Introduction

Sextus' life and works

The two books Against the Logicians are part of a larger work by Sextus Empiricus, the best known ancient Greek skeptic and the only one from whom we possess complete texts, as opposed to fragments or second-hand summaries. About Sextus Empiricus himself we know virtually nothing. He identifies himself as a member of the Pyrrhonist skeptical tradition, on which more in the next section. He occasionally refers to himself in the first person as a medical practitioner (PH 2.238, M 1.260, cf. M 11.47). His title would suggest that he was a member of the Empiricist school of medicine. This is confirmed by Diogenes Laertius (9.116), who refers to him as "Sextus the Empiricist"; it would anyway not be surprising, given that we know the names of several other Pyrrhonists who were also Empiricists. Sextus at one point addresses the question whether medical Empiricism is the same as Pyrrhonist skepticism (PH 1.236-241), and unexpectedly replies that another school, the Methodist school, has closer affinities with skepticism. However, it is possible to read this passage as expressing suspicion towards a certain specific form of Empiricism, rather than towards the school as a whole. 1

Such indications as there are concerning where Sextus was born, or where he worked in his maturity, are too slender to bear any significant weight. The evidence suggests that he lived in the second century CE, but it is not clear that we can fix his dates with any more precision than that. 2 In any case he appears to be curiously isolated from the philosophical currents of his own day. In the second century there were flourishing Aristotelian and Platonist movements, yet Sextus shows no awareness of them whatever; his focus is invariably on the Hellenistic period (that is, roughly, the last three centuries BCE) and earlier. His immediate influence appears to have been virtually non-existent; we hear of a student of his, Saturninus, but for the rest of antiquity interest in skepticism seems to have been extremely limited. It is a very different story when Sextus' works were rediscovered in the early modern period; and this belated influence makes his writings of interest to students not only of ancient but also of modern philosophy.

Sextus' voluminous surviving oeuvre comprises three distinct works. The best known is Outlines of Pyrrhonism (commonly referred to by PH, the abbreviated form of the title in Greek), which survives complete in three books. Of these the first is a general summary of the Pyrrhonist outlook, and the other two deal with the theories of non-skeptical philosophers in each of the three standardly recognized areas of philo- sophy in the post-Aristotelian period, namely logic, physics, and ethics; the discussion of logic occupies the whole of Book 2, while Book 3 is shared between physics and ethics. 3 Another work, Against the Learned (Pros Mathematikous - also referred to by the Latinized title Adversus Mathematicos, or by the abbreviation M), is complete in six books, and is quite different in subject-matter. It addresses a variety of specialized sciences (one per book); in order, the subjects are grammar, rhetoric, geometry, arithmetic, astrology, and musical theory. 4 This work is of interest for many reasons, but of only marginal relevance to this volume.

It is the third work to which the books translated in this volume belonged. Surviving from this work, in addition to the two books Against the Logicians (and in this order), are two books Against the Physicists and one book Against the Ethicists. 5 But it is all but certain that there was originally more. The final sentence of Against the Ethicists clearly signals that the entire work has come to an end. But the opening sentence of Against the Logicians refers back to a just-completed general treatment of Pyrrhonism. This was long thought to be a reference to PH. But that cannot be correct, since PH is not, as a whole, a general treatment of Pyrrhonism; the reference must rather be to a lost portion that discussed Pyrrhonism in general terms, as does Book 1 of PH. It appears, then, that this work as a whole covered the same broad subjects, in the same order, as PH, but at considerably greater length. Sextus himself calls this entire work Skeptical Treatises (Skeptika Hupomnemata); he makes several references, using this title, to what are clearly passages from Against the Logicians and Against the Physicists (M 1.29 [26], 2.106, 6.52). That this is not the title by which the work is now generally known is due to an egregious error committed at some point in the manuscript tradition. The manuscripts represent the five surviving books as a continuation of the six books of M; as a result, Against the Logicians, Against the Physicists, and Against the Ethicists are generally referred to collectively as M 7-
11.6 There is reason to believe that the complete work was ten books long - that is, that the lost general portion occupied five books. 7 In the manuscripts, the two books of Against the Physicists and the single book Against the Ethicists are labeled (either at the beginning or the end) as the eighth, ninth, and tenth books respectively of Sextus' Skeptika, or of his Hupomnemata, both clearly abbreviations of Skeptika Hupomnemata; and Diogenes Laertius (9.116) refers approvingly to a ten-book work of Sextus entitled Skeptika, which is presumably the same work. If this is not all the product of some other, now inexplicable, error, the entire original work must have been very extensive indeed. Even in its current, incomplete form, it is roughly twice as long as either of the other two complete works.

