connection Philop. Aet. . – Rabe on why the world is not a god, which includes mereological materials similar to what one finds in Simplicius and – on Theseus’s ship in particular.  In Tim. ii. . –.  = Theophr. fr.  S& cf. Theophr. Metaph. , b ff.  ulian, Or.  (). , . –.  Rochefort = Theophr. fr.  S& see : Xenarchus the careful analysis in . alcon, Aristotelianism in the First Century of Seleucia (ambridge, ), –. Created on 7 December 2013 at 21.30 hours page 324  Taneli Kukkonen . ristotle’s universe n one sense, this is enough to situate Simplicius’ comments on lexander.  comprehensive account of the kosmos as an ‘intelligent and ensouled god’, as per Proclus’ formulation, was available in the Timaeus, but for its pre-eminence to emerge with clarity, lexander’s claims towards having uncovered an analogous ristotelian cosmology in On the Heavens had to be shown for the misguided exaggerations they were. nd this is what Simplicius sets out to do. Simplicius’ verdict, taken in conjunction with Proclus’ putdowns, hardly seems fair. There is no reason why ristotle should have aspired to meet the criteria set down by the cademy postSyrianus, or to think that lexander ever set out to prove that he did. t is more likely that lexander’s use of the phrase peri kosmou in his commentary on De caelo reflects Stoic vocabulary. ut even if Simplicius misunderstood lexander’s intentions when it comes to positioning the treatise On the Heavens, it is still worth investigating whether there is anything in the ristotelian corpus that could contribute to the construction of something cosmopoietic in the late antique sense of the word. Several candidates present themselves their relative standing tells us something about the pressures faced by the Peripatetic school at various points in its history. irst let it be stated that none of the preserved school treatises treat the kosmos in any sustained fashion, nor do we know of any lost treatise that would have carried that title. ut an examination of isolated mentions of the word kosmos in the school corpus reveals even more, and this is a task best executed through a simple tally.  count forty-eight references to kosmos in the authentic corpus in onitz’s Index, one to kosmika panta as referring to the universe, plus five to kosmopoiein. ut of these instances, seven (all in a single passage  In Tim. i. . –. , fusing Tim.  – with the description of the kosmos as a ‘god-in-becoming’ at Tim.  .  See . ansfeld, ‘Peri kosmou:  ote on the istory of a Title’ [‘Peri kosmou’], Vigiliae Christianae,  (), –.  t bears noting that when a late antique philosopher of the likes of Proclus approaches the way in which a Presocratic thinker ‘makes the world’, he sees in it the way in which that thinker sees the rational ordering of the universe as being laid out. The implicit corrective applied here is that even what appear to be genuinely genetic accounts of the world’s genesis in early reek philosophy are in fact to be taken as historiai in the more timeless sense of that term. Created on 7 December 2013 at 21.30 hours page 325 On Aristotle’s World b  b in the Politics: . ,  – ) in fact refer to the retan magistrates (kosmoi) and can thus be dismissed out of hand, while a further thirteen have nothing to do with the universe but instead refer to some particular instance of positive ordering in the omeric sense. ourteen further instances derive from citations of Presocratic or Platonic usage (this also takes care of every example of kosmopoiein and the kosmika panta), while eight in the Meteorology refer to some part of the physical universe, either the ‘lower world’ or the ‘upper world’ but never ‘the world’ as a whole. Similar anomalous uses are encountered also in Metaphysics Κ (a) and in the Nicomachean Ethics (. , b), once each. This leaves only four instances in the entire corpus where ristotle of his own initiative uses the term kosmos in anything like the sense of the universe. our! ut there is more: none of these four instances amounts to anything much at all in the philosophical sense. n the Politics (. , b) ristotle says that the gods and the world have no external actions, only internal activities, which because of the parallelism and the political context can be taken to be mere metaphor (see Section ). n De caelo . , b, ristotle avers that the kosmos is spherical, while at . , b, he speaks offhandedly about what is meant by transversing the kosmos (i.e. the physical universe), neither a particularly pregnant statement. This leaves only the principle stated at De caelo . , a, that ‘the order of the world is eternal’ (tou kosmou taxis aidios estin). This is certainly a deeply felt ristotelian sentiment, and one that finds ample corroboration elsewhere. ut standing on its own, stranded in a less-appreciated chapter of a less-appreciated treatise (De caelo .  treats the position of the earth within the universe), it does not yet amount to anything at all. Strictly in the terminological sense, any impetus the Peripatetics may have had for talking about the kosmos they must have received from sources other than ristotle.  have excluded from the above tally the pseudepigraphic Economics (which at a says nothing of significance) and the De mundo (on which more below). ut a word should be said about the lost dialogue On Philosophy. n this exoteric work ristotle, if we are to believe reports, made more liberal use of the term kosmos. The treatise appears to have included among its strands of argumentation an appeal to the essential goodness as well as imperishability of the current world order. iven how the dialogue by all accounts  The principal testimony is found in Philo, Aet. – ff. see . Effe, Studien zur Created on 7 December 2013 at 21.30 hours page 326  Taneli Kukkonen adopted a more popular tone and how it presented a more or less unified front with the Platonic tradition, the evocation of the term kosmos is perhaps not that surprising. owever, as has been noted by David Runia, when it comes to just these points, the vocabulary we find in Philo’s testimony in particular may have undergone substantial revision. onsequently, Philo’s work On the Indestructibility of the World forms a less than trustworthy guide to the precise preoccupations of ristotle’s original treatise. To say that On Philosophy would have provided a substantial account of the kosmos such as was missing from De caelo and the rest of the school treatises, or even that it made copious reference to the concept, must therefore remain an unsubstantiated conjecture. ext comes the pseudo-ristotelian treatise On the World (Peri kosmou). This work, which most likely stems from the early imperial period—that is, more or less concurrently with Philo’s authorship—announces its concern with the kosmos in its very title. nd it is noticeable how the author everywhere seeks to tease out, establish, and underscore the theological and ethical implications of the Peripatetic philosophical school’s natural philosophy. There is talk here of harmonia (a, a ff.), of theologizing (theologizein, b), and of course of kosmos as opposed to akosmia (a–). The world is even defined explicitly in terms of the ranking and ordering of the whole, by and through god. t is plain that all this is said in competition with both the Platonists and the Stoics, in an attempt to pre-empt any criticism concerning a reputed insufficient piety in the ristotelian worldview. Still, the work provides no hint of its author perceiving there to be any sort of gap when it comes to explaining () how an immaterial Prime over is supposed to instigate motion in the Kosmologie und Theologie der Aristotelischen Schrift ‘Über die Philosophie’ (unich, ).  D. T. Runia, Philo of Alexandria and the Timaeus of Plato (eiden, ), –.  Well attested in the manuscripts with minor variations: see W. . orimer, Aristotelis De mundo (Paris, ), , n. Titulus.  tōn holōn taxis te kai diakosmēsis, hupo theou te kai dia theon, b–. f note here is that another, more mundane definition of kosmos as consisting of the heaven and the earth put together, along with the natures they contain (b–), looks like a good candidate for a predecessor and possible inspiration to the rabic custom of calling Peri ouranou by the compound name Fī al-samāʾ wa-l-ʿālam (see n.  above)—presuming that ʿālam there is to be construed in an equally mundane sense as designating, roughly, ‘earth’. Created on 7 December 2013 at 21.30 hours page 327 On Aristotle’s World  outermost sphere () how exactly this single motion translates into multiple motions throughout the kosmos and, most crucially, () how it is that the many motions arising in the universe as a whole amount to an ordered totality. On the World also gives us little to go on when it comes to unifying the efficient model of causality in Physics  with the final causality of Metaphysics Λ moreover, its author sees no problem in asserting that the Ruler of the ll resides in resplendent solitude, ignorant of much of what lies beneath its dignity, and that perhaps individual events in the sublunary world do not touch upon the divine majesty at all. On the World, in other words, presumes much and promises more, but its flowing rhetoric is not matched by corresponding explanatory power. oving past the ristotelian and pseudo-ristotelian materials, the next proper port of call is lexander of phrodisias. lexander’s contributions form the focus of the next section. ut before this, it is worth taking a moment to consider just why the kosmos does not figure more largely in ristotle’s esoteric corpus—why it is relegated to the margins and the pseudepigrapha. Why do cosmological perspectives provide such an ill fit for the overall ristotelian pattern of explanation and understanding (a) The first and most fundamental stumbling block,  submit, is that for ristotle there simply is no world, conceived of as a single object, such as would admit of a unified investigation. The physical universe just is not a single being rather, it is a collection of beings—a collection, moreover, that is divided into two highly dissimilar groups, sublunary and celestial entities. While ristotle never defines kosmos anywhere—he would not, since he has scarcely any need for the concept—the more conscientious Peripatetic philosophers beginning with lexander agree on this point, and many of them address it directly. s verroes records lexander’s authoritative statement on the matter, ristotle never believed that ‘the ll’ (al-kull) would form a single continuous totality (jumla wāh.ida muttas.ila), nor—so the implication goes—should the faithful ristotelian. The comment is made all the more revealing by emanat to  : An f anything, R. W. Sharples, Peripatetic Philosophy  Introduction and Collection of Sources in Translation (ambridge, ), –, shows the paucity of reliable or truly informative materials beyond those relating to either ristotle, pseudo-ristotle, or lexander.  