Abstract
In this paper I discuss Sextus’ account of sophisms in Outlines of Pyrrhonism, II 229–259. After a few introductory remarks on the sceptic capacity as a rational capacity, I analyse the sceptic’s general attitude towards sophisms and draw attention to the sceptic reaction to the way in which dogmatic logicians deal with insolubilia or aporoi logoi. I provide some context for the Pyrrhonian attitude towards sophisms, emphasizing a radical difference between the sceptics and the dogmatists: while the dogmatists tend to deal with insolubilia as exceptions, the sceptic questions the exceptionality of such cases. I suggest that the Pyrrhonian approach to sophisms is based on the impossibility for the sceptic to detach pragmatic considerations from logical issues. Finally, I draw some conclusions on the Pyrrhonian attitude towards dealing with arguments without logical theory.
1 Introduction
In this paper I argue that Sextus’ discussion of sophisms in his Outlines of Pyrrhonism (henceforth: PH) II 229–259 prompts a series of instructive questions about the Pyrrhonian point of view on rationality and, therefore, on the nature of Pyrrhonian scepticism as a rational activity. Section 2 provides a few introductory remarks on the sceptic capacity as a rational capacity. Against that background, section 3 analyses the sceptic’s general attitude towards sophisms based on PH II 229–259. At the end of this section, I draw attention to the sceptic’s reaction to the way in which dogmatic logicians deal with insolubilia or aporoi logoi. Section 4 attempts a contextualisation of the Pyrrhonian attitude towards sophisms by briefly rehearsing a debate between Stoic and Academic sceptics on insolubilia. This debate brings to light a radical difference in perspective between the sceptics and the dogmatists: while the dogmatists tend to deal with insolubilia as exceptions or localised and self-contained phenomena, the sceptic questions the exceptionality of such cases and wonders whether the dogmatist’s attitude towards insolubilia is more telling about human reason than dogmatic logicians are willing to concede. I suggest that this might be the background for the Pyrrhonian approach to sophisms, which is based on the impossibility for the sceptic to detach pragmatic considerations from logical issues and logical theory. Finally (section 5), I draw some conclusions on the Pyrrhonian attitude towards dealing with arguments without logical theory, and I submit a few final considerations on how the sceptic attitude compares to that of other varieties of mistrust of logic in antiquity.
2 On the Sceptic Capacity, Rationality, and Logical Theory
The sceptic1 is characterised by the capacity to set up equipollent arguments against each other in such a way as to provoke suspension of judgement and, eventually, tranquillity (PH I 8). The sceptic activity, which is supposed to keep the sceptic from assenting to what is unclear, seems to be a rational activity, and, crucially, this is the activity the sceptic needs to keep on practising if he wants to remain in the state of mind in which judgement is suspended.2 How does the sceptic regard this activity?
The sceptic seems ready to concede that he has the (irresistible) impression of being naturally endowed with (so-called) perception and thought, or reason (PH I 11, 24: phusikōs aisthētikoi kai noētikoi). This is the natural guidance (PH I 11, 23–24: huphēgēsis phuseōs or huphēgēsis phusikē) the sceptic can rely on in his non-dogmatic preservation of life. We can presume that the skeptikē dunamis and its exercise are also manifestations of the “natural” reason that the sceptic (non-dogmatically) ascribes to himself. Both sensation and reason can produce in him impressions he cannot help being affected by (PH 19–20;3 cf. Against the Professors4 VIII 301, 309, 316, 320). One way in which the sceptic feels affected is that he sometimes feels compelled to assent to some impressions. And yet, there is a use of reason that allows the sceptic to resist this compulsion, at least in some important cases: whenever opposite arguments seem compelling, the sceptic resists the compulsion in both directions and suspends judgement.
Whether there really is any such faculty of reason and what it is supposed to be is, of course, a question the sceptic will resist from addressing. But the sceptic at least registers what appears to him as his own rational activity.5 He relies on concepts that are passively formed (PH II 10), and he uses natural language to express himself – even if some precautions must be in place in order to use language in a non-dogmatic way.6 Presumably, the sceptic could also concede that reporting about the way he is affected at the time of the report (i.e., the only thing the sceptic can report about: PH I 15) is one way in which (so-called) reason (whatever that is) is used.
Whether the journal of the sceptic, reporting how he is affected, is philosophically interesting reading or not depends, at least to some extent, on whether we think that the attitude of the sceptic captures any basic and ineliminable features of human existence – which is, as is well known, a controversial issue. In his report, the sceptic cannot make any universal quantifications, since the report can at most cover all data the sceptic has scrutinised to date (PH I 197–200, 202–205). One consequence of this is that the sceptic cannot make any statements about the universal features of anything, including logical matters and, more specifically, arguments.7 Arguments may seem to the sceptic to be products of so-called reason,8 but, strictly speaking, the only thing the sceptic can say about arguments is how those he has scrutinised appear to him at the time of his report. What the sceptic can say about the way in which arguments appear to or affect other people is based on other people’s reports. Even if there were any truth of the matter or any “objective” features of arguments “out there” (be this in terms of arguments as lekta or as anything else), the sceptic does not know and suspends judgement. He may even report that this attitude brings him peace of mind.
If this picture is correct, the sceptic will not be able to give a fully general account of the distinction between valid and invalid arguments. Can he still be sensitive to the distinction in the individual arguments he considers in his activity? While scholars have addressed the issue of the sceptic’s rationality with reference to the sceptic’s relation to truth,9 the issue of the sceptic’s relation to validity is even more basic. If there is no way for the sceptic to make distinctions between good and bad arguments and any argument goes, what is the ongoing scrutiny of arguments supposed to contribute to the sceptic practice? Can the sceptic not suspend judgement by default, without even bothering going through the arguments? Wouldn’t that be an even easier life? Notice that apparently the sceptic is supposed to be sensitive to at least some distinctions obtaining among arguments, if he can calibrate the strength of the sceptic therapy to the seriousness of the dogmatic disease (PH III 280–281): the sceptic will intentionally resort to arguments that are more cogent (tois embrithesi) and capable of removing a claim vigorously (eutonōs anaskeuazein dunamenois) to contrast especially severe dogmatic affections (i.e., deep self-conceit), whereas milder (kouphoterois) arguments will be required to contrast less severe conditions. What is this distinction about? Does the distinction rest on logical features of the arguments? If so, how does the sceptic identify such features? If not, what is the basis of the distinction? In this paper I suggest that there might be something to learn about the sceptic point of view on these issues by looking at the sceptic attitude towards so-called sophisms.
3 The Sceptic and Sophisms
Sextus’ most extensive discussion of sophisms is in PH II 229–259. There is no systematic treatment of the topic in M VII–VIII. The passage in PH concludes the discussion of the dogmatists on the logical part of philosophy in book II. The general point is that the capacity to distinguish between truth and falsehood is what logicians invoke as a service to life in helping people defend themselves against the tricks of the sophists,10 but the highly abstract technicalities of the dialecticians are unnecessary to deal with whatever sophisms one may come across in real life.
Sextus’ report starts with a general account of sophisms, which he certainly borrows from the dogmatists. On the account he refers to, a sophism is ‘an argument that is persuasive and deceptively devised in such a way as to induce acceptance of the conclusion, which is either false or similar to what is false or unclear or otherwise unacceptable’ (PH II 229). Two points are key:
(i) the conjunction of persuasiveness and deception;
(ii) the emphasis on some problem in the conclusion.