Ancient Greek skepticism before Sextus

Contemporary scholarship recognizes two traditions of Greek skepticism, Academic and Pyrrhonist. It was only the Pyrrhonists who actually called themselves skeptics. But already in antiquity the two traditions were widely seen as having certain crucial features in common, so that the term "skepticism" is readily applied to the Academics as well. The word skeptikos literally means "inquirer." As Sextus explains it at the beginning of PH 1, the skeptic is someone who is still searching for the truth, as opposed to believing either that he has found it or that it is undiscoverable. Sextus regularly refers to members of the first non-skeptical group as dogmatists; by analogy, members of the second group are today sometimes called negative dogmatists. It is important to note, then, that skepticism as understood in the ancient Greek world did not consist in a denial of the possibility of knowledge (or, for that matter, a denial of anything else). In modern philosophy this is precisely what skepticism has generally been taken to be; but from the ancient skeptical perspective this position is just as much anathema as are dogmatic positions that claim to be in possession of the truth. The skeptic's attitude is rather one of open-mindedness, of not thinking that one has discovered the truth, but not ruling out the possibility of its discovery either; the skeptic neither affirms nor denies, but suspends judgment. Suspension of judgment (epoche) is, then, a key term in the self-description of both the Academic and Pyrrhonist skeptical traditions.

Pyrrhonism takes its name from Pyrrho of Elis, a little-known figure from the late fourth and early third centuries BCE. Pyrrho attracted an immediate following, notably including his biographer Timon of Phlius, who is undoubtedly the most important source of our meager evidence about him. But it looks as if this early Pyrrhonism died out after a generation or two. Meanwhile, in the early to mid-third century the Academy, the school founded by Plato, was taken in a skeptical direction - a direction apparently encouraged by elements in Plato's portrait of Socrates - under the leadership of Arcesilaus. The skeptical Academy persisted until the early first century BCE, when the skepticism softened and the school itself fragmented. But around the same time, in part as a reaction against the softening of the Academy's skepticism, a new skeptical movement, claiming inspiration from Pyrrho, was started by another little-known figure, Aenesidemus of Cnossos, himself apparently an Academic at first. It is this revived Pyrrhonist movement to which Sextus later belonged. We know the names of several other Pyrrhonists, but virtually nothing about their thought.

I spoke of suspension of judgment as the hallmark of ancient Greek skepticism, both Academic and Pyrrhonist. But it should not be thought that skepticism in the period was entirely uniform, either between the two traditions or within each of them. The most obvious difference between the two traditions is that the Pyrrhonists consider suspension of judgment to have a very significant practical effect. According to them, suspension of judgment frees one from the tremendous turmoil, both intellectual and emotional, that is associated with the holding of definite beliefs about how things really are. The result of suspension of judgment is therefore ataraxia, freedom from worry. This theme does not appear in Against the Logicians; it is concerned with the marshaling of arguments designed to generate suspension of judgment, not with the further outcome for someone in that condition. However, ataraxia does receive considerable attention in PH, and also in Against the Ethicists. The Academic skeptics, on the other hand, give no indication of holding that suspension of judgment has any particular practical benefit. Both Arcesilaus and Carneades, his greatest successor, took pains to show that choice and action were possible in the absence of definite beliefs; a passage from Book 1 of Against the Logicians (150-189) is our most substantial evidence of this. But there is no suggestion that one is better off withdrawing from definite beliefs, other than in terms of intellectual respectability.

Quite apart from this major difference, Sextus does not consider the Academics to be genuine skeptics. That is, he does not consider the position they have adopted to be genuine suspension of judgment. This could perhaps be gathered from the passage of Against the Logicians just referred to. Arcesilaus and Carneades are examined in the course of Sextus' review of thinkers who accepted the existence of a criterion of truth - a central tenet in any dogmatist philosophy. Now, this may seem to be unfair of Sextus. For the criteria that he attributes to Arcesilaus and Carneades are criteria to be used in practical decisions; yet he himself has earlier distinguished a criterion of truth (on the existence of which he will suspend judgment) from a criterion of action, which even the skeptic inevitably employs (1.29-30). However, it is clear from a passage of PH
1, where Sextus emphasizes the distinction between Pyrrhonism and Academic thinking, that he takes the specific character of the Academics' practical criteria, as well as other features of their thought, to commit them to dogmatism, both positive and negative (PH 1.226-234). He allows that the Academics (especially Arcesilaus) say many things that sound like Pyrrhonism. But in their mouths, unlike those of the Pyrrhonists, these things are in his view delivered in the guise of definite beliefs, and therefore disqualify them from the title of skeptics. It is open to serious question whether Sextus is right about this - which is why the notion of Academic skepticism can be upheld in modern scholarship. But the fact remains that Sextus does not regard Arcesilaus and Carneades as kindred spirits; for him, Pyrrhonism is something quite different, and not only because of the place it assigns to ataraxia.