f we take the rabic title of lexander’s Fī mabādīʾ al-kull to reflect a reek original of the likes of Peri tōn archōn tou pantos, then lexander will have been mindful of the fact that kosmos is less of an ristotelian term than is to pan. Created on 7 December 2013 at 21.30 hours page 328  Taneli Kukkonen ing from lexander’s prooemium to Metaphysics Λ, the treatise in which according to lexander’s estimation ristotle’s account of the ultimate structure of reality culminates (see the next section). nterpreting verroes’ testimony is complicated somewhat by the fact that whereas ristotle in the reek talks about a hypothetical treatment of the ll as a whole (ei hōs holon ti to pan, Metaph. Λ , a), verroes’ rabic for the same Metaphysics passage has ‘the ll as a totality’, al-kull ka-l-jumla, which is not quite the same thing. ut whichever term one prefers, the fundamentally composite nature of the ll comes through with clarity. ater Peripatetics largely follow lexander. or example, bū as.r al-ārābī, a tenth-century aghdādī philosopher, explains in his Principles of Beings that the world (al-ʿālam) is a collection made up of six kinds of bodies in total (al-jumlat al-mujtamaʿat min hādhihi l-ajnās al-sittat min al-ajsām): these are, in descending order of nobility, () the celestial spheres, () the rational and () irrational orders of animals, () plants, () minerals, and—on the simplest level—() the four sublunary elements. n a piece of polemic directed against ohn Philoponus, al-ārābī furthermore claims that in his treatise On the Heavens ristotle ‘intended to explain that the world is made up of bodies that possess different substances and that the world is not a homogeneous thing’. The comment is occasioned by al-ārābī’s desire to reiterate against Philoponus how the celestial region follows a different set of rules from the sublunary domain—by now, a familiar point—but the broader lesson stands: the world is not any one thing, nor can it be treated as such. inally, in the Philosophy of Aristotle al-ārābī straightforwardly states that what is meant by the world is the totality of bodies. This last formulation is the same we find in bn Sīnā’s (the atin vicenna, –) Book of Definitions.  In Metaph. Λ, comm.  = Tafsīr, iii. . –. verroes’ testimony intimates that ristotelian categorical theory plays into the discussion of why all reality is not of a piece: what is underlined is the primacy of substance, as befits an introduction to Metaph. Λ .  bū as.r al-ārābī, Al-siyāsat al-madaniyyat al-mulaqqab bi-mabādīʾ almawjūdāt, ed. . ajjār, nd edn. (eirut, ), . –.  rabic original in . ahdi, ‘The rabic Text of lfarabi’s Against John the Grammarian’, in S. . anna (ed.), Medieval and Early Modern Studies in Honor of Aziz Suryal Atiya (eiden, ), – at – English translation in . ahdi, ‘lfarabi against Philoponus’, Journal of Near Eastern Studies,  (), – at .  l-ārābī, Falsafat Arist.ūt.ālis, ed. . ahdi (eirut, ), , §§ – bū Created on 7 December 2013 at 21.30 hours page 329 On Aristotle’s World  This usage is further echoed by verroes in his Commentary on the De caelo. lthough verroes’ take on ristotle differs somewhat from the other commentators—for him, De caelo treats extended body—he is clear on the point that there is no real sense in which the world could be taken to be a single being. Extended body as such is as if the genus (ka-l-jins) under which the simple bodies fall: but after a preliminary treatment of some of the common features of such extension, one can only proceed to treat each of them separately, since the nature of heavenly and sublunary body is radically different. The very structure of On the Heavens, then—the way it opens with several chapters charged with the task of establishing once and for all the utter dissimilarity of the celestial and subcelestial regions—works against any desire to treat the universe as a single whole. enturies later, the Parisian rt master ean uridan (d. ) still makes much the same observation as al-ārābī does. n his Questions on De caelo uridan states several times that whenever the term ‘the world’ (mundus) is evoked, it in fact supposits for an aggregate of beings (aggregatum ex entibus). bstracting from the nominalist vocabulary, uridan’s point is fairly simple: ‘the world’ is not a name for a single being, rather, it stands for some determinate set of entities, either all beings universally (universitas omnium entium) or else everything physical, excepting from the picture od and the separate substances. nd this verdict seems correct when it comes to ristotle: even on the rare occasion when ristotle ʿlī bn Sīnā, Kitāb al-h.udūd, in Tisʿ rasāʾil (onstantinople, ), . vicenna interestingly adds that one speaks also about ‘the world of nature’, ‘the world of soul’, and ‘the world of intellect’ in the sense that these are ‘whole totalities’. The explanation underlines how in this second sense of ‘world’, a fundamental homogeneity is assumed regarding the things that make up a particular world.  See bn Rushd, In De caelo , comms. – [–], in Averrois Commentaria magna in Aristotelem: De celo et mundo [Commentarium magnum], ed. . . armody,  vols. (euven, ). The rabic original for these comments is missing: see bn Rushd, Commentary on Aristotle’s Book on the Heaven and the Universe. Sharh. kitāb al-Samāʾ wa-l-ʿālam, facsimile of the manuscript produced by . Endress (rankfurt a.., ), .  bn Rushd, Commentarium magnum, comms. – Talkhīs. al-samāʾ wa-l-ʿālam, ed. . al-Dīn al-ʿlawī (ez, ), –.  This, of course, is the standard complaint against ristotelian cosmology in entry-level textbooks on the history of science to this day, although there the comparison is with early modern science.  ean uridan, Quaestiones super libris quattuor De caelo et mundo, ed. E. . oody (ambridge, ass., ), bk. , qq. , , and . Created on 7 December 2013 at 21.30 hours page 330  Taneli Kukkonen speaks, for example, of the nature of the whole (tou holou phusis, Metaph. Λ , a) or the being of the ll (tou pantos ousia, a), this is quickly brought back to the level of beings in the plural (ta onta, a). n saying that ristotle in this sense has no concept of a world— that is, no robust belief in the physical universe as a single entity— am deliberately going against an argument made recently by ohan atthen and im ankinson that aims to show how ristotle treats the whole physical universe precisely as a hylomorphic compound. atthen and ankinson base their claim on a subtle interpretation of ristotle’s argument for the universe’s completeness in De caelo .  and on the explanatory role played by natural places in the motions of the four elements. This points in the direction of a limited universal teleology which, however, atthen in a follow-up piece is careful to denude of any providential or animist connotations. atthen cautions against treating ristotle’s world as anything akin to a single living being even as he advocates recognition of certain holistic presuppositions in ristotle’s cosmology, as the title of his latter essay has it. Space does not permit a complete analysis of atthen and ankinson’s thesis, nor is it the purpose of this article to provide a refutation. n the present context, let it just be said that  share Theophrastus’ scepticism—which, it should be said, is conscientiously recorded by atthen and ankinson—when it comes to whether even a weak teleological conception of a universal order can be decoupled from the organicist metaphor, as atthen wants to do. Theophrastus in his Metaphysics picks out precisely the doctrine of natural places for critical investigation when it comes to probing the outer limits of teleological explanation within an ristotelian framework. ccording to Theophrastus, the reason it makes sense to talk about natural places in the context of animal organs and limbs is that each of these is conducive to the continued  . atthen and R. . ankinson, ‘ristotle’s Universe: ts atter and orm’, Synthese,  (), –.  . atthen, ‘The olistic Presuppositions of ristotle’s osmology’ [‘olistic Presuppositions’], Oxford Studies in Ancient Philosophy,  (), –.  inimally,  should say that  find some of atthen and ankinson’s textual evidence unconvincing. The sunholon ouranon at De caelo . , a, for example, seems to me a collective term, based both on the formulation itself and on the argumentative context.  atthen and ankinson, ‘ristotle’s Universe’, –.  