Bryn Mawr Classical Review 2001.07.22
David Sedley (ed.), Oxford Studies in Ancient Philosophy, Volume XVII. Oxford: Oxford University Press, 1999. Pp. 364. ISBN 0-19-825019-3. $60.00.
Reviewed by Michael Pakaluk, Philosophy, Clark University, Worcester, MA 01610 (firstname.lastname@example.org)
Word count: 3037 words
In this volume, the series Oxford Studies in Ancient Philosophy, under the editorship of David Sedley, once again brings together serious and well-argued papers representing the best work in ancient philosophy. The concentration of thought and scholarship here is so great that a brief review can hardly do justice to the volume; it is as if one needs to review four or five books rather than one. I shall therefore simply state the main line of thought of each contribution and, where this seems appropriate, offer a brief criticism. I hope, however, to say enough to make clear the significance of each.
Pieter S. Hasper, in "The Foundations of Presocratic Atomism", offers a new account of the argument for atomism that Aristotle gives at GC I.2, 316b21-34. The argument there ostensibly develops a reductio, based on the supposition that there are no atoms: 1. suppose there are no atoms, i.e. that nothing is indivisible; 2. then everything is everywhere divisible; 3. but from a possible state of affairs nothing absurd follows; 4. suppose, then, that a thing is in fact divided everywhere; 5. then what results would be nothing; 6. yet this would be absurd (since something would consist of nothing; 7. moreover, since there are cycles of generation and destruction, and things come to be from that into which they perish, then things would come to be from nothing); 8. hence it is not the case that everything is everywhere divisible; 9. thus some things are not everywhere divisible; 10. thus, there are indivisible constituents of things, viz. atoms.
Aristotle replies, correctly, that the argument is fallacious because it makes a mistake when it supposes that the presumed possibility be realized: what we ought to suppose, is not that a thing is in fact divided everywhere, but rather that it is divided somewhere, since to say that it is everywhere divisible is to say that it is possible, at any point, that a single division be made there.
Hasper however thinks the argument is sound because, he claims, it depends upon what he calls the Atomistic Principle (AP): 'No unity from a plurality, no plurality from a unity'. The upshot of AP is that there are no genuine potentialities as regards separation or combination: that which is divided is always and necessarily divided; that which is integral is always and necessarily such. Given AP, then, the claim that a thing is divisible is simply the claim that its already existing division can be made manifest. (Note, however, the problem in claiming that divisions here can be made manifest, sc. the apparent unity has the real potential to appear to be a plurality.) Hence, 'everything is everywhere divisible' amounts to 'everything is everywhere divided', which is absurd; thus not everything is everywhere divisible (i.e. divided)--and there are atoms.
There are two problems with Hasper's interpretation. The first is that AP is so strong a principle that it is indistinguishable from atomism; it cannot, without circular argument, be treated as a premise in an argument for atomism. Thus it is incorrect to speak of it as providing the argumentative 'foundation' for atomism, as Hasper does. Second, Aristotle does not attribute this argument to the atomists. Hasper claims that "AP is necessary to turn the argument of GC I.2 into a valid one" (12), yet he gives no textual grounds for taking the argument there to be relying upon it, nor can this be done, since nothing like AP appears in the argument Aristotle provides. In short, Hasper gives a rational reconstruction of a reported argument for atomism, which has the twofold disability of being circular and without textual foundation.
In "Platonic Pessimism and Moral Education", Dominic Scott first observes that Plato, in his characterization of Polus, Callicles, and Thrasymachus, is very much concerned with the problem of what makes some persons resistent to being persuaded by moral arguments that they are nonetheless unable to refute. Scott then contends that the psychology of the Republic is intended in part to explain this resistence. However, Plato's explanations there, at least as Scott construes them, turn out not to be especially enlightening: we can become resistent to moral argument either (a) from hearing the wrong sorts of stories in childhood (392 A13-B4), or (b) by being enticed by pleasures into adopting, without rational support, principles that would justify our pursuit of those pleasures (558C8ff.). Scott does not take up the reasonable question of whether such things are indeed adequate to explain the behavior of Plato's own characters. And it would seem that they are not: the truculence of a Thrasymachus, or Polus' concern with how things appear and what people generally think, seem to be unaccountable on (a) and (b) alone.
