Printer Friendly

On the coherence of Pyrrhonian Skepticism.

Early in Outlines of Pyrrhonism (PH) Sextus Empiricus writes,
   [A]s regards none of the things that we are about to say do we firmly
   maintain that matters are absolutely as stated, but in each instance we are
   simply reporting, like a chronicler, what now appears to us to be the case.
   (1.4, my emphases)

   The Skeptic Way is a disposition to oppose phenomena and noumena to one
   another in any way whatever, with the result that ... we are brought first
   to epoche [suspension of belief] and then to ataraxia [intellectual
   tranquillity]. (1.8)

   [W]e say that the criterion of the Skeptic Way is the appearance--in effect
   using that term here for the phantasia [impression]--for since this
   appearance lies in feeling and involuntary pathos it is not open to
   question.... Holding to the appearances, then, we live adoxastos [without
   beliefs]. (1.22-23, my emphases)

   Not even in putting forward the Skeptic slogans ... like "Nothing more" ...
   or any of the others ... [does the Skeptic dogmatize].... He considers that
   ... the "Nothing more" slogan says that it itself is no more the case than
   its opposite, and thus it applies to itself along with the rest. (1.14-15)

These are among the most important of the many passages, in Sextus and elsewhere, that have given rise to what has been called the received view of Pyrrhonian Skepticism, or Pyrrhonism. (1) On this view the Pyrrhonist, inter alia, makes a completely general distinction between how things are and how they appear to be--so that not only is it true that
   music may sound, and hence appear, loud; sandpaper may feel, and hence
   appear, rough; but equally an argument may appear valid, a statement may
   appear true, an action may appear unwarranted; (2)

he claims to hold no beliefs concerning how things are, not even beliefs about the truth of the Skeptic slogans themselves, and to assent only to propositions expressing what appears to him to be the case; (3) he claims to be guided in his actions solely by what appears to him to be the case; and he opposes phenomena and noumena to one another in any way whatever in order to bring about suspension of belief since, as he reports, it appears to him that ataraxia is intrinsically desirable, and that epoche is a means of achieving it.

At first blush, it must seem surprising that such a noncommittal view should be charged with incoherence or inconsistency; yet in recent years numerous such charges have been brought against it, and even in the ancient world it faced charges of self-refutation. In this paper I defend Pyrrhonism against all of the most important of these charges of which I am aware. My principal focus is on the modern ones, in countering which I occasionally go beyond what is explicit in ancient writings to draw on what I take to be the implicit resources of the Pyrrhonists' position. If I am successful in these efforts, a more complete understanding of Pyrrhonism will have emerged, in the light of which the ancient charges of self-refutation will also be seen to fail.

I do not here address either the plausibility of Pyrrhonism as it is understood on the received view, or the rightness of so understanding it, since each of these questions would require a paper of its own; but if my efforts are successful I will have removed at least some of the objections, deriving from the principle of charity, to answering the second of them affirmatively.

1. Hume on Pyrrhonism

The best known formulation of two fundamental charges against Pyrrhonism is David Hume's:
   [A] Pyrrhonian cannot expect, that his philosophy will have any constant
   influence on the mind.... [H]e must acknowledge, if he will acknowledge
   anything, that ... were his principles to prevail ... [a]ll action would
   immediately cease; and men remain in total lethargy.... It is true; so
   fatal an event is very little to be dreaded. Nature is always too strong
   for principle.... [A]ll his objections ... can have no other tendency than
   to show the whimsical condition of mankind, who must act and reason and
   believe; though they are not able, by their most diligent enquiry, to
   satisfy themselves concerning the foundation of these operations, or to
   remove the objections, which may be raised against them. (4)

Here Hume charges that a person who governed his life by Pyrrhonian principles would (could?) engage in no action at all, and that there could be no such person, Nature being too strong for principle. Notably, however, he does not charge that the Pyrrhonist is inconsistent or incoherent, but only that his objections are impotent and tend only to show "the whimsical condition of mankind."

In this respect he differs strikingly from two of the contemporary writers with whose work we shall principally be concerned. Michael Frede has written,
   [O]n this interpretation [of the ancient Skeptics, viz., the received view]
   the skeptical position turns out to be inconsistent. For it is generally
   assumed that ordinary, everyday life is simply not possible without any
   beliefs or views; and so it is generally assumed that the skeptic refutes
   himself, when he insists on total suspension of judgment, while, at the
   same time, constantly relying on all sorts of judgments in his actual life.
   Hume's version of this objection is perhaps the most familiar; without a
   doubt, it has contributed substantially to the standard picture of the
   skeptic as a person who, if only he took his own views seriously, would be
   completely helpless in ordinary life. (SB, 180)

While Frede here adequately conveys the content of Hume's charge, his description of it as a charge of inconsistency or self-refutation threatens to obscure the issue; in the usual senses of the terms, there is a world of difference between a view's being inconsistent or self-refuting and our being unable to conform to its requirements. (Compare the view that one ought to believe all of the logical consequences of what one believes, which is surely neither inconsistent nor self-refuting.)

We can underscore the nature and importance of this difference by considering Myles Burnyeat's two-pronged attempt to defend a comparable conception of Hume's charge against the Pyrrhonists. He begins by emphasizing Hume's central contention (H) that "[O]ur nature constrains us to make inferences and to hold beliefs which cannot be rationally defended against skeptical objections" (CSL, 118; my emphases) and adds that "It is very clear that Hume appreciated" as well the following:
   [a] ... in the normal case, that which we think we should not believe we do
   not believe: it takes rather special circumstances to make intelligible the
   idea that a man could maintain a belief in the face of a clear realization
   that it is unfounded. [b] If skepticism is convincing we ought to be
   convinced, and [c] that ought to have a radical effect on the structure of
   our thought. (CSL, 119)

The radical effect Burnyeat has in mind is our abandoning the beliefs targeted by the (hypothesized) convincing skeptical arguments,just as we normally abandon a belief in the face of a clear realization that it is unfounded; thus, he takes Hume's appreciation of this line of thought to explain his "press [ing] the Pyrrhonist ... on the question whether he can stop holding the beliefs which his arguments show to be unreasonable" (CSL, 119).

Hume certainly does press the Pyrrhonist on the question whether he can give up his beliefs, but the clear implication of Burnyeat's discussion is that he does so in order to cast doubt on the soundness of the Pyrrhonist's arguments. In my view, this constitutes a serious misunderstanding of Hume's position. The heart of the problem lies in the fact that there is, so to speak, no Hume who holds, or "appreciates," (H) and [a]--[c]; rather, there are Hume-at-backgammon, who appreciates [a], but has no views concerning (H), [b] or [c], and Hume-in-his-study, who holds (H) and [b], but rejects [c]. So, at least, I shall argue.

I take [b] to be a platitude to the effect that if the arguments for skepticism are convincing, then we ought to be convinced by them--convinced, that is, that there are no rational bases for our contingent, objective beliefs. I take it, further, that Hume accepted not only [b], but its antecedent (which is implicit in (H)) and, hence, its consequent; Hume was convinced of the soundness of the skeptical arguments.

[c] is an altogether different matter. How could Hume, the arch-proponent of (H), hold that the skeptic's objections/arguments ought to have the effect of our abandoning those beliefs? As Burnyeat rightly emphasizes, nothing is more central to Hume's thought than the view that our beliefs concerning all manner of contingent, objective matters of fact are not the products of reason but of such things as custom, habit, and imagination. If so, however, then a demonstration that our beliefs are not supported by reason, or even that they fly in the face of reason, should not be expected to have--it is not true that they "ought" to have--the effect of our abandoning them. The idea that the soundness of the skeptic's arguments can be tested by determining whether they have appropriate effects on his, or his audience's, beliefs is entirely alien to Hume's fundamental view of the sources of those beliefs.

But how, then, can we attribute to him [a], which is clearly closely related to [c]? Only, as I have already indicated, by attributing it to Hume-at-backgammon; the "normal case" Burnyeat has in mind is the sort of case considered and responded to by Hume-at-backgammon, understood as employing our quotidian standards--the backgammon-player's standards--of well-founded belief in some fairly ordinary situation. That Hume would find as baffling, perhaps even "unintelligible," as do the rest of us someone who says, for example, "I've recently come to the conclusion that there's not a shred of evidence to support the idea that megadoses of vitamin C prevent colds; I still believe that they do, but I now realize that I'm completely unjustified in doing so."

It may be that Burnyeat has confounded the special case of what I shall call the akratic Pyrrhonist (5)--one whose life belies his claim to live without belief--with that of the sort of individual just imagined. Be that as it may, however, it will be worth our while to consider a bit further how problematic is his attempt to carry over our ordinary assessment of individuals who hold admittedly unjustified beliefs to the special case of the akratic skeptic. Suppose Jones believes that his wallet is in the house, even though he can't find it, since he is quite certain that he hasn't been out all day; suppose further, however, that he suddenly remembers having made a quick run to the corner grocery store this morning, and concludes that it is as likely as not that he left his wallet there, or dropped it on his way home; suppose finally that he nevertheless continues to believe that it is in the house (and add, if you like, that losing the wallet would be disastrous for him). This is the "ordinary" sort of case we have trouble understanding. We (with Hume-at-back-gammon) want to ask, "How can he still believe it's in the house, when he recognizes that it's just as likely to be at the store"? The general type of situation illustrated here is this: someone believes one among a set of recognizably relevant and mutually incompatible candidate answers to some reasonably clear question, even though he acknowledges that the available evidence, while it provides some support for some of those answers, provides no more for the one he believes than for at least one of the others.

Clearly, however, this is not the situation of the akratic Pyrrhonist. His position is surely not unintelligible, and in any case whatever difficulty we may detect in it arises not from his believing one among several alternatives that he acknowledges to be equally likely, but from his holding any contingent, objective beliefs at all, given that the evidence appears to him to provide no support for any such beliefs. Suppose that there is an apple on the table at which the akratic skeptic is sitting, that the table is well lit, he is not blindfolded, and so forth, and that he believes that there is an apple on the table (and suppose we occupy whatever privileged position is necessary in order for us to know this last). We do not find ourselves wanting to ask, "How can he believe that there is an apple on the table, given that it appears to him that there is no basis for that or any other belief?." On the contrary, we think (if we have any relevant thought at all), "Well of course he believes there's an apple on the table; how could he not?" It is not much of an overstatement to say that the only thing we find baffling about him is his skepticism, his insistence that there appears to be no basis for holding any beliefs at all. That he believes that there is an apple on the table contributes not a whit to rendering him unintelligible.

