The Dissoi :ogoi and early greek skepticism

Chapter 10 The Dissoi Logoi and Early Greek Skepticism Thomas M. Robinson Dissoi Logoi (“Contrasting Arguments”) is the name of a little-known anonymous document that is found appended to the works of Sextus Empiricus in the manuscripts containing his writings. It was first published in Paris in 1570, and since that date there has been dispute, among its handful of readers, concerning its date, authorship, purpose, and overall philosophical commitments.’ That it was perceived by some, at any rate, of its earliest readers as generally “skeptical” in tone seems clear enough from the fact that it is invariably found in conjunction with the works of Sextus Empiricus, but the precise nature of its skepticism is less clear. From various pieces of internal evidence it seems to have been written long before the time of Sextus, in fact in the first or second decade after the end of the Peloponnesian War (404 BCE), and most people now believe it is the work of a Sophist teacher of the day.’ For purposes of this chapter I shall assume the general correctness of both of these beliefs and confine my discussion to the nature and purpose of the work’s contents. I begin with a brief outline. The 4000-word treatise, which ends abruptly, and looks unfinished, consists of nine short chapters. The structure of the first four is strictly antilogical. Chapter 1, for example, entitled “On Good and Bad,” begins: On the matter of what is good and bad contrasting arguments are put forward in Greece by educated people: some say that what is good and what is bad are two different things, others that they are the same thing, and that the same thing is good for some but bad for others, or at one time good and at another time bad for the same person. 1. In Stephanus, Diogenes Laertius Opera 1 (Paris 1570, 470-482). For the most recent edition of the work see T.M. Robinson (1979), from which the various translations in this chapter are taken. 2. See especially 1.8, “And in the matter of war (I shall speak first of the most recent events) the Spartan victory over the Athenians and their allies was good for the Spartans, but bad for the Athenians and their allies...” 3. For the various views on authorship that have been put forward over the years, see T.M. Robinson (1979, 41-54). A notable recent exception to the currently accepted view is that of Conley (1985), which upholds the possibility of anonymous Byzantine authorship.

The author then goes on to offer a series of reasons why good and bad might reasonably be thought to be one and the same, and then immediately afterward a series of reasons why they might reasonably be thought to be different. An example of the first runs: “Illness is bad for the sick but good for the doctors. And death is bad for those who die, but good for the undertakers and gravediggers.” Examples of the second have a uniformly hypothetical cast to them, each one amounting to an assertion of the (presumably self-evident) contradiction it would entail if the identity thesis in question were true; for example, “Being sick is bad for the sick and also good for them if what is good and what is bad are the same thing.” Chapter 2, “Seemly and Shameful,” follows a similar pattern to chapter 1, as does chapter 3, on “Just and unjust.” As in chapter 1, the author has an easy time making out a case for the identity thesis, but in maintaining the difference thesis once again does little more than simply assert the (self-evident) contradiction it would entail were the identity thesis in question true. Chapter 4, likewise, “Truth and Falsehood,” offers a list of uncontroversial examples in support of the identity thesis, though this time, in his defense of the difference thesis, he bolsters the familiar charge of self-contradiction on the part of those maintaining the identity thesis by the interesting deployment of what looks like an early version of the Liar Paradox (4.6). It is also said that the false statement is different from the true statement; as the name differs, so likewise does the reality. For if anyone were to ask those who say that the same statement is false and true which of the two their own statement is, if the reply were “false,” it is clear that a true statement and a false statement are two different things, but if he were to reply “true,” then this statement is also false. Chapter 5 is untitled, and the familiar phrase dissoi logoi is missing from it, but the basic structure of the treatise as seen so far is maintained: a conglomerate identity-thesis is propounded, and then a refutation of it offered. The thesis in question consists of two parts: 1. People possessed of contrary properties may in fact act and speak identically (various common speech acts and activities of the sane and the demented, the wise and the ignorant are offered as examples; e.g., “The demented, the sane, the wise and the ignorant both say and do the same things. First of all they call things by the same name: “earth,” “man,” “horse,” “fire,” and everything else. And they do the same things: they sit, eat, drink, lie down, and so on, in the same way”). 