BMCR 2001.11.16

Pyrrho, his Antecedents, and his Legacy

, Pyrrho, his antecedents, and his legacy. Oxford: Oxford University Press, 2000. x, 264 pages ; 24 cm. ISBN 0198250657 $60.00.

The testimonies for early Pyrrhonism are few and often obscure so it is to be expected that modern interpretations differ and conflict with each other. Commentators disagree as to what Pyrrho’s idea was, what Aenesidemus wanted to say, what the relationship was between these two, the relationship between them and Sextus Empiricus, and how the sceptical Academy fits into the picture. Richard Bett has been eminent in the debate concerning Pyrrhonism, with a number of articles and a commentary on Sextus’ Against the Ethicists.1 He now offers a comprehensive interpretation of the fortunes of early Pyrrhonism, a summation of much that he has written and more. The book is immensely valuable to all students of ancient thought since Bett addresses most of the key issues of early Pyrrhonism in an orderly, learned and intelligently argued manner. It is informative and thought-provoking from start to finish, and his answers to the many problems of Pyrrhonism are always illuminating, although not such as to preclude dissent, as he acknowledges.

The book is divided into four parts. In the first part Bett lays down his interpretation of the central idea of Pyrrho, which he bases on a close reading of the single most important testimony for his views, that of Aristocles (preserved in Eusebius’ Praeparatio Evangelica 14.18.1-5). The basic claim is that Pyrrho was far from advocating a sceptical stance as understood by Sextus Empiricus, involving suspension of judgement in the face of undecidable disagreement, but rather that he maintained that reality is inherently indeterminate and hence that no statement about it is true or false. Bett labels this idea the ‘indeterminacy thesis’. The ideas advanced in this part are controversial and have already aroused rather fierce criticism.2 The second part, concerned with the practical side of Pyrrho’s philosophy, his views on how one should lead one’s life and attain tranquillity, is less contentious, although Bett uses the indeterminacy thesis as his interpretative basis. The third part continues the discussion of the indeterminacy thesis in that Bett offers an account of its motivation, both philosophical and historical. Having considered the philosophical background, most importantly Plato’s discussion of relativism and Heracliteanism in the Theaetetus and sensibles in Republic V, Bett finds an overlap between the ideas of Plato and Pyrrho. So, the ideas of Plato displace Aristotle’s discussion of scepticism in Metaphysics IV as the preferred context for Pyrrho’s ideas. He also discusses at length other possible historical antecedents and influences, the Eleatics, Xenophanes, Protagoras, Democritus and others. In the fourth and last part Bett seeks to explain the connection between Pyrrho, Aenesidemus and, to some extent, Sextus. Aenesidemus comes across as a closer kinsman of Pyrrho (of the indeterminacy thesis) than of Sextus. It follows that Aenesidemus’ philosophy, at least as presented in the summary of Photius ( Bibliotheca 169b18-171a4), is quite different from the scepticism of Sextus. Aenesidemus emerges as a relativist, and traces of his relativism may be discerned in Sextus as well as in the account of Diogenes Laertius.

This is a short and selective account of the main ideas presented in the book. They form a network with the indeterminacy thesis at its center; Bett’s interpretation of Pyrrhonism stands or falls with his construal of the indeterminacy thesis. At issue is little more than the meaning of two sentences: “things are in equal measure indifferent [or indifferentiable] and unstable [or unmeasurable] and indeterminate [or undeterminable/undecidable]; for this reason neither our sensations nor our opinions tell the truth or lie”. Hardly a word of the passage has escaped detailed scholarly scrutiny, and no wonder, since it holds the key to understanding Pyrrho’s ideas. As the brackets indicate, the first sentence can be understood in two quite different ways. It can be taken to refer to our inability to know what a thing by nature is, or it can amount to a metaphysical claim according to which things by nature are indifferent etc. regardless of our epistemic abilities. There is no consensus as to which reading to adopt. Bett maintains that the only way to settle the issue is to consider the logic of the passage as a whole, and that demands a ‘metaphysical’ reading, for only then can one make sense of the latter sentence which clearly states that sensations and opinions are neither true nor false, and not that one cannot tell whether they are true or false. All this depends on keeping the explanatory connective ‘for this reason’ ( διὰ τοῦτο) and not changing it into ‘because’ ( διὰ τὸ), as has been suggested and quite often accepted. In effect, Bett presents us with two options. Either we accept the metaphysical reading of the first sentence or we are left with the following reading: ‘we cannot differentiate, measure or decide how things by nature are, and therefore our sensations and opinions are neither true nor false’. The latter sentence seems to be equivalent to ‘therefore things are not as we perceive them or think they are nor yet contrary to that’; we may perceive them as F or not-F, but they are neither by nature F nor not-F. Bett dismisses this reading, apparently as nonsensical (22).