See atthen, ‘olistic Presuppositions’, –. Created on 7 December 2013 at 21.30 hours page 331 On Aristotle’s World  life and well-being of the overall organism. n the case of animals, it is also easy to see how the whole plays a determining, indeed definitive, role with regard to the parts (see rist. Metaph. Ζ , b– PA .  etc.). owever, in the case of the movements of the elements it is more difficult to see how these would in any way promote the ‘being of all things taken together’ (tou sumpantos ousia). The overall impression is that there is no easy way even to determine what the whole is whose life or flourishing the parts are supposed to serve (Theophr. Metaph. a–). Despite Simplicius’ efforts to portray Theophrastus’ ruminations in the light of a eoplatonic emanative hierarchy, and notwithstanding modern suggestions of a reading of Theophrastus that would include an organicist and indeed holistic cosmology, the staunchly aporetic way in which he approaches the doctrine of natural places sufficiently shows to my mind what the Peripatetic problem is with any attempt to view the universe as a whole. Without a notion of the universe being for the sake of something, it is hard to see how the constituents that make up its set could ever be seen to form a unity. (b) This ties in with my second, much less controversial thesis, which is that the universe for ristotle does not form a single teleologically oriented and ordered whole. Unlike Plato, who in the Timaeus as well as in the Statesman ( –) is happy to make use of the image of the world as a single animal or organism—one with a single soul and a shared life—ristotle resists mightily the notion that everything in the world would serve a single purpose. The difference, famously, is reflected even in the Politics, where ristotle roundly mocks the Republic’s postulation of maximal unity as a good for the state (Pol. . –). ut then, this is revealing also from  Simpl. In Phys. . –.  Diels . van Raalte, ‘The dea of the osmos as an rganic Whole in Theophrastus’ Metaphysics’, in W. W. ortenbaugh and R. W. Sharples (eds.), Theophrastean Studies on Natural Science, Physics and Metaphysics, Ethics, Religion and Rhetoric [Theophrastean Studies] (ew runswick, ), –.  See R. Sorabji, ‘s Theophrastus a Significant Philosopher’, in . . van phuijsen and . van Raalte (eds.), Theophrastus: Reappraising the Sources (ew runswick and ondon, ), – at –.  See further . . ennox, ‘Theophrastus on the imits of Teleology’, in W. W. ortenbaugh, P. . uby, and . . ong (eds.), Theophrastus of Eresus: On his Life and Work (ew runswick, ), –, repr. in ennox, Aristotle’s Philosophy of Biology (ambridge, ), – . Ellis, ‘The porematic haracter of Theophrastus’ Metaphysics’, in ortenbaugh and Sharples (eds.), Theophrastean Studies, –. Created on 7 December 2013 at 21.30 hours page 332  Taneli Kukkonen a cosmological standpoint: underlying the Platonic ideal of the Republic is the notion of a form of the ood, and for the systematizing Platonist this is what the Demiurge of the Timaeus also acknowledges as a regulating principle when fashioning the visible universe. ristotle, by contrast, takes it as axiomatic that each natural kind—really, each individual representative of each natural kind— has its own good to pursue, which makes the term ‘the good’ as multivalent as the term ‘being’. The impasse itself hints at one last path left open for the ristotelian to explore. Perhaps in ristotle, the very notion of being— more specifically, being as actuality, and in some sense fullness of being—could be explanatory somehow of the very shape and contents of the universe atthen has put forward a version of this argument, albeit in elliptical form. That it is such is not to my mind coincidental: for while there are certainly hints to the effect of such a belief lurking in the background of ristotle’s work, these are so few and so loosely joined that one must bring a whole host of outside assumptions to the school texts in order to make the conception work. ost famous in this regard is probably the chapter in On Coming-to-Be and Passing-Away (. ) in which ristotle posits that perennial coming-to-be is the next best thing to true immortality and that this is why the generation of the elements from one another is organized the way that it is. What appeals about this notion is that it joins an account of the make-up of the universe with the notion of metaphysics as first philosophy. owever, the assignment of this arrangement—the interchange of the elements being guided by the heavens and specifically by the sphere of the sun—to ‘the god’ (ho theos, b) is problematic to say the least, as there is little indication as to which god could possibly be meant thereby.  NE . , a–. n the limits of ristotelian teleology overall see . R. ohnson, Aristotle on Teleology (xford, ) on the issues surrounding universal teleology in particular, ibid. –.  atthen, ‘olistic Presuppositions’, –.  The first book of the Meteorology can be viewed as a technical account of how this might be thought to happen on the mechanical level (as per Solmsen), although many of the details remain unclear.  few passages in ristotle’s biology argue analogously to GC .  that the perpetuation of the species is the next best thing to individual immortality: but in none of these does ristotle discuss any implications on the level of the kosmos rather, the point seems to be simply that living beings seek to extend the continuation of their own existence through procreation. See e.g. DA . , a–b GA . , b–a. Created on 7 December 2013 at 21.30 hours page 333 On Aristotle’s World  The Prime over hardly acts intentionally and craftsman-like in this manner, and no other candidates readily present themselves. The majority of modern commentators have consequently passed over the question with evident embarrassment. Even Richard odéüs, who takes an otherwise robust view of ristotle’s remarks concerning the gods, ultimately explains away the reference to divine world-fashioning in On Coming-to-Be and Passing-Away in terms of a metaphorical transference, with ‘god’ standing in for ‘nature’. s  hope will have become clear by now,  am not satisfied that the move represents anything more than an evasive manœuvre. Phusis in ristotle stands for the nature of an individual thing, and unless we simply assume that the universe as an individual—as kosmos—has its own phusis, then it is hard to say what the nature is that is supposed to fashion one part of the universe (the heavenly motions) in the light of the needs of another (the sublunary existents). n the other side, to assume such a cosmic nature would beg the question, in addition to which we would be forced to regard ristotle’s world as itself a divine entity generative of—more being f such an inspirational view really were in ristotle’s sights, one would rather expect him to spell it out. hristopher olmo in an otherwise often confounding study of al-ārābī has some intriguing observations on the history of the problem. e notes that al-ārābī, in a book entitled The Philosophy of Aristotle, seems to advocate a pursuit of knowledge concerning the purpose of the world as a whole, where reflection on the whole would somehow disclose the purpose of the parts. Such a universal teleology would certainly make of the universe a single being (the word used by al-ārābī is al-kull, the ll, which translates to pan), yet al-ārābī’s treatment ends up looking vague and far from conclusive. The Philosophy of Aristotle traces the line of discussion about parts in service of the whole until we reach metaphysics, which as the divine science would disclose the ‘purpose of the totality of the world’: but ristotle’s Metaphysics makes only a tentative start towards this, which is the reason al-ārābī famously concludes The Philosophy of Aristotle on a despondent note, admit So both oachim and Williams. See R. odéüs, Aristotle and the Theology of the Living Immortals (lbany, Y, ), – and .  . . olmo, Breaking with Athens: Alfarabi as Founder (anham, d., ), –.  Created on 7 December 2013 at 21.30 hours page 334  Taneli Kukkonen ting that ‘we do not possess metaphysics’. ut olmo also points to how aimonides in the Guide of the Perplexed (. ) appears to put forward one possible answer: aimonides says that the world’s ultimate purpose is the actualization of all that is possible, and that consequently the being and perfection of each [type of] existent is constitutive of the world’s perfection as a totality. These are bold proclamations to make on the basis of ristotle, and olmo expresses doubt that they are made entirely in earnest, or at least on predominantly philosophical grounds. ut the least we can say is that this is at once a notion towards which an ristotelian might feel drawn, and at the same time one for which wholly insufficient explicit evidence exists in the set of school texts we have. one of this is very new, of course. The whole line of thought is already present in Theophrastus’ Metaphysics, where Theophrastus first hints that the issue of the Prime over might connect somehow with how the parts of the universe cohere in one whole (Metaph. a–), then hastens to add that the mechanics by which this happens remain wholly unclear—this despite the topic’s evident centrality to the whole project of establishing first principles (a–). We may conclude that it is the hinted-at yet never substantiated links between sublunary, superlunary, and immaterial existence that are decisive of whether ‘the world’ in the end is a coherent concept to use in the context of ristotle, and hence the demonstrability of such links that either makes or breaks the prospects of a true ristotelian cosmology. s we shall see from the example of lexander, the matter can prove hard to decide. . The principles of the ll s has already been mentioned, lexander’s most comprehensive treatment of the questions surrounding the cosmic order is found in his treatises On the Principles of the All, extant only in an rabic  See al-ārābī, Falsafat Arist.