2. The same person/thing is himself/itself possessed of contrary properties “The talent,” he says, for example, “is heavier than the mina and lighter than two talents; the same thing then is both lighter and heavier.” And “what ex- The Dissoi Logoi and Early Greek Skepticism 189 ists (is the case) here [sc. in Greece] does not exist (is not the case) in Libya; nor does what exists (is what is the case) in Libya exist (turn out to be the case) in Cyprus... Consequently, things both exist (are the case) and do not exist (are not the case).” As far as the difference thesis is concerned, this is for the first time in the treatise defended with a set of arguments far surpassing the simple assertion of apparent self-contradiction on the part of identity-theorists, the first of them interestingly distinguishing identical speech acts by the varying appropriateness of the occasion of utterance (5.9-10), the final one (5.15) drawing a careful distinction, two generations before Aristotle, between existence simpliciter and existence secundum quid. In the author’s own words: As for the affirmation that the same man exists and does not exist I ask, “Does he exist in some particular respect or in every respect?” Thus, if anyone denies that the man in question exists, he is making the mistake of asserting “in every respect.” The conclusion is that all these things exist in some way. Chapter 6, “On Whether Wisdom and Moral Excellence are Teachable,” preserves the structure of the first five, setting out first the one case (in this instance the case that wisdom and moral excellence are in fact unteachable) and then the counter-case with arguments remarkably similar to a number of those found in Plato’s Protagoras (328c, 324dff) and Meno (93b-94e). At 6.3-4, for example, we read: Another (proof that wisdom and moral excellence are unteachable) is that, had wisdom and moral excellence been able to be taught, there would have existed recognized teachers of them—the way there have been recognized teachers of the arts. A (further) proof is that those men in Greece who became wise would have taught this wisdom to their children and their friends. Chapters 7, 8, and 9 are untitled, and the structure of case and counter-case has now apparently started to break down. The first of the three chapters propounds the single logos that it is absurd that offices should be—as many were in Athens—assigned by lot; the essence of the counter-case consists of a single statement at the end asseverating the value of electing qualified people for such offices. The second chapter consists of a little essay on the characteristics of the paradigmatic sophist/orator/politician, and defends among other things the remarkable view that such individuals are by definition both omniscient and omnicompetent. The treatise ends with a short discussion of the value of memorization and with what looks like the beginning of a list of a few specific mnemonic techniques.

JA A Opinions on the author’s own stance in the treatise have varied. For Diels and Kranz‘ he is simply “talentlos,” and his views-—if he has any—unworthy of further investigation. For others,” he is simply a paradigmatic Protagorean, and as such a common-or-garden relativist, convinced that on any topic contrary arguments can be proffered, and offering examples of such arguments in his treatise. For others he is Protagorean in a further, tougher sense, not only proffering contrary arguments on various topics but attempting while so doing to make the weaker of each pair the stronger. And the matter is further complicated by items here and there in the treatise that look to be possibly Gorgian or Hippian or even in a couple of places possibly Socratic? in influence, suggesting to some that the author is more of an eclectic than a proponent of any specific philosophical Weltanschauung. That the author