The second charge that Burnyeat takes Hume to bring against the Pyrrhonist is the following:
   [W]hat the skeptic invalidates when his arguments are successful, and hence
   what he would take from us if such arguments could have a "constant
   influence on the mind," is nothing less than reason and belief. (CSL, 118;
   my emphasis)

Now part 2 of section 12 of the Enquiry does indeed begin as follows:
   It may seem a very extravagant attempt of the sceptics to destroy reason by
   argument and ratiocination; yet is this the grand scope of all their
   enquiries and disputes. They endeavor to find objections, both to our
   abstract reasonings, and to those which regard matter of fact and

But Hume sharply divides his discussion into two parts, devoted respectively to abstract reasonings and those concerning matters of fact and existence, and Pyrrhonism figures only in the final paragraph of the latter, the entire substance of which appears in the passage cited earlier. The most exacting scrutiny of that passage will reveal no suggestion on Hume's part that Pyrrhonism undermines reason in the sense of the capacity to make or assess inferences, or detect or demonstrate consistency or inconsistency; nor does Hume ridicule the Pyrrhonist for attempting, incoherently, to use reason to undermine reason, as he earlier, in his discussion of abstract reasoning, ridicules Descartes for his contrasting attempt to use his "faculties" to defend those faculties. Rather, he faults Pyrrhonism only for attempting to undermine our ability to use our reasoning capacity to guide our lives, by undermining any factual beliefs from which we might reason to any useful conclusion.

In sum, Hume does not charge the Pyrrhonist with incoherence, does not accuse him of attempting to offer a reasoned attack on our capacity to reason, and does not find the akratic Pyrrhonist "unintelligible"; rather, he simply contends that, for a reason entirely independent of his principles--namely, that Nature is too strong--the Pyrrhonist is unable to conform to them. In sharp contrast, both Burnyeat and Frede argue that the Pyrrhonist's inability to avoid belief is internal to his position as understood on the received view; they argue that one or another of his contentions is unintelligible except on the supposition that he holds certain beliefs, which contradicts his claim to live without belief, and therefore renders his comprehensive position incoherent. I turn now to those arguments.

2. Achieving Epoche

Frede, in the course of arguing that it is impossible to understand how a Pyrrhonist could ever discover that epoche yields ataraxia, says,
   For the skeptic to make this discovery ... he must either bring it about
   that he is in a state in which he has no beliefs [epoche] or somehow be put
   into such a state. What, though, could bring him into such a state, on what
   grounds could he bring himself into such a state; they could only be
   dogmatic ones. (SB, 185; my emphases) (6)

Of the two possibilities he mentions, Frede rightly focuses on the question how Pyrrhonists could "bring themselves" into a state of epoche, and ignores the possibility that they could he "put into" that state by some unspecified means or external agency; not because the latter possibility is unimaginable, or absurd, but because it does not come to grips with Sextus's own account of how the epoche/ataraxia relation was discovered:
   [T] he Skeptics were hoping to achieve ataraxia by resolving the anomaly of
   phenomena and noumena, and, being unable to do this, they suspended
   judgement. But then, by chance as it were, when they were suspending
   judgment, ataraxia followed, as a shadow follows the body. (PH 1.29, my

On this account, the Skeptics brought themselves into a state of epoche in the most direct possible way--they suspended judgment.

One difficulty with Frede's objection is his presumption that in order to bring oneself into a state of epoche one must do so "on" some "grounds." There is no suggestion in Sextus that the first Skeptics suspended judgment "on the (dogmatic) grounds" that such-and-such was true; rather, he portrays their act as one of simple frustration, comparable to Apelles's throwing a sponge at his painting.

A second difficulty is that Sextus's account of how it was discovered that epoche yields ataraxia is an account of how certain thinkers--at PH 1.12 he refers to them as "talented people"--became Pyrrhonists, not an account of how some thinkers who were already settled Pyrrhonists made that discovery. Consequently, even if the process by which they made their discovery involved reliance on one or another dogmatic belief, that would reveal no defect, to say nothing of any incoherence, in their resulting position.

There are, however, complex and important matters that Sextus glosses over in his laconic account, and in relation to which Frede's charge finds some purchase. To begin with, while the thinkers who gave up on resolving the anomaly with which they were confronted suspended belief about the particular issues that gave rise to it (and experienced ataraxia with respect to them), it does not follow that they abandoned all belief concerning how things are or, therefore, that they experienced ataraxia with respect to how things are in general, or, a fortiori, that they became full-fledged Pyrrhonists on that very occasion. Their journey to Pyrrhonism may have been quite complex, and it may well have included stages at which they engaged in just the sort of process Frede has in mind--stages at which they invoked various grounds on which to suspend belief about certain matters.

Nor does Frede's worry find purchase only in the context of speculative biography. At PH 1.198-203, Sextus addresses at length the facts that not even a settled Pyrrhonist, who has suspended belief about every dogmatic claim he has considered, will have considered every possible dogmatic claim, and that since his past considerations of such claims have yielded only judgments about how they appeared to him at the time, they would need to be reconsidered should the question of their status arise again. Consider the following:
   Both of the slogans "I am non-apprehensive" and "I do not apprehend"
   express a personal pathos, in accord with which the Skeptic declines for
   the presort to take an affirmative or negative position on any of the
   non-evident matters of inquiry, as is evident from what we have previously
   said about the other slogans. (PH 1.201; my emphases)

Among the things he has just said concerning other slogans is the following:
   Whenever the Skeptic says "Everything is indeterminate," he uses "is" in
   place of "appears to me to be," and with "everything" he does not refer
   simply to all there is, but rather to the Dogmatists' non-evident objects
   that are under his consideration;.... [T] he one who is saying "Everything
   is indeterminate" means ... "as related to me" or "as appears to me";
   consequently what is said comes down to this: "All the matters of dogmatic
   inquiry that I have considered appear to me to be such that not one of them
   seems to me superior." (PH 1.198-99, my emphases)

The following question therefore arises concerning how the skeptic is to proceed when a new (old) claim is presented for his (re)consideration: How can he argue that the considerations for and against it are equally balanced (and that he ought therefore to suspend belief concerning it) without appealing to any beliefs?

According to Burnyeat he cannot:
   [Its appearing] to the skeptic that any dogmatic claim has a contrary
   equally worthy or unworthy of acceptance ... is the result of a set of
   arguments designed to show ... that this is in fact the case (my emphases).
   Such arguments can compel him to suspend judgment because they compel him
   to accept their conclusion--to accept, that is, that in each and every case
   dogmatic claims are indeed equally balanced and hence that one ought to
   suspend judgment.... But accepting the conclusion that p is true on the
   basis of a certain argument is hardly to be distinguished from coming to
   believe that p is true with that argument as one's reason.... Certainly it
   appears to him that dogmatic claims are equally balanced, but this
   appearance, so called, ... is only to be made sense of in terms of reason,
   belief and truth--the very notions the skeptic is most anxious to avoid.
   (CSL, 138)

This is a serious challenge, and it will require a good deal of work to see why it is ultimately unsuccessful. (7)

We may begin by considering a crucial premise of Burnyeat's argument, namely, that its appearing to the Skeptic that any dogmatic claim has a contrary equally worthy or unworthy of acceptance is--because, presumably, it could only be--the result of a set of dogmatic arguments designed to show that this is true. But such an assertion raises, only to ignore, the crucial question whether, and if so how, it is possible for a Pyrrhonist to offer genuine arguments without engaging in dogmatism. Most commentators seem to think that there are two unproblematic options when it comes to construing the arguments that fill Sextus's pages. When he seems to offer a straightforward dogmatic argument designed to establish the truth or falsity of something, either he is offering a reductio ad absurdum argument that utilizes as premises only borrowings from one or another dogmatist, or he is simply "recording the appearance" that a certain argument is sound (and its conclusion, therefore, true). It is of course the latter sort of case that concerns us here, since Burnyeat's contention is in effect that such an appearance of soundness can be acquired by Sextus only as a result of his being persuaded of the actual soundness of the argument in question. By raising, if only indirectly, the question of how a Skeptic could acquire an appearance of soundness without endorsing, or being persuaded by, some dogmatic argument, Burnyeat has put his finger on a critical issue.

Let us have before us a schematic example of the sort of argument presumably attributed to Sextus by those who understand him simply to be recording the appearance that some argument is sound: The (dogmatic) argument for the conclusion that the (dogmatic) arguments for and against (dogmatic) claim C are balanced appears to be sound; hence it appears that the arguments for and against C are balanced. Now there is admittedly strong reason to suppose that many of Sextus's arguments must be understood on this model, on pain of attributing to him clearly dogmatic arguments. Consider, for example, this argument from early in the Outlines:.
   [W]hen somebody brings up an argument we cannot refute, we say to him:
   "Just as [i] before the birth of the person who introduced the system which
   you follow, the argument supporting that system did not yet appear sound
   although [ii] it really was, so also [iii] it is possible that the opposite
   of the argument you now advance is really sound despite [iv] its not yet
   appearing so to us, and hence [v] we should not yet assent to this argument
   that now seems so strong." (PH 1.33-34)

This looks for all the world like a dogmatic argument, offered in Sextus's own voice, for the conclusion that he should not assent to a certain argument (alternatively, an argument schema, the premises of any instance of which he is saying he is prepared to endorse); certainly its premises are not borrowed from any dogmatist. How, then, are we to avoid concluding that he here incomprehensibly violates the very principles he has just set forth by offering a dogmatic argument? Only, so far as I can see, by invoking his overarching proviso at PH 1.4, and taking his argument to have something like the form indicated just above. It should be understood, in other words, to run as follows: A certain (dogmatic) argument for the conclusion that we should not yet assent to argument X appears to be sound; hence, it appears that we should not yet assent to X.

Further, even in his reductio arguments in which he borrows premises from various dogmatists, Sextus almost invariably employs, in addition, premises that appear to be his own. To take but one example, in the course of rebutting Democritus's assertion that human beings are "what we all know," he says,
   [S]ince ... no human being is known by all human beings, no one, according
   to him, will be a human being. (PH2.23, my emphasis)

Here it seems that we must understand Sextus to be asserting, of the italicized premise, which seems clearly to be his own, only that it appears to be true, which means that, once again, his argument should be understood as follows: "The following argument appears to be valid, and its premise appears to be true: `No human being is known by all human beings; therefore, if human beings are what we all know, there are no human beings;' hence, it appears that if human beings are what we all know, there are no human beings." (8)

There are, then, strong reasons for supposing that this is how we should construe all those of Sextus's arguments that seem to be offered "in his own voice." But there are severe difficulties with doing so, the first of which is implicit in Burnyeat's argument. An essential part of that argument can be expressed more fully as follows: In many, if not most, cases it would be absurd to construe a report by Sextus that some argument appears to be sound as a report of a more or less instantaneous "appearance of soundness" on a par with a report, say, of an appearance of being written in black ink. Rather, an appearance of soundness must be the end product of some sort of temporally extended process by which one moves from some premise-like considerations to some conclusion-like terminus via some inference-like steps. But no intelligible account of such a process can be given except in terms of normal dogmatic arguments from premises to conclusion, that is, arguments having something like form (A): p; p implies q; therefore, q.