ūt.ālis, §§ , ,  on al-ārābī and ristotle see further T.-. Druart, ‘l-ārābī, Emanationism, and etaphysics’, in P. orewedge (ed.), Neoplatonism and Islamic Thought (lbany, Y, ), –.  olmo, Breaking with Athens, .  . . ovejoy, The Great Chain of Being (ambridge, ass., ), provides ample further testimony to the powerful draw of this idea, of course, and aptly points out that the true philosophical engine driving its adoption by monotheist intellectuals is Platonic (Tim.  – ) rather than ristotelian. Created on 7 December 2013 at 21.30 hours page 335 On Aristotle’s World  translation as Fī mabādīʾ al-kull. n addition to this, several minor self-standing works address different aspects of the same problematic, enough so that we may say that the od–world relationship forms a running theme in lexander’s independent authorship. n two interlocking articles, ob Sharples has provided a comprehensive overview of lexander’s efforts to craft an effective Peripatetic counterpart to the theologies of the Stoics and the Platonists. n the present context,  shall content myself with remarks that have a bearing on the notion of kosmos in these discussions. To establish first what lexander was up against, consider, for example, the treatise on astronomical matters (either meteōra or kuklikē theōria) by the Stoic leomedes, written around the time lexander took on the mantle of diadochos. leomedes in his introduction maintains that the term kosmos is used in many senses however, its most apposite meaning has to do with the orderly distribution (diakosmēsis) of the universe’s constituent parts. This could yet be meant in a fairly mundane sense, but it becomes clear that leomedes has something grander in mind when in what follows he cites the administering of order by ature, adducing as evidence the ordering of the parts within it the orderly succession of what comes into existence the sympathy of the parts in it for one another the fact that all individual entities are created in relation to something else and, finally, the fact that everything in the cosmos renders very beneficial services. (Cael. . – Todd trans. owen and Todd) This designates ature as a providential force and universal teleology as an ordering that benefits humanity most of all. oth are crucial Stoic tenets, both illustrate the way in which the ellenistic schools sought to exalt the visible cosmic order: both grasp with both hands ideas for which at best meagre hints can be found in ristotle. leomedes’ presentation closely echoes the way in  R. W. Sharples, ‘ristotelian Theology after ristotle’, in D. rede and . aks (eds.), Traditions of Theology: Studies in Hellenistic Theology, its Background and Aftermath (eiden, ), – id., ‘lexander of phrodisias and the End of ristotelian Theology’, in T. Kobusch and . Erler (eds.), Metaphysik und Religion: Zur Signatur des spätantiken Denkens (eipzig, ), –.  See R. . Todd, ‘The Title of leomedes’ Treatise’, Philologus,  (), –.  or the notion that nature always goes for the optimum see De caelo . , a–  . , a– . , a–b . , b but in each of these cases the more plausible interpretation is that the nature referenced is simply the nature of each thing, seeking its peculiar perfection. nly De caelo . , a–, appears to inti- Created on 7 December 2013 at 21.30 hours page 336  Taneli Kukkonen which Diogenes aertius discusses Stoic definitions of the kosmos (. –), together with the reasonableness and providential care that it manifests (. –). Telling,  think, is the way Diogenes claims the Stoics’ first referent for kosmos to be god himself, or the power permeating the universe, whence everything derives and into which everything periodically reverts. This makes it relatively easy to see what alcidius (In Tim. ch. ) might mean when he says that the Stoics regarded the body of the world (corpus mundi) as not only one and whole, but also a substance. To borrow ristotelian terminology, the unity of the Stoic world issues from both the material and the formal aspects of its being.  Stoic might even go so far as to say that the world’s parts only have their being as parts of the whole. nd even if Proclus would probably remain unimpressed, such a Stoic perspective on the world could even be reconciled with the Platonic: arsilio icino says that the proper subject matter of the Timaeus is that ‘universal nature’ which acts as the seminal vivifying power of the whole world, subdued by the world soul but presiding over matter. Seen in the light of the late antique demand to have a single subject matter for a single treatise icino’s suggestion is rather ingenious: if there is a single kind of thing that ties together the mundane and the supercosmic, then such a logos, precisely as phusis, would be what the Timaeus investigates. ut for an ristotelian, none of this would work, since the proximate species and their natures are many and incommensurable, mate that nature could have ordained the properties of higher things (the heavens) for the sake of the lower (the stability of sublunary existence), but ristotle’s choice of words is telling: it is as if (hōsper) nature had foreseen this, which when read closer appears to be a counterfactual. or anthropocentrism in ristotle the evidence is even scantier, but see Pol. . , b–, and for a spirited defence, D. Sedley, ‘s ristotle’s Teleology nthropocentric’, Phronesis,  (), –.  In Tim. ch. . or the Stoic view see P. Scade, ‘Stoic osmological imits and their Platonic ackground’, in V. arte, . . cabe, R. W. Sharples, and . Sheppard (eds.), Aristotle and the Stoics Reading Plato (ondon, ), –.  to tou kosmou merē tō pros to holon pōs echein kai mē kath’ hauta einai: hrysippus, as reported by Plut. Stoic. repugn.  –  for comments see R. Sorabji, Matter, Space and Motion: Theories in Antiquity and their Sequel (ondon, ), –.  Proclus, In Tim. i. . –. , points out that from a Platonic point of view the Stoic philosopher—here, hrysippus—fatally fails to distinguish between transcendent and immanent causes and therefore remains unable to cover adequately all the phenomena that an account of the kosmos must, from the world’s intelligibility to the workings of divine providence within it.  icino, Opera (asel, ), intro, ch. , ed. P. . Kristeller,  vols. (Turin, ), ii. . Created on 7 December 2013 at 21.30 hours page 337 On Aristotle’s World  as ristotle’s remarks against elissus illustrate. The bottom line for the student of nature, and indeed for the student of being, are the many and disparate substances that inhabit the universe, of the likes of horses and human beings, and these are not united in having a single substantial form (Phys. . , a–). or will an appeal to prime matter do when it comes to defining an overall subject that would unite natural philosophy, as Simplicius notes in his comments ad loc. against the Eleatics (In Phys. . –. ). lso, contrary to what On the World would like to suggest, ristotle’s manifest repudiation of a world soul (DA . , a–) deprives the conscientious Peripatetic of any straightforward way of presenting all worldly phenomena as issuing from a preordained harmony such as was available either to the Platonist interpreter of the Timaeus or to the Stoic philosopher endowed with the notion of an all-pervasive logos.  Peripatetic of the likes of ritolaus will not have helped the school’s reputation with his argument that the world is eternal due to it being the cause of its own existence. So what was a Peripatetic philosopher to do when attempting to flesh out an ristotelian account of the kosmos lexander’s response in the Principles is instructive on several levels. (a) irst of all, lexander lays down as axiomatic that an ristotelian investigation into first principles will proceed from effects to causes, since demonstrative knowledge regarding the first principles is unattainable (Mabādīʾ, § ). The methodological principle of course is of fundamental importance to ristotelian science in commenting on ristotle’s second and third aporiai in Metaphysics Β, lexander elaborates on its significance for the foundations of metaphysics. ut whatever the details of lexander’s concep See T. K. ohansen, ‘rom Plato’s Timaeus to ristotle’s De caelo: The ase of the issing World-Soul’, in . . owen and . Wildberg (eds.), New Perspectives on Aristotle’s De caelo (eiden, ), –.  Philo, Aet.  = ritolaus, fr.  Wehrli. ritolaus’ seeming equation of god with both intellect and aether, meanwhile, brushes up uncomfortably close to Stoicism: see Sharples, ‘ristotelian Theology after ristotle’, . or the tangled interpretation of icero’s De natura deorum . . , which similarly questions whether ristotle’s god is intellect, the world, or the heaven, see . P. os, Cosmic and MetaCosmic Theology in Aristotle’s Lost Dialogues (eiden, ),  ff.  lex. phr. In Metaph. .  