Ataraxia as the ultimate product of one's intellectual activity appears to be a constant in the history of Pyrrhonism, from Pyrrho himself through Aenesidemus to Sextus. But it is by no means so clear that the precise nature of that intellectual activity, or of the suspension of judgment that results from it, was the same at every stage of the tradition. It is questionable whether Pyrrho practiced any full-scale suspension of judgment at all. While he is reported as recommending that we not trust our sensations and opinions as guides to the nature of things, the basis for that recommendation appears to be either a metaphysical thesis that things are inherently indefinite (which would make him a dogmatist) or an epistemological thesis that the nature of things is unknowable (which would make him a negative dogmatist).8 Certainly Sextus does not appeal to Pyrrho's thought in any detail; he simply says that Pyrrho seems to have been closer to skepticism than any of his predecessors (PH 1.7). Indeed, he rarely even mentions him (never, in Against the Logicians). Aenesidemus is reported as claiming to "philosophize in the manner of Pyrrho" (Photius, Bibl. 169b26-27 = LS 71C3), but this too can be understood as implying a general similarity of approach rather than a detailed correspondence of doctrine. So Pyrrho may well have served more as an inspiring prototype than as a source of specific arguments or ideas. Uniquely in Greek philosophy (before the later Pyrrhonists), he claimed to have arrived at tranquillity by way of a certain kind of mistrust or withdrawal of belief - rather than by gaining an understanding of the detailed workings of the universe; this may have been enough to make Aenesidemus adopt him as a kind of founding father.

Leaving aside Pyrrho himself, it also appears that the Pyrrhonism of Aenesidemus was in important respects different from the version to be found in most of Sextus, including Against the Logicians. But since this difference is not unconnected with the way we read Against the Logicians itself, it will be convenient to touch on it later (see pp. xix-xxiv).

The general character of Against the Logicians

The status of logic as one of the three major parts of philosophy has already been mentioned. But it is important to note that logic, in this context, covers considerably more than what we would normally understand by this term. The Greek word logos can mean both "speech" and "reason" (among other things), and the scope of logic, as conceived in the Hellenistic period, reflects this duality. For the Stoics, whose philosophical taxonomy was by far the most complete and systematic of any at the time, logic included rhetoric and the study of language, as well as the study of the means for determining what is true and what is false (and what is neither). And under the latter heading came not only the study of the components, structure, and validity of arguments - that is, material that we would call logic - but also the study of whether and how we can tell the way the things really are - in other words, material that for us would fall under epistemology. 9 The Epicureans rejected many of these topics as useless, including the study of argument forms for its own sake, and this led some to claim that they rejected logic itself. But they were certainly interested in methods for determining what is true, and this, as both Sextus (Against the Logicians 1.22, cf. 14-15) and Seneca (Letter 89.11) point out, means that they did in fact make contributions to logic as understood in antiquity, whatever label they or others might give to it.

Sextus' Against the Logicians does not cover all the areas included in the Stoics' conception of logic. The more purely linguistic aspects of the subject, as they conceived it, are addressed in Against the Grammarians and Against the Rhetoricians (M 1-2).10 But Against the Logicians certainly discusses epistemological matters in addition to - in fact, far more than - logical ones in our narrower sense. The whole of the first book is occupied with the question whether there is a criterion of truth. The second book then tackles the topics of truth itself, sign, and demonstration. 11 With the partial exception of truth, all of these topics have to do with methods for settling what is the case. In the second book there is a fair amount of discussion of logical matters, in our sense, along the way; but much, if not most, of the time this is ancillary to the broadly epistemological themes that constitute the basic outline. Both sign and demonstration, for example, are defined in terms that require reference to logical notions such as premises, consequences, and conditionals
(2.244-256, 300-315); but both of them are means for discovering truths about unobservable things, and Sextus' overriding question is whether there are any reliable means for doing this. The first book, by contrast, contains virtually no discussion of what we would call logic. This is because Sextus (most of the time, at any rate)12 understands a criterion of truth as a means for grasping immediately observable truths, rather than for inferring to unobservable ones; see 1.25 for the initial distinction in these terms between criteria on the one hand, and signs and demonstrations on the other. Hence in the discussion of the criterion, questions about the reliability of inferences, or more generally about the logical relations between distinct propositions in an argument, remain in the background.