The second serious difficulty is that the form of argument that I have suggested we are almost forced to attribute to Sextus--It appears that p; therefore, it appears that q--is blatantly invalid. Casting such arguments in a form parallel to (A) yields the invalid (B): It appears that p; it appears that p implies q; therefore, it appears that q.

If we are going to find a cogent form of reasoning available to the Pyrrhonist--one that does not involve normal dogmatic argument--we will have to look further afield. One possibility is suggested by the following consideration. If a dogmatist were challenged to explain why he believes q, he might well say that he does so because he believes both p and p implies q. In so saying, he would not be endorsing, or instantiating, the invalid form of argument (C): I believe that p; I believe that p implies q; therefore, I believe that q. He would not be making an inference at all, but offering his reasons for believing q.

Might we then understand Sextus, when he appears to be offering straightforward arguments, to be offering instead his reasons for doing something analogous to believing? There is something to this suggestion, and we shall eventually return to it, but for the moment we need only notice that the most obvious analogy will not do. The obvious analogue to believing q would be assenting to "It appears that q," and it is far from clear that for Sextus the idea of having, or offering, reasons for such assent would be so much as intelligible; such assent is, according to him, involuntary; under appropriate conditions, one simply cannot fail to assent to such judgments. (9)

We need yet another perspective on the problem. The account I shall propose may not be entirely new to the literature, but I hope to show, by making it fully explicit, that there is indeed a significant form of reasoning and argumentation available within the constraints of the Pyrrhonian framework. I shall begin by offering my conception of the structure of a typical fully explicit Pyrrhonian argument conceived on the model of (B):

(i) It appears to me that it appears to S that p; (10)

(ii) It appears to me that it appears to T that not-p;

(iii) It appears to me that all appearances are equally authoritative;

(iv) It appears to me that there are no other relevant considerations;


(v) It appears to me that the considerations in favor of p and not-p, respectively, are equally balanced;

(vi) It appears to me that if the considerations in favor of p and not-p, respectively, are equally balanced, then one should suspend judgment concerning p;


(vii) It appears to me that one should suspend judgment concerning p.

Of course both inferences here--of (v) and (vii)--are invalid. But consider now a modern philosopher who regards absolute certainty as a will-o'-the-wisp and the possibility of error as omnipresent, or one who acknowledges that none of his statements is immune to revision (meaning that none is free of the possibility of error), or a Bayesian who insists that any contingent proposition should be assigned a nonzero prior probability. How are we to understand an argument of form (A), whose premises and conclusion are asserted without qualification, when it is offered by such a philosopher? By his own admission, after all, his premises may--no matter how remote the prospect--be mistaken, which seems to suggest that he has no business asserting them categorically.

Suppose, in particular, that Quine's Word and Object (11) had begun as follows: "No judgment is immune to revision. As regards none of the things that I am about to say, therefore, do I firmly maintain that matters are absolutely as stated, but in each instance I am simply stating what I now judge to be true." This would have had no effect whatever on the proper understanding of Quine's text, but would merely have made the implicit explicit. But that does not mean that the typical Quinean argument has the invalid form (D), "I judge that p and that p implies q; therefore, I judge that q"; rather, it has form (A): "p; p implies q; therefore, q." As Quine says, "Within our own total evolving doctrine, we can judge truth as earnestly and absolutely as can be; subject to correction, but that goes without saying" (WO, 25; my emphasis).

Well, it goes without saying for some, but hardly for all; what we may call the extreme dogmatist is precisely one who regards his views and arguments as immune to criticism, as not subject to correction (perhaps because he thinks they have been vouchsafed him by some deity, which belief he also takes to be immune to criticism). I take it that both Quine and the extreme dogmatist offer arguments of form (A), and that the relevant difference between them lies in their respective attitudes toward those arguments. This suggests the possibility of seeing the Pyrrhonist's distinctive relationship to his arguments analogously. Whereas Quine believes both that his arguments are sound and that that belief is fallible, the Pyrrhonist's attitude toward his arguments is more complex. Although he has no beliefs at all about their soundness, he puts them forward for consideration, that is, he offers them, because they appear to him to be sound; he withholds belief in, and offers no defense of, their soundness because, inter alia, it also appears to him that an argument's appearing to him to be sound constitutes no reason whatever for supposing it to be so (and because it appears to him that there is no other reason for supposing it to be so). (12)

The upshot is this. Initially, at least, we should see the Pyrrhonist as simply offering, or putting forward, straightforward arguments such as the following, while claiming nothing for them other than that they appear to be sound, that is, that each premise appears to be true and each inference valid:

(i') It appears to S that p;

(ii') It appears to T that not-p;

(iii') All appearances are equally authoritative;

(iv') There are no other relevant considerations;


(v') The considerations in favor of p and not-p, respectively, are equally balanced;

(vi') If the considerations in favor of p and not-p, respectively, are equally balanced, then one should suspend judgment concerning p;


(vii') One should suspend judgment concerning p.

But of course this cannot be the end of the matter, since we must now ask what possible point there could be to merely "putting forward" such arguments. There are, in fact, several. First, it is here that the analogue of "giving reasons" comes to the fore. The difficulty we noted earlier was that whereas we are often in a position to cite our reasons for believing or judging that p, it is far from clear that Sextus would find intelligible the idea of having reasons for assenting to "It appears that p." But although there are no reasons for which he assents to it, Burn-yeat is right to insist that in this sort of case his assent comes about as the result of a certain temporally extended process, and that there is a reason that--an explanation of the fact that--he has come to assent to it. Thus, I suggest that one point of his putting forward such an argument is to offer such an explanation--an account, which appears to be correct, of how it has come to appear to him that p is true. We should see him as reporting, for example, that it appears to him that its appearing to him that (vii') is true is a result of its appearing to him that (vii') follows from (v') and (vi'), which appear to him to be true.

Second, by putting forward such arguments Sextus places his dogmatic opponents in the position of having to find some flaw in them if they are not to be forced to concede their soundness, hence to consider, at least, suspending judgment. Finally, putting forward his many and multifarious arguments in written form serves the self-directed purpose of trying to arrange the large and heterogeneous collection of arguments and theses that appear to him to be, respectively, sound or true into a coherent whole.

Returning now to Burnyeat, his charge was that it can appear to Sextus that the considerations for and against any dogmatic claim are equally balanced only because certain dogmatic arguments compel him to accept the conclusion that they are equally balanced. But if my proposal is right, we should see Sextus as reporting that it appears to him that the reason that it appears to him that (v') above is true is not that he has been compelled by argument to accept it, but rather that it appears to follow from (i')-(iv'), all of which appear to be true; and that the reason that it (subsequently) appears to him that (vii') is true is that it appears to follow from (v') and (vi'), both of which (at that point) appear to be true. (13) Finally, we should note that since (vii') appears to Sextus to be true, and since he is committed to guiding himself by appearances alone, he now has everything he needs for suspending judgment concerning p.

Three other advantages of this proposal are worth noting. First, it offers us a new way of understanding Pyrrhonian arguments for the possibility of life without belief. Frede, having noted that we cannot justifiably take such arguments to be dogmatic, suggests that in offering them "the Skeptics are simply following their usual strategy of providing equipollent arguments on both sides of every issue" (184). But we are now in a position to see them as "taking a position" on the issue without making any dogmatic commitment; we can understand them to be (i) reporting that acting without belief appears to be possible, (ii) reporting that there appear to be sound arguments for its possibility, and (iii) putting forward those apparently sound arguments for consideration.

Second, the proposal enables us to respond to a range of charges of self-refutation brought by the Stoics. (14) These are to the effect that the Skeptic refutes himself if he (a) uses a criterion to assert that there is no criterion, (b) offers an argument (one sort of "sign") for the nonexistence of signs, (c) purports to prove that there is no proof, or (d) cites a reason why there is no such thing as a reason. The idea is not that the Skeptic's views are falsified by virtue of their own content (which one might consider the "pure" form of self-refutation), but rather, in Burnyeat's words, that "it is the Skeptic's undertaking to establish his thesis by reason that falsifies it, for his thesis is that there is no such enterprise to undertake." (15)

On my view, however, the Skeptic does not undertake to establish any theses, and Burnyeat, in what seems to be partial agreement, suggests that the Skeptic's reply to such charges is that "within the dispute the Skeptic attempts a straightforward refutation of notions like criterion and proof ... [but] he is not actually committed to his conclusion: it is enough for him to have shown it to be as well supported as his opponents' view, so that the right attitude is to suspend judgment." (16) It is the phrase `within the dispute' that suggests that this proposal is in partial agreement with my own. However, on my proposal Sextus does not claim to have shown his conclusions to be supported to any degree at all; rather, he merely puts forward arguments for them that appear to him to be sound, and that appear to balance the arguments for competing conclusions incompatible with them. The point is to challenge his opponent to explain how he can continue to maintain his position in light of the existence of such arguments--arguments to whose soundness he is presumed to be committed.

In the end, Burnyeat concludes that the Skeptic's position is self-refuting in the sense he has characterized: "[W]hat the Skeptic is saying is: `There are no true propositions of the form "p because q" because'...." (17) On my proposal, however, what the Skeptic is saying is, "It appears to me that there are no true propositions of the form `p because q', and that the reason that this appears to be so is that there appears to be a sound argument for the proposition that there are no true propositions of the form `p because q'." Not only would he not say that there are no true propositions of this form, but he would insist that it appears to him that the fact that there appear to him to be no such true propositions is no reason whatever for supposing that there are none.

The final advantage of my proposal is that it makes clear how the Pyrrhonists could "continue to inquire"; they would be perfectly free to consider new arguments against things that appeared to them to be true--for example, that a life without belief is possible--including attempts to show that various apparently sound arguments supporting such theses are in fact flawed. If such a counter-argument came to appear to them to be sound, they might find themselves in any of several new positions: it might continue to appear to them that a life without belief is possible, even though there appeared to be no sound arguments for its possibility; or such a life might fail to appear to be possible, and also fail to appear to be impossible, just as the number of stars neither appears to be even nor appears to be odd; or, of course, it might eventually come to appear to them that such a life is not possible, which eventuality might or might not be a result of there coming to appear to them to be a sound argument for its impossibility.