ff. n his comments on the first chapter of the Metaphysics lexander concludes that metaphysics aims at a non-demonstrative knowledge of the first principles of everything, with the ultimate first principle being the good that is the final cause: In Metaph. . –. . Created on 7 December 2013 at 21.30 hours page 338  Taneli Kukkonen tion, his basic contention runs directly counter to the school of thens’s approach and Proclus especially, to whom the deductive and demonstrative mode of philosophizing provided by Plato and a presumed Pythagorean tradition formed a point of pride. The famous tradition according to which ristotle tended to broach questions in theology from the direction of nature (or metaphysics from the direction of physics), whereas Plato talked even of nature in a theological register (phusiologein theologikōs), can in fact be seen as a half-defensive way of acknowledging this methodological contrast. ccording to the more generous interpretation, Plato preferred the top-down mode of examination because it better concurred with the underlying structure of reality, while ristotle’s bottom-up way of proceeding could be justified as being part and parcel of the natural philosopher’s normal viewpoint (and perhaps something to which ristotle had become unduly but none the less excusably accustomed). Yet such an attempt at harmonization disregards the bluntness of lexander’s testimony. ccording to lexander, apodeixis simply is not available for first principles and, there being no higher method of enquiry for the committed Peripatetic (Platonic dialectic manifestly does not count), a bottom-up methodology by consequence really is the best that any conscientious philosopher can hope for. t is in the light of things better known to us that we move on to things better known by nature, and only in so far as the explanation of physical phenomena warrants that we postulate the existence of immaterial entities. This already separates the Peripatetic approach to theology from the Platonist one in one important respect: ristotelian philosophy ventures onto theological terrain only where mundane explanations are found to be insufficient, while the line of explanation in the Platonic philosophy of Plotinus and Proclus may advance freely from the direction of the intelligible and the supraintelligible principles towards the sensible. The implications of this for cosmology are immediately apparent in the corresponding dearth of references to the kosmos in the Peripatetic vocabulary.  Though of course this higher form of proof will be called Platonic dialectic: see artijn’s comments in Proclus on Nature as well as D. . ’eara, Pythagoras Revived (xford, ), –.  The apologetic tone in Philop. In Phys. . – is quite evident. Simpl. In Cat. . – Kalbfleisch talks about how for Plato, natural things participate in the things above, while ristotle by contrast considers even higher things with an eye towards their relation to nature. Created on 7 December 2013 at 21.30 hours page 339 On Aristotle’s World  acking a god’s-eye point of view, the ristotelian philosopher will only rarely, if at all, have occasion to view the world as a unitary object of study. (b) Second, it is significant that lexander, when it comes to providing an ristotelian account of the ordered universe to rival that of the Stoics and the Platonists, reaches not for On the Heavens, nor to the Physics, but to the Metaphysics, specifically the latter half of book ambda. t is as an object of desire and—an important lexandrian innovation—imitation that the pure actuality of the irst over can provide a measure of order, actuality, and perfection to all things. ccordingly, it is in Metaphysics Λ that ristotle in lexander’s and many subsequent Peripatetics’ view establishes the way the principles of the universe are reflected on every level of reality, and in Λ  that the underlying unity of everything is put forward. Tellingly, lexander’s use of the term ʿālam (rabic for kosmos), which is largely absent from the first two-thirds of the treatise On the Principles of the All, really picks up in a climactic finale that essentially recasts Λ  in a conspicuously theological mould. There is providence aplenty in the ristotelian universe, albeit that the good world order is eternal and everlasting in its universal aspects: or the make-up of this universe [s.anʿa hādhā l-kull] and the natural bounty which the creator [al-khāliq] put into it, and the mutual agreement, harmony, and conformity of its parts with one another according to their relation with the whole, evince such an order and harmony that,  This of course applies only in so far as one writes solely as an ristotelian philosopher: the enduring popularity of the De aeternitate mundi literature in the iddle ges testifies to how outside concerns might still lead ostensible Peripatetics to adopt the extra-ristotelian practice of treating the world as if it were a unified object of od’s actions. ut then this is part of my point: from the point of view of how ristotle constructs his philosophy, such questions would never really enter the picture, and the concept of kosmos would accordingly also be redundant, or nearly so.  Mabādīʾ, §§  ff. see the lossary, s.v. ʿ -L-M. The only other instances of ʿālam occur where lexander defends the natural indestructibility of the present world order, on the lines of what we find in On the Heavens . – and On Philosophy: see Mabādīʾ, §§ , –. The equivalence is so close that  take it lexander is consciously following ristotle and his essentially polemical use of the term kosmos in these passages.  enequand ad loc. remarks that this theological designator may derive from De mundo, where genetōr is used at b and a or it may represent a conflation in the rabic translation process with the Platonic Demiurge or again it may be a monotheist incursion. t any rate, given lexander’s insistence on the irst ause only being a final cause (and certainly not a eoplatonic aition poiētikon), the locution sticks out somewhat. Created on 7 December 2013 at 21.30 hours page 340  Taneli Kukkonen if you should assume one of them to be abolished by hypothesis, none of the remaining things could possibly remain in its state. (Mabādīʾ, § , trans. enequand) omparing the tone here with leomedes above is sufficient,  think, to establish that the Principles was written in an explicit effort to provide a Peripatetic counterpart to the Peri (tou) kosmou literary genre established by the Stoics. (c) lexander has a fairly sophisticated account of how the various celestial rotations recognized by Λ  play into all this. Essentially, the purpose of the celestial mechanics is to put some distance between the irst over and sublunary occurrences through postulating a series of mediating mechanisms that allow for differentiation in the make-up of the sublunary domain. lexander famously holds that the ultimate expression of divine providence lies in the perpetuation of the sublunary species, which happens through sublunary motions being regulated by the uniform celestial rotations. t is in this connection also that On the Heavens is allowed to make a minor contribution, since ristotle argues in De caelo . —in a fashion that is explicitly flagged up as uncertain and tentative—for the view that some variation in the celestial motions is needed for there to be an interchange of sublunary elements. This, indeed, is one of the few places where the extant fragments of lexander’s lost commentary speak of the kosmos in a setting where the corresponding vocabulary is missing in ristotle (see Simpl. In De caelo . –). Simplicius commends lexander for his manliness in admitting the need to argue for a degree of divine governance and ordering in this instance, instead of being content merely with natural or material necessity, as was presumably usually the case.  az.h.ar-ta bi-bāli-ka: this is an unusual translation choice for hupotithenai— wahama bi-l-fard. and cognate expressions were much more common—but enequand’s English is undoubtedly correct none the less. n Simplicius’ differences with lexander when it comes to reasoning per impossibile in this fashion see Simplicius’ testimony, In De caelo . –.  further on the topic, see T. Kukkonen, ‘lternatives to lternatives: pproaches to ristotle’s rguments per impossibile’, Vivarium,  (), –.  See . . odnár, ‘lexander of phrodisias on elestial otions’, Phronesis,  (), – S. azzo and . Wiesner, ‘lexander of phrodisias in the Kindī-ircle and in al-Kindī’s osmology’, Arabic Sciences and Philosophy,  (), –.  Mabādīʾ, §  see further the texts collected in lexander of phrodisias, La provvidenza. Questioni sulla provvidenza, ed. and trans. S. azzo and . Zonta (ilan, ).  andrizomenos . . . kata tina theian dioikēsin te kai diataxin apologizesthai, In De Created on 7 December 2013 at 21.30 hours page 341 On Aristotle’s World  ueller in his translation of the passage takes Simplicius’ tone to be sarcastic, but if  am right, then this is a notably gentle form of sarcasm: it amounts to a compliment that may be slightly backhanded, but is none the less genuine enough. The fact that, as far as  can tell, both Simplicius and lexander misread ristotle’s original argument in the same way only serves to underscore how both are essentially in agreement here. (d) or all this, we may take account of how limited is the unity achieved by the ristotelian world on lexander’s telling. This is all the more remarkable given how, on the face of it, lexander grants the Stoic party rather more than seems necessary. e calls the single world (ʿālam wāh.id) a single body (jism wāh.id), for instance, at the same time that he underlines how it envelops and encompasses the totality of things (jamīʿ al-ashyāʾ). lexander even describes the divine potency that unites the world in distinctly Stoicizing terms, as a spiritual power spread throughout all its parts (quwwa rūh.