is, in general terms, Protagorean in the weaker sense seems undeniable. He clearly does believe that on various topics contrary arguments can be proffered, as he spends the greater part of his time offering examples illustrating the proposition’s truth. Whether relativism on his part can be inferred from this is however debatable, both in his own case and in that of Protagoras. As is well known, Plato argued in the Theaetetus that relativism was the logical conclusion of Protagoras’s “man the measure” doctrine. As is less well known, and as Levi has pointed out,’ there was no suggestion on Plato’s part that Protagoras himself ever drew any such conclusion. What we do know is that, according to Diogenes Laeftius,® Protagoras was the first to say that there were contrary arguments on every topic, and also “the first to institute contests between arguments” (protos . . . logon agonas epoiesato). One natural inference from this is that the “contest” Protagoras saw as basic to philosophical disputes was one between arguments, not people; any further inference is hazardous. Not long after Protagoras’s floruit, however, we find Aristophanes— followed later by Aristotle—attributing to him the view that he set out to “make the weaker argument the stronger.”” But there is no independent evidence to support such a claim, which has all the appearances of popular prejudice based on (legitimate) irritation at the verbal game-playing—apparently for its own sake— practiced by sophists, such as Plato’s Dionysodorus and Euthydemus, claiming Protagoras as their inspiration. Protagoras himself, by contrast, when . Diels and Kranz (1951) ad loc. . For details, see Robinson (1979), 55ff. . Robinson (1979), 59ff. . Levi (1940), 292-306. . 9.50ff. = Protagoras Al, B6a DK6. . Aristophanes, Clouds 882-884; Aristotle, Rhet. 1402a 23 = Protagoras B6b DK6. The Dissoi Logoi and Early Greek Skepticism 191 placed on the stage by Plato, comes across, not as a relativist, but as a man adhering to the straightforward and uncomplicated beliefs of his day. If this picture of Protagoras is the correct one, all that the author of the Dissoi Logoi, if he saw himself as in general terms a follower of Protagoras, should be credited with is the view that on a large number of topics contrasting arguments can be and, in terms of sound philosophy, should be proffered; the quality of the individual arguments however remains open for discussion and assessment. Before looking at those arguments, we should first examine briefly two general claims about them that have been thought to offer evidence of the author’s own philosophical stance. 1. The author makes it clear that, in the opening chapters at least, he backs the difference thesis.!° In addition to his own statements on the matter, we have the facts that: a) the identity thesis is invariably placed first in sequence, and b) the difference thesis is in chapters 1-4 the one spelled out at greater length. 2. The difference thesis invariably attacks a version of the identity thesis wholly different from the one that has just been defended." Let us look at these claims and their supposed implications. 1. (The author invariably backs the difference thesis.) In fact, however, the author on a number of occasions expresses clear support for the identity thesis (1.2, 2.20, 3.7, 4.2), and this evidence must be weighed in the balance along with the evidence (1.11, 2.26, 5.6, 6.7, 6.13) that suggests a commitment to the identity thesis. One possible conclusion from this is that the author has little idea of what he is doing, and is indeed as Diels suggested “talentlos.” There are, however, other possible conclusions, which I shall discuss later. a) (The identity thesis is invariably placed first in sequence.) It has been argued from this that it is consequently the difference thesis and its supporting arguments that are invariably left ticking over in the reader’s mind, and that this is a suasion in support of the view that the author himself supports such a difference thesis. There may be something in this, but against it one should point out that the identity thesis invariably propounds views, the factual accuracy of which is open to relatively simple testing and observation. This being the case, and something readily ascertainable by even the most unsophisticated reader, the author may well, it seems to me, have intended the reader to notice that the difference thesis 10. For proponents of the view, see Robinson (1979), 73. 11. Robinson (1979), 76.