3. Achieving Ataraxia

But we have not yet plumbed the depths of Burnyeat's objections concerning epoche and ataraxia. For Sextus, epoche is not an end in itself, but an apparent means to the apparently desirable state of ataraxia, and even if he is in a position to provide himself with everything he needs for suspending belief, it does not follow that we can make sense of the idea that he would thereby achieve ataraxia. Once again, Burnyeat does not think we can:
   [I]f tranquillity is to be achieved, at some stage the skeptic's questing
   thoughts must come to a state of rest or equilibrium. There need be no
   finality to this achievement; the skeptic may hold himself ready to be
   persuaded that there are after all answers to be had.... But ataraxia is
   hardly to be attained if he is not in some sense satisfied--so far--that no
   answers are forthcoming, that contrary claims are indeed equal. And my
   question is: How can Sextus then deny that this is something he believes?

   I do not think he can. Both the causes (reasoned arguments) of the state
   which Sextus calls appearance [namely, the appearance that no answers are
   forthcoming] and its effects (tranquillity and the cessation of emotional
   disturbance) are such as to justify us in calling it a state of belief.
   (CSL, 139-40)

The previous section was devoted to showing that the causes (reasoned arguments) of "the state which Sextus calls appearance" do not justify us in calling it a state of belief; in this section we shall consider whether its effects do so.

It is not entirely clear what stance Burnyeat takes in issuing his challenge to Sextus on this score. On the one hand, he appears to acknowledge at least the possibility that something other than belief might have the effects in question; that is, he does not seem to be importing into the discussion, and relying on, some theory of mental states according to which a belief just is whatever state has the sorts of (causes and) effects in question. This is surely the stance he must take if he is not simply to beg the question against Sextus and, indeed, Pyrrhonism ab initio. On the other hand, he also seems summarily to reject that possibility, since in place of any argument against it he offers only the essentially rhetorical question, "How can Sextus then deny ...?"

Whatever his stance may be, the only question he raises that is of interest in connection with Pyrrhonism is whether, given that it is theoretically possible for ataraxia to be brought about by something other than a belief state, Sextus can provide any coherent and credible account of what that other something might be. I shall argue that he can. My only reservation about my proposal is that Sextus himself does not clearly articulate it, and even contributes in a small way toward obscuring it; both of its crucial elements, however, are prominent in his work.

I take it to be a desideratum, though not a strict condition of adequacy, for any account of how Sextus comes to experience ataraxia that it explain as well how his Skeptical progenitors--the first Skeptics--did so. Hence I shall begin by considering his claim that they experienced ataraxia as a result of suspending judgment (PH 1.26-29). (Here he strongly suggests that ataraxia resulted solely from suspending judgment, or at least that no other comparably important factors were at work, and it is this suggestion that, in my view, tends to obscure the full account.)

Two other elements in Sextus's sketch of the early Skeptics need to be borne in mind. First, they had originally hoped to achieve ataraxia by determining the truth about the issues that concerned them (PH 1.12, 26), but did not succeed in determining it, either before or after suspending judgment. (Note the role of their own psychological hypothesis that discovering the truth would bring about the desired state.) Second, these thinkers were clearly open-minded inquirers before suspending judgment (more cautiously, we have no reason to suppose that they were not), and, given Sextus's very definition of a Skeptic, they must be presumed to have remained so after suspending judgment. But then the absence of ataraxia prior to their suspending judgment can be explained neither by their not having determined the truth nor by their having been open-minded inquirers (nor by the combination of the two). What, then, could explain it? The answer, clearly, must lie in something that was true of them before, but not after, they suspended judgment, and the obvious candidate is their holding those beliefs--those judgments, whatever they were--the giving up of which constituted their suspending judgment. (This is presumably the explanation for Sextus's having attributed their experiencing ataraxia solely to their suspending belief.)

But how could their holding beliefs account for their failing to enjoy ataraxia? Why would they not have regarded their beliefs as being the very truths whose discovery they had hoped would yield ataraxia (and why did they not, therefore, already enjoy it)? As it happens, Burnyeat describes the relevant phenomenon vividly, though in connection with quite a different issue. Speaking of Sextus and his fellow settled Skeptics, he says that we cannot avoid the conclusion that their working through the Skeptical arguments left them holding the belief that no answers were forthcoming by supposing instead that it left them in a state of bafflement:
   Bafflement could be the effect of arguments for and against; you are pulled
   now this way, now that, until you just do not know what to say.... The
   problem is to see why this should produce tranquillity rather than acute
   anxiety. (CSL, 139)

There can be little doubt that the first Skeptics, prior to becoming Skeptics, were pulled now this way, now that; more importantly, however, there can be no doubt at all that they held beliefs--made judgments. Hence, it is at least plausible to see them not only as having been pulled this way and that, but also as having believed first this, then that, in their efforts to arrive at the truth. But then they might well, at some point, have come to feel anxiety about their latest--that is, their contemporaneous--beliefs. Why should they have expected those beliefs to fare any better than had their predecessors? It is more than plausible, it seems to me, that it was precisely this sort of anxiety that led these thinkers to suspend judgment. If that hypothesis is right, then what was responsible for their anxiety was the combination of their holding whatever beliefs they held at the time and their recognition of the possibility that those beliefs might suffer the fate of earlier ones--rejection in favor of yet further apparently superior beliefs.

Here, then, is the hypothesis I propose on Sextus's behalf: his predecessors' experiencing ataraxia was the result of (at least) two things: their abandonment of their anxiety-producing beliefs (their suspending judgment) and their retention of their openness to the possibility that they might discover the truth (their not leaping to the conclusion that no answers were forthcoming). The reason that this combination was capable of producing ataraxia is this: First (the element that Sextus focuses on), both the anxiety produced by their vulnerable beliefs and the discomfiting prospect of experiencing further anxiety if they were to continue their practice of adopting beliefs were removed; second, their prospects for discovering the truth, and thereby achieving ataraxia, were retained. To put it slightly differently, we should take it that it appeared to them that: Either the truth could be determined or it could not; if it could, then since they remained open-minded they might yet succeed in determining it; if it could not, then by suspending judgment they had at least avoided the danger of mistakenly thinking that they had determined it (which, I take it, is a part, if not the whole, of the "precipitancy" Sextus attributes to the Dogmatists), as well as the anxiety associated with being aware of the possibility that they had succumbed to that danger. It is quite plausible, it seems to me, that this complex appearance should have induced ataraxia. (18)

The case of Sextus, or of any settled Skeptic, is of course quite different. He suspends belief not out of anxiety or frustration, but as a result of having brought to bear on the question at hand one or more of the vast battery of arguments of his tradition, with the result that it appears to him that no answer to the question is forthcoming. Burnyeat's question is how that process could bring about ataraxia except on the assumption that the arguments have led the thinker to believe that no answer is forthcoming.

The answer, I submit, is the very one we have just encountered. We should take it that it appears to Sextus that: Either the truth can be determined or it can not; if it can, then since he remains open-minded he may yet succeed in determining it; if it can not, then by suspending judgment he has at least avoided the danger of mistakenly thinking that he has determined it, as well as the anxiety associated with being aware of the possibility that he has succumbed to that danger.

It is perhaps worth underscoring the diametrical opposition between Burnyeat's and my proposals: whereas he locates the genesis of ataraxia in belief, I, with Sextus, locate it (in part) in the absence of belief. In my view, the belief he attributes to Sextus--that contrary claims are equally balanced--would have been a source not of ataraxia but of anxiety; the fact that such propositions--versions of the Skeptic slogans--were as threatening as any other nonevident propositions in this respect helps to explain Sextus's emphatic insistence on their self-applicability and his consequent suspension of belief concerning them (PH 1.187-209). (19)

Burnyeat has lodged the following preemptive objection to any such account as the one I have proposed:
   The skeptic goes on seeking ... in the sense that he continues to regard it
   as an open question whether p or not-p is the case, at least for any
   first-level proposition concerning real existence. But this should not mean
   that he is left in a state of actually wondering whether p or not-p is the
   case, for [a] that might induce anxiety. Still less should he be wondering
   whether, in general, contrary claims are equally balanced. For if it is a
   real possibility that they are not, that means it is a real possibility
   that there are answers to be found; and [b] it will be an immense worry to
   him, as it was at the very beginning of his skeptical education, that he
   does not know what these answers are. (CSL, 139)

There seems little to be said for [a]. Why should mere wondering whether p is the case (tend to) induce anxiety? (Wondering whether there is life on other planets, for example, surely has little tendency in and of itself to produce anxiety.) It might induce one to inquire into the matter, to consider relevant arguments, and the like, but why should such activities, or dispositions to engage in them, induce anxiety? The only support Burnyeat offers for this contention is to be found in [b], whose "the very beginning ..." I construe, for lack of any discernible alternative, as a reference to the condition of the first Skeptics prior to their suspending judgment. But as I have already indicated, it seems plausible that the source of their "immense worry" was their having apparently vulnerable beliefs rather than their failing to have knowledge. If so, then their experience provides no basis for supposing that lack of knowledge is, in and of itself, a source of anxiety.

According to Burnyeat, only Sextus's belief--and not the mere non-epistemic appearance--that no answers are forthcoming is capable of explaining his ataraxia. On the account I have offered, however, neither that belief nor that appearance explains it; that is, "the state which Sextus calls appearance"--whatever it may be--does not produce ataraxia. Simplifying somewhat, that state is produced solely by the combination of his freedom from belief and his open-mindedness.

One sympathetic to Burnyeat's perspective might raise the following objection: At best, you have shown only that its appearing to Sextus that no answers are forthcoming is not the proximate cause of his ataraxia; but isn't it the proximate cause of his lack of belief (by virtue of being the proximate cause of his suspending belief) which in turn is (one part of) the proximate cause of his ataraxia? No; that appearance does not cause (or "compel") him to suspend belief; it merely provides the occasion for him deliberately to suspend belief in accordance with his practice of guiding his actions by the appearances (in this case the appearances that no answers are forthcoming and that under such conditions one ought to suspend judgment). (20) This is the locus of the difference between him and his predecessors; whereas they suspended judgment out of frustration and anxiety, he does so on principle, out of his determination to be guided by the appearances.

Finally, one to whom it seems that belief must at some point enter into this process might try to push the argument back one step further, and ask how the arguments responsible for its having come to appear to Sextus that no answers are forthcoming could produce that effect unless he believed them to be sound. But of this matter we have already treated.