It is true that this reading is strange, and Bett’s reading is the more convincing for it. But early Pyrrhonism is full of wonders, even on Bett’s interpretation. Can we simply dismiss the claim that ‘if it is undecidable that a thing is F (or not-φ then it is not by nature F (or not-φ as non-sensical? Can it not rely on a conception of what it is for a thing to be by nature somehow, namely that ‘if a thing is by nature F, then it is decidable that it is F’? In fact, Bett attributes a similar conception to Sextus (and perhaps ultimately to Aenesidemus), a qualified one, admittedly, and in a context restricted to ethics, in his commentary on an argument for the claim that nothing is by nature good or bad ( AM 11.69-78). Bett labels that conception the ‘recognition requirement’: “in order for something really to be good (or bad) for a certain person, that thing must be recognized as good (or bad) by that person”.3 In Bett’s commentary, that requirement plays an important role in his discussion of early Pyrrhonian arguments in Sextus.

Turning back to the indeterminacy thesis, according to Bett, Pyrrho’s concern is not with the impossibility of knowledge, but rather with how things are. In the third part, Bett suggests that the grounds Pyrrho had for advancing his indeterminacy thesis were “the differing impressions of the same objects” (114), the variability and conflict of appearances. Pyrrho’s reasoning would then go as follows: “things present themselves to us in various and conflicting ways; so there is no single, fixed way things are; so reality is indeterminate” (117). It is probably incontestable that Pyrrho started out by noting the conflict of appearances, as Bett maintains. And this observation is epistemological. Therefore it is somewhat surprising to find Bett stating, apparently quite generally, in a criticism of Long and Sedley,4 that we “do not need to attribute to Pyrrho any dubious slide from epistemological to metaphysical considerations, since we have no reason to assume that his thinking must have started out on an epistemological plane” (116n9). Bett is here against moving from epistemological observations to metaphysical claims, at least in a dubious way. That may also be part of the reason why he dismissed the ‘epistemological’ reading of the passage under consideration. But even on Bett’s reasoning, the slide is there and the starting point is epistemological, although it is not the same slide and the same starting point as others have found. Long and Sedley claim that since there is no reason to prefer one appearance over another, the world is indeterminate, according to Pyrrho. Thus undecidability plays a role. Bett’s account is different.

He attributes to Pyrrho a conception of what it is for a thing to be by nature somehow, the invariability condition: ‘if a things is by nature F, then it is invariably F’. This condition, Bett argues convincingly, is not special to Pyrrho. Armed with it, and the variability of appearances, Pyrrho has no need for undecidability to reach the indeterminacy thesis, variability suffices, Bett continues. But what has this condition got to do with the conflict of appearances? The invariability condition, as stated, makes no mention of appearances, but it is appearances that conflict. The Stoics, if Bett is right, assumed the invariability condition (at least for the concepts of good and bad), but they were not impressed by the conflict or variability of appearances, for this conflict could be explained by pointing out that some appearances were true while others were false; appearances could deceive. It seems that, if the variability of appearances is to hook up with the invariability condition, we would need: ‘if a thing is by nature F, then it invariably appears as F’. In that way, the fact that appearances of the same thing are variable, F and not-F, would get Pyrrho to the claim that the thing is not by nature F or not-F. Thus it would simply be ruled out that variability could be explained away by pointing out that one appearance of an object is true while a conflicting one is false. So, in effect, Pyrrho uses a very strong version of the invariability condition, which consolidates the conflict of appearances and dismisses the possibility that appearances may deceive. If a thing appears sometimes as F and sometimes as not-F, these conflicting features are contingent and not essential; it is ruled out that one of the appearances can be deceptive and that the other appearance can indicate an essential feature.