Described in outline, Sextus' method is simply to subject to scrutiny the views of the dogmatists in these areas. This means that, in addition to his own criticisms and counter-arguments, Against the Logicians (like most of Sextus' works) contains a considerable amount of summary of other people's views. The most extensive case of this is the long historical survey that makes up roughly the first half of the discussion of the criterion of truth (1.46-260). Sextus describes all the earlier views that might be thought to bear on this subject (even though the Hellenistic term "criterion of truth" postdates most of the thinkers in question). As a result, this passage is a mine of information about ideas that in many cases are otherwise poorly recorded. But there are numerous other passages of the same kind throughout both books. Outside the first half of Book 1, it is the Stoics, always for Sextus the preeminent dogmatists, whose views receive the most scrutiny, and therefore the most summary. This is particularly true on technical logical matters, where the Stoics are almost the only school represented; the most obvious exception is the views on the truth-conditions for conditionals held by Philo and Diodorus (2.113-117, cf. 265) - but even these are closely associated with the Stoics, since the founder of Stoicism, Zeno of Citium, studied logic with them
(Diogenes Laertius 7.16, 25). It is striking that Aristotle and the Peripatetics, whose formal logic was the main rival to that of the Stoics (and was far more influential beyond antiquity), are never mentioned in this context in Against the Logicians. Nor, for that matter, does Aristotle's theory of demonstration, elaborated in the Posterior Analytics, receive any attention in the section of Book 2 devoted to that subject. But these are just extreme cases of a general phenomenon - namely, Sextus' comparative lack of interest in Aristotle's philosophy and his heavy concentration on Stoicism. Again, one fortunate by-product of this preference is that Sextus supplies many details about Stoic philosophy that we would not otherwise have.

A cursory reading of Against the Logicians might leave one with the impression that Sextus' goal is to show that the dogmatists are wrong about the issues addressed: for example, that there is no such thing as a criterion of truth. But this would be a mistake. It is true that a great deal of the argumentation takes the form of undermining the dogmatists' pretensions to knowing the answers in these areas. But Sextus several times takes the trouble to make clear that his aim is something other than this might suggest (1.443, 2.2, 159-160, 298, 476-477). In keeping with what was explained in the previous section, he intends to bring us to a position of suspension of judgment on the topics in question, such as whether or not there is a criterion of truth. This is to be accomplished by juxtaposing the positive arguments of the dogmatists with the critical arguments supplied by himself, resulting in a situation of "equal strength" (isostheneia) between the opposing arguments. "Equal strength" is best understood as a psychological notion; it is not that both or all of the opposing positions are rationally justified to an equal degree (which would require endorsement of theoretical notions that would themselves be objectionably dogmatic), but simply that one is supposed to find them equally persuasive - in which case, according to Sextus, suspension of judgment inevitably results. Sextus does not, then, identify with the critical arguments, even though we may presume them to have been largely devised by the skeptics; they are offered as a counter-weight to the dogmatists' arguments, the eventual outcome being that one identifies with no particular set of arguments. 13

This strategy is a further reason, besides clarification of what is to be attacked, for the lengthy summaries of dogmatic views. As Sextus says (2.476-477, cf. 160), it actually suits his purpose for the dogmatists' arguments to be presented as strong ones - strong enough, that is, to balance his own counter-arguments, but no stronger; equally, then, it suits his purpose to present these arguments fully and sympathetically. Another, similar argumentative purpose may perhaps be discerned in the way he structures the opening review of positions on the criterion of truth; this begins by listing those he takes to have denied the existence of such a criterion, and continues with the believers in a criterion - who in turn differ among themselves in significant ways. The effect is to balance a great many dogmatic arguments against each other; from a skeptic's perspective this is ideal, since, if the arguments are of comparable strength, it may be calculated to generate suspension of judgment without the skeptic himself having to lift a finger. Sextus does not actually say that this is what he is doing, but his mention, at the close of this section, of having just laid out the "disagreement" about the criterion
(1.261) may suggest such an agenda. Regardless of Sextus' own purposes, it is plausible that this was the intention of whoever originally compiled the material in this way.

Earlier sources and an earlier phase of Pyrrhonism

The last comment raises the question of Sextus' relation to the Pyrrhonist tradition that preceded him. It has long been understood that Sextus draws to a very considerable extent on earlier sources in the Pyrrhonist tradition and probably elsewhere. As noted earlier, there are no other Pyrrhonists besides Sextus whose work has survived intact. But there are correspondences between passages of Sextus and passages of Diogenes Laertius' summary of Pyrrhonism (9.74-108) that are too close for coincidence; they extend beyond similarities of subject-matter to parallels in argumentative structure, and even detailed correspondences in vocabulary and sentence-structure. They also occur at numerous different places in Sextus' work, as opposed to being confined to a single book. (There are occasional parallels between Sextus and other authors as well, but I shall ignore these; the parallels with Diogenes are by far the most wide-ranging.) Since Diogenes mentions Sextus, and also Sextus' pupil Saturninus (of whom nothing more is known), he is clearly the later of the two, and one might suppose that he is simply copying his material from Sextus. But there are also sufficiently many differences between the two authors to make this highly unlikely. In addition to some stylistic differences, Diogenes very often treats material in a different order from Sextus, and some of his material does not correspond to anything in Sextus (but this is interspersed with material that does). Diogenes is quite explicit about using earlier sources, and he could hardly have made up this non-corresponding material. The conclusion therefore seems inevitable that Sextus and Diogenes are both drawing on the same earlier (but now lost) source or sources, either directly or at one or more removes. 14