āniyya tasrī fī jamīʿ ajzāʾi-hi: Mabādīʾ, § ). ut the first characterization immediately appears much less robust if we understand lexander in this context to use kosmos in the same sense in which ristotle evokes ouranos in De caelo . , b–, as the outermost heaven, as indeed the argument seems to require: the outermost heaven as one perfectly uniform (and uniformly active) body bears the same regulatory function to the rest of the universe as the single ruler bears to the city. The second point, meanwhile, upon closer inspection merely acknowledges that all of physical reality as ristotle sees it is infused with potentiality striving towards actualization. ccording to lexander, it is because the divine power is diffused caelo . – cf. similarly In De caelo . –, commenting on De caelo . , a– (on which see above, n. ).  Simplicius, On Aristotle’s n the eavens . –, trans. . ueller (ondon and thaca, Y, ),  n. .  s  read the passage, De caelo .  makes no reference to providence instead, it treats both the existence of earth at the centre of the universe and the interchange of the elements as brute facts and then proceeds to postulate a variation in the celestial motions as the simplest explanation for these two phenomena. lexander apparently telescoped the change in the elements to the emergence of more complex forms of actuality, pointedly including living beings, in his commentary, thus transforming ristotle’s argument to one that conveys a belief in providence (understood in the lexandrian manner: see Simpl. In De caelo . –): but this is projection on lexander’s part.  enequand translates tasrī as ‘penetrates’: this is possible, but seems to me needlessly Stoicizing, as the rabic does not necessary imply any active agency on the part of the spiritual power evoked, which is why  prefer ‘spread across’ or ‘dis- Created on 7 December 2013 at 21.30 hours page 342  Taneli Kukkonen throughout all the world’s parts that they follow, turn towards, and otherwise imitate the most noble of beings what is more, ‘all things which share in it do so according to the state and position of each one of them in relation to it’ (Mabādīʾ, § ). ll this means is that each thing, or else each natural kind, enjoys a unique relationship to the irst ause, and this because each approaches it in its own way. The point about directionality is in fact crucial to lexander. The irst ause does not reach out to physical things in order to be in contact with them rather, what is common to all things in the world and to the things which are manifestly distinct from one another is to aim at being in contact with this first substance according to what is proper to each one of them in the nature appropriate to it. This is the cause of their duration and permanence, and of their remaining in the place proper to them. (Mabādīʾ, § , trans. enequand) n other words, the things in the world reach out to the irst ause, not the irst ause to them: their desire is for it (or its perfection, or—to be yet more precise—that exact form of perfection which is appropriate to a thing’s nature), not its for them. ut this also means that there simply is no one world such that it would have a direct relationship with the irst ause: rather, the god–world relationship, such as it is (both terms should be used advisedly in an ristotelian context), consists of a series of one-on-one relations between individual beings striving for whatever share they can have of perfection per se, which is actuality as such, and the being which best exemplifies that perfection. To reiterate, then, the irst ause does not persed throughout’. t is the term ‘spiritual’ itself (rūh.ānī) which appears much more irretractably Stoic, as rūh. often stands for pneuma in the raeco-rabic vocabulary. enequand in the introduction to his translation (Mabādīʾ, ) suggests that rūh.ānī is merely an alternative translation of theios, which is certainly possible, given how theios becomes rūh.ānī in the rabic translations of On the Heavens. owever, the analogy is incomplete, since in the translations of De caelo what needed to be suppressed was the notion that the heavenly body would be divine (see . Endress, ‘verroes’ De caelo: bn Rushd’s osmology in his ommentaries on ristotle’s On the Heavens’, Arabic Sciences and Philosophy,  (), – at – likewise Endress, Übersetzungen, –), whereas the translator of the Principles will have had no reason to downplay the divine nature of the providential force reaching all parts of the universe.  The expression tanh.ū nah.w afd.al al-mawjūdāt is remarkably polyvalent: enequand in his translation has both ‘follow’ and ‘turn towards’, but nah.ā also admits of ‘imitate’ as a translation, which certainly fits the context. Created on 7 December 2013 at 21.30 hours page 343 On Aristotle’s World  relate to the world in any way—indeed, there is no world to which it could relate in this manner rather, each of the universe’s entities relate to it, and out of this somehow a world (a kosmos, as an ordered whole whose parts are mutually supportive) emerges. onsequently, whatever bounty, harmony, and mutual agreement one can discern in the orderly arrangement of the universe’s parts, to pick up on lexander’s vocabulary, the explication of these in terms of the congruence of the heavenly motions is not immediately transparent. or one thing, one still needs to explain what makes the various unmoved movers give rise to precisely that set of celestial motions which results in an orderly and beneficial sublunary system. What makes this puzzling is that in the whole cosmic system, all entities without exception look only upwards, as it were, in their quest for individual perfection, and never down (that is, nothing exercises providence intentionally and primarily). The challenge is already laid down in Rose’s purported fragment  of On Philosophy, which appears, if nothing else, at least to adopt quite effectively the open-ended and quizzical tone of many front-line Peripatetics. (This observation should not be regarded as an endorsement of the fragment as authentically ristotelian.) The fragment sets up a series of disjuncts that take Metaph. Λ  as their starting point, but incorporate principles from the Physics as well.  multiplicity of disordered principles would result not in a world (mundus) but in chaos but this cannot be, since plainly things [here on the sublunary plane] do happen according to nature, either always or for the most part, rather than against it therefore even the assumption of multiple principles will implicitly include a presupposition regarding their ordering which in turn presupposes a first principle, either among the ones just mentioned or outside their recognized set. The end result is the recognition, wholly in line with the closing words of Metaphysics Λ, that there should be one ruler: but the issue of what could account for the way its supreme state is transferred onto the sensible plane is left wholly unanswered, as indeed it remains in lexander (and ristotle). f the later Peripatetics, verroes appears to have gone the furthest in attempting to answer this question and to construct a viable ristotelian cosmology in the process. verroes very clearly builds on lexander but even here, a fair amount of work is needed to connect the dots. n brief, it appears that verroes’ answer hinges on the notion that each of the unmoved movers has an incomplete share Created on 7 December 2013 at 21.30 hours page 344  Taneli Kukkonen in the intelligible content of the irst and that this share translates into a kind of kinetic code which is passed on to the celestial spheres, making the comprehensive actuality of the irst ause into the (indirect) cause of everything coming to cohere that results from the celestial motions inspired by the separate intelligences. ut even if we were to accept that in this fashion all the mechanical aspects of the story can successfully be accounted for, one may still recall ëtius’ complaint (. . ) that in ristotle’s view the sublunary domain is well ordered only by accident, not primarily (kata sumbebēkos ou proēgoumenōs), and that consequently the world for ristotle is neither ensouled, nor rational, nor intellective through and through (holon di holōn). To this criticism, which is essentially Platonist in character,  do not think that an ristotelian would have a ready answer. (e) oming back to lexander, it is worth noting, finally, just how deliberately lexander positions his treatise in relation to the foregone Peripatetic tradition. n the one hand, lexander unhesitatingly proclaims that the Principles constitutes a disclosure of his own view (kashf raʾyī) on the other, he insists that everything he puts forward is in accordance with ristotle’s outlook (bi-h.asb raʾy Arist.ūt.ālis: Mabādīʾ, § ). lexander claims to have taken what he took from the ‘divine ristotle’—an eyebrow-raising epithet, to be sure—by way of principle and summary. This way of framing the Principles signals that lexander is well aware of a lacuna on the subject of cosmic order in ristotle’s extant works, one that extends roughly from an exposition of the nature of the irst ause (al-ʿillat al-ūlā) to the effects it has on the sublunary domain. oreover, lexander is careful to preface his exposition with some important epistemic qualifiers. The Principles is an exercise in speculation, in that it means to spell out what he (that is, lexander) believes can be said on these topics in accordance with ristotle’s stated views. (Mabādīʾ, § ) urthermore, even if we assume with oraux, enequand, and others that lexander was acquainted with On the World when he  See T. Kukkonen, ‘verroes and the Teleological rgument’, Religious Studies,  (), – R. . Taylor, ‘verroes on Psychology and the Principles of etaphysics’, Journal of the History of Philosophy,  (), – for the physical side of things, see D. . Twetten, ‘verroes’ Prime over rgument’, in .-. renet (ed.), Averroès et les averroïsmes juif et latin (Turnhout, ), –.  akhadhnā-hū ʿan al-ilāhiyy Arist.ūt.