(ii) turns out to be a prominent feature in Sextus’ discussion of sophisms, but (i) is an important descriptive criterion for identifying sophisms. It does not rely on any clear logical features of arguments. A later comment makes clear that deception can go either way: it can induce acceptance of a conclusions that seems dubious, but it can also make someone hesitate to accept something that is obvious, which will also count as a case of sophism: ‘it seems to divert towards what is false in such a way as to make those who are not paying attention hesitate in assenting to it, while it concludes what is true’ (PH II 233). In both cases, a sophism is an argument that diverts, distracts (apagein: 233 and 234), and, in this way, puzzles. All that matters on Sextus’ account is how the argument impacts on someone who is exposed to the argument; no mention is made of any logical feature of the argument as such.
Sextus’ emphasis on (i) to identify sophisms is indicative of a more general attitude that can be better understood with reference to other features of the passage. In particular, Sextus mentions that, in addition to what he reports, people also say ‘other things’. We can only speculate as to what this refers to, but clearly Sextus is selecting what he wants to mention. Examples of such ‘other things’ that dogmatists would say about sophisms around Sextus’ time can be found e.g. in Alexander, On Aristotle’s Topics, Wallies 20.2–22.6. One distinction Alexander refers to and Sextus does not mention is the one between eristic (or sophistic) arguments based on their matter (i.e., valid arguments whose premises seem to be endoxastic11 but are not endoxastic) and eristic arguments based on their form (i.e., invalid arguments). Alexander’s distinction is based on Aristotle, Top. I 1, 100b23–101a4, and introduces two very different ways in which sophistic arguments can be deceptive: there can be a problem with the premises, or there can be a problem with the logical structure of the argument. No such distinction is drawn in Sextus’ account. Similarly, we do not find in Sextus any systematic discussion of different kinds of fallacies. This is important, too, because the way in which dogmatic logicians think of logic as a tool to defend ourselves from the tricks of the sophists is that logic gives us the tools to identify fallacies. Through this diagnosis, we can expose the pretense of the sophists and “solve” their arguments, thereby setting ourselves free from the puzzlement deriving from being exposed to sophistic arguments. As we shall see, Sextus’ advice to set ourselves free from the puzzlement and troubles deriving from sophisms is radically different.
For Sextus, sophisms are, quite generally, arguments ‘spoiling truth with apparent persuasiveness (phainomenais pithanotēsin)’. As is well known, pithanon can indicate persuasiveness or probability, and there is an issue as to whether persuasiveness or probability better captures Carneades’ views on the pithanon.12 There is also an issue as to how much Academic scepticism is included in Sextus’ scepticism. However, based on the picture sketched in section 2, it seems that the only way in which the Pyrrhonian sceptic can speak of pithanotēs is with reference to the way in which an argument (or, more generally, an impression) affects him.13 So, the persuasiveness of sophisms has to do with the way in which such arguments appear and affect someone who is exposed to them. Whether an argument is a sophism depends on whether, on top of being persuasive, it is also devised with the purpose of deceiving people. One implication of this account seems to be that the sophist is someone who is capable of devising arguments in such a way that they appear persuasive, where the emphasis on the ‘apparent persuasiveness’ invites a distinction between what merely appears to be persuasive and what genuinely is (or ought to be) persuasive. But can the Pyrrhonian sceptic draw such a distinction?
The special inclination towards mere appearance is, at least since Plato, a Leitmotiv in the characterisation of sophists, and one which also figures in dogmatic accounts of sophisms (see e.g. Aristotle, Sophistic Refutations 165a19–31; cf. Metaphysics IV 2, 1004a24–25). Importantly, at least some dogmatists have a way to characterise what Sextus calls the ‘apparent persuasiveness’ of sophisms as due to the logical inexperience (apeiria: see Aristotle, Sophistic Refutations 164b25–165a3) of those who are exposed to sophisms and do not have the tools to disentangle themselves from their snares. The dogmatists think that they can rely on their theories of valid and invalid arguments to expose deceptive persuasiveness. But this road is not an option for the sceptic: it is unclear whether the sceptic can distinguish between the appearance of persuasiveness deriving from ignorance, i.e. from the incapacity to detect some specific objective mistakes in an argument, and any other appearance of persuasiveness. The sceptic does not have a general theory of valid arguments and finds himself in the position of not really being able to distinguish between “objective” features of arguments (such as the necessity with which the conclusion follows from the premises) and the “subjective” reaction to such arguments and their unfolding. The sceptic may well raise the question: “How do I distinguish cases in which I feel compelled to accept a conclusion that actually follows from the premises, from cases in which I feel compelled (by a deceptive argument) to accept a conclusion that does not really follow from the premises?” The answer of the sceptic is that he cannot distinguish: all he has to go on with is his affection of compulsion, and that affection is there in both cases. In the case of sophisms, the affection of compulsion is also accompanied by an affection of puzzlement. It is unclear to the sceptic whether there is anything one can rely on in order to answer that question in the first place. This is why, as I will presently suggest, the sceptic needs to focus on the conclusion (point (ii) in Sextus’ account).
In order to better appreciate the Pyrrhonian point of view, it might be helpful to distinguish two approaches to arguments. One approach consists in looking at arguments as complex formal objects, composed of one or more premises and a conclusion, in a sort of temporal and epistemic vacuum, independently of whether and how an argument can be or has been actually formulated (in time, by a speaker, etc.). The Stoic understanding of arguments as complex lekta certainly gave a significant contribution to this way of thinking of arguments,14 but the assumption that such lekta are ‘something’ and something worth investigating15 is quite far from what most ancient philosophers, and certainly from what the sceptic might have ever thought about arguments. The alternative and more common way of thinking about arguments16 consists in taking arguments as complex logoi that unfold in time, usually in a dialogical exchange (real or fictitious) between two interlocutors. The way in which we can interact with a logos in progress, as it were, and the way in which we can interact with a logos once it has already reached its conclusion, are different: for example, we can prevent a logos in progress from reaching a conclusion, but we can solve a logos that is already out there by pointing out a logical mistake in it.17
The distinction between these two ways of looking at arguments is quite crucial to understand the sceptic strategy: the sceptic, instead of engaging in debates in which he is expected to concede or reject premises, should wait until the whole argument has been formulated, and only then consider his attitude towards that argument. This points to a fundamental difference between the Stoic and Peripatetic approach to sophisms on the one hand and the Pyrrhonian approach on the other hand. While for the Peripatetics and the Stoics philosophical resistance to deception in arguments fundamentally relies on the capacity to expose falsehood or invalidity by spelling out where it resides and what the mistake consists in, for the sceptic philosophical resistance to deception consists in refraining from engaging with any argument that looks deceptive as a whole (PH II 250). Peripatetics and Stoics engage in dialectical exchanges with an opponent because they believe that they can rely on their logical expertise. What the sceptic sees here is unwarranted confidence in dealing with unclear matters (in this case: logical theories) manifesting in the rash behaviour of the dogmatists.