There is room for debate as to how much Sextus modified the material that he took from these unknown predecessors. Many scholars have seen him, like Diogenes, as little more than a copyist of previous material. But this seems unduly patronising. For one thing, as noted earlier, Against the Logicians, Against the Physicists, and Against the Ethicists cover roughly the same ground as PH 2-3. Here again there are a great many parallel passages in the two works, and in some of these cases, too, there is a very close similarity of thought and language. Clearly one of these works is a revised version of the other; either Sextus wrote PH first and then expanded it into the work of which M 7-11 is the surviving portion, or he wrote the latter work first and then condensed it into PH. I shall return in the next section to the question of which work came first. But either way, it must be allowed that Sextus shows some initiative in the way he organizes and reworks his material. For despite the many close parallels, there are also significant differences; entire topics are treated in one work and ignored in the other, and the language and approach do sometimes differ considerably. Besides, it is fair to say that a consistent authorial personality comes through in Sextus' works; however little we know of Sextus the man, his writing has a characteristic voice (the precise tone of which I will leave it to readers to discover for themselves). His extensive use of preexisting material is not to be doubted. However, it looks as if he does not just passively appropriate this material, but molds it into a product that is distinctively his own.

One likely source of material for Sextus, either directly or indirectly, is Aenesidemus. We know from Sextus himself and from others that Aenesidemus wrote a work in eight books called Pyrrhonist Discourses
(Purroneioi Logoi); given Aenesidemus' position as the originator of the later Pyrrhonist tradition, we may plausibly assume that this work was treated as seminal by at least some in that tradition. In Against the Logicians (1.345) Sextus refers in passing to the Ten Modes, one of the several sets of standardized forms of skeptical argumentation (summarized in PH 1), as the Ten Modes of Aenesidemus. But it would hardly be surprising if much more of what Sextus borrows from the tradition derived ultimately from Aenesidemus, even though he mentions him only relatively infrequently. But if this is so, then Sextus is apparently using material that originally belonged to a version of Pyrrhonism somewhat different from the version his own works mostly espouse. For there is good reason to believe that, at some point between Aenesidemus and Sextus, Pyrrhonism underwent a change.

For Sextus, as we have seen, suspension of judgment is reached by the juxtaposition of opposing arguments of "equal strength," so that one withdraws assent from either
(or any) of these arguments. The dogmatists are thereby exposed as misguided for trusting in the truth of their arguments. But the goal is not to show that the items in which they believe, such as criteria of truth, signs, and demonstrations, do not exist, or that their beliefs about the nature of these items are false; rather, it is to generate equally powerful arguments on either side, thus relieving one of the burden of beliefs on these topics either way. 15 But there is evidence of an earlier form of Pyrrhonism, associated with Aenesidemus, in which endorsement of conclusions to the effect that certain things (in which the dogmatists believe) do not exist was quite acceptable skeptical procedure.

Our most substantial piece of evidence on Aenesidemus' thought is a summary of his Pyrrhonist Discourses by Photius, the ninth-century patriarch of Constantinople (Bibl.
169b18-170b35 = LS 71C + 72L). It appears that the first book expounded the Pyrrhonist outlook in general terms. The other seven books then dealt with particular topics addressed by the dogmatists; Photius only gives us a sentence about each, but his report is nonetheless striking. Among the topics included was that of signs, and on this topic Photius tells us, "In the fourth book he asserts that signs (in the sense that we call things that are clear signs of things that are unclear) do not exist at all, and that those who think they do exist are deceived by a vain attraction" (170b12-14). Contrary to Sextus' careful preface to his arguments against the sign (2.159-161), where he makes clear that these arguments are not to be endorsed but to be balanced against the dogmatists' positive arguments, Aenesidemus apparently did endorse such arguments, and in no uncertain terms. Photius reports the same kind of conclusion, delivered with similar degrees of outspokenness, in the case of Aenesidemus' discussions of causes (170b17-22) and of the ethical end (170b30-35).

Photius might, of course, be accused of misunderstanding Aenesidemus. No doubt Aenesidemus' discussion of signs (and of the other topics) did include arguments against their existence. Indeed, Sextus reports an argument from Aenesidemus' fourth book, an argument to the conclusion that signs are not apparent things (2.215, 234). Since the very concept of a sign is of something observable that licenses an inference to something unobservable, this is essentially equivalent to concluding that signs do not exist. But Sextus has no trouble using this argument as part of a strategy of generating suspension of judgment about the existence of signs; and Aenesidemus himself, one might say, could just as well have done the same thing. The impression Photius gives of vigorous denial might be explained as simply the product of an unsympathetic reading; it is clear from the largely dismissive criticisms following his summary (170b36-171a4) that he does not take Aenesidemus particularly seriously.