Paolo Crivelli in "Aristotle on the Truth of Utterances" argues that (i) among linguistic items, sentence tokens ('utterances'), not types, have truth-values, according to Aristotle, and that (ii) the truth value of an utterance can change. Crivelli develops his points meticulously and with great clarity, so that we can easily assess their strength. He gives four arguments for (i), none of which, however, is compelling. I shall mention only the first, which is philosophical rather than textual and interpretative. Crivelli maintains that utterances in relation to types are analogous to individuals in relation to species and genera, and that, just as individuals are said by Aristotle to be primary (Cat. 5), so utterances are more appropriately regarded as the primary bearers of truth or falsity. Yet, as against this, the disanalogies between utterances and individuals are equally great, so that it is unclear which should be regarded as decisive: for instance, an utterance is always an utterance of or by someone, that is, it necessarily belongs to someone, and in this respect it seems more like an accident than a substance. Yet if we want to say that an utterance has a truth-value as expressing the belief of an individual, then it is unclear whether, among linguistic items, the utterance or the sentence type is better regarded as giving the content of the belief.
As regards (ii), Crivelli thinks that the plausibility of the claim that a discrete utterance can change in truth-value is made clear by cases such as the following. In 1992 Jim shouts loudly "Labour represents the majority of the British nation"; in 1997, Jim looks back and says, correctly, "That loud remark was false then but it is true now." Crivelli argues that, since what is referred to here is described as 'loud', it must be an utterance, not a sentence type--yet it has changed in truth-value. However, one might maintain, in response, that Jim in 1997 might more naturally have referred back to his earlier remark by saying "What I said loudly was false but is true now", using only an adverb, which would suggest rather that it is the content of the remark that is true or false.
Justin Broackes' wide-ranging "Aristotle, Objectivity, and Perception" maintains that Aristotle tended to regard color as having an objective basis in inanimate nature, independent of its being perceived and prior to its relationship to a perceiver--in fact, Broackes holds that Aristotle was a reductionist as regards color (103). The piece may be understood as a reply to an essay by Sarah Broadie,1 which argues in contrast that, for Aristotle, colors are as basic as any features of nature and should be understood as eternally built into nature precisely to guide living things in their perception. Broackes would trace Aristotle's reductionist tendencies to his conviction that relatives must rest upon absolutes (89-91) and also his attachment to the principle of the 'Priority of Inanimates'(91-96), viz. that the activity of animate beings must admit of explanation in terms of principles that suffice to explain inanimate things. (The label comes from Broadie, who in fact denies that Aristotle accepts the principle.) Yet that colors, as powers to produce perceptions in living things, and therefore as relative existents, must rest upon absolutes, implies only that colors must at least exist in something (as Broadie would of course also admit: colors are qualities), not that colors must be reducible to something else. Also, it is hard to see how the Priority of Inanimates is consistent with the Aristotelian view that the first cause is living and that every other thing, in its own way, tries to imitate this living being.
In "The Unity of the Virtues in Aristotle", Edward Halper tries to defend Aristotle's acceptance of the Unity of the Virtues (UV), viz. that a person cannot truly have one virtue without having them all, by drawing a distinction between 'proper' and 'psychic' virtues. A proper virtue is what is displayed in the actions that Aristotle identifies as paradigmatic of that virtue, e.g. proper courage is displayed in our finding the mean relative to feelings of confidence and fear on the field of battle; proper justice is displayed in a judge's distributing awards in conformity to merit as required by the law; etc. A psychic virtue consists rather in the subordination of a passion to the rational part of the soul. (It is defined by reference to the soul only, not by reference to actions, hence the term, 'psychic'). Thus, psychic courage consists simply in someone's feeling confidence and fear only to a reasonable degree, whether he is on the field of battle or not; psychic justice is found whenever any judgement involving fairness is not deflected by greed, whether the person involved is an appointed judge or not; and so on. Halper's view is that UV does not hold for proper virtues (since it is implausible to hold that these paradigmatic activities must all be bound up with one another) but that it does hold for psychic virtues, since, Halper thinks, any virtuous action involves the exercise of every psychic virtue.
