The third problem matters most: that of keeping a balance between a black-letterlegalistic approach and one which focusses on the objectives of the disputing partiesand on Isaeus’ forensic techniques. Here I have qualms, for while C.G. o¶ers much onthe values of the oikos, the crucial steps in this long-drawn-out inheritance dispute,though of course taken within a framework of law and procedure, all comprisedbehavioural choices which have to be understood socially, not just explained legally.This applies even to the initial adoption in 412/11, which only makes sense as anextra-legal bargain (thus rightly Wyse, Isaeus, p. 414): the process of epidikasia, ofwhich C.G. makes much, was simply a legal μgleaf. Likewise the adoptee’s attemptto split the opposition by buying o¶ its main spokesman (§ 13), rightly termed‘extragiudiziale’ (p. 168), does indeed follow a judicial act, a conviction for falsewitness (§ 12), but one best viewed as part of a sequence of trials of strength andpublic opinion, not within the legal requirement to convict more than half thewitnesses. Likewise, even more forcibly, with the out-of-court fudge narrated sodisingenuously in §§ 14–19, where Isaeus’ prime task is to camou·age the fact that ‘theall-important clause was not in the document’ (Wyse, Isaeus, p. 404, citing § 25).Explicit recognition that ‘Athenian law courts were a public stage on which privateenmities were played out’ (R. Osborne, JHS 105 [1985], 53) would have helped totemper C.G.’s over-legalistic approach in this otherwise thorough and praiseworthyedition.

University of Liverpool J.K. DAVIESj.k.davies@liverpool.ac.uk

PLATONIC COMPOSITION

Harte (V.) Plato on Parts and Wholes: the Metaphysics of Structure.Pp. x + 311. Oxford: Clarendon Press, 2002. Cased, £48. ISBN:978-0-19-823675-7.doi:10.1017/S0009840X06003015

Under what circumstances do many things compose one thing? How can a singlewhole arise from its many parts in such a way that it is a uniμed individual and not amere collection? According to H., there are two strands of thinking to trace in Plato’sattempts to answer such questions. The μrst problematises the idea that compositionjust is the identity of a whole with its parts. The second develops and defends a modelof composition according to which wholes are structures, and the parts of suchwholes exist only as parts, only in so far as they belong to the structures theycompose. The problematic account of composition is on display primarily in theTheaetetus and Parmenides. The alternative model is developed in the Sophist,Philebus and Timaeus. H. does not defend a developmental thesis. The problematicmodel is not endorsed by Plato at some points only to be rejected later; it is there tobe problematised all along.

At the heart of Plato’s ruminations lies the question of whether or not compositionis ontologically committing. According to the rejected model, composition isinnocent. The one just is the many. But the model borders on incoherence. How canmany things be one thing? The Platonic alternative requires that composition yields orcreates a one from many. A whole just is a complete, single entity created from manyparts. For Plato, composition requires ontological commitment. Of course, not just

the classical review 33

The Classical Review vol. 57 no. 1 © The Classical Association 2007; all rights reserved

any relation among many things is su¸cient to compose a single whole. Compositionis restricted. There are wholes only when certain types of structural, individuatingrelations obtain between distinct parts and between those parts and the whole theycompose.

The picture H. sets out is attractive and well-defended. The argument to show thatPlato problematises the model of composition as identity is detailed and convincing.The unveiling of the metaphysics of structure via careful attention to the di¸cultPhilebus and the slightly more accessible Timaeus is impressive. Much as I admire H.’se¶orts, however, I believe that an important phase in Plato’s thinking aboutcomposition goes unrecognised in her exegesis. There is a third alternative to whichPlato sometimes seems attracted. On her view, Platonic composition always proceedsfrom the top down: ‘wholes come μrst; and parts only thereafter’ (p. 277). But there isevidence that composition sometimes occurs when many things, which are identiμableprior to their composing a whole, enter into certain kinds of relations to produce anew unity. Sometimes composition is from the bottom up.

The key dialogue is the Sophist. H. interprets Plato’s examples of the mixing andweaving of letters, kinds and syntactic elements to illustrate – though not obviously toadvocate – a metaphysics of structure according to which parts are identiμed only invirtue of the wholes they compose. But the metaphysics of structure so conceivedseems less at home in the Sophist than in the Philebus or Timaeus (wheremathematical blending principles are explicitly employed). Consider the example ofcomposition from letters. According to the example, vowels run through the otherletters like a bond; without vowels it is impossible for the other letters to μt together(Soph. 253a5). H. resists the idea, defended by Moravcsik and others, that vowels actas glue-like elements, special parts of the wholes they help to compose. According toH., vowels are necessary for the combination of letters, but not su¸cient.