But Bett in fact seems to reach the same conclusion. He adds to the invariability condition a supplementary assumption the function of which is to replace ‘appears’ with ‘is’ (120-21). This assumption is that apparent features of a thing are themselves not a matter of obscurity; there is no reason to doubt the way things appear. Thus, if x appears as F in some circumstances, then x is F in these circumstances. “Otherwise, there would be room for the speculation that some feature that seemed to us merely temporary and contingent might in fact belong to an object invariably and unqualifiedly, and so be part of that object’s nature” (121). And as Bett says later, “…the slide between ‘presents itself’ and ‘is’ depends on the assumption that the characteristics of a thing in specific circumstances—as opposed to their real natures—are not themselves…a matter for doubt” (216n55). The supplementary assumption is clearly designed to rule out the possibility that the conflict of appearances can be solved by maintaining that one of them is true and the other false.

Pyrrho may indeed have used this assumption, as Bett argues. What I find surprising, however, is Bett’s appraisal of such an assumption, which rejects the possibility that some appearances may be deceptive. In the case of relational and some evaluative features the condition may seem reasonable, and that is how Plato may have employed it, according to Bett (136-37). But if generalised, this is a rather sweeping supposition, and one which most philosophers would reject even if they accepted the invariability condition itself. Bett says: “Pyrrho may mistrust the senses in one way; but he cannot subject ordinary experience to any truly hyperbolic doubt without cutting off the route to his conclusion. However, in not entertaining hyperbolic doubts, he would hardly be doing anything very adventurous or surprising” (121). Some would no doubt find it both adventurous and surprising to maintain that we have no reason to doubt the way things appear, even if we do not demand hyperbolic doubt. Mere conflict and disagreement as to a thing being good or bad would, on this reasoning, suffice to show that it was by nature neither; if we can possibly find two contrary features of one and the same object, then neither feature can belong to the object by nature. This is a remarkable stance.5

Bett ends this important discussion of the indeterminacy thesis by undermining his own opposition to sliding from epistemological observations to metaphysical claims, or so it seems to me: “[…] Sextus does not allow us to make any inferences from the appearances to the nature of things. Pyrrho, according to my reconstruction, is considerably less cautious; from the variable appearances, he infers that reality is indeterminate. But then, given that he does hold the indeterminacy thesis, some such audacious move was perhaps only to be expected” (123). What was a dubious slide is now an audacious move.

Bett’s argument for the claim that Pyrrho did employ some version of the invariability condition is compelling, but I do not thereby feel forced to accept his construal of the indeterminacy thesis, assuming I have understood it correctly. Although an invariability condition is a bridge that allows passage between epistemological observations and metaphysical claims, that does not mean that one has to take the first sentence of the Aristocles passage as a metaphysical claim. But although Bett does not dispel all doubt, the cogency of the presentation casts in relief the important issues facing the student of Pyrrhonism. The Aristocles passage is longer, and Bett discusses it all in detail in the first chapter of the book. He does claim that his version of the indeterminacy thesis is supported by the rest of the passage, but one could say with equal justification that the rest of the passage is interpreted in the light of the thesis. That does not lessen the excellence of the discussion, for example the different uses of the ‘no more’ phrase in Pyrrhonism. His discussion of other testimonies relating to the fundamental ideas of Pyrrho, be they epistemological or metaphysical, is also instructive. He is surely correct in detecting contamination of later Pyrrhonian ideas in accounts of Pyrrho. Here, Bett does much needed work in clarifying individual testimonies.

His interpretation of Pyrrho’s practical philosophy, in the second chapter, will probably be less controversial. In fact, most of the chapter discusses the reliability of testimonies, often strange, in particular the biographical tradition. Although Bett uses the indeterminacy thesis as a point of reference, especially in order to explain Pyrrho’s acquisition of tranquility, one could agree with many of his conclusions but still reject the indeterminacy thesis. This is true for his discussion of the role of appearances in Pyrrho’s scheme of things. Bett argues cautiously that Pyrrho, or Timon, in fact offers appearances as a practical criterion. If so, Pyrrho’s view would be both akin to that of later Pyrrhonians and that of the Academic sceptics, even if these offered a practical criterion as a dialectical ploy. But in this connection, Bett does not discuss the Academic tradition of answering the inactivity argument by offering a practical criterion, which may have influenced later Pyrrhonians just as much as Pyrrho’s view, and maybe even Timon’s exposition of Pyrrho’s view. In fact, as far as I can see, Bett does not often discuss, at least in any detail, possible Academic influences on or connections with the Pyrrhonian tradition, for example with regard to the practical criterion. He does, however, mention that the Academics argued that life was possible for a sceptic (220n62), to differentiate this stance from that describing happiness as the sceptic’s reward. He also points out (200n20) that the Academics used equipollence, without pursuing possible connections. But, then, Bett is not arguing against Academic influences on Aenesidemus, but rather for the influences of Pyrrho.