Various remarks in Sextus go in this direction. Firstly, Sextus objects to the dogmatic approach to sophisms that it is superfluous and vain (PH II 235), because it is unclear that one can draw a neat distinction between obviously valid arguments and invalid arguments.18 Secondly and more specifically, the level of generality and the far-fetched cases in the discussions of sophisms by dogmatic logicians do not seem to add much to what experts in the various ordinary arts, crafts, and disciplines can already do without engaging in logical theory (PH II 251).19 Thirdly, it is unclear why, when an apparently persuasive argument ends with a conclusion that is in opposition to evidence and what is apparent (enargeia/phainomena), reference to such evidence and appearances would not be enough to reject the argument as a whole. On this last point, Sextus’ remarks are puzzling. He claims that it is unclear that the dogmatists’ strategy of engaging in more or less complicated arguments in support of the opposite (obvious) conclusion is worth the effort. For it is unlikely that, whatever argument they will come up with, that argument will be less controversial than the evidence and the appearances that made one doubt the original argument in the first place: do we really need arguments concluding that there is change to see that there is something odd about Zeno’s arguments? Do we need arguments concluding that the snow is white to raise doubts about Anaxagoras’ argument that snow is black? (PH II 244–246) These remarks are puzzling because, unless Sextus takes the pieces of reasoning the dogmatists produce in order to expose a fallacy as arguments in support of the opposite obvious conclusion, it does not look like the dogmatists’ approach to a fallacious argument concluding that P (where P is paradoxical) consists in producing an argument concluding that not-P (where not-P is obvious or uncontroversial). Perhaps Sextus’ point simply is that the solution will still be less clear than the opposite of the conclusion of the sophism, and in order to see that the opposite of the conclusion of the sophism is clear one need not resort to the logical technicalities of the solutions of sophisms.20
Against this background, there is still a question concerning part (i) of Sextus’ account of sophisms. How can the sceptic detect deception? My answer is that the sceptic can only detect his own puzzlement, and that is taken as a guide to potentially deceptive situations. Perhaps we can say that the ‘apparent persuasiveness’ of sophisms is experienced by the sceptic as an affection of feeling compelled to concede a conclusion because of an argument together with an affection of puzzlement at the conclusion: the puzzlement is due to the fact that the sceptic feels compelled by the argument to accept the conclusion, whereas the claim expressed in the conclusion is clearly nothing the sceptic would feel inclined to accept in the absence of the accompanying argument.21 I suggest that the very broad account of sophisms in PH II 229–259 is meant to emphasize precisely this: there may or may not be a mistake in the argument, but there is something puzzling about the argument, and the puzzling bit is that it seems to compel one to accept a conclusion the opposite of which seems clear to the sceptic (or not to accept a conclusion which seems clear). In order to appreciate this situation, the sceptic must look at the whole argument, including premises and conclusion, and cannot simply interact with the premises as they are submitted. If this is right, then one implication is that the sceptic, just to be on the safe side, will not engage in dialectical exchanges: the sceptic does not concede or deny any premise; he sits and waits until the argument has been fully formulated, and then looks at the result.
Interestingly enough, resistance to the conclusion even in the absence of a clear diagnosis of a mistake in the argument is a strategy Aristotle himself suggests pursuing when facing arguments such as Zeno’s paradoxes (Top. VIII 8, 160b6–10). For Aristotle this is clearly a suboptimal situation to find oneself in in a dialectical exchange. In the context of the passage, Aristotle makes the non-trivial point that, in these cases, not conceding the conclusion despite not having any clear objection to the argument, is not a case of unqualified misbehaviour (duskolainein) in dialectical exchanges. It is impossible for us to tell whether such behavioural rules had any direct influence on Sextus, but the somewhat ethical assessment of behaviour in logoi is certainly something Sextus is willing to emphasize (see e.g. PH II 251–252, on the sceptic prohairesis in the sceptic approach to logoi). As we shall see, the possibility of relying on strategies recommended by dogmatic philosophers in dialectical exchanges may well be one of the sceptic’s major weapons in defending their own attitude – or, rather, in trying to undermine the dogmatic attitude.
One extremely important place to see the sceptic in action is PH II 251–255. In the first part of this passage (251–253), Sextus is pointing out that sophisms encompass not only apparently persuasive arguments with a false conclusion, but also apparently persuasive arguments with an absurd or paradoxical conclusion. Absurd or paradoxical claims are identified by Sextus with reference to what is agreed upon or is uncontroversially absurd (252: homologoumenōs atopon). All arguments, Sextus says, can be brought under two main cases: either the argument leads to an unacceptable conclusion, or it leads to a conclusion that one ought to accept. Either way, it does not really matter how the argument leads to the conclusion: the sceptic will see how to react and how to deal with the argument once the whole argument has been presented, i.e., once the conclusion has been expressed. Until then, the sceptic stays away “from the road” leading to that conclusion. He suspends judgement on all premises and, in this way, does not participate in the process of unfolding the argument.
There are several difficulties in the way in which Sextus describes the attitude of the sceptic. One difficulty is that Sextus does not seem to make any room for a distinction between a concession for the sake of the argument and assent. One could concede a premise for the sake of the argument without believing in the truth of the premise and assenting to it. It is unclear whether Sextus thinks that there might be an important distinction to draw here, but from the way in which he goes on to emphasise that even the Stoics in some cases (on which I shall return presently) ‘stop answering’, i.e. suspend judgement and stop conceding the premises (doing what sceptics do all along) suggests that the sceptic tends to identify a concession withing a dialectical exchange with an act of assent.22 The sceptic waits until the whole argument has been produced, and only then considers how to interact with the argument.
I mentioned that Sextus’ overall strategy is reminiscent of Aristotelian advice. This is not explicit in Sextus. Sextus, however, explicitly refers to some Stoic moves in dialectical exchanges. He suggests that the sceptic’s attitude of suspending judgement all along and from the start might be more consistent than the Stoic attitude of stopping answering only in some cases and without a clear explanation of why they stop when they stop. The sceptic attitude to suspension of judgement, unlike the Stoics’ attitude to the exceptional cases in which they have to resort to it, is consistent with the sceptic’s style of life, following the ordinary path (PH II 254 and 257), and is a matter of choice (prohairesis: PH II 251–252).23 But what are the cases in which the Stoics suspend judgement? In PH II 253–255 Sextus spells out one24 such case, i.e., when they face the sorites. In the next section I spell out why this might be a crucial detail revealing the more general sceptic attitude towards logic and reason.
4 Boundaries and Exceptions: When the Dogmatists are Puzzled and Troubled
Sextus’ report is not the only report on the Stoic way of dealing with the sorites. Cicero, Academica (Ac.) 2.49–50 and 91–94, is another fundamental piece of evidence, and in Cicero we get a better sense of a debate between Stoics and Academic sceptics on this point. In this section I shall briefly rehearse a few main points of that debate before suggesting that this might provide the philosophical background to understand Sextus’ account of the Pyrrhonian attitude.
Ancient sources call ‘sorites’ what we should regard as two different types of argument. To disambiguate, I shall call ‘sorites-proper’ arguments built on numerical series, whereas I shall reserve the label ‘little-by-little’ to arguments which were also called ‘sorites’ (e.g., Cicero Ac. 2.49–50) but are not built on numerical series: they are rather built on a series of premises progressively departing from an initial conceded premise, where the departure from the initial premise is done “little by little”. Sources make clear that the Stoics dealt with these two types of argument in different ways, which clearly shows that Stoic logicians were not ambiguous or confused about the two. The fact, however, that the two went under the same name puts some emphasis on some (superficial) similarities between the two which may have been exploited by the adversaries of the Stoics.