But surely Plato is looking to isolate the cause of combination, not simply anecessary condition, when he suggests that the expert (the one who can provide acausal explanatory account of letters and their combinations) is the one who willidentify the vowels and distinguish them from the other letters. If letters can beidentiμed independently of the items they compose, and if vowels are glue-likeelements acting as unifying principles, then the vowel example probably modelscomposition that occurs from the bottom up. Composition, in this case, would treatsyllables and words (and ultimately kinds) as wholes composed of parts in the way ahouse might be a whole composed of bricks and mortar.

In a second example from the Sophist, a logos is formed when a name and a verbare woven together. Composition occurs when names (signs for things performingactions) and verbs (signs for actions performed) μt together. The statement ‘manlearns’ is a complex unity. H. interprets this example as an account of structuredsyntactic space. This syntactic space has name-shaped slots and verb-shaped slots,and those slots are deμned by the functions their occupants must carry out. Syntacticspace is structure abstractly conceived. Well-formed sentences like ‘man learns’ arestructured wholes.

Though it is appealing to view this example as highlighting structure, it is less clearthat it is an example of top-down composition rather than bottom-up composition.Names and verbs seem readily identiμable as such before they become parts ofsentences. Indeed, the examples of failed attempts to produce a logos (e.g. ‘man lionhorse’) depend on the possibility of identifying names or verbs independently of thecomplex unities they can, in successful cases, come to compose. The composition of alogos seems better modelled on the assembling of a jigsaw puzzle. Only certain

34 the classical review

attempts to μt various pieces together will yield the whole puzzle. Still, the pieces canbe identiμed independently of their roles as parts. On this model of composition,elements can be sorted by means of intrinsic di¶erences (e.g. shape) which thenground di¶erent but complementary combinatorial powers. In the syntactic case,names and verbs would be sorted by their capacity to act as di¶erent kinds of signs.The di¶erence in signifying capacity would then ground di¶erent, but complementarysyntactic capacities.

Though Plato experiments with the bottom-up models outlined, I do not mean tosuggest that he strictly embraces either one. H. is right, I think, that Plato’s favouredmetaphysics of composition is ultimately to be found in the operations of the unifyingmathematical proportions of the Philebus and Timaeus. Fortunately, the challengingand rewarding insights into composition to be found in those dialogues are keenlyexplicated and masterfully explored in this excellent book.

Dartmouth College CHRISTINE J. THOMASchristine.j.thomas@dartmouth.edu

THE PHAEDO AND TIMAEUS

Karfik (F.) Die Beseelung des Kosmos. Untersuchungen zurKosmologie, Seelenlehre und Theologie in Platons Phaidon undTimaios. (Beiträge zur Altertumskunde 199.) Pp. 293. Munich andLeipzig: K.G. Saur, 2004. Cased, €89. ISBN: 978-3-598-77811-7.doi:10.1017/S0009840X06003027

This book is concerned with Plato’s cosmology, and explores Plato’s conceptionof the relation between soul, world and god. In seeking to show that thereexists an intimate connection between the Platonic psychology, cosmology andtheology, K. selects the dialogues Phaedo and Timaeus. At μrst sight, this wouldsuggest: (1) that the book’s contents are as rich as its systematic scope is vast, and (2)that the Phaedo and the Timaeus will play more or less equal roles. But a closer lookat K.’s work reveals his title and preface to be somewhat misleading in these tworespects. It should also be noted that the book is not a monograph, but a collectionof essays.

The book is divided into four main sections, the μrst two on the Phaedo, the lasttwo on the Timaeus. These sections take the form of long individual papers. Versionsof the two studies on the Phaedo were previously published, and the μrst study on theTimaeus is an interpretation of one – extraordinarily di¸cult – line of Greek text, Ti.41a7. They are followed by a bibliography, an index locorum and an index nominum.The fourth and last section, on the doctrine of motion in the Timaeus, has threeappendices, making it by far the longest (it makes up more than a third of the entirebook). In contrast, the two studies of the Phaedo take up only a little over a quarter ofthe space allotted to all four studies. Hence, one might quibble with the fact that thecentral theme, the theory of motion, is not adequately re·ected in the book’s title.Moreover, the theme of motion and the other topics treated by K., such as aspects ofthe Timean theology, are certainly of wide-ranging importance, but more limited thanthe announced scope. Lastly, this book on Plato’s cosmology is for the most part

the classical review 35

The Classical Review vol. 57 no. 1 © The Classical Association 2007; all rights reserved