But this reaction would be a mistake. For, leaving aside the question of Photius' own credibility, Photius is not the only author to describe Pyrrhonists as denying the existence of things. Diogenes Laertius' summary of Pyrrhonism also includes numerous reports of Pyrrhonists arguing to conclusions of the form "there is no such thing as X" - signs are just one example (9.96) - and also reports of Pyrrhonists "doing away with"
(anairein) various things believed in by dogmatists, which appears to amount to the same thing. And, if Diogenes too might be impugned as a philosophically naive reporter, the same phenomenon can be observed in one book of Sextus himself, namely Against the Ethicists. Here Sextus argues for the conclusion that nothing is good or bad by nature. And here it is not open to us to claim that he means these arguments to function as one side of an opposition, with the dogmatists supplying the other side. Not only does he not say that this is what he is doing (as he does in Against the Logicians). He also tells us several times that it is the skeptic's acceptance of the conclusion that nothing is by nature good or bad that produces the desired state of tranquillity (M 11.118, 130, 140).

There was, then, a phase of Pyrrhonism - a phase that, given Photius' report, it is plausible to trace to Aenesidemus - in which arguing that the dogmatists were mistaken, and that the entities in which they believed did not exist, without any juxtaposition of those arguments against the dogmatists' own positive arguments, was normal and accepted Pyrrhonist procedure. This, of course, raises the question how such a procedure could be considered compatible with any form of suspension of judgment. The issue is somewhat complicated, 16 and not really germane to our present concerns. But very briefly, one possible answer centers around a certain conception of what it is for something to be by nature a certain way. According to this conception, the nature of something is fixed and invariable. Hence, to take two examples already mentioned, to say that something is by nature good, or by nature a sign, is to say that it is invariably and in all circumstances good, or a sign. And to deny that anything is by nature good, or by nature a sign, is to deny that anything is invariably and in all circumstances good, or a sign. Now, a denial of this kind does not offer any positive characterization of the nature of anything; to say that nothing is invariably good, or invariably a sign, is not to assert that anything is invariably (and therefore by nature) of any particular character. And this suggests a way in which such denials could be understood as compatible with a certain form of suspension of judgment: a suspension of judgment, that is, that consisted in refusing any attempt to specify the nature of anything. 17

But let us leave this issue aside. The important point for our purposes is simply that a version of Pyrrhonism that seems to precede Sextus himself (but that survives intact in one of his books) allowed a method of argumentation that, by Sextus' usual standards, would qualify as negative dogmatism. Now, given this state of affairs, as well as Sextus' undoubted reliance on earlier sources, it is natural to wonder whether Against the Logicians contains any traces of this earlier phase of Pyrrhonism. One obvious possibility is that the long stretches of argument against the dogmatists - stretches of argument that, as I said, look on superficial inspection as if Sextus intends them to show that the dogmatists are wrong - derive from this earlier phase, in which that was precisely the intention. As we saw, Sextus does explicitly appeal to Aenesidemus in one part of the discussion of signs; and the debt may well be more extensive. 18 Again, it is not that Sextus does not make clear his own intentions in employing these destructive arguments. But one may well wonder whether, had he approached these topics with a clean slate instead of adapting already existing materials, he might have structured his discussion differently, so as not to give even an impression of negative dogmatism.

Another possible indication of the same thing is Sextus' periodic use, in Against the Logicians, of the word anairein, "do away with," to describe the skeptic's activity. As we saw, this word occurs a number of times in Diogenes Laertius' summary of Pyrrhonism, where it is interchangeable with "argue for the non-existence of." For the Pyrrhonists to "do away with" things, in this sense, was normal in the earlier phase of Pyrrhonism.
19 But in Against the Logicians Sextus' strategy, as he several times reminds us, is different from and indeed incompatible with this. Nevertheless, in numerous places (1.299, 371, 2.1, 142, 157-158, 290, 338) he describes himself as "doing away with" certain kinds of objects posited by the dogmatists. In one place (1.26) he even uses the term "do away with" in the same context as "suspend judgment"; to "do away with" a set of objects posited by the dogmatists is, according to this passage, sufficient for putting us into a state of suspension of judgment about them. Now again, this is far from conclusive. One can perhaps understand "do away with X" as shorthand for "offer arguments against the existence of X, which will then be juxtaposed with arguments for the opposite conclusion." But even if this is correct, the possibility remains that the repeated occurrence of this word is due to Sextus' use of material that had its original home in a version of Pyrrhonism where the Pyrrhonist could quite straightforwardly, and without any resort to shorthand, claim to "do away with" the entities that he discussed. Once again, Against the Logicians, unlike Against the Ethicists, is not an instance of the earlier variety of Pyrrhonism. 20 But it would hardly be surprising, given Sextus' use of preexisting sources, if it contained traces of that earlier variety.