The ‘heap’, from which the sorites-proper takes its name, is based on three incompatible assumptions:
(1) 1 seed is not a heap.
(2) If n is not a heap, then n+1 is not a heap.
(3) There is some m such that m seeds are a heap.
Sextus refers to Chrysippus’ recommendations as to how to behave if one is ‘asked’ the sorites-proper. The situation Sextus has in mind is one in which a logos comes about through dialogical interaction between two interlocutors, a questioner and an answerer. In this case, the Stoic is the answerer. The questioner will start by asking whether one seed is a heap, and with each question he will increase the number of seeds by one; when will the Stoic start to answer that the amount is a heap?
In these arguments, the initial steps may seem uncontroversial, but one soon comes to steps in which it is unclear how one ought to answer the question – because it is unclear what the right answer would be, or whether there is a right answer in the first place. Chrysippus recommends that one stop answering the questions when things are still clear, in order to avoid assent in unclear cases. As far as we know, Chrysippus’ strategy probably included the introduction of logical distinctions making room for the special status of the declarative sentences corresponding to the questions of the sorites-proper. On Chrysippus’ account, at least some of these declarative sentences would not express a proposition, but rather correspond to impressions that can be both persuasive and unpersuasive (M VII 242–243).25 Impressions can be not only persuasive and unpersuasive, but, unlike Stoic propositions, they can also be true and false, or neither true nor false (whereas a proposition, or axiōma, for the Stoics is a complete lekton that is either true or false).26 A similar strategy was probably used to account for the status of the declarative sentence figuring in the Liar: ‘I am lying’.27 This move allows the Stoics to say that the principle of bivalence holds for propositions (axiōmata), but not all declarative sentences express propositions: at least some of the declarative sentences in the sorites-proper are declarative sentences that do not express a proposition. This distinction allows the Stoics to provide pragmatic recommendations on how to approach the sorites-proper or the liar without undermining any piece of their logic.28 But, as is often the case in controversies against the Stoics, their opponents would be eager to emphasise the ad hoc flavour of the Stoic approach and retort the move against the Stoics as undermining their own logical principles.29
Little-by-little arguments are not based on numerical series but are rather built in such a way that question n+1 slightly departs, “little-by-little”, from question n. Carneades was famous for resorting to such arguments in his controversies with the Stoics (see e.g. Cicero, De natura deorum 3.43–52, and Sextus, M IX 182–184). While the Stoics may have condemned the use of such arguments as sophistic, they had a fairly straightforward way of dealing with these arguments by simply not conceding some of the premises. The Stoic strategy is clearly illustrated in one famous passage in Cicero:
‘And the first thing to criticise is their [scil.: of Carneades’ followers] use of an extremely sophistic form of argument – one that usually meets with very little approval in philosophy – in which something is added or subtracted little by little and progressively. (They call this the ‘sorites’, because they produce a heap by the addition of single grains.) This is clearly a fallacious and sophistic form of argument. You build it up like this: if god has presented a sleeper with an impression that’s persuasive, why not also one that’s extremely truth-like? Next, why not one that’s difficult to discriminate from a true impression? Next, one that can’t even be discriminated? And finally, one that doesn’t differ from a true impression at all? If you reach this far because I have conceded each successive point to you, it will be my fault; but if you proceed on your own authority, it’s yours. Who will grant you the assumptions that god can do everything or that he would act like this if he could?’
Cicero, Ac. 2.XVI 49–50, trans. Brittain
Carneades’ strategy in using these arguments against the Stoics presumably consisted in pointing out that the only way the Stoics had to block the series of questions was by relying on some unjustified assumptions. Famous examples of assumptions the Stoics would rely on in order to block little-by-little arguments are the assumption that god cannot do everything (the Stoics deny that god can do wrong), or the assumption that there cannot be two identical (and, therefore, completely indiscernible) entities.
While the Stoics had different ways to deal with the sorites-proper and little-by-little arguments, I take it that the point of Academic criticism here is that the Stoics may think that they can deal in different ways with the two types of argument, while in fact they rely on unwarranted assumptions in both cases. In the case of little-by-little arguments – the Academics may argue – they explicitly do so (relying, e.g. on the assumption that god cannot do anything wrong, or on the assumption that there do not exist two identical entities); in the case of sorites-proper, they rely on (what the Academics take to be) peculiar distinctions between propositions and sentences. If so, (the Academics would continue) perhaps the Stoics ought to feel equally uneasy in both cases.
On this (polemical) account, the Stoics’ reaction to sorites-proper shows at least two things: (1) it shows that the Stoics themselves acknowledge that there are at least some cases in which suspending judgement is the right thing to do; (2) their ad hoc strategies to deal with sorites-proper and other insolubilia show that their logic is not really powerful enough to deliver the systematic and secure distinction between valid and invalid arguments, truth and falsehood, in which there lies its philosophical value. From the Stoic point of view, (1) is a fair point, but not one the Stoics would be upset about: a pragmatic solution to paradoxes does not undermine any piece of their logical system. And (2) sounds worst than it is: it is true that insolubilia are very specific arguments requiring a dedicated treatment; but the treatment is not ad hoc in the sense of being out of place in their system or in the sense of not being the object of serious and systematic investigation (as is shown by the vast amount of works that the Stoics dedicated to the topic).
This being said, what could the Pyrrhonian sceptic do with this wealth of logoi between Stoics and Academics? I suggest that the omnivorous Pyrrhonian sceptic can not only rely on the disagreement between Stoics and Academics as such, but he might also find something interesting, from the Pyrrhonian point of view, in this whole debate. One feature of the Stoic approach to the sorites-proper seems to be that it was regarded as a fairly self-contained topic of investigation, both in the sense that it does not look as if the Stoics thought that the problems posed by the sorites should lead to a revision of their broader theory of logic,30 and in the sense that the Stoics themselves (at least, Chrysippus) may have taken up the challenge of thinking more carefully about the validity of logical principles (such as the principle of bivalence) with reference to the special and exceptional case of some insoluble sophisms.31
The Stoic resort to the exceptional status of some sentence(s) is not an isolated move in ancient Greek philosophy.32 But one important feature of the Stoic strategy in the present case is that it seems to presuppose that one can distinguish between ‘normal’ cases, providing the rule, and deviations from the normal cases. This is a presupposition that clearly informs not only the discussion of particular logical problems, but also and perhaps more perspicuously the Stoic position in the debate with the Academics about cognitive impressions as the criterion of truth.33 I would like to suggest that, in the same way in which the Academics challenge the possibility of distinguishing between cognitive and non-cognitive impressions, the Pyrrhonians might challenge the Stoics on the possibility of distinguishing between valid and invalid arguments by following at least partially similar strategies. Insolubilia occupy a grey area between valid and invalid arguments. The Stoics isolate insolubilia as a sort of logical peculiarity, as an exception, which they can approach in a pragmatic way without undermining their logical theory: but, to the Pyrrhonian sceptic, it is unclear that these are exceptions that can be isolated because the Pyrrhonian sceptic cannot really distinguish between pragmatics and logic. The Pyrrhonian sceptic approach to all arguments is pragmatic. From a Pyrrhonian point of view, the very fact that Stoics resort to the “sceptic strategy” in some cases which they acknowledge as challenging is telling because the Pyrrhonian does not see the Stoic justification for treating such cases as isolated. But if there is no principled way to contain the snares of insolubilia within clear boundaries, it becomes unclear when exactly we can trust reason and when we cannot, and this, from a Pyrrhonian point of view, should undermine the Stoic confidence in the power of reason.