Bett’s discussion of those notorious elegiac couplets of Timon preserved by Sextus in AM 11.20 that purport to offer Pyrrho’s view on the nature of the divine and the good rigorously refutes attempts to make sense of it. But it is still somewhat disappointing in that they are dismissed as worthless. Some might find Bett’s conclusion a bit severe: “since we are in no way forced to accept that Pyrrho is the one who speaks about ‘the nature of the divine and the good’, we need not trouble ourselves with the impossible task of trying to reconcile this fragment with the outlook expounded in the bulk of this chapter and the previous one” (102). Because we are not forced to regard the mysterious couplets as a testimony we should not do so. Even granting the impossibility of interpreting the couplets consistently with a general outlook, at least Bett’s outlook, I would have expected a more aporetic attitude. Bett uses this same somewhat dogmatic methodology in his treatment of Aenesidemus’ Heracliteanism in chapter four, where he dismisses it as a red herring in an examination of Aenesidemus’ Pyrrhonism. But, then, I have no better suggestion.

Bett’s treatment of the nature and background of the indeterminacy thesis, which he offers in the third part, has already been discussed. Again, the treatment is very useful, since the merits of possible antecedents are meticulously weighed. And again, the indeterminacy thesis is used as a point of reference. In keeping with his views, Bett finds that the closest connection is between Plato and Pyrrho, although he refrains, unlike many in a similar position, from postulating a direct influence. If one accepts Bett’s construal of Pyrrho’s indeterminacy thesis, his discussion of Plato’s ideas adds depth to it, but it is another matter whether his discussion of Plato offers a reason for accepting his construal of the indeterminacy thesis. I very much doubt that it does, and Bett does not claim that it does. The differences are marked, as Bett acknowledges, and Timon is hostile to Plato. Nevertheless, he does point out that Plato adopted the invariability condition (although obviously not across the board like Pyrrho), and thereby offers a historical justification for attributing the condition to Pyrrho.

Bett’s discussion of the connection between Aristotle’s treatment of scepticism in Metaphysics IV and the Aristocles passage carry the usual hallmarks of precision and lucidity. He points out all the problems of regarding Pyrrho as in some way responding to Aristotle, but also as indebted to the philosophers that Aristotle is criticising. What remains striking, however, are the similarities between Aristotle’s discussion and the Aristocles passage, although it is, as Bett shows, extremely difficult to give a coherent account of the similarities; everything depends on how one interprets the Aristocles passage.

In the fourth and last part Bett offers his view on the development of Pyrrhonism from Pyrrho himself down to Sextus. Most of the chapter is devoted to Aenesidemus and based on a close reading of Photius’ testimony. According to Bett, Aenesidemus took up Pyrrho’s observation that appearances conflict, naturally enough, but also his invariability condition, according to which a thing is by nature F if it invariably is F, as well as the supplementary assumption that allows the slide from ‘appears’ to ‘is’. The variability is then depicted in relative assertions: ‘a thing is variably F’ is equivalent with ‘a thing is relatively F’. These locutions indicate that the thing in question is not by nature F. In this way, Aenesidemus has clear links with Pyrrho, but few with Sextus. Then Bett suggests that when Aenesidemus claims that there are no causes or signs or other objects of dogmatic belief, he is claiming only that nothing is by nature a cause etc. but only variably, or relatively. Thus his negative assertions are readily intelligible. What emerges is a philosophy closer to Pyrrho than to Sextus, who would never have asserted anything about a thing’s nature except that he did not know what it was. Still, suspension of judgement can be attributed to Aenesidemus, but a substantially different one from that of Sextus. Bett seeks support for his interpretation in the accounts of Diogenes Laertius and especially Sextus’ Against the Ethicists, which is, he claims, intelligible if one has a background such as this in mind.