Against the Logicians compared with PH 2

The question of similarities and differences between PH 2-3 and Against the Logicians, Physicists, and Ethicists was introduced in the last section. A comparison between Against the Logicians and its counterpart PH
2 reveals some notable differences; each contains a considerable amount of material that the other omits. The largest and most obvious portion of text in Against the Logicians having no parallel in PH 2 is the long survey of previous positions for and against the criterion of truth (1.46-260). Since PH is designed as an outline account, it is not surprising that Sextus would have decided not to include this material in any form in PH
2; although, as we saw (pp. xvii-xix), it plays a valuable role in Against the Logicians, it is not strictly necessary for a skeptical treatment of the criterion of truth, and is easily detachable from the rest of Sextus' account. What is more surprising is that PH 2, although it is less than a third the length of Against the Logicians, discusses a number of topics that are not explicitly dealt with at all in the longer work. After the end of the section on demonstration (134-192), which corresponds to the final portion of Against the Logicians, there are chapters on deduction, induction, definition, division, "division of a name into things signified" (214), whole and part, genus and species, common attributes, and sophisms, none of which has any counterpart in Against the Logicians. Thus, despite being much shorter, PH 2 is in a certain sense more comprehensive than Against the Logicians. And while, as we saw, Against the Logicians is devoted (at least if one looks at its broad structure) largely to epistemological topics, with logic in our narrower sense being treated most of the time as ancillary to these, PH 2 focuses explicitly on a number of subjects that are clearly (in our sense) logical in nature. There is even a hint in these chapters of some of the linguistic concerns that the Stoics also classified under logic.

This just underscores the extent to which, when it comes to the topics that both works do include, Against the Logicians is lengthier in its treatment than PH 2. The former contains numerous arguments that the latter omits, and, even where both works have versions of what is recognizably the same argument, Against the Logicians regularly develops the argument in a much more leisurely, and frequently more rambling, fashion. "Lengthier" in this context, therefore, does not necessarily mean " better." Indeed, there is a diffuse, everything-but-the-kitchen-sink quality to much of Against the Logicians, which often makes it hard to keep track of the main thread of the discussion. I have tried to compensate for this by including an outline of the argument, both as a complete whole immediately before the translation and in the form of headings within it. It will be noticed that the outline of the second book is considerably longer than that of the first, even though the two books themselves are not very different in length; this is because the general structure, and the place of each passage within it, takes even more effort to grasp in the second book than in the first.

So the greater extent of Against the Logicians compared with PH 2 does not obviously work to its advantage. PH 2 is more concise and therefore easier to follow, and this is generally not at the expense of any cogency in argument - often the reverse. In addition, even in the areas covered by both works, there are some cases where Against the Logicians does not include material that it might profitably have included, and that PH 2 does include. Both works mention a series of "indemonstrable" arguments - that is, basic argument-forms not admitting of justification in still more basic terms - that play an important role in Stoic logic. The Stoics held that there were five such forms of argument; but while PH 2 mentions all five (157-158), Against the Logicians mentions only three (2.223-226). This is partly due to the different roles these summaries play in the two works. In PH 2 the indemonstrable arguments are introduced as examples of redundancy in argument. If all five such arguments can be shown up as redundant, then, it is claimed, "all of dialectic is overturned"
(156); so it clearly suits Sextus to introduce all five. In Against the Logicians, on the other hand, the indemonstrables are introduced as part of a lengthy digression analyzing Aenesidemus' argument about signs, and for this purpose only the first three need to be mentioned. But this is just another example of PH 2's generally more adroit handling of its material. For the digression in Against the Logicians is really not necessary for the purpose at hand; the validity of Aenesidemus' argument is obvious without any excursus on indemonstrable arguments - as, indeed, Sextus has already made clear before the excursus begins
(2.217-222). Yet a more complete treatment of indemonstrable arguments might well have been useful in some other place.

Again, PH 2 gives a more complete summary of the variety of views about the truth-conditions for conditionals. Both works include the views of Philo and Diodorus
(Against the Logicians 2.113-117, PH 2.110-111). But PH 2 then adds a view centered around a notion of "connectedness" (sunartesis) between the antecedent and the consequent of the conditional - mentioned in Against the Logicians only in a later and wholly unexplained reference (2.265)21 - and a further view as well (PH 2.111-112). Finally, PH 2 includes at least some passing mention of Peripatetic logic (PH 2.163-166,
193-198), of which, as I noted earlier, Against the Logicians appears to be wholly unaware.