ālis ʿalā .tarīq al-mabdaʾ wa-l-ikhtis.ār: Mabādīʾ, § . Created on 7 December 2013 at 21.30 hours page 345 On Aristotle’s World  crafted On the Principles of the All and that he made use of aspects of that treatise when he fashioned his own ristotelian response to the theological and cosmological challenge issued by the Platonist and Stoic parties—both suggestions which  find plausible—the fact that lexander never once refers to On the World testifies to his at least entertaining doubts when it comes to the authenticity of that treatise. We know from Proclus’ testimony (In Tim. iii. . – ) that concerns about De mundo’s provenance were raised early on, and we may assume the same conclusion negatively from the scant use made of the treatise in antiquity (as compared to its popularity in the rabic and atin iddle ges). When aap ansfeld drily remarks regarding On the World that ‘lexander and his late eoplatonist fellow commentators, with Philoponus and David as exceptions of very minor importance, apparently preferred to say as little as possible about it, or even nothing at all’, this seems about right. . onclusion n this article  have made the case that the concept of kosmos as it was understood in reek philosophy from Plato onwards ill fits ristotle’s principal philosophical commitments and that it accordingly occupies only a marginal place in his writings. ut how much does this matter, ultimately ertainly  do not expect contemporary scholars to cease talking about ristotle’s cosmology in a casual way, nor is it my purpose to denigrate the efforts of later ristotelians in coming up with a more satisfactory presentation (from their point of view) of the ristotelian kosmos as a whole or to rank them according to some presumed scale of purity or orthodoxy. ll  ased on lexander’s Quaestio . , Sharples accepts oraux’s earlier hypothesis that lexander regarded De mundo as genuine: see lexander of phrodisias, Quaestiones . –. , trans. R. W. Sharples (ondon and thaca, Y, ),  n.  but see the reference to Kupreeva in n.  above.  ansfeld, ‘Peri kosmou’, . The very minor exceptions mentioned by ansfeld are Philop. Aet.  ( and  Rabe) and David, In Cat. . – usse, both of which refer to On the World in incidental fashion: neither author postulates an overarching ristotelian cosmology on the basis of De mundo.   mean here something like Tamar Rudavsky’s characterization of cosmology as ‘that enterprise which describes what the universe looks like’ (reported by Y. T. angermann in ‘rabic osmology’, Early Science and Medicine,  (), – at ). Created on 7 December 2013 at 21.30 hours page 346  Taneli Kukkonen  have wanted to demonstrate is that from a certain point of time onwards—perhaps as early as pseudo-ristotle’s and lexander’s encounters with the Stoics, certainly by the heyday of the schools of thens and lexandria—the more robustly theological and ‘cosmic’ portrayals of the unity of physical reality that were put forward by the Platonic and Stoic schools began to make it seem as though there was something missing from the ristotelian picture. The efforts of various Peripatetics to meet this challenge are reflected in part in their increased appeals to the concept of kosmos, just as the shortcomings of the source materials are reflected in their struggles to incorporate the notion into a framework that scarcely has a need for it. There can be no doubt that the term kosmos possesses a significance in Plato, and especially in the Timaeus, that far outstrips anything found in ristotle. The way in which the sensible universe is a product, something generated by technē, coupled with the uniformity of what is aimed at in the act of creation, accounts for the ‘Demiurge’s monomania’, to borrow ames ennox’s phrase: that is, it accounts for the way in which the Timaeus portrays the world as a single creation. or Plato even the forms constitute a kosmos (Rep.  ), which means that they must be studied as members of a unified whole. ristotle’s use of kosmos, by comparison, is essentially dialectical and opportunist. e is willing to evoke the concept, but he does so mainly in contexts where this can serve as a useful reminder of the conceptual commitments to which earlier thinkers are beholden, as in a reference to those who hold spontaneity to be the cause of both our heaven and ‘all the worlds’. n such a context, an appeal to the kosmos can help to point out how the presupposition of a beneficent ordering pervades the thinking of even those thinkers supposedly willing to entertain that there is no ordering on the macro (astronomical) level at all. n intriguing citation in Philoponus from lexander’s lost commentary on De caelo (Philop. Aet. , . –.  Rabe) should be read in the same light,  think. ccording to  See . Vlastos, Plato’s Universe (xford, ). See . . ennox, ‘Plato’s Unnatural Teleology’, in Aristotle’s Philosophy of Biology, – at .  or a history of the expression see D. T. Runia, ‘ rief istory of the Term kosmos noétos from Plato to Plotinus’, in . . leary (ed.), Traditions of Platonism: Essays in Honour of John Dillon (ldershot, ), –.  tines hoi kai touranou toude kai tōn kosmōn pantōn aitiōntai to automaton, Phys. . , a–.  Created on 7 December 2013 at 21.30 hours page 347 On Aristotle’s World  Philoponus’ testimony, lexander frames ristotle’s investigation of the world (peri tou kosmou), specifically its imperishable and ungenerated nature, by saying that ristotle begins from the views of his predecessors (tas doxas tas tōn pro hautou), so that it is the Platonic and Presocratic usage of kosmos that is allowed to guide ristotle’s choice of terminology. Yet another example comes from the Politics, in which ristotle points to the parallel cases of the internally organized activities of city, world, od, and virtuous person in order to establish that it is not always necessary to have an outside partner in order to lead a full and eudaimonic life (Pol. . , b–). There can be no doubt that kosmos is here used as a synonym for to pan: it is in fact the only way for the argument to make sense (the world does not look outside itself for something to relate to, just because there is nothing outside). Still, apart from the simple act of setting side by side individual, societal, universal, and divine activity, there is no suggestion that these different-scale orderings would enjoy any sort of causal relation, on the lines of those portrayed in the Republic and the Timaeus. ll in all, one would have to say that ristotle appears reluctant even to evoke the term kosmos, let alone to endow it with any real systematic import. Even so, concepts are one thing, while conceptions are another. What are we to say about ristotle’s cosmology at the end of the day, understood now in the broader sense  believe that Simplicius has it fundamentally right. f by cosmology we mean simply an accounting of all the kinds of things there are in the physical world, and perhaps their positioning relative to one another, then this is to be found in all of ristotle’s works on nature put together. nd if by such physical things are meant primarily the simplest kinds of bodies into which embodied entities may break down, the units whose relative positioning moreover gives us an approximate layout of the physical universe, then we have arrived at what is essentially quinas’ take on On the Heavens. ut if by cosmology is meant  ompare how icino in the preface to his synopsis of the Timaeus says that the justification for ascending from the natural to the divine in this Pythagorizing manner is found in the way that all natural things are, after all, effects and images of divine things. t is because of this that the Timaeus treats of the world triply, as it were: the divine, celestial, and human aspects are all considered as the nature of the universe is related to the higher causes (Opera, ii. –).  The Index Aristotelicus lists only four instances of kosmopoiein and cognate expressions, all of which are related to reports of Presocratic cosmogonies.  n quinas see further . . Weisheipl, Nature and Motion in the Middle Ages (Washington, ), –. Created on 7 December 2013 at 21.30 hours page 348  Taneli Kukkonen the more demanding enterprise of showing how all things in the world—or all things equally—rely on higher causes, how certain fundamental forces govern all of embodied reality, and what grants the world a unity such that it becomes possible to speak of it as ‘the world’, then it is hard to see what in the ristotelian corpus would satisfy this requirement. Metaphysics Λ is probably the closest ristotle comes to providing this kind of unified theory of everything: but the account is sketchy in the extreme and requires considerable extrapolation on the basis of other materials in order to function as a full-blown cosmology.  am, then, basically in sympathy with elen ang’s assessment that Metaphysics Λ aims at showing how a discussion of substance will at the same time provide the best account of ‘the ll’, as shown by Metaph. Λ , a– but  do not share in her optimistic view that the treatise operates with ‘remarkable efficiency’, nor would  say that it ‘may be offered as a perfect example of an investigation of substance’. The strenuous efforts of the later commentators (including a borderline hostile Theophrastus) testify to the seams that show and the lacunae that remain in ristotle’s expressed views, and to the unavoidable need for creative extrapolation. t is in these creative acts that the true history of ristotelian cosmology is told. University of Otago       RP Y lexander of phrodisias, La provvidenza. Questioni sulla provvidenza, ed. and trans. S. azzo and . Zonta (ilan, ). On the Cosmos [Mabādīʾ], ed. . enequand (eiden, ). Quaestiones . –. , trans. R. W. Sharples (ondon and thaca, Y, ).  