4. Assent and Belief

Having argued that its causal role justifies us in regarding the state of its appearing to one that p as the state of one's believing that p (where p is a philosophical proposition), Burnyeat goes on to argue that there is a logical connection between the two:
   The source of the objection I have been urging is that the skeptic wants to
   treat "It appears to me that p but I do not believe that p," where p is
   some philosophical proposition such as "Contrary claims have equal
   strength," on a par with perceptual instances of that form such as "It
   appears (looks) to me that the stick in the water is bent but I do not
   believe that it is." The latter is acceptable because its first conjunct
   describes a genuine experience--in Greek terms, a pathos, a phantasia,
   which awaits my assent. And it is important here that assent and impression
   are logically independent. For they are not independent in the
   philosophical case. In the philosophical case, the impression, when all is
   said and done, simply is my assent to the conclusion of an argument, assent
   to it as true. (CSL, 140; my emphases)

A defense of the concluding identity claim here could of course be based on various sorts of considerations, including identity of causal roles, but clearly Burnyeat's contention is, at a minimum, that in the philosophical case "It appears to me that p" is not logically independent of "I do not believe that p"; rather, the former implies the denial of the latter.

There are numerous difficulties with this argument. To begin with, it seems that many philosophical, or theoretical, cases share the characteristic that Burnyeat takes to be distinctive of perceptual ones. Consider, for example, the use of paradoxical arguments to bring out the theoretical disasters awaiting those who inadvertently employ illicit self-reference, or attempt division by zero. Such arguments can appear to be sound even though it is obvious that they are not. More directly relevant, perhaps, are cases in which philosophers and other theoreticians struggle to detect flaws in arguments that appear to them to be sound, but that militate against their own convictions. In such cases the appearance, or impression, of soundness is clearly not tantamount to assent. In short, whatever the role of phantasia in perceptual cases, there seems to be something that corresponds to it in philosophical, or theoretical, cases. Further, given that such a distinction clearly obtains in cases where appearance and belief do not coincide, the burden is on Burnyeat to show that it does not obtain generally, just as he implicitly supposes that the corresponding distinction obtains generally in perceptual cases, that is, both in ordinary cases in which phantasia and belief coincide, and in unusual cases in which they do not.

A second reason for doubt runs in the opposite direction; many perceptual cases appear to share the characteristic that Burnyeat takes to be distinctive of philosophical ones, namely, that one's impression that p just is one's assent to p, "assent to it as true." If I have become familiar with the stick-in-water illusion, then it is simply false that when I look at a straight stick immersed in water I will have the impression that the stick is bent; rather, I will typically have the impression that it is not bent, since it will look just the way straight sticks normally look when immersed in water.

Equally problematic is the suggestion that the reason for the alleged asymmetry between perceptual and philosophical instances of "It appears that p" is that any instance of the former, such as "It appears to me that the stick in the water is bent," describes a pathos. It is one thing to suppose that the truth of such a statement depends on nothing other than the nature of the subject's current pathe, but notoriously quite another to suppose that the reason for this is that it describes them in any straightforward sense. Thus, one would be quite unwarranted, for example, in attributing to Burnyeat the idea that a subject to whom it appears that the stick is bent has a pathos that is bent, or that a subject experiencing the Muller-Lyer illusion, in which two lines of equal length appear to be of unequal length, has a pathos having two unequal lines as constituents. But without having in hand at least the rudiments of an alternative account of what it might be for "it appears to me that p" to describe my pathos, we are in no position to judge whether an adequate account would underwrite Burnyeat's wanted distinction.

Finally, appearances do not in any case divide at all neatly into perceptual and philosophical (or theoretical) cases. How, for example, are such remarks as "His playing appears to be effortless," "It sounds like Mozart" or Paul Grice's "The pudding looks indigestible" to be classified?

At most, however, what these considerations show is only that Burnyeat has failed to delineate clearly a certain alleged perceptual/philosophical asymmetry, not that the conclusion he intends it to support is mistaken. That conclusion is that with respect to such central Pyrrhonian "slogans" as that contrary claims have equal strength, assenting to their appearing to be the case implies believing them to be true--that is, that "It appears to me that contrary claims have equal strength" implies "I believe that contrary claims have equal strength."

Now we have already considered and rejected (under "Achieving Ataraxia") one argument that, as I understand him, Burnyeat takes to support this conclusion, and I shall consider and reject yet another below (under "Detachment from Oneself"). But I want to note here a significant obstacle that he does not address, but would need to overcome, in order to succeed. It is of the utmost significance that a Pyrrhonian assents to one appearance that is utterly alien to virtually all participants in ordinary discourse, namely, that how things appear to him--or to anyone else--is no indication whatever of how things are. Thus, Burnyeat would need to show that Sextus could not consistently maintain the following position: "It appears to me that contrary claims are equal; but I do not mean to imply that I believe that contrary claims are equal, since it appears to me that how things appear to me is no indication at all of how things are; hence I do not believe, in spite of the present appearance, that contrary claims are equal (or, of course, that they are not)." The nature and magnitude of this challenge are best appreciated by framing it in Grice's terms. (21) What Burnyeat must show, on the supposition that someone who says "It appears to me that p" thereby implies that he believes that p, is that the implication is logical or semantic (what Grice calls "conventional"), that is, that what he says implies that he believes p. What he must rule out is the alternative possibility that the implication is nonconventional, that is, that it is either the speaker or his saying what he says that implies that he believes p. In short, he must show that the implication that the speaker believes p is not "cancelable"--that the speaker cannot remove it simply by saying that he does not mean to imply it.

Burnyeat sometimes goes so far as to suggest that a Skeptic cannot consistently acknowledge even an inclination to believe, as, for example, in the following:
   The skeptic wants to say something of the form "It appears to me that p but
   I do not believe that p," with a nonepistemic use of "appears," but it
   looks to be intelligible only if "appears" is in fact epistemic, yielding a
   contradiction: "I (am inclined to) believe that p but I do not believe that
   p." (CSL, 138; my emphases)

With these comments Burnyeat raises, though perhaps inadvertently, a question I want briefly to pursue, namely, whether the Skeptic could freely admit an inclination to believe. Let me begin by noting the oddity--given that there is presumably some difference between believing and being inclined to believe--of his describing the assertion he attributes to Sextus as a contradiction. The most likely explanation of this, it seems to me, is that Burnyeat here understands "I am inclined to believe p" in something like the way a typical utterance of "I am inclined to agree with you," said in response to some assertion, is to be understood. Such a remark usually expresses either outright agreement or a belief that the assertion is very likely true. If this is how he is to be read, then his contention is that Sextus's "It appears to me that p" implies that he believes p to be true or at least likely, either of which counts as an epistemic use of "appears." In that case, however, his contention is subject to the objection we have just canvassed.

It is at least possible that Burnyeat has in mind something quite different, namely, that Sextus could not consistently acknowledge that he has inclinations, in the sense of tendencies or dispositions, to believe, which tendencies he successfully resists. It is one thing for Sextus to claim to abstain from belief, but it would be a radical step beyond that for him to claim to be free even of any (natural) tendencies or dispositions to believe, and thus free of any need to overcome them. Whether or not this is what Burnyeat has in mind, however, it is worth at least brief consideration and, as we shall see, he has himself indicated lines along which Sextus could respond to it.

I shall come at the issue indirectly, by considering first a difficulty in understanding Sextus's discussion of his "fourfold regimen of life." At PH 1.23-24, he says,
   Holding to the appearances, then, we live life without beliefs but in
   accord with the ordinary regimen of life, since we cannot be wholly
   inactive. And this ordinary regimen of life seems to be fourfold: one part
   has to do with the guidance of nature, another with the compulsion of the
   pathe, another with the handing down of laws and customs, and a fourth with
   instruction in arts and crafts. Nature's guidance is that by which we are
   naturally capable of sensation and thought; compulsion of the pathe is that
   by which hunger drives us to food and thirst makes us drink; the handing
   down of customs and laws is that by which we accept that piety in the
   conduct of life is good and impiety bad; and instruction in arts and crafts
   is that by which we are not inactive in whichever of these we acquire. And
   we say all these things without belief. (My emphases)

We can make a very rough distinction between the first and second pairs of elements of Sextus's "ordinary regimen of life": the former seem concerned with aspects of life over which we have relatively little control, the latter with ones over which we have a good deal of control. To some extent, Sextus's own terminology reflects this; with the first pair, he speaks of (sheer) capability (of sensation and thought) and compulsion, and here it would not be unreasonable to read him as saying that we are by nature bound to sense (perceive) and think, to seek food and water. But with the second pair, he speaks of accepting and being active, and it seems clear that we are not bound to accept the local laws and customs or to practice our arts in accordance with the instruction we have been given; rather, it seems that we have some significant latitude in these matters.

Now the above passage is offered as an explication of what Sextus has just said at PH 1.21-22:
   [T]he criterion of action [of the Skeptic Way], by attention to which in
   the conduct of daily life we do some things and not others, ... is the
   appearance--in effect using that term here for the phantasia--for since
   this appearance lies in feeling and involuntary pathos it is not open to
   question. (My emphasis)

Focusing on the second pair of Sextus's four elements, the question is this: If the criterion of conduct lies in feeling and involuntary pathos, how are we to understand, for example, his assertion that the handing down of customs and laws plays a role in determining conduct? I propose the following: we should understand him as saying, roughly, (i) that it appears to him that conforming to what appear to be the laws and customs appears to be good, and violating them bad, and (ii) that this leads him to act in a certain way on a given occasion only if acting in that way appears to conform to what appear to be the laws and customs. (Compare his remark on the art of medicine: "[T] he Methodic physician is led by the pathe to what is appropriate" (PH 1.238; my emphasis).) I shall not pursue the interpretive question further, since my principal interest is in some comments that Burnyeat has made on the matter, and these, I think, could readily be recast in terms of my proposed reading. He understands Sextus to be saying,
   [O]f course the skeptic will have his preconceptions, the result of being
   brought up in certain forms of life, and these will prompt him to act one
   way or the other. But the point is that he does not identify with the
   values involved. He notes that they have left him with inclinations to
   pursue some things and avoid others, but he does not believe there is any
   reason to prefer the things he pursues over those he avoids. (CSL, 132; my

If we bring this perspective to bear on the special case of believing, and substitute "not accepting the truth-indicativeness of the appearances" for "not identifying with the values involved," it seems that Burnyeat should accept the following:
   Of course the skeptic will have a (natural) inclination to believe p
   whenever it appears to him that p, but the point is that he does not regard
   its appearing to him that p, or his having a (natural) inclination to
   believe p, as any indication that p is true.