PH 2 seems, then, to be in various respects superior to Against the Logicians. Another glaring organizational example is this. Both works include discussions of the Stoic distinction between the truth (he aletheia) and what is true (to alethes). But whereas the discussion in PH 2 belongs where one would expect, in the course of the discussion of truth (81-83), in Against the Logicians it is placed very awkwardly between an introductory section on the criterion and the review of historical positions on the criterion
(1.38-45). The effect is to interrupt the discussion of the criterion and to insert material that has no connection with anything else in the first book. Scholars have also pointed to Sextus' treatments of the criterion and of the sign as cases where Against the Logicians is inferior, in respect of structure, cogency of argumentation, or both, to PH 2.22 Moreover, it must be admitted that in some places in Against the Logicians the writing, or the transition of thought, is just very ungainly (and that PH 2 is not comparable in this respect); I have indicated the most extreme cases in my notes. These defects do not by any means render Against the Logicians valueless or uninteresting. But they do make it in some ways difficult reading, which I hope my notes and outline will do something to mitigate.

They also put into sharp focus the question which of the two works came first. The traditional view has been that PH was Sextus' first work, and that Against the Logicians and the larger work to which it belongs are the result of his revising and expanding the material that went into PH. But this view was based on comparisons of style and vocabulary in the two works that, while of considerable interest for various reasons, are worthless for establishing their chronology. 23 The question therefore needs to be considered afresh.

I have mentioned the existence of a large number of parallels between PH 2-3 and the larger work. The specific case of PH 2 and Against the Logicians is no exception. These parallel passages are listed in a special section at the end of the volume. 24 A glance at this list reveals that in both works, the discussion of the topics in logic that they share unfolds in roughly the same order; inspection of the passages themselves shows that there are a great many instances of the same specific arguments in both, and even some close verbal similarities. My impression is that there are fewer of the latter than in the case of Against the Ethicists and the corresponding ethical section of PH 3.25 Nonetheless, the nature and extent of the common material makes it evident that one of these treatments of logic is a revised version of the other, just as in the case of ethics (and, for that matter, physics). But now, if this is the case, the superiority of PH 2 over Against the Logicians, in the numerous respects just mentioned, would seem to favor the view that Against the Logicians came first, and that PH 2 is a later, cleaned-up version of roughly the same material; one normally expects revision to result in improvement, not deterioration. Of course, this is no more than a general rule. It is not inherently impossible that Sextus became more inept in his style of composition and more sloppy in his argumentation as he got older, or that he had more trouble with works on a larger scale. But it is difficult to imagine why anyone would have made some of the specific changes that we would have to suppose he made, if PH 2 was the earlier work. Why, for example, would one remove the discussion of the Stoic distinction between the truth and what is true from its natural place in the section on truth (where it belongs in PH 2), and put it in the section on the criterion, where it does nothing but interrupt the flow? Or why, in a discussion designed to emphasize disagreement about the truth-conditions for conditionals (Against the Logicians
2.112-117), would one limit oneself to just two views on this issue, suppressing any mention of two other views that the earlier, shorter work had already included? It is much more natural in such cases to suppose that the shorter and more neatly composed work is the later one. 26

Another consideration points in the same direction. I mentioned that Against the Logicians sometimes uses the word anairein, "do away with," to describe what the skeptic does with the dogmatists' views, and that this seemed to be a relic of an earlier version of Pyrrhonism distinct from Sextus' own. But it is striking that PH never uses the word anairein to describe the skeptic's own procedure. In PH the word is sometimes used to refer, as one might expect, to the demolishing of someone's view. But it is never suggested that this is something the skeptic does; on the contrary, it is several times stated that this is precisely what the skeptic does not do (1.193, 196, 197). It would be very surprising if Sextus first used the word in a way appropriate to his version of Pyrrhonism and then, in revising the work, started using it in a way that conformed to an earlier version incompatible with his own. It is far more likely that he uncritically reproduced this earlier usage in his first work, which stuck more closely to its sources, and then, in revising this material, adjusted his vocabulary so as to make it conform better to his own position.

These brief remarks certainly do not settle the question. But if they are on track, they point to the conclusion that Against the Logicians was Sextus' first attempt at the subject-matter of logic, and that many of the awkwardnesses of that first attempt were ironed out in the subsequent revisions and improvements that led to PH 2. I have argued elsewhere that Against the Ethicists was composed before the ethical section of PH
3; the evidence in that case is similar in kind to that appealed to just now, but much more extensive. 27 If we assume that each work was written in its entirety at separate periods of Sextus' life, then the priority of Against the Ethicists would lead us to infer the priority of Against the Logicians as well. Unfortunately, that assumption is clearly not beyond question. Differences of style and vocabulary between the two works, considered in their entirety, may perhaps support it. But stylistic considerations are a notoriously shaky basis on which to construct arguments about order of composition; the last century or more of scholarship on Plato has made this all too clear. What we can say, though, is that the idea that Against the Logicians is a revised and expanded version of PH 2, representing his mature thinking on the subject, is at least open to serious question. Once again, to think of Against the Logicians as Sextus' first attempt in this area, rather than as his final word, does nothing to deprive it of historical and philosophical interest. But it may result in our regarding it, flaws included, in a somewhat different light.