The point about the exceptional and ad hoc treatment of insolubilia is, of course, already an Academic point (see Cicero, Ac. 2.97). But it is at least unclear that someone like Carneades would or could adopt the same deflationary attitude towards the distinction between good and bad arguments which, as I suggested, seems to be in line with the Pyrrhonian profile. If all I can say about arguments is how each of them affects me, what basis is there for me to draw a general boundary between valid and invalid ones? Anaxagoras doubted the senses because they are not capable of discerning differences little-by-little (kata mikron: M VII 90–91) from white to grey to black. Does the Pyrrhonian have any reason to think that the spectrum between valid and invalid is different from the spectrum between white and black, allowing for differences kata mikron that we cannot detect?
The confidence in reason of those who endorse logical theories is built in a similar way as their confidence in their criterion of truth. In the same way in which the Stoics assume some special type of impression as the criterion of truth, so dogmatic logicians (Peripatetics and Stoics alike) assume some patterns of arguments as immediately evident (the indemonstrables).34 The very fact that dogmatic logicians themselves admit that there are aporoi logoi, for which they might even invoke what to the sceptic may look like an exceptional suspension of their logical rules and assumptions, should show the dogmatists that their confidence in their logical theories and their general applicability is unjustified. The premises of aporoi logoi end up having the special status of sentences that do not express a proposition, but rather correspond to impressions that can be both persuasive and unpersuasive, true and false, or neither true nor false.35 But this is precisely how declarative sentences in dogmatist disputes appear to the sceptic all along. Is there any reason to restrict that special status to the premises of insolubilia?
On the sceptic account, insolubilia should trouble the dogmatic logician as a haunting reminder of the weakness of their logic, and, perhaps, of reason more generally.36 As the fictional Carneades emblematically says to his Stoic interlocutor: ‘As far as I am concerned […] you can snore if you like as well. But how does that help you? There’s someone coming after you who’s going to wake you from your sleep and keep asking you the same questions. Whatever number it was that you stopped at, if I add one to it, will that make ‘many’? Carry on further, for as long as you like. What more do I need to say? After all, you admit it: you can’t specify the last time it is ‘few’ and the first time it is ‘many”.37
From a Pyrrhonian point of view, this is a clear case in which dogmatism, rather than providing the rock of certainty on which the dogmatists can rely and rest, ends up troubling their mind. Logical theory does not seem (quite literally) worth the trouble for the Pyrrhonian sceptic.38 It seems not only useless whenever we really need to diagnose a problem in an actual argument, but also harmful to those very people it was supposed to help.
Stoic recommendations on how to behave when they face aporoi logoi such as the sorites correspond to the sceptic default attitude. The language of aporia is spread across Sextus’ writings, but perhaps it is not a matter of chance that aporos is used also to indicate a logos39 on which the sceptic suspends judgement. The sceptic ‘school’ or way of life is also called aporētikē, not only in the sense that it consists in raising aporiai, but also in the sense that it takes its name ‘from being at a loss with respect to assent or denial’. Those who cannot assent or deny, do not answer ‘yes’ or ‘no’: they suspend judgement and just stop answering. If one does not answer, as the passage in Cicero emphasises, the whole responsibility of the argument falls on the side of the questioner, if the questioner decides to keep on going. The option of disengaging from the argument in dialectical exchanges is open to both the sceptic and the dogmatist, but while the dogmatist (on the sceptic account) is and should be troubled by the fact that he cannot justify when exactly he should disengage, the sceptic remains untroubled and confirmed in his choice of life.
5 Some Conclusions: the Sceptic and Varieties of Mistrust of Logic
Of the various philosophical traditions Sextus interacts with, only two (i.e., Stoics and Peripatetics) purposefully and systematically developed tools for the explicit analysis of arguments as part of their logical investigations. Despite the differences between the Stoic and the Peripatetic approach (including their different views on the status of logical investigations as a part or a tool of philosophy,40 and on the type of arguments that count as sullogismoi),41 these two traditions are relatively isolated in championing the importance of the systematic analysis of arguments and their structure as a valuable contribution to philosophy. In fact, they both needed to come up with arguments to defend their position.
This does not imply that no contribution to the study of arguments came from other traditions or single authors.42 Even less does the Peripatetic and Stoic monopoly of logical theory imply that other philosophers did not use and discuss arguments in their philosophical activity: obviously, they all did. But the Pyrrhonian sceptic is not alone in dismissing logical investigation as such. Mistrust of logic in antiquity comes in a spectrum and it might be helpful to compare the Pyrrhonian stance to those of others.
I cannot do justice to all figures and nuances that would be worth considering, but one important overarching consideration is that, in dealing with other varieties of logical mistrust, we need to introduce qualifications, whereas it is unclear that the Pyrrhonian sceptic will want to qualify his scepticism towards logic in any clear way. One consequence of this is that, while others reject some aspects of logical theory but do not radically challenge the idea that the activity of reason complies with some general rules, it is unclear that the Pyrrhonian sceptic wants to make room for this idea in the first place – at least, in any way that would imply commitment to any logical principles. To illustrate this point, let us briefly consider a couple of cases.
Epicurus, often represented in later sources as a detractor of ‘dialectic’,43 was clearly concerned with epistemological issues, including methods for reliable inferences from what is clear to what is unclear. Closer investigation suggests that his attitude towards ‘dialectic’ was one of general diffidence towards the advantages of any abstract analysis of arguments.44 The very exercise of philosophy is based on the use of reason and arguments, but all we need in order to make progress is already embedded in the natural formation of pre-conceptions and in the appropriate use of natural language. While people can use the natural capacity of reasoning more or less well, and it is certainly helpful to have ‘rules’ on how to do this best, one does not need any sophisticated logical theory in order to see that there is something wrong with the paradox of the Veiled and take a laugh at it.45 This “natural” use of reason is all we have and all we need in order to make progress in knowledge and free ourselves from fear.46
From the opposite end of the philosophical spectrum, for Plotinus the soul of the philosopher is capable of reaching through to the real essence of things and does not deal with ‘naked theorems and rules (psila theōrēmata […] kai kanones)’ (Enn. I.3, 5 l. 11; see, more generally, Enn, I.3, 4 ll. 18–23; 5, ll. 10–23). The philosopher, Plotinus goes on, will know ‘incidentally’ what arguments are sophisms, and what arguments are good. Plotinus’ point seems to be that the philosopher can “naturally” distinguish between good and bad arguments without any need for the theoretical investigation of technical argument analysis (which is left, quite dismissively, to others). Plotinus was probably comforted by the fact that Plato himself never bothered to distinguish the different types of syllogisms or of fallacies, but he was nonetheless capable of using different types of logoi – according to Plotinus, in a rather effective way.47
Despite the obvious differences, both Epicurus’ and Plotinus’ confidence in reason without logical theory is based on what the Pyrrhonian sceptic would regard as dogmatic assumptions about the structure of the world and/or of the nature of the human soul. This route to confidence in the practice of reasoning is not available to the Pyrrhonian sceptic. A more complex, but possibly more informative comparison, might be that between the Pyrrhonian attitude and the attitude of the Academics. Again, in the case of the Academics, qualifications seem to be crucial.48 Both Arcesilaus and Carneades were, to some extent, detractors of logical theory in the sense that they were detractors of a specific logical system (i.e., the Stoic logical system), which, by the time, had come to be regarded as the logical system. However, both Arcesilaus and Carneades were outstanding dialecticians and must have found some value in logical training, including the capacity to detect coherence and validity. But it is at least very dubious that the same attitude can be detected in the Pyrrhonian Sextus. Lists of arguments are piled up against each other, without any clear assessment or criticism. Is the Pyrrhonian sceptic sensitive to validity? In this paper I have argued that the sceptic simply cannot be sensitive to validity understood as a universal and objective feature of arguments. The Pyrrhonian sceptic engagement with arguments can only be piecemeal, without theory and, for all relevant purposes, pragmatic.