possesses whatever strength it does thanks to a witting or unwitting misrepresentation of the identity thesis, and the point will be made with particular clarity if the identity thesis is placed first in sequence. b) (The difference thesis is in all four of the first four chapters spelled out at greater length.) This is as it happens simply false; it is in fact the identity thesis that in those chapters is the longer, as a glance at the text makes clear. Even if it were true, however, nothing of any clarity concerning the author’s commitments, pace Untersteiner (1961-1962), can be inferred, except by those who find a causal connection between an argument’s length and its plausibility. Of much greater substance (and possible implication) is the point ([2] above)— glossed over or not noticed by commentators—that the difference thesis invariably attacks a version of the identity thesis different from the one that has just been defended. I say “defended,” rather than “proffered and defended,” because typically the identity thesis will be proffered as an (extreme) identity statement but actually defended as a (very moderate) predicative statement. Among many examples, one might take chapter one of the treatise. The opening lines, as we have seen, read as follows: On the matter of what is good and bad contrasting arguments are put forward in Greece by educated people: some say that what is good and what is bad are two different things, others that they are the same thing, and that the same thing is good for some but bad for others, or at one time good and at another time bad for the same person. What is interesting here is the fact that the identity thesis is first expressed as an identity statement (“[what is good and what is bad] are the same thing”) and then immediately afterward as a predicative statement (“the same thing is good for some but bad for others,” etc.). The thesis is then defended (1.2-10), but only in its predicative version (“death is bad for those who die, but good for the undertakers and the grave-diggers,” etc.). When it comes to the difference thesis however (1.11ff), the author defends it by reference to the extreme version of the identity thesis—i.e., as an identity statement—that had been mentioned but in no way defended. He has as a result a very easy time pointing out that such a thesis is self-contradictory. It is indeed, and would no doubt have been admitted to be such by the proponents of the predicative version of the thesis who had just had their arguments aired. But he has of course knocked down a straw man, and we are no further forward as to his views on the value of the arguments that uphold a moderate—and perfectly defensible—version of the identity thesis, as against the value of arguments in defense of the extreme version. The Dissoi Logoi and Early Greek Skepticism 193 What we have instead is’a constant reiteration that he favors the difference thesis (i.e., that x and y are, as he puts it, as different in reality as they are in name [1.11]) but with no hard arguments in defense of that thesis other than the claim that its opposite leads to self-contradiction, and—as we have seen—with the occasional asseveration that he does in fact support various identity-theses. Two major interpretations of all this seem possible. The first is that of Diels, to wit, that our author is simply devoid of talent. Though Diels does not spell the matter out, presumably he means either that the author has failed to notice that the version of the identity thesis he attacks is never the version of it the arguments for which he has just propounded; or, more seriously, that he could not in fact recognize the difference between identity statements and predicative statements in the first place. The result in either case is a lamentable failure to realize that the supposed difference thesis he supports is wholly compatible with if not in fact the very same as the moderate “identity thesis he has just in fact outlined. The second interpretation leaves open the possibility that the author is fully aware of the difference between an identity statement and a predicative statement, but for propaedeutic reasons not unsimilar, perhaps, to those underlying Plato’s aporetic dialogues, constructs arguments predicated upon their supposed sameness. If this is the case, the treatise will consist of more than rote-lists of arguments and counterarguments for the use of sophists. It will in fact be a genuine teaching manual for sophists, with arguments displayed in such a way as to challenge readers (or hearers) to hone their rhetorical skills by working out for themselves the relative worth of such arguments. I am inclined to accept this latter interpretation, on the grounds that it fits the facts as well as the first and has at the same time the merit of crediting our author with a modicum of philosophical intelligence. It does not, however, tell us anything specific about the author’s own stance on particular issues, and more generally whether those who, in appending the treatise to the works of Sextus Empiricus, clearly thought he was reasonably classifiable as a skeptic, even if only in some fairly moderate sense of the term. As far as content is concerned, if the interpretation outlined immediately above is correct, it might be thought that we may be forever precluded from knowing the author’s own philosophical stance, for that stance was irrelevant to his objective in writing the treatise. There is, however, a case for saying that his position emerges regardless. As we saw earlier, while for the most part strongly backing the difference thesis he also on occasion makes affirmative noises concerning the identity thesis. This would on the face of it constitute evidence that he is putting up the best case he can for both sides, and thereby blocking any rash attempt on the reader’s part to waste time attempting to unmask his (the author’s) own private opinions. But this could be mistaken. The affirmative noises are all, as it turns