The distinction between proper and psychic virtue seems wrongly drawn and would appear to be incapable of doing the work Halper wishes it to do. Briefly, Halper's argument requires that 'proper virtue' be a species or kind of virtue, but Aristotle in fact uses only the adverb, 'properly' (κυρίως)2, and it is a mistake to reason from claims of the sort, "An F is properly G", to the conclusion that there is a certain species of F, consisting of those Fs that are G. e.g. from "A goalkeeper is properly agile" it does not follow that there are 'proper goalkeepers' who are distinguished from other kinds of goalkeeper by their agility. It is better to take Aristotle's use of κυρίως to signal simply a central- or focal-case analysis; but UV seems no more plausible for non-central than for central cases of a virtue.
Halper thinks that UV is indeed more plausible for psychic virtues because he thinks that Aristotle believes that:
(A') "the exercise of any one virtue requires the appropriate exercise of every passion" (119). If the 'exercise' of every passion enters into any virtuous action, but the passions are kept in line by the various psychic virtues, then all of the psychic virtues are involved in any action. Halper's reason for accepting (A') is that any wayward impulse whatsoever is capable of keeping us from doing what we should; thus all wayward impulses must be ruled out in any virtuous act. But it is a mistake to reason that, if some impulse felt at time t would have deflected P from doing his duty, then P at time t was displaying the virtue that appropriately governs impulses of that sort. If at this moment I were deathly afraid that a meteor was going to burst through my roof and strike me at my desk, I would get up and flee, and thus I would fail to complete this review; it does not follow, on that account, that in completing this review I am in fact displaying 'psychic' or any other sort of courage.
Victor Caston's "Something and Nothing: The Stoics on Concepts and Universals" artfully combines logical analysis and careful scholarship to give a reconstruction of the initial Stoic response to Plato's theory of Forms in the competing theory of concepts (ἐννοήματα). Caston maintains that only the earliest Stoics, Zeno and Cleanthes, held a theory of concepts, and that this theory was quickly rejected, under the pressure of Chrysippus' logical criticisms, and replaced by the theory of expressibles (λεκτά). Caston views early Stoic talk of concepts as a primitive recognition of the realm of intentional being, hence the confusion in the sources as to whether concepts are something or nothing at all.
There seems to be a lack of congruence, however, between Caston's description of the theory of concepts and his appraisal of it. Caston argues, no doubt rightly, that Zeno and Cleanthes intended not to identify Platonic Forms with concepts but rather to make any appeal to Forms unnecessary: theirs was an eliminativist program. However, as Caston points out, the Forms were meant to do two things: first, they were meant to underwrite our practice of predicating the same term of multiple subjects (the term refers to a single thing, the Form, in which all of those subjects participate); second, they were meant to account for the objective existence of natural kinds (each member of any species is what it is, and is the same as any other member of that species, by participating in the Form). Yet Caston praises the Stoic theory of concepts on the grounds that "...what the Stoics have shown [sic] is that what is identical in each case need not be what makes a thing the sort of thing it is, and this is a genuine conceptual advance" (213). That is, concepts accomplish the first task without accomplishing the second. But how, then, would concepts make an appeal to Forms unnecessary? One suspects metaphysical regression here, not a conceptual advance: but at least more needs to be said.