But what about the distinction drawn in PH III 280–281 between stronger and weaker logoi as the sceptic remedy for different stages of dogmatic disease? The passage is problematic both because (a) it suggests that the sceptic is supposed to be able to draw such a distinction, and because (b) it suggests that the sceptic can intentionally rely on the distinction and resort to weak arguments. (a) may commit the sceptic to the endorsement of some logical theory, and (b) may turn the sceptic into a sophist.49 We can only speculate on what the sceptic would have to say about these accusations against the backdrop drawn in this paper, but in principle the sceptic seems to have some replies to both (a) and (b). As for (a), the sceptic could say that the distinction concerns the way in which arguments appear to individuals and is not based on any universal account of validity. As for (b), even when the sceptic uses (what appear to him or to others to be) weaker arguments, the sceptic (i*) acts out of his philanthropic motivation and not with the aim of deceiving and (ii*) does not produce arguments in order to induce acceptance of a conclusion in the ways distinctive of a sophism. So, whatever the sceptic is doing, the sceptic activity does not produce sophisms under Sextus’ account of sophisms since (i*) and (ii*) set apart the sceptic arguments from sophisms with respect to both (i) and (ii) (on (i) and (ii), see section 3). Quite to the contrary, Sextus occasionally suggests that it is rather dogmatic logicians that, through their ingenious arguments, can deceive people.50
Even if the sceptic can defend scepticism from accusations (a) and (b), the picture drawn in the paper raises a further and more radical question. The attitude towards sophisms and the emphasis on the conclusion may suggest that all the sceptic needs to look at are single sentences (the conclusions): if they are clearly false or not clear, they should be rejected or one should suspend judgement; if they are clear, the sceptic may accept them and live with them independently of whether there is any argument in support of them. So, what is the continuous scrutiny of arguments pro and contra any given thesis supposed to deliver? Couldn’t the sceptic simply go for suspension of judgement or endorsement of what seems clear without even engaging in this rather time-consuming activity of setting up arguments against each other?
To conclude, I make two complementary suggestions on this point. Let me stress once again that, on the account given in section 3, the conclusion cannot be detached from the argument in order to bring about that combination of compulsion and puzzlement which is characteristic to the sceptic reaction to sophisms. One suggestion is that the sceptic cannot give up on the sceptic activity for the simple reason that suspension of judgement is not the default attitude for any human being: the sceptic needs some practice that keeps him in that modality – a modality that, as Hume quite emblematically characterised, is in contrast with the pull of nature. However, what the sceptic activity brings to the sceptic is perhaps not necessarily an active engagement with those arguments, but rather a (quite complex) continual reminder of how he came to suspension of judgement. There are some passages (e.g., PH II 103; 130; 177) where Sextus explicitly speaks of arguments pro and contra as ‘reminders’, ‘commemorative signs’ of the equipollence of the two opposite views. Since the sceptic has no objection to commemorative signs, this use of argument seems (at least superficially) in line with other aspects of his attitude.
The second suggestion is that it could still be the case that individual Pyrrhonian sceptics may feel psychologically inclined towards the conclusion of an argument rather than to the opposite; in this respect, individual Pyrrhonian sceptics will differ from each other. But all Pyrrhonian sceptics can rely on the modes, the use of which is not supposed to lead to a conclusion the sceptic assents to, but rather to the act of suspension of judgement (suspension of judgements is what the modes ‘conclude’).51 This act is the outcome of the practice of the sceptic activity informed by the modes, a practice for which psychological peculiarities and inclinations of the individual sceptic will be irrelevant. So, after this whole reflection, the sceptic is left with the appearance, which he does not question, of what people call ‘reason’ to deal with logoi without any logical theory and without any robust logical distinctions. This reason, however, apparently can be trained, in a non-dogmatic way, to practise suspension of judgement methodically, where a piecemeal pragmatic approach to arguments is all is left once the sceptic mistrust of logic is fully spelled out.52
Acknowledgements
I would like to thank James Allen, Katerina Ierodiakonou, Máté Veres and two anonymous readers for all their comments on earlier drafts of this paper.
Issue and Editors
This article is part of the special issue “Scepticism and Argument: Sextus Empiricus on Logic”, edited by Máté Veres and Katerina Ierodiakonou.
Bibliography
Primary Texts
[Aristotle]. Aristotelis Topica et Sophistici Elenchi. Ross, W. D. (ed.), 1958. Oxford: Clarendon Press. (= Top.; SE)
[Alexander of Aphrodisias]. Alexandri Aphrodisiensis in Aristotelis Topicorum libros octo commentaria. CAG II.2. Wallies, M. (ed.), 1891. Berlin: Reimer.
[Cicero]. Fasc. 42: Academicorum Reliquiae Cum Lucullo. Plasberg, O. (ed.), 1922. Stuttgart: Teubner. (= Ac.)
[Cicero.] On Academic Scepticism. Brittain, C. (tr.), 2006. Indianapolis, Cambridge: Hackett.
[Sextus Empiricus]. Sexti Empirici Opera vol. I.
ΠΥΡΡΩΝΕΙΩΝ ΥΠΟΤΥΠΩΣΕΩΝ . Mutschmann, H. (ed.), 1912 & Mau, I. (ed.), 1958. Leipzig: Teubner. (= PH)[Sextus Empiricus.] Sexti Empirici Opera vol. II. Adversus dogmaticos libros quinque (adv. mathem. VII–XI) continens. Mutschmann, H. (ed.), 1914. Leipzig: Teubner. (= M)
Secondary Literature
Allen, J. 1994. Academic Probabilism and Stoic Epistemology. The Classical Quarterly 44(1), 85–113.
Barnes, J. 1997. Logic in Academica I and the Lucullus. In: Inwood, B. & Mansfeld, J. (eds.), Assent and Argument: Studies in Cicero’s Academic Books. Leiden, New York, Copenhagen: Brill, 140–160.
Barnes, J. 1999. Aristotle and Stoic Logic. In: Ierodiakonou, K. (ed.), Topics in Stoic Logic. Oxford, New York: Oxford University Press, 23–53.
Barnes, J., Bobzien, S., & Mignucci, M. 1999. Logic. In: Algra, K., Barnes, J., Mansfeld, J. & Schofield, M. (eds.), The Cambridge History of Hellenistic Philosophy. Cambridge: Cambridge University Press, 77–176.