At any rate, the burden is on him to explain why believing and pursuing should not be treated in parallel fashion.

Returning, then, to his original charge, he has provided no basis for regarding "I am inclined to believe that p but I do not believe that p" as inconsistent, or for thinking that the Skeptic could not consistently acknowledge a natural inclination to believe. The Skeptic could quite consistently maintain that "Although it appears to me that p, and--apparently by virtue of that fact and my own constitution--I do have a (natural) inclination to believe p, I do not believe p," to which he might well add, "A principal point of the Tropes and other Skeptical arguments is precisely to enable one to resist such inclinations."

5. Detachment from Oneself

A second argument concerning assent, also due to Burnyeat, is to the effect that the Pyrrhonian's refusal to consider his own appearances as more weighty or authoritative than others' constitutes a kind of "detachment from himself." The first of two versions of this argument runs as follows:
   [The skeptic] notes the impressions things make on him and the contrary
   impressions they make on other people, and his own impressions seem to him
   no stronger, no more plausible, than anyone else's.... When a thing appears
   in a certain light to him, that no more inclines him to believe it is as it
   appears than would the fact of its so appearing to someone else.... Thus
   the withdrawal from truth and real existence becomes, in a certain sense, a
   detachment from oneself.... [I]t is here that I would locate the ultimate
   incoherence of the skeptical philosophy. (CSL, 129)

Now first, if the argument of the previous section is right, then there is no obstacle to Sextus's acknowledging that its appearing to him that p does incline him to believe that p, while claiming to resist that inclination successfully. Second, he is not in a position to be influenced directly by both his own and others' appearances, since how things appear to others is not accessible to him in the same way as how things appear to him; hence, Burnyeat's claim that the two are equally efficacious (or inefficacious) in inclining him toward belief finds no purchase. Of course, it can appear to him that it appears to others that p, but that is quite another matter, and in any case it is surely true that such a (nested) appearance will do less in the way of inclining him to believe p than will its appearing to him that p.

But all of this fails to come to grips with what I take to be Burnyeat's real point, which we may express counterfactually: even if, somehow, the fact of its appearing to others that p were as directly available to Sextus as is its appearing to him that p, the latter would "seem to him no stronger, no more plausible," than the former. But even this does not get Sextus quite right; even under the envisaged counterfactual supposition, the question for him would not be whether his own impressions "seemed stronger or more plausible," but whether there would be any reason to suppose that they are so. But then the burden of proof would be on Burnyeat to address the following question: What exactly is supposed to be wrong with regarding others' impressions as being on a par with one's own with respect to their strength or plausibility, that is, the likelihood of their being an indication of how things are? Surely it takes little in the way of either humility or objectivity to make one willing to acknowledge that there is no reason to suppose that one's own appearances are, as a general matter, more likely than others' to be reliable indicators of truth; a fortiori, it is hard to see in such an acknowledgement any "radical detachment from oneself."

It may be that in advancing this first version of the argument Burnyeat has in mind the second as well, which he offers in the course of expanding on the consequences of his "literal inconsistency" charge against Sextus:
   [T]he danger of allowing talk about appearances or impressions of thought
   [is that] it comes to seem legitimate to treat states which are in fact
   states of belief ... as if they were independent of assent in the way that
   sense-impressions can be. For if ... the philosophical impression includes
   assent, it ought to make no sense for the skeptic to insist that he does
   not assent to it as true.... If the skeptic does insist, if he refuses to
   identify with his own assent, he is as it were detaching himself from the
   person (namely, himself) who was convinced by the argument, and he is
   treating his own thought as if it were the thought of someone else. (CSL,
   140; my emphasis)

The trouble with this version is that it depends on the supposition that assenting to an impression-that-p, where p is a philosophical proposition, includes assenting to the truth of p. If that supposition were true, then of course the Skeptic would be mistaken in claiming to have the one without the other. But I have contended that Burnyeat's argument for that supposition fails, and if it is false, then the charge that the Skeptic "refuses to identify with his own assent" to the truth of p, and thereby detaches himself from himself, is insofar baseless, since there need not be--and according to the Skeptic, of course, there is not--any such assent with which to identify or refuse to identify himself. (22)

6. Assenting to Appearances

Benson Mates has raised the last of the questions we shall consider: whether, from his own perspective, Sextus can consistently assent even to the appearances. (23) Sextus clearly thinks that our having some concept of, for example, motion is a necessary condition of our so much as investigating or raising questions about either the concept, or the nature, of motion. If so, Mates argues, then it would seem also to be a condition of its appearing to us that, for example, a ship is in motion. The latter idea, which in any case is difficult to gainsay, is that it cannot appear to one that a ship is in motion unless one understands what it would be for a ship to be in motion, and that the latter requires that one have some concept of motion. Now when Sextus investigates motion ("local transition"), he concludes that "if we go by what the dogmatists say" we can form no (coherent) concept of motion; hence, if we are to find his assent to "It appears to me that the ship is in motion" intelligible, we must attribute to him some nondogmatic concept of motion, and the question is what sort of concept that might be.

Mates proposes a straightforward solution to the problem: Sextus should invoke his own distinction between ordinary, or "pre-philosophical," concepts (prolepseis) and dogmatic, or "philosophical," concepts--the former free of, and the latter laden with, philosophical implications concerning, for example, the nature of place and time--and insist that the appearances to which he assents employ, or are couched in terms of, the former. This would seem to me to be a perfectly adequate proposal, were it not for the fact that on at least one occasion Sextus strongly suggests that the Skeptic has his own distinctive concepts:
   I suppose the Skeptic is not precluded from a conception that arises during
   the discussion itself from clear appearances affecting him passively, and
   that does not at all imply the existence of its objects; since, as the
   Dogmatists say, we conceive not only of things and states of affairs that
   exist, but also of those that do not. Hence, the person who suspends
   judgment remains in the Skeptic state both when he is inquiring and when he
   is forming conceptions; for it has been shown that he assents to whatever
   things affect him in accord with a passively received phantasia, insofar as
   that phantasia appears to him. But notice that even in this case the
   Dogmatists are precluded from inquiry. For inquiring about objects and
   states of affairs is not inconsistent in those who agree that they do not
   know how these things are in nature, but only in those who think they have
   accurate knowledge of them. (PH 2.10; my emphases) (24)

Sextus makes these remarks in the course of advancing a somewhat intricate argument to the effect that the dogmatist is incapable of engaging in inquiry, but we need not be concerned with the merits of that argument.

On my proposed reading of this passage, which I take to be at least plausible, three things are worth noting about it. First, and foremost, Sextus strongly suggests that the Skeptic can avoid the dogmatic concepts which in his view make it impossible for Dogmatists to engage in inquiry by forming, and thinking in terms of, his own concepts. Two considerations support this reading: his contention that the Skeptic can remain in the Skeptic state while forming his concepts, which strongly suggests that the same is not true of other concepts, and his contention that "even in this case the Dogmatists are precluded from inquiry," with its clear implication that it is the Dogmatists' concepts that stand in their way. Second, he seems to suggest that the distinctive features of these concepts are their arising from passive appearances and their lacking existential import. For my purposes, it is not important whether he is in fact suggesting this, or whether the suggestion is defensible; my interest lies in his more fundamental point that the Skeptic can and does form distinctive concepts. Third, it is quite striking that in the course of explicitly distancing himself from the Dogmatist, Sextus proposes to utilize concepts that arise during the discussion rather than relying on the ordinary concepts he already possesses. Whatever his reasons for proceeding in this fashion, his doing so strongly suggests that he recognizes, or is at least prepared to recognize, three distinct sets of concepts--the ordinary man's, the Dogmatist's, and the Skeptic's.

It seems to me, then, that in addressing Mates's problem on Sextus's behalf, we should distinguish three sets or, better, three types of concepts. The reason for distinguishing types is that there is no such thing as "the" dogmatic concept of motion, for example, since various dogmatists have various concepts of motion; similarly, surely, for Skeptical and ordinary concepts. Second, we should insist that in formulating the propositions to whose apparent truth he assents Sextus utilizes his own concepts. Finally, we should say enough about the differences among the three to provide a basis for determining the relative vulnerability to Skeptical arguments of propositions couched in their respective terms.

Beginning, then, with dogmatic concepts, little needs to be said; for Sextus, it seems to me, the distinctive feature of such concepts is simply their explicit incorporation of various conditions--necessary, sufficient, or both--for something's falling under them.

Getting clear about the nature of Skeptical concepts is quite another matter. My proposal concerning how these might be characterized is inspired in part by Donald Davidson's contention that
   what a person's words mean depends in the most basic cases on the kinds of
   objects and events that have caused the person to hold the words to be
   applicable.... [W]hatever [a person] regularly does apply [her words] to
   gives her words the meaning they have and her thoughts the contents they
   have. (25)

As it stands, of course, this proposal cannot be brought to bear directly on Sextus, since he never holds his words to be applicable to objects in the ordinary way. But in developing an appropriate analogue, I shall begin by considering some remarks in which he may seem so to apply them. Consider the following:
   [W]hen we question whether the external object is such as it appears, we
   grant that it does appear, and we are not raising a question about the
   appearance but rather about what is said about the appearance; this is
   different from raising a question about the appearance itself. For example,
   the honey appears to us to be sweet. This we grant, for we sense the
   sweetness. But whether it is sweet we question insofar as this has to do
   with the [philosophical] theory, for that theory is not the appearance, but
   something said about the appearance. (PH 1.19-20)

Here, in "[T]he honey appears to be sweet," Sextus may seem to apply the term "the honey" to something in the ordinary way, and he says explicitly that he grants that that external object appears. Must we read him, then, as asserting the existence of this object, and thus assenting to something beyond the appearances? It seems to me, for three reasons, that we need not and should not (and here I do not pretend to be addressing the larger question of the correctness of the received view). First, such a reading would flatly contradict things he says elsewhere, as, for example, when he compares Arcesilaus with the Pyrrhonists:
   Arcesilaus ... does indeed seem to me to share the Pyrrhonean arguments, so
   that his Way is almost the same as ours. For one does not find him making
   any assertion about the existence or nonexistence of anything. (PH 1.232)

Second, there are several occasions (PH 1.135, 198, 200, 202) on which Sextus says that he sometimes uses even the unqualified "p" as an abbreviation for "It appears to me that p," the latter being the canonical form of his reports of appearances, and this strongly suggests that in the absence of any clear counterindication, "The honey appears to be sweet" should be taken as similarly intended, that is, as a variant of "It appears to me that the honey is sweet." Finally, there is a different, quite plausible, reading of this passage available to us; we should read Sextus as saying that when the question under discussion is whether some external object is such as it appears, then he grants, for the sake of discussion, that there is an external object that appears. This would leave us free to see him as refusing to assert that anything external exists or does not exist, and would fit well with his remarks at PH 2.10 concerning the kind of concept of a non-evident thing that the Skeptic is "not precluded" from forming, with its attendant suggestion that such a concept is less robust than the ordinary man's.