out, in favor of the straightforward, easily verifiable predicative version of the identity thesis in question. And in view of the fact that this version is, on the face of, comfortably compatible with the difference thesis that is its supposed contrary, it seems eminently possible that the author’s “own” stance will in fact coincide with this thesis. If this is true, our Sophist-author will turn out to be—not surprisingly— an empiricist, and one firmly convinced of the worth and importance of the wealth of empirical evidence he uses in support of the predicative statements of various identity theses throughout the treatise. And the type of skepticism appropriately thereby attributed to him will consist of the skepticism of all those who, before and since, have demanded evidence before drawing general conclusions. N Whether such empiricism and skepticism was an all-embracing feature of his thinking is a conclusion less easy to draw. At 1.17, for example, the author disclaims any intention of telling us “what the good is” (ti esti to agathon), a phrase that would not have been surprising on the lips of Socrates. And at 4.5 he writes: “It is clear, then, that the same statement is false when the false is present to it, and true when the true is present to it.” The doctrine of “presence” of universal to particular is, as is well known, thought by many to be characteristic ofthe Platonic/Socratic Theory of Forms and indeed a specifically Platonic/Socratic contribution to philosophical speculation. This may well be the case and our author— given the presumptive date of publication of the treatise—may be one of possibly many influenced by Socratic argumentation. On the other hand, it seems equally possible that such an argument was “in the air” in sophistic circles at this time, and actually drawn by Socrates from such circles. In view of the lack of evidence either way, the matter must be left in tantalizing doubt. However that may be, it seems beyond question that the author is strongly affirming in this instance at any rate a position more appropriately characterized as essentialist than empiricist. In combination with the empiricism that appears to characterize so much else of the treatise it constitutes a view of the real and of language used to describe the real that Socrates (if not Plato) would almost certainly have found respectable, and for that matter Aristotle too. Two further pieces of evidence, already noticed in passing, to support such a view should be mentioned. 1. In the chapter on truth and falsehood we read the following statement, as a defense of the difference thesis begins (4.6): It is also said that the false statement is different from the true statement; as the name differs, so does the reality. For if anyone were to ask those who say that the same statement is false and true which of the two their own statement is, if the reply were “false,” it is clear that The Dissoi Logoi and Early Greek Skepticism 195 a true statement and a false statement are two different things, but if he were to reply “true” then this statement is also false. The argument is a shrewd one, and incorporates elements of the so-called Liar Paradox. In its fully articulated form this was first set out a generation later by Eubulides, a disciple of Euclides of Megara (who was himself a contemporary of the author of the Dissoi Logoi). Perhaps for present purposes more significantly, an analogous argument is, with some effect, set out by Socrates in Plato’s Theaetetus (170eff), as part of his criticism of Protagoras’s “man the measure” doctrine. 2. At the end of chapter 5 the author concludes: As for the affirmation that the same man exists and does not exist I ask: “Does he exist in some particular respect or in every respect?” Thus, if anyone denies that the man in question exists, he is making the mistake of asserting “in every respect.” The conclusion is that all these things exist in some way. The simpliciter/secundum quid distinction employed in this argument is remarkable to come across two generations before the appearance of Aristotle’s Sophistical Refutations, and suggests an expertise on the author’s part far greater than is usually attributed to him. We also know, from Aristotle himself, that little manuals of sophistic arguments were in circulation in Athens and elsewhere in his day. That opens up the intriguing possibility 1) that one such manual was the Dissoi Logoi, 2) that it was actually read by Socrates/Plato and Aristotle, and 3) either that its use of features of the Liar Paradox influenced Socrates/Plato or that its use of the simpliciter/secundum quid distinction influenced Aristotle, or both. As in the earlier case of the origin of the doctrine of “presence,” no evidence is unfortunately available either way, and the matter must once again be left in tantalizing doubt. To return to the question at hand: Can the author of the Dissoi Logoi be meaningfully described as a skeptic? The answer it seems to me must be Yes, on two counts, though hardly in the technical sense of the term that Sextus would himself have recognized. 1. As far as the content of his beliefs is concerned, he is committed to the adducing of empirical evidence in areas where such evidence is either actually or potentially available. In the area of true and false propositions, by contrast, he seems more essentialist in his analysis. The total package might reasonably be described as one of moderate empiricism and moderate skepticism, in that empirical evidence, when adduced, will serve to dispel the septicism that fueled the investigation, but not all problems it seems are deemed to be amenable to the adducing of such empirical evidence.