Bett, R. 1990. Carneades’ Distinction Between Assent and Approval. The Monist 73, 3–20.
Bobzien, S. 2002. Chrysippus and the Epistemic Theory of Vagueness. Proceedings of the Aristotelian Society 102, 217–238.
Bobzien, S. 2003. Logic. In: Inwood, B. (ed.), The Cambridge Companion to the Stoics. Cambridge: Cambridge University Press, 85–123.
Bobzien, S. 2012. How to Give Someone Horns. Paradoxes of Presupposition in Antiquity. History of Philosophy & Logical Analysis 15, 159–184.
Bobzien, S. 2020. Demonstration and Indemonstrability of the Stoic Indemonstrables. Phronesis 65, 355–378.
Brittain, C. 2005. Common Sense: Concepts, Definition, and Meaning In and Out of the Stoa. In: Inwood, B. & Frede, D. (eds.), Language and Learning: Philosophy of Language in the Hellenistic Age. Cambridge: Cambridge University Press, 164–209.
Burnyeat, M. 1997. Antipater and Self-Refutation: Elusive Arguments in Cicero’s Academica. In: Inwood, B. & Mansfeld, J. (eds.), Assent and Argument: Studies in Cicero’s Academic Books. Leiden, New York, Copenhagen: Brill, 277–310.
Burnyeat, M. & Frede, M. (eds.), 1997. The Original Sceptics: A Controversy, Hackett.
Castagnoli, L. 2018. Dialectic in the Hellenistic Academy. In: Bénatouïl, Th. & Ierodiakonou, K. (eds.), Dialectic after Plato and Aristotle. Cambridge: Cambridge University Press, 168–217.
Cavini, W. 1993. Chrysippus on Speaking Truly and the Liar. In: Doering, K. & Ebert, T. (eds.), Dialektiker und Stoiker: Zur Logik der Stoa und ihrer Vorläufer. Stuttgart: Steiner, 85–109.
Frede, M. 1974. Stoic vs. Aristotelian Syllogism. Archiv für Geschichte der Philosophie 56, 1–32.
Gourinat, J.-B. 2018. Stoic Dialectic and its Objects. In: Bénatouïl, Th. & Ierodiakonou, K. (eds.), Dialectic after Plato and Aristotle. Cambridge: Cambridge University Press, 134–167.
Ierodiakonou, K. 1998. Aristotle’s Logic: an Instrument, not a Part of Philosophy. In: Avgelis, N. & Peonidis, F. (eds.), Aristotle: Logic, Language and Science. Thessaloniki: Sakkoulas Publications, 33–53.
Ierodiakonou, K. 2018. Dialectic as a Subpart of Stoic Philosophy. In: Bénatouïl, Th. & Ierodiakonou, K. (eds.), Dialectic after Plato and Aristotle. Cambridge: Cambridge University Press, 114–134.
Machuca, D. 2009. Argumentative Persuasiveness in Ancient Pyrrhonism. Méthexis 22, 101–126.
Machuca, D. 2011. Pyrrhonism and the Law of Non-Contradiction. In: Machuca, D. (ed.), Pyrrhonism in Ancient, Modern, and Contemporary Philosophy. New York: Springer, 51–78.
Machuca, D. 2017. Again on Sextus on Persuasiveness and Equipollence. Archiv für Geschichte der Philosophie 99(2), 212–228.
Machuca, D. 2021. Can the Sceptic Search for Truth? Elenchos 42, 321–349.
Marchand, S. 2010. Le Sceptique Cherche-t-il Vraiment la Vérité. Revue de métaphysique et de morale 1, 125–141.
Mignucci, M. 1999. The Liar Paradox and the Stoics. In: Ierodiakonou, K. (ed.), Topics in Stoic Philosophy. Oxford, New York: Oxford University Press, 54–70.
Morison, B. 2015. What is a Perfect Syllogism? Oxford Studies in Ancient Philosophy 48, 107–166.
Obdrzalek, S. 2006. Living in Doubt: Carneades’ Pithanon Reconsidered. Oxford Studies in Ancient Philosophy 31, 243–280.
Perin, C. 2006. Pyrrhonian Scepticism and the Search for Truth. Oxford Studies in Ancient Philosophy 30, 337–360.
Perin, C. 2010. The Demands of Reason: An Essay on Pyrrhonian Scepticism. Oxford: Oxford University Press.
Pullman, G. L. 1994. Reconsidering Sophistic Rhetoric in Light of Skeptical Epistemology. Rhetoric Review 13.1, 50–68.
Rescher, N. 2013. Cosmos and Cognition. Studies in Greek Philosophy. Berlin: De Gruyter.
Sedley, B. 2018. Epicurus on Dialectic. In: Bénatouïl, Th. & Ierodiakonou, K. (eds.), Dialectic after Plato and Aristotle. Cambridge: Cambridge University Press, 82–113.
Striker, G. 2001. Scepticism as a Kind of Philosophy. Archiv für Geschichte der Philosophie 83, 113–129.
Svavarsson, S. H. 2014. Sextus Empiricus on Persuasiveness and Equipollence. In: Lee, M.-K. (ed.), Strategies of Argument: Essays in Ancient Ethics, Epistemology, and Logic. New York: Oxford University Press, 356–374.
Veres, M. 2020. Keep Calm and Carry On: Sextus Empiricus on the Origins of Pyrrhonism. History of Philosophy and Logical Analysis 23, 100–122.
Vogt, K. 2011. The Aims of Skeptical Investigation. In: Machuca, D. (ed.), Pyrrhonism in Ancient, Modern, and Contemporary Philosophy. New York: Springer, 33–50.
Vogt, K. 2022. Vagueness and Kataleptic Impressions. Aristotelian Society Supplementary Volume 96(1), 165–183.
Williams, M. 1988. Scepticism Without Theory. Review of Metaphysics 41, 547–588.
For the sake of brevity, I omit the qualification ‘Pyrrhonian’ unless the qualification is essential to distinguish between Academic sceptics and Pyrrhonian sceptics.
For a particularly clear account of this view, see Vogt 2011, 42: “It is only through the focus on argument that the sceptic can arrive at this balanced state of mind”; cf. 43 “Sextus’ skeptic does not move away from the use of reason. The skeptic is a reasoner, and must be a reasoner.”
The passage suggests that logoi can be so forceful as to compel assent even to what is contrary to the evidence of the senses.
Henceforth: M.
Cf. M VIII 301, 309 and the sceptic reply to the dogmatists’ accusation that the Sceptic resorts to reason to come up with arguments and is affected by arguments (M VIII 275–298).
Cf. M VIII 321–322 and 337–336a, on the possibility of discussing proofs based on a concept (or preconception) of proof that is not an apprehensive impression of it. On the sceptic precautions for using language in such a way that it does not convey a dogmatic position, see PH I 187–209. On concepts, see Brittain 2005.
On this point and how it relates to the sceptic views on logical principles such as the principle of contradiction, I am sympathetic to Machuca 2011.
See, perhaps, M VIII 301.
On sceptic ‘investigation’ as a rational procedure see e.g., Machuca 2021; Marchand 2010; Perin 2006; Perin 2010, especially Ch. 1; Veres 2020; Vogt 2011. For a reading of Pyrrhonian scepticism on the side of irrationalism, see Striker 2001.