Bett’s ideas concerning Aenesidemus, which he has already advanced in his commentary on Against the Ethicists, owe much to Paul Woodruff.6 Their explanatory force is of course great, although perhaps not sufficient to displace the view according to which Aenesidemus bases his version of suspension of judgement on undecidability and equipollence just as Sextus does (even, as I believe, in Against the Ethicists), and as the sceptical Academy did. Bett’s ideas present an intriguing alternative. Aenesidemus’ assertions to the effect that there is no cause etc. can be explained much like Sextus’ analogous claims, namely as counter-balancing the claim that something is a cause. Again, Aenesidemus’ Academic credentials would make him likelier to adopt undecidability arguments than the invariability condition of Pyrrho; he is bound to have known them. Indeed, as Bett says, “[…] the invariability condition […] looks like a surprisingly unsceptical item for anyone to accept who claimed to be embarking on a radical new anti-dogmatic path” (236). Further, by depriving Aenesidemus of undecidability arguments one has little scope in explaining their emergence in Sextus, except by pointing at the usual suspect, Agrippa, who would then have introduced Academic scepticism (undecidability arguments) into Pyrrhonism. Bett is hesitant, and he does not want to attribute the innovation to Sextus himself, and he is content to suggest that we “imagine some successor of Aenesidemus who begins as an adherent of Aenesidemus’ variety of Pyrrhonism, but who then comes to feel […] that the invariability condition is unacceptable” (237). In addition, although Bett does succeed in explaining why Aenesidemus called himself a Pyrrhonist, that may also be explained differently and as more a political move than a philosophical one; thus, Aenesidemus’ Neo-Pyrrhonism may, at least in part, have been the revival of the radical scepticism of the Academy dressed up in Pyrrhonian garb. Finally, the apparent vestiges of Aenesideman relativism (as construed by Bett) in Sextus’ and Diogenes’ accounts of the modes of Aenesidemus, although problematic for anyone wishing to challenge Bett’s idea, can be and have been explained away.7

Thus, the crucial issues in the history of Pyrrhonism, in particular the meaning of Aristocles’ testimony and the nature of Aenesidemus’ scepticism, continue to offer scope for disagreement, entirely in keeping with the subject-matter. I saw no blemishes on the book. The standards of production are as high as those of composition.


1. Sextus Empiricus: Against the Ethicists, Translated with an Introduction and Commentary (Oxford: Clarendon Press, 1997). Of the articles one might mention ‘Aristocles on Timon on Pyrrho: The Text, its Logic and its Credibility’, Oxford Studies in Ancient Philosophy 12 (1994), 137-81, which is the basis of the first part of the present book, and ‘What did Pyrrho Think about ‘The Nature of the Divine and the Good’?’, Phronesis 39 (1994), 303-37.

2. In his ‘Pyrrho on the Criterion’, Ancient Philosophy 18 (1998), 417-34, T. Brennan is highly critical of Bett’s idea.

3. Bett (1997 xvi; cf. xvii, 101-102).

4. See A. A. Long & D. N. Sedley, The Hellenistic philosophers, volume 2 (Cambridge: Cambridge University Press, 1987), 6.

5. As Bett formulates the invariability condition, usually as “in order for an object to be a certain way by nature, it must be that way invariably or without qualification” (118), it seems to be left open that a thing be invariably F but still not by nature F. This leaves open the possibility that, even if a thing is invariably F, it may nevertheless be contingently F. Apparently, this would not change anything for Pyrrho, on Bett’s account, since he postulates conflicting features for all objects, so that nothing is anything specific by nature, and reality is indeterminate. But the invariability condition was accepted by, for example, the Stoics, Bett claims, and “the reason why virtue and vice are the only things … that are genuinely good and bad respectively … is that virtue is the only thing that is good, and vice is the only thing that is bad, in all circumstances and without qualification” (197). Here, the invariability of virtue’s goodness seems to imply its reality. Perhaps Bett has in mind the equivalence of invariability and reality, so that, only if a thing is invariably F, is it by nature F (see 119n11).

6. ‘Aporetic Pyrrhonism’, Oxford Studies in Ancient Philosophy 6 (1988), 139-68.

7. See J. Annas & J. Barnes, The Modes of Scepticism, (Cambridge: Cambridge University Press, 1985).