See . S. ang, ‘The Structure and Subject of Metaphysics ambda’, Phronesis,  (), – at .  s for the comprehensive sidelining of On the Heavens in this development, telling,  think, is the way in which even a contemporary thinker such as Remi rague chooses to tackle the task of answering ‘the question of the world in ristotle’. ar from adducing evidence from De caelo, rague instead builds a complex case on the intersection of metaphysics (interpreted as theology), noetics, and ethics. f course, this may reflect a eideggerian preoccupation on rague’s part more than it does anything in ristotle. ut it is none the less remarkable how very little On the Heavens contributes to the overall picture. Even the materials we would normally call ‘cosmological’ in the loose sense of the word relate to the proofs for the Prime over and the actus purus of the Physics and the Metaphysics, respectively. See R. rague, Aristote et la question du monde (Paris, ), chs. –. Created on 7 December 2013 at 21.30 hours page 349 On Aristotle’s World  llan, D. . (ed.), Aristotelis De coelo libri quattuor (xford, ). ‘n the anuscripts of the De caelo of ristotle’, Classical Quarterly,  (), –. altussen, ., Philosophy and Exegesis in Simplicius: The Methodology of a Commentator [Philosophy and Exegesis] (ondon, ). altzly, D. (trans.), Proclus: Commentary on Plato’s Timaeus, iii. Book , Part : Proclus on the World’s Body (ambridge, ). odéüs, R., Aristotle and the Theology of the Living Immortals (lbany, Y, ). odnár, . ., ‘lexander of phrodisias on elestial otions’, Phronesis,  (), –. onitz, ., Index Aristotelicus, nd edn. (erlin, ). os, . P., Cosmic and Meta-Cosmic Theology in Aristotle’s Lost Dialogues (eiden, ). On the Elements: Aristotle’s Early Cosmology (ssen, ). rague, R., Aristote et la question du monde (Paris, ). uridan, ean, Quaestiones super libris quattuor De caelo et mundo, ed. E. . oody (ambridge, ass., ). olmo, . ., Breaking with Athens: Alfarabi as Founder (anham, d., ). Druart, T.-., ‘l-ārābī, Emanationism, and etaphysics’, in P. orewedge (ed.), Neoplatonism and Islamic Thought (lbany, Y, ), –. Effe, ., Studien zur Kosmologie und Theologie der Aristotelischen Schrift ‘Über die Philosophie’ (unich, ). Elders, ., Aristotle’s Cosmology: A Commentary on the De caelo (ssen, ). Ellis, ., ‘The porematic haracter of Theophrastus’ Metaphysics’, in ortenbaugh and Sharples (eds.), Theophrastean Studies, –. Endress, ., ‘verroes’ De caelo: bn Rushd’s osmology in his ommentaries on ristotle’s On the Heavens’, Arabic Sciences and Philosophy,  (), –. Die arabischen Übersetzungen von Aristoteles’ Schrift De caelo [Übersetzungen] (diss. Ph.D., rankfurt a.., ). alcon, ., Aristotelianism in the First Century : Xenarchus of Seleucia (ambridge, ). al-ārābī, bū as.r, Al-siyāsat al-madaniyyat al-mulaqqab bi-mabādīʾ almawjūdāt, ed. . ajjār, nd edn. (eirut, ). Falsafat Arist.ūt.ālis, ed. . ahdi (eirut, ). azzo, S., and Wiesner, ., ‘lexander of phrodisias in the Kindī-ircle and in al-Kindī’s osmology’, Arabic Sciences and Philosophy,  (), –. Created on 7 December 2013 at 21.30 hours page 350  Taneli Kukkonen icino, arsilio, Opera (asel, ), ed. P. . Kristeller,  vols. (Turin, ). inkelberg, ., ‘The istory of the reek Kosmos’, Harvard Studies in Classical Philology,  (), –. ortenbaugh, W. W., and Sharples, R. W. (eds.), Theophrastean Studies on Natural Science, Physics and Metaphysics, Ethics, Religion and Rhetoric [Theophrastean Studies] (ew runswick, ). raham, D. W., Explaining the Cosmos (Princeton, ). (trans. and ed.), The Texts of Early Greek Philosophy: The Complete Fragments and Selected Testimonies of the Major Presocratics,  vols. (ambridge, ). uthrie, W. K. ., Aristotle: n the eavens (ambridge, ass., ). bn al-adīm, Kitāb al-fihrist, ed. . lügel,  vols. (eipzig, ). bn Rushd, bū l-Walīd, Averrois Commentaria magna in Aristotelem: De celo et mundo [Commentarium magnum], ed. . . armody,  vols. (euven, ). Commentary on Aristotle’s Book on the Heaven and the Universe. Sharh. kitāb al-Samāʾ wa-l-ʿālam, facsimile of the manuscript produced by . Endress (rankfurt a.., ). Tafsīr mā baʿd al-t.abīʿa [Tafsīr], ed. . ouyges,  vols. (eirut, –). Talkhīs. al-samāʾ wa-l-ʿālam, ed. . al-Dīn al-ʿlawī (ez, ). bn Sīnā, bū ʿlī, Kitāb al-h.udūd, in Tisʿ rasāʾil (onstantinople, ). ohansen, T. K., ‘rom Plato’s Timaeus to ristotle’s De caelo: The ase of the issing World-Soul’, in . . owen and . Wildberg (eds.), New Perspectives on Aristotle’s De caelo (eiden, ), –. ohnson, . R., Aristotle on Teleology (xford, ). Kerchensteiner, ., Kosmos: Quellenkritische Untersuchungen zu den Vorsokratikern (unich, ). Kranz, W., ‘Kosmos’, Archiv für Begriffsgeschichte,  (), – and –. Kukkonen, T., ‘lternatives to lternatives: pproaches to ristotle’s rguments per impossibile’, Vivarium,  (), –. ‘verroes and the Teleological rgument’, Religious Studies,  (), –. Kupreeva, ., review of . enequand (ed.), Alexander of Aphrodisias: n the osmos, in Ancient Philosophy, . (), –. ang, . S., ‘The Structure and Subject of Metaphysics ambda’, Phronesis,  (), –. angermann, Y. T., ‘rabic osmology’, Early Science and Medicine,  (), –. ennox, . ., Aristotle’s Philosophy of Biology (ambridge, ). Created on 7 December 2013 at 21.30 hours page 351 On Aristotle’s World  ‘Plato’s Unnatural Teleology’, in Aristotle’s Philosophy of Biology, –. ‘Theophrastus on the imits of Teleology’, in W. W. ortenbaugh, P. . uby, and . . ong (eds.), Theophrastus of Eresus: On his Life and Work (ew runswick, ), – repr. in ennox, Aristotle’s Philosophy of Biology, –. orimer, W. . (ed.), Aristotelis De mundo (Paris, ). ovejoy, . ., The Great Chain of Being (ambridge, ass., ). ahdi, ., ‘lfarabi against Philoponus’, Journal of Near Eastern Studies,  (), –. ‘The rabic Text of lfarabi’s Against John the Grammarian’, in S. . anna (ed.), Medieval and Early Modern Studies in Honor of Aziz Suryal Atiya (eiden, ), –. ansfeld, ., ‘Peri kosmou:  ote on the istory of a Title’ [‘Peri kosmou’], Vigiliae Christianae,  (), –. Prolegomena: Questions to be Settled before the Study of an Author or a Text (eiden, ). and Runia, D. T., Aëtiana: The Method and Intellectual Context of a Doxographer, ii/ (eiden, ). artijn, ., Proclus on Nature (eiden, ). ‘Theology, aturally: Proclus on Science of ature as Theology and the ristotelian Principle of Metabasis’, in . Perkams and R.-. Piccione (eds.), Proklos: Methode, Seelenlehre, Metaphysik (eiden, ), –. atthen, ., ‘The olistic Presuppositions of ristotle’s osmology’ [‘olistic Presuppositions’], Oxford Studies in Ancient Philosophy,  (), –. and ankinson, R. ., ‘ristotle’s Universe: ts atter and orm’, Synthese,  (), –. oraux, P., Aristote: Du ciel (Paris, ). ’eara, D. ., Pythagoras Revived (xford, ). Rescigno, ., Alessandro di Afrodisia: Commentario al De caelo di Aristotele. Frammenti del primo libro (msterdam, ). Runia, D. T., ‘ rief istory of the Term kosmos noétos from Plato to Plotinus’, in . . leary (ed.), Traditions of Platonism: Essays in Honour of John Dillon (ldershot, ), –. Philo of Alexandria and the Timaeus of Plato (eiden, ). Scade, P., ‘Stoic osmological imits and their Platonic ackground’, in V. arte, . . cabe, R. W. Sharples, and . Sheppard (eds.), Aristotle and the Stoics Reading Plato (ondon, ), –. Sedley, D., ‘s ristotle’s Teleology nthropocentric’, Phronesis,  (), –. Created on 7 December 2013 at 21.30 hours page 352  Taneli Kukkonen Sharples, R. W., ‘lexander of phrodisias and the End of ristotelian Theology’, in T. Kobusch and . Erler (eds.), Metaphysik und Religion: Zur Signatur des spätantiken Denkens (eipzig, ), –. ‘ristotelian Theology after ristotle’, in D. rede and . aks (eds.), Traditions of Theology: Studies in Hellenistic Theology, its Background and Aftermath (eiden, ), –. Peripatetic Philosophy  to  : An Introduction and Collection of Sources in Translation (ambridge, ). and Sheppard, . (eds.), Ancient Approaches to Plato’s Timaeus [Ancient Approaches] (ondon, ). Simplicius, On Aristotle’s n the eavens . –, trans. R. . ankinson (ondon and thaca, Y, ). On Aristotle’s n the eavens . –, trans. . ueller (ondon and thaca, Y, ). Siorvanes, ., ‘Perceptions of the Timaeus: Thematization and Truth in the Exegetical Tradition’, in Sharples and Sheppard (eds.), Ancient Approaches, –. Solmsen, ., Aristotle’s System of the Physical World (thaca, Y, ). Sorabji, R. (ed.), Aristotle Transformed: The Ancient Commentators and their Influence (ondon, ). ‘s Theophrastus a Significant Philosopher’, in . . van phuijsen and . van Raalte (eds.), Theophrastus: Reappraising the Sources (ew runswick and ondon, ), –. Matter, Space and Motion: Theories in Antiquity and their Sequel (ondon, ). The Philosophy of the Commentators – ,  vols. (thaca, Y, ). Steel, ., ‘Why Should We Prefer Plato’s Timaeus to ristotle’s Physics Proclus’ ritique of ristotle’s ausal Explanation of the Physical World’, in Sharples and Sheppard (eds.), Ancient Approaches, –. Taylor, R. ., ‘verroes on Psychology and the Principles of etaphysics’, Journal of the History of Philosophy,  (), –. Todd, R. ., ‘The Title of leomedes’ Treatise’, Philologus,  (), –. Twetten, D. ., ‘verroes’ Prime over rgument’, in .-. renet (ed.), Averroès et les averroïsmes juif et latin (Turnhout, ), –. van Raalte, ., ‘The dea of the osmos as an rganic Whole in Theophrastus’ Metaphysics’, in ortenbaugh and Sharples (eds.), Theophrastean Studies, –. Vlastos, ., Plato’s Universe (xford, ). Weisheipl, . ., Nature and Motion in the Middle Ages (Washington, ). Westerink, . ., Anonymous Prolegomena to Platonic Philosophy, X, nd edn. (Dilton arsh, ). Wright, . R., Cosmology in Antiquity (ondon, ).