Stephen Menn's complementary "The Stoic Theory of Categories" aims to explain the origin of the four Stoic highest kinds: ὑποκείμενον, ποιόν, πως ἔχον, πρός τί πως ἔχον. Briefly, Menn's story is that the first two categories were appropriated from Plato (Timaeus 49a6-50c) by Zeno and Cleanthes, who shared with Plato and Aristotle the view that, when a thing is appropriately called F, it is so on account of something in that thing, which causes it to be F, say, its F-ness. But this simple view of the basis of predication cannot be sustained when one holds, with the Stoic, that all causes are corporeal: for instance, Socrates is taller than Theatetus, until Theatetus grows, at which point "Socrates is taller" becomes false, yet evidently not on account of any corporeal difference in Socrates. Thus, the original two categories must be expanded, to include now predications that are true, not on account of something in the subject of predication which makes them so. As the example suggests, this expansion might have been motivated originally by difficulties in relative predications: so Menn believes, and he gives evidence that it was Aristo who first introduced a third category, πρός τί πως ἔχον, with Chrysippus later adding πως ἔχον, to cover cases of non-relative predication in which there appear to be no simple corporeal basis for the predication.
Menn thinks the development was occasioned by a dispute in ethics over the unity of virtues. Aristo wanted to hold that the names for the various virtues signify simply the same thing in different circumstances. That is, each virtue-name indicates something πρός τί πως ἔχον. Menn speculates that Chrysippus at first attacks this view in his That the Virtues are ποιά, insisting that a different expertise (or branch of 'physics') is needed for each virtue. But he later comes to see that his own position threatens the unity of the virtues, so he maintains, rather, that the different virtue-names indicate different dispositions of the ἡγεμενικόν in a sage. They indicate, then, something πως ἔχον.
This is a fascinating story, and one would like it to be true, but given that it has no direct textual support, Menn overstates his case when he judges that "...this was probably how the Stoics first came to formulate the third category" (236).
I shall deal briefly with the last four contributions to the volume.
Spyridon Rangos in "Proclus on Poetic Mimesis, Symbolisms, and Truth" explains Proclus' resolution of the quarrel between philosophy and poetry: Homer is not inferior to Plato, since poetry can imitate the Ideas directly (the origins of allegoresis); and even philosophy is imitative, since all symbolic activity is imitative.
Christopher J. Martin in "Non-Reductive Arguments from Impossible Hypotheses in Boethius and Philoponus" explores arguments in the named philosophers in which reasoning begins from something presumed to be impossible, yet where the aim is not to derive a contradiction (showing the original assumption false), e.g., Boethius assumes that God does not exist, which he regards as impossible, in order to establish some results about the character of the goodness of creatures. This sort of reasoning ought not to be informative if "anything follows from a contradiction", as in standard contemporary formal logic. Martin begins to explore the conditions for such arguments' being significant.
The volume ends with two critical notices of books. Edward Hussey gives a good introduction to recent scholarly work on the so-called Derveni Papyrus in a review of André Laks and Glenn W. Most (eds), Studies on the Derveni Papyrus. Finally, Charles Kahn looks at The Routledge History of Philosophy, volume I, edited by C.C.W. Taylor, which covers Greek philosophy through Plato, and recommends the volume as the basic secondary text in an introductory course.
I should note the remarkable coherence of the volume, for a collection of this sort. Taken together, the contributions trace the history of ancient philosophy, from the atomists to the Greek commentators, and several pieces are interrelated in interesting ways, e.g. the unity of the virtues is an important theme in Halper and Menn; principles of logic in Caston and Martin; Stoic predication in Caston and Menn. Moreover, several essays (Crivelli, Halper, Caston, Menn) give very fine examples, though differing in style and method, of how one may make use of rational reconstruction, in doing history of philosophy, to arrive at genuine philosophical insight.
1. Sarah Broadie, "Aristotle's Perceptual Realism", Southern Journal of Philosophy, 31, suppl. (1993), 137-59.
2. Aristotle in fact uses κυρίως only in his discussion of courage. Halper somewhat misleadingly writes (126-7) as if κυρίως functions as a kind of technical term and is deployed consistently throughout Aristotle's discussion of the various virtues.