2. His methodology, by contrast— and here he draws directly on Protagoras — is uncompromisingly empirical, in that no investigation is deemed to have been properly conducted unless both sides of the question—and there always are two, at any rate in the matter of the particular questions he raises— have been aired. Whether the above arguments alone would account for the work’s invariably being found in Sextus Empiricus manuscripts is certainly open to question, given the broader and more searching range of Sextus’s own brand of skepticism. But three further features of its style and contents may have served as suasions in favor of such a collocation. 1. If the author was understood (wrongly, as I have argued, but not surprisingly, given what we know of the way in which Protagoras’s views on logos/ counterlogos came to be interpreted in antiquity) as himself adhering to the view that in his various chapters logos and counter-logos were equipollent, a compiler could be forgiven for seeing an affinity between this view and the doctrine of equipollence that so characterizes the writings of Sextus. 2. A more specific suasion is found at 5.13-14, where, after mentioning instances of changes in meaning brought about, not by the subtraction of anything but by a mere change of accent, he goes on: Since, therefore, there is such a difference when nothing is taken away, what if in that case somebody does either add something or take something away? I shall give an example of the sort of thing I mean. If someone were to take away one from ten <or to add one to ten, Diels>, there would no longer be ten or even one, and so on in the same way in all other instances. As I have put it elsewhere (Commentary, ad loc.): The argument turns on a particular use of the word “ten.” The form can be used collectively (ta deka—cf. words like “score,” “dozen,” etc., in English) or distributively (deka), leaving open the possibility for the author to argue (as he apparently does here) that to (supposedly) take away from “the ten” (ta deka) that I have is to leave me without my ten, and so apparently without even one of them! (In the same way, one might argue, one cannot take one away from a “rugby fifteen” and get a “rugby one” plus a “rugby fourteen;” one either has a rugby fifteen or one has not.) Similar addition-subtraction puzzles crop up in Plato (Crat. 432a-b) and Aristotle (SE 22.178a29-37),* but are as it happens a particular feature of the 12. For discussion of the passage, see Sprague (1990), 20. The Dissoi Logoi and Early Greek Skepticism 197 writings of Sextus; see Pyrr. Hyp. 3. 85, Adv. Math. iv. 23-24, Adv. Phys. 1. 303330, 2. 308-309. 3. A final possible specific suasion is to be found at 2. 20, where the author is arguing that the difference between seemly and shameful actions is that the former are performed “at the right moment” (kairoi), the latter “at the wrong moment” (en akairiai). An analogous—though far from identical— argument is expounded by Sextus at Adv. Eth. 64-67 and credited to the Stoic Ariston of Chios; and, possibly tellingly, an analogy is drawn from the different placing of letters in the spelling of names (cf. Dissoi Logoi 5.12 for the same point), which is said to depend upon the force of hoi kairoi (cf. ibid. 2.20, kairoi). Such suasions, small though they may be as individual items, could well, it seems to me, have been deemed to have sufficient collective force to per‘ suade a hesitating compiler to place the Dissoi Logoi alongside the works of Sextus. Whether we were meant to infer from such a collocation that the work should be attributed to Sextus himself or simply to his philosophical entourage we cannot tell, and in any case the compiler was wrong. Where he was less obviously wrong was in his appreciation of the fact that there are features of its style and content which, as I hope I have shown, the master might well have interpreted as significant gropings towards the light.