My proposal, then, is that Sextus should characterize his concept of honey as a concept of the sort of thing--if such there be--that appears (has appeared) to be thus and so, or as the sort that normally appears thus and so. He could use such a concept to identify the particular honey presently before him by thinking of the latter as, say, the sample of honey--if such there be--that appears to be in a bowl before him. Properties and relations would be amenable to similar treatment. For example, his concept of the property of being in motion would be the concept of the property--if such there be--certain of whose instances have appeared so and so, and his concept of the relation of betweenness would be the concept of the relation--if such there be--certain of whose instances have appeared thus and thus.

Finally, we come to the ordinary man's concepts. These, it seems clear, incorporate the Skeptic's concepts, in that they too are concepts of sorts of things that appear to have certain properties and to stand in certain relations. But this by no means exhausts their content; the ordinary concept of honey, for example, clearly includes its being sweet, independently of its appearing to anyone in particular to be so.

Given this tripartition of types of concepts, Sextus should respond to Mates's question by saying that the appearances to which he assents are couched in, or framed in terms of, his own Skeptical concepts, so that his "It appears to me that the ship is in motion" could be expressed more fully as follows: "It appears to me that a thing of the sort that has earlier appeared thus and thus--if such there be--is now before me, and that it has the property common to those things that have earlier appeared so and so--if such there be." So far as I can see, this has no consequences that Sextus would find philosophically problematic, and there is no obstacle to his assenting to it. (26)

Further, he could now say quite simply what it would be for things to be as they appear to be; for a ship to be in motion would simply be for there to be a thing of the sort--given that there is such a sort--that has earlier appeared thus and thus, and for it to have the property--given that there is such a property--common to those things that have earlier appeared so and so.

The last question to be considered is how propositions deploying, respectively, these three sorts of concepts differ in their vulnerability to Skeptical arguments. Broadly speaking, these arguments are of two types: those that present obstacles to determining the truth or falsity of propositions whose coherence is not challenged, and those that challenge the coherence of some proposition by challenging the coherence of one or more of the concepts deployed in it.

Sextus advances many arguments of the first sort; fortunately, we need not canvass all of them. The crucial point can be made quite simply, provided only that at least one such argument can be applied completely generally--can be brought to bear, that is, on any non-evident proposition. Again, there are many such arguments, and I shall note two that, as Gisela Striker has persuasively argued, seem to be the fundamental arguments at work in the Pyrrhonists' Ten Tropes--Undecidability and Relativity. (27) Undecidability runs as follows: Since the appearances conflict with respect to each non-evident proposition, and there is no principled way of choosing among them, we must suspend judgment as to which of them, if any, corresponds to how things are. Relativity--better, perhaps, Relationality--runs as follows: Since every appearance is relational, consisting of something's appearing to some particular person who is in a certain internal state or condition and in certain circumstances, it can reveal nothing about the nonrelational properties of the thing that appears, that is, it can provide no information about the truth or falsity of any non-evident proposition saying what that thing is like in itself, so that once again we must suspend judgment concerning how things are.

Either of these arguments can be brought to bear on any non-evident proposition, even ones deploying Skeptical concepts. (28) Consider, for example, Undecidability. Sextus would argue that even with respect to a non-evident proposition, S, deploying Skeptical concepts, it appears to some that S, but to others that not-S (or that S', or that S", or ..., each incompatible with S), and that since there appears to be no principled way of choosing among these appearances, it appears that one should suspend judgment concerning S.

Sextus's charges of apparent incoherence are an entirely different matter. First of all, he levels such charges only against dogmatic concepts, or propositions that deploy them. Even with respect to these, however, he never suggests that there is something about the type of concept dogmatists defend that insures that any such concept is bound to be incoherent, or that any specification of the conditions of something's falling under a given concept is doomed in principle to failure; rather, his arguments turn without exception on the specific features of individual (proposed) concepts. Of course, he often says such things as, "If we go by what the Dogmatists say, there is no (coherent) concept of X, or X does not exist, or ..., "but clearly this must be understood to mean, "If we go by what the Dogmatists have so far said, ..."; as befits a self-described inquirer, Sextus leaves open the possibility that someone may eventually develop apparently coherent dogmatic concepts.

The upshot is that the three types of nonevident propositions do not differ at all with respect to their vulnerability to Skeptical arguments. Every non-evident proposition is vulnerable to such arguments as Undecidability and Relationality, and may be vulnerable to the charge of incoherence as well, but no such proposition is vulnerable to the charge of incoherence simply by virtue of its type.

7. Hume and Pyrrhonism

Since his "Pyrrhonian" is not the Pyrrhonist of the received view, Hume never directly challenges the latter's claim that (non-epistemic) appearances appear to suffice to guide action, and, indeed, that they appear to guide, and therefore to account (in part) for, his own actions. But Hume is hardly alone in this rather spectacular omission; to my knowledge, no one has directly challenged the Pyrrhonists on this score.

Only two possible explanations occur to me. First, Sextus's critics may have failed to distinguish two questions: (a) whether, in the absence of belief, appearances could guide the Pyrrhonist's action (or account for his occupying various states he claims to occupy), and (b) whether they do guide his action; but there would be little excuse for attributing such a conflation to them in the absence of any direct evidence. Easily the more likely explanation is that the critics regard (a) as so wildly hypothetical as to be of no interest. If one thinks that Pyrrhonists, along with the rest of our species, hold beliefs as a matter of necessity--either (with Hume) natural or (with Burnyeat) conceptual necessity, or both--then one will find it absurd, even perhaps incomprehensible, to doubt that those beliefs guide their actions. From such a perspective, (a) would come to this: "If, per impossibile, someone were to succeed in divesting himself of all belief, could the appearances suffice to guide his actions"? Hardly a burning question.

The dialectical situation is this. Sextus and Hume are at a standoff, since Hume offers no argument for his claim that belief is naturally necessary, which claim appears to Sextus to be mistaken; further, if the arguments of this paper are correct, then Sextus and the "conceptual necessity" and other "internal incoherence" critics are also at a standoff, since to date the critics' attacks have failed; in a word, it remains an open question whether a life lived without belief, and guided only by the appearances, is possible.

There is no denying, however, that the sympathies of those who have considered the matter are overwhelmingly with Hume, if not necessarily with the modern critics. Let us suppose for a moment, then, that Hume is right, so that the only possible sort of Pyrrhonist is an akratic Pyrrhonist. What of any philosophical interest follows concerning his principles? Nothing. What does follow, however, if Hume is also right in thinking that there are no satisfactory responses to his arguments, is that the condition of mankind is whimsical.

In writing this essay, I have incurred great debts to two distinguished philosophers: Benson Mates, having agreed only to alert this complete stranger to any manifest blunders, entered instead into an extensive correspondence with penetrating insights; John Cooper read the penultimate draft with meticulous care and offered acute suggestions. I am deeply indebted as well to the Review's editors and referees for virtually collaborative efforts of the highest order that prompted further improvements. As always, I have benefited early and late from Roger Wertheimer's peerless eye for every sort of infirmity that can afflict philosophical work.

In quoting from Outlines of Pyrrhonism I rely throughout on Mates's translation in The Skeptic Way (Oxford: Oxford University Press, 1996), which I refer to hereafter as SW.

(1) I follow Michael Frede in describing this view as "received," but both he and Gisela Striker have vigorously defended a sharply different interpretation that has gained numerous adherents.

The first of the quoted passages deserves special emphasis. Since the received view is an interpretation of the Pyrrhonists' avowed position, endorsing it is consistent with thinking that no self-described Pyrrhonist succeeded in adhering to his stated position. In defending Sextus's claim to have articulated at least a coherent position, I take him absolutely at his word at PH 1.4; it is not easy to see how he could have emphasized this point more strongly than he did by affording it such a prominent position. It cheered me, therefore, to find Mates insisting on the point; see his "Bolzano and Ancient Pyrrhonism," in Bolzano's Wissenschaftslehre (Florence: Olschki, 1992), 124-25, and SW, vii-viii.

(2) Julia Annas and Jonathan Barnes, The Modes of Scepticism (Cambridge and London: Oxford University Press, 1985), 23. As far as I know, Arne Naess was the first modern writer to argue for the complete generality of this Pyrrhonian distinction; see his Scepticism (London: Routledge and Kegan Paul, 1968), 16ff.

Though the point does not substantially affect the claim of complete generality for this distinction, it seems to me that Sextus does not in fact assent to the truth of any proposition. Mates has noted that Sextus never indicates his assent to a sentence of the form "It appears to me that P is true," or even "It appears to me to be true that P," and has plausibly suggested that this is a matter of principle, since for Sextus the concept of truth seems to be, at a minimum, the concept of some sort of correspondence with some external reality, of which there seems to him to be no coherent conception (SW, 59). Myles Burnyeat has argued, similarly, that for the Skeptic "truth ... is closely tied to real existence as contrasted with appearance" and "belief is the accepting of something as true." See his "Can the Skeptic Live His Skepticism?" (CSL), in The Skeptical Tradition, ed. M. F. Burnyeat (Berkeley: University of California Press, 1983), 117-48; 121.

Nevertheless I shall often, for brevity's sake, speak of Sextus's assent to the apparent truth of some proposition p (or the apparent soundness of some argument); on such occasions, I should be understood to be saying that he assents to "It appears to me that p" (or to "It appears to me that X" for each premise X of the argument in question). See also note 3.

(3) In dissenting from the received view on this score, and arguing that the ancient Skeptics willingly held all sorts of quotidian beliefs, Frede argues that
   [T]he skeptic ... will not think that it only seems as if things were so
   and so; for that thought presupposes that he believes ... that, in reality,
   things are quite different from the way they seem to be. For him ...
   nothing rules out the possibility that ... things should be exactly as they
   appear to be.... When [he] speaks of what seems to him to be the case ...
   he cannot be speaking of something which he thinks only seems to be the

But in assenting to appearances only the Pyrrhonist is not judging that they are only appearances. Of course, he cannot consistently hold that things are not as they appear, but neither can he consistently hold that things are as they appear; for him, nothing rules out either possibility. See "The Skeptic's Beliefs" (SB, hereafter) in Frede's Essays in Ancient Philosophy (Minneapolis: University of Minnesota Press, 1987), 179-200; 191.