Cf. Alex., in a. pr., Wallies 8.19–29 on why the philosopher deals with sophisms; D.L., Lives VII 46–48 on why logic is a necessary part of philosophy for the Stoics.
There is an issue as to how to translate the Greek endoxon. Some qualifications are required, but, broadly speaking, an endoxastic premise is a premise which is endorsed by all, or by most, or by some wise people. The convergence of consensus on the premise makes it ‘reputable’.
See Allen 1994; Obdrzalek 2006.
Machuca 2009, Svavarsson 2014 and Machuca 2017 discuss persuasiveness in Sextus in relation to the issue of whether the label ‘persuasive’ is supposed to indicate what actually persuades the sceptic, or ought to persuade the sceptic (Svavarsson), or affects the sceptic in a certain way without necessarily actually compelling the sceptic to endorse what seems persuasive (Machuca). I cannot do justice to the whole discussion, but I agree with Machuca 2017 that persuasiveness refers to the way in which the sceptic registers a certain pathos. Machuca 2017 introduces a further distinction, between epistemic persuasiveness and psychological persuasiveness, to make room for the possibility that the sceptic may feel (psychologically) more inclined towards a certain view than to another, even if, epistemically, he suspends judgement on both. I shall briefly return to this point at the end of section 5.
On this score the Stoics have a clearer account than the Peripatetics about the subject matter of logic: see Ierodiakonou 1998. The ontological status of arguments for the Peripatetics is at least to some extent unclear.
On the objects and value of stoic logic, see Ierodiakonou 1998 and 2018; Gourinat 2018.
An explicit reference to this way of thinking about arguments can be found, quite tellingly, in Sextus, M VIII 302.
Cf. Arist., Top. VIII, 160b23–161a15.
This is presumably a reference to the discussion of proof in PH II 144 ff. and 156 ff.
On this point, see Johanna Schmitt’s paper in this volume.
An anonymous reader suggested to me that perhaps Sextus’ account here is somehow influenced by the tradition of dissoi logoi. This is certainly possible. More generally, the Pyrrhonian way of thinking in terms of opposite equipollent arguments may suggest to Sextus that any two contrasting views can be expressed in terms of a pair of arguments with opposite conclusions.
Similar considerations would apply mutatis mutandis to the case in which the sceptic is led to doubt a clear conclusion because of the accompanying argument (PH II 233).
This is, of course, unfortunate. However, the sceptic may have good reasons to raise doubts about assent as a distinct mental act: see Bett 1990 for a thought-provoking discussion of this point with reference to Carneades. Since the Pyrrhonian sceptic can only rely on individual reports (see sextion 2), what people say and concede might indeed be all the Pyrrhonian can say about other people’s assent.
I cannot fully explore this point, but the importance of prohairesis in the attitude towards arguments is not a trivial point: cf. Aristotle, SE 165a30–31, Met. IV 2, 1004b24–26.
Since the Stoic sage is only supposed to assent to apprehensive impression, he will not assent to any impressions that are not apprehensive.
On this feature of insolubilia, see Bobzien 2002, 220; 223 n. 25.
Cf. D.L., Lives, VII 65; Sextus, M VIII 74.
Cf. Cicero, Ac. 2.95. On the Stoic approach to the Liar see Cavini 1993; Mignucci 1999.
I do not have the space to expand on this point, but the view that the Stoics offer a pragmatic solution to paradoxes is substantiated by Cavini 1993 and Bobzien 2012.
For a discussion of this strategy as it is portrayed in Cicero’s Academica, see Barnes 1997. More generally on Academic dialectic and Hellenistic debates see Castagnoli 2018.
Cf. Bobzien 2002, 218.
Bobzien 2002, 220–222.
For a fascinating discussion of exceptional sentences in philosophy, starting with the Socratic: ‘The only thing I know is that I do not know’, see Burnyeat 1997. Cf. also Aristotle, Met. IV 8, 1012b18–22; Sextus, M VIII 479–481.
On the sorites in this context, see Vogt 2020.
On the meaning of the designation ‘indemonstrable’, see M VIII 223; cf. PH II 146 ff., 156. For a thorough discussion, see Bobzien 2020. Notice that why it is exactly that some syllogisms are designated by Aristotle as “perfect” is a controversial issue: see e.g. Morison 2015.
On the different types of impressions, and the explicit mention of aporoi logoi in this context, see M VII 242–243.
The section on sophisms in Cicero, starting in Ac. 2.91, is meant to show the weakness of reason more generally. It does this by showing the weakness of Stoic logic, but it could be objected that the weakness of a particular logical system need not show the weakness of reason as a whole. However, the Pyrrhonian sceptic can only speak about reason with reference to the way in which people report about it, and the Stoic reports are certainly among the most impressive. Notice that what the Pyrrhonian sceptic regards as haunting and troubling could just be an exciting challenge for the Stoic.
Cicero, Ac. 2.93, trans. Brittain.
There are other places where Sextus points at the troubling aspect of logical investigations: see e.g., M VIII 292.
See e.g., M III 82, 115; M IX 436; M X 122, 245, 292.
On the status of logic as a part or a tool of philosophy, see Ierodiakonou 1998 and 2018.
On the differences between Peripatetic and Stoic sullogismoi, see e.g., Frede 1974; Barnes 1999, Barnes, Bobzien and Mignucci 1999.
Sextus himself (M VII 2–19) gives an overview of all those who gave some contribution of the logical part of philosophy.
See e.g., Cicero, Ac. 2.97; Fin. I 22; II 18; cf. D.L., Lives X 31.
Sedley 2018, 88: “One may detect here in Epicurus a general distrust of argument-analysis. How do we know it is a sophism? Less because of any lack of formal soundness than because we already know, pragmatically, that its conclusion is not one that we can live with.”
I rely on the evidence discussed in Sedley 2018, 87.
Cf. Cicero, Fin. I 63.
Cf. how Alcinous in his Didaskalikos ascribes to Plato the Peripatetic distinction of syllogisms into scientific, dialectical, rhetorical, and sophistic (ch. 3,1–2; chs. 5–6; see in particular ch. 6, 3–6, where the author spells out in what Dialogues Plato used the different types of syllogisms).
For a detailed reconstruction of the complex attitude of the Academics towards dialectic and a thorough discussion of the main sources, see Castagnoli 2018.
There have been studies stressing the similarities between sceptics and sophists: see e.g., Pullman 1994; Rescher 2013, ch. 4.
See the use of the language of heuresilogia (‘inventing ingenious arguments’, also in a derogatory way) to indicate the logical part of dogmatic philosophy, e.g. PH II 84. See also PH II 61–62 on the deceptive inclination of the clever mind (cf. PH II 42).
For the terminology of ‘concluding’ (sunagein), see e.g., PH I 36, 99, 123, 124, 135. Alternative formulations include: suspension of judgement will ‘follow’ (akolouthein): PH I 140; suspension of judgement is ‘introduced’ (eisagesthai): PH I 79, 89, 117; we are ‘compelled’ (anankazesthai) to suspend judgement: PH I 78; 121.
At least since Burnyeat & Frede 1997, it is well known that there are different ways to read the radicality of Pyrrhonian scepticism. I do not have the space to expand on this point, but the Pyrrhonian sceptic I outline in this paper may end up being a „sceptic without theory“ as in Williams 1988.