Gail Fine, too, understands Sextus to be saying that Skeptics hold beliefs, though only "beliefs about how they are appeared to." I take this to imply that according to Sextus Skeptics believe, and therefore assent to, the truth of some propositions of the form "It appears to me that p," a reading that I resist for the reasons cited in note 2, among others. See "Descartes and Ancient Skepticism: Reheated Cabbage?" Philosophical Review 109 (2000): 195-234; 207-8.

It is important to be clear about the nature of the appearances to which a Pyrrhonist assents. They are not subjective experiences, sensory or otherwise; rather, an appearance is, in Mates's words, "whatever seems or appears to be the case ...; it is properly expressed by a `that'-clause (as, for example, in `It appears that the good fare ill and the bad fare well' ...)" (SW, 6). I thank John Cooper for urging me to underscore the point.

(4) David Hume, part 2 of section 12, "Of the Academical or Sceptical Philosophy," of An Enquiry Concerning Human Understanding (London, 1748), in Enquiries Concerning the Human Understanding and Concerning the Principles of Morals, ed. L. A. Selby-Bigge, revised P. H. Nidditch (Oxford: Oxford University Press, 1975).

Given Hume's nominal topic in this Section, and the fact that he speaks so readily and without qualification of the "excessive principles" of skepticism, it seems likely that his "Pyrrhonian" is not the Pyrrhonist of the received view, but rather a partisan of some theoretical position antithetical to the possibility of our acquiring knowledge, or even rational bases for belief. For our purposes, however, the difference does not matter, since Hume's objections are focused exclusively on the question of the possibility of life without belief.

(5) An akratic subject, as normally understood, is one who is aware of his inability to control certain aspects of his behavior, but for lack of any concise alternative I use "akratic Pyrrhonist" to refer to a Pyrrhonist who, contrary to his claim not to do so, does in fact hold beliefs.

(6) A Pyrrhonist would not, of course, claim to have discovered that a life without belief brings peace of mind, but only report that it appears to do so.

(7) Though the issues are independent of my concerns in this paper, it should be noted that the ideas that (i) an argument can compel one to accept its conclusion, and that (ii) one who is convinced that the arguments on two sides of a question are balanced--and to whom it appears that one ought to suspend judgment in such cases--is compelled to suspend judgment, are at best dubious. As for (i), it is by now a commonplace that one man's modus ponens is another man's modus tollens, and (ii) seems to depend on the rather quixotic idea that human beings are compelled to act in accord with their principles, rather than revise or abandon them. It may be significant that the ancient example of our inability to form a belief concerning the question whether the number of stars is odd or even seems to be Burnyeat's favorite illustration of our being compelled to suspend judgment. (See CSL, 132, and "Idealism and Greek Philosophy: What Descartes Saw and Berkeley Missed," Philosophical Review 91 (1982): 3-40; 24.) In the context of Pyrrhonian thought, at any rate, this is at best a degenerate case of balanced arguments, since what is most striking about it is the transparent absence of arguments on either side of the question.

For a recent discussion of (i) and related matters see the papers in part 1 of Gilbert Harman's Reasoning, Meaning and Mind (Oxford: Oxford University Press, 1999).

(8) Of course when Sextus, in such contexts, states things that appear to him to be the case, he states only such things as he thinks will also appear to his dogmatic opponents to be so; my point is only that some of these originate with him, and are not borrowed from the dogmatists. For some discussion of the bearing of this point on the reading of the Tropes, see SW, 57-59.

(9) It seems to me that "involuntary" does not quite capture Sextus's thought here, and it is for this reason that I have added "one simply cannot fail to assent to such judgments." I take Sextus to have in mind something like this: that there are occasions on which, if the question arises whether it appears to one that p--whether, for example, it appears to one that the honey is sweet--one cannot fail to respond affirmatively. (That is, one cannot fail to do so in foro interno; one can, of course, misrepresent to others how things appear to one.) It is such unavoidable responses that constitute, for Sextus, acts of assenting to the appearances, and it is their unavoidability that he means to express by describing them as involuntary.

In contrast, there are no occasions on which, if the question arises whether p is true, where p is any non-evident proposition, one cannot fail to respond affirmatively--one can always withhold, or suspend, judgment.

(10) Here I incorporate a provision that should be noted explicitly, though it is implicit in the received view: for a Pyrrhonist, how things appear to others is a matter of how things are, not of how they appear.

(11) W. V. Quine, Word and Object (Cambridge: MIT Press, 1960), referred to hereafter as WO. A striking similarity between Quine and Sextus is worth noting here. Although he does not make the point explicitly, Quine's "No judgment is immune to revision" is clearly to be understood as including itself within its scope; compare Sextus's "[T] he Skeptic considers that ... the `Nothing more' slogan ... applies to itself" (PH 1.14-15).

(12) Note Sextus's own formulation: "[I]n putting forward [his own slogans the Skeptic] is saying what seems to him to be the case" (PH 1.15, my emphases). In Sextus Empiricus I (Harvard: Harvard University Press, Cambridge, 1976) R. G. Bury translates this passage as "[I]n his enunciation of these formulae he states what appears to himself" (my emphases).

A reader familiar with Burnyeat's essay might object that Sextus is not entitled to a certain presupposition that is in play here, namely, that an argument's appearing to him to be sound is distinct from his believing it to be sound. We shall examine this crucial question in section 4 below; my present purpose is only to show that, given that presupposition, there is a significant form of argumentation available to Sextus.

(13) I suggest that Sextus could offer such an account of the stages by which he arrived at his present condition in order to show that he could do so even if we understood him to be saying, at PH 1.4, that each of his statements is to be understood as expressing only how things appeared to him at the moment at which it was written. If we allow him (as we surely should) to speak in the "specious present," then he can say simply that the apparent reason for the conclusion's appearing to be true is the argument's appearing to be sound.

(14) For a thorough discussion of these, see Myles Burnyeat, "Protagoras and Self-Refutation in Later Greek Philosophy," Philosophical Review 85 (1976): 44-69.

(15) Ibid., 51.

(16) Ibid., 52, my emphasis.

(17) Ibid., 53.

(18) Though Sextus does not explicitly attribute to his progenitors the particular pre-suspension belief that it is possible to determine the truth, such an attribution is surely plausible. Making it would not, however, substantially alter the picture, and would not affect my proposal, if we also assumed that that belief was among those which they suspended. First, its pre-suspension presence would be seen as having contributed to their anxiety, since they would have seen themselves as having failed to achieve something that was definitely possible, and not only something that might be possible. Second, while abandoning that belief might have resulted in diminution, or even loss, of their hope that they might at some time determine the truth, it would have left untouched their openness to the possibility that they might do so.

(19) It is worth noting also that on Burnyeat's account Sextus does after all turn out, though for reasons quite other than those he adduces, to be one of those people who "maintain a belief in the face of a clear realization that it is unfounded" (if, that is, we take him at his word, with the exception of his claim to live without belief). Let N be any non-evident proposition that Sextus has considered. Then it appears to him, as a result of bringing his arguments to bear on N, that the considerations for and against N are equally balanced, so that no answer is forthcoming (NEB); accordingly, he suspends judgment concerning N, and experiences ataraxia with respect to it. Since on Burnyeat's account this last can be explained only on the hypothesis that Sextus believes NEB, we may infer that he believes NEB. But NEB, too, is a non-evident proposition, so that it also appears to Sextus that the considerations for and against NEB are equally balanced, so that no answer is forthcoming (NEBEB); accordingly, he suspends judgment concerning NEB, and experiences ataraxia with respect to it. Again, we may infer that he believes NEBEB. Thus, Sextus believes NEB, and he believes that that belief is unfounded--that there is no more reason to believe NEB than there is to believe not-NEB.

(20) I am not here invoking some libertarian conception of free agency, or suggesting that there is no causal story to be told about how Sextus comes to suspend belief. I am insisting only that suspending belief is for him an act which is under an agent's control rather than a mere (passive) occurrence. See also footnote 7.

(21) Paul Grice, Studies in the Way of Words (Cambridge: Harvard University Press, 1989); see especially part 1, "Logic and Conversation," a revised version of his William James Lectures given at Harvard in 1967. Grice focuses most of his attention on his theory of "conversational implicatures," but these are a subcategory of nonconventional implicatures, and for our purposes it does not matter whether the implicatures we are considering are conversational.

(22) Michael Williams makes this point in "Scepticism Without Theory," Review of Metaphysics 41 (1988): 547-88, though he arrives at it on the basis of a radically different reading of Sextus.

(23) See SW, 26-29.

(24) I intend no distinction between my "concept" and Mates's "conception"; I have simply left his translation intact.

For some discussion of the significance of this passage for understanding Sextus's treatment of "the problems Scepticism is faced with when trying to deal with the conceptual legacy it inherited," see Jacques Brunschwig's "Sextus Empiricus on the kriterion: the Sceptic as Conceptual Legatee," in his Papers in Hellenistic Philosophy (Cambridge: Cambridge University Press, 1994).

(25) Donald Davidson, "Knowing One's Own Mind," Presidential Address delivered at the Sixtieth Annual Pacific Division Meeting of the American Philosophical Association; Proceedings and Addresses of the American Philosophical Association 60 (1987): 441-58; 456.

(26) I have emphasized that Sextus could characterize his concepts in terms of his own appearances because they are the only ones that are strictly speaking available for the purpose; what is crucial for my argument, however, is only that they be characterizable in terms of appearances in general, since Sextus says nothing to suggest that such concepts, or propositions couched in terms of them, might be incoherent.

(27) See her "The Ten Tropes of Aenesidemus" in Myles Burnyeat, The Skeptical Tradition, 95-115. For a penetrating discussion of the nature and role of another set of universally applicable arguments, namely, the so-called Agrippan Modes, see Jonathan Barnes's The Toils of Scepticism (Cambridge: Cambridge University Press, 1990).

(28) A non-evident proposition couched in Skeptical terms will of course rarely, if ever, be asserted, since those given to making dogmatic assertions--namely, Dogmatists--will rarely, if ever, deploy Skeptical concepts, and those who deploy Skeptical concepts--namely, Skeptics--will rarely if ever make dogmatic assertions. But that fact is neither here nor there for our purpose, which is simply to understand the applicability of Skeptical arguments to such propositions.

University of Houston
COPYRIGHT 2001 Cornell University
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Johnsen, Bredo C.
Publication:The Philosophical Review
Geographic Code:1USA
Date:Oct 1, 2001
Previous Article:Visual experience and motor action: are the bonds too tight?
Next Article:Scanlon's contractualism: critical notice of T. M. Scanlon, What We Owe to Each Other.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters