by Tim O'Keefe
A different reply is offered by the founder of Academic Scepticism, Arcesilaus. This reply eschews even the weak sort of assent to something as plausible that Cicero and Carneades allow. The sceptic, says Arcesilaus, will still have sense-impressions: they arise in him involuntarily. So he will see the cliff in front of him. Likewise, he will have impulses to act. Instinctively, he will have an aversion to stepping off the cliff. These impressions and impulses will suffice for the sceptic to move around in the world, avoiding cliffs, stepping through doors and eating peanut butter sandwiches. The only thing that the sceptic avoids is assent. That is, the sceptic never decides to believe that there is a cliff in front of him, and likewise with the propositions that his other impressions furnish him. But he should not assent to these things, as doing so would be rash, and he need not assent to them, as impulses to act will arise anyway (Plut. Adv. Col. 1122a–f, IG III-12, LS 69A). Sextus Empiricus gives a similar account of how a sceptic can act. The sceptic has no beliefs about how things are. But the sceptic does not abolish the appearances, including sense-impressions and feelings. So it will still seem that there is a cliff in front of him, he will feel fear, it will look as if there is a peanut butter sandwich in front of him, it will seem tasty, he will feel hungry and so on. These impressions will move the sceptic around here and there without any beliefs (Sext. Emp. Pyr. I xi 21–4). So the sceptic has no criterion for belief, but he does have a criterion of action: the appearances.3
The Epicureans would probably have two objections here. The first is that the Arcesilean (and Sextan) reply implicitly assumes a faulty model of belief. Arcesilaus, for example, claims that he has no beliefs when he avoids the cliff and reaches for the sandwich. But Epicurus would say that Arcesilaus does really have beliefs, whatever he might say to the contrary, and that he has beliefs is evinced by his behaviour. Arcesilaus is assuming a Stoic model of cognition, not coincidentally, as the Stoics were the main dialectical opponents of the Academic sceptics. The Stoics sharply distinguish between human and animal action. Animals have involuntary impressions (“fire ahead”) and the impulses that follow from them (“flee!”). For human beings, however, the process whereby impressions lead to impulses is mediated by assent. It is up to us whether to assent or not to the content of our impressions, and it is only when we assent that fully fledged beliefs are formed. For Arcesilaus, the sceptic can steer himself around as animals do without having the voluntary assents that lead to belief. But Epicurus would not make such a sharp divide between animal and human cognition. Animals have beliefs, as we do, and that some belief arises in us involuntarily and irresistibly does not mean that it is not really a belief.
Secondly, the Epicureans could repeat the point above about using the “plausible” as a criterion, if plausibility means merely subjective convincingness. Let us grant that animals act without thinking about whether or not their impressions are true. Nonetheless, their actions are effective because the senses of animals provide them with information about their environment. But the sceptic does not believe that the senses are reliable guides to the way the world is. So the sceptic seems to have no good justification for acting as he does.
TEN
The canon
The truth of all sensations
The first criterion of truth is sensation. But sceptics cast doubt on the veridicality of the senses, often by bringing up examples of conflicting appearances. The same wall that seems white to me will (supposedly) appear yellow to a man with jaundice. An oar seems bent when stuck partway under water but straight after being pulled out, and a tower seems round in the distance but square close up. The common-sense position is that some of these sensations are true, others false. Since the wall is really white, the oar really straight and the tower really square, the sensations that report these things are true, while the ones that report otherwise are false. But the sceptical response to this common-sense position is that we have no criterion to distinguish reliably between true and false sensations. And since we have no good way of deciding which sensations are true and which are false, we should suspend judgement.
Although Epicurus himself did not confront the sceptical challenge as formulated in precisely this way – it was put this way by later sceptics, the Academics and the Pyrrhonians – the general problem of conflicting appearances was featured prominently in earlier disputes about the reliability of the senses, and many people concluded on the basis of sensory variability that the senses could not be relied on. The Epicureans agree that if the truth of any sensation is cast into doubt, then there can be found no criterion whereby to distinguish the true from the false ones. Once we start to doubt the reliability of the senses, we cannot use reason to determine which sensations are true and which false, as reason itself is a product of sensation (KD 23, DRN IV 480–85). But as we saw in Chapter 9, the Epicureans regard such scepticism about the senses as self-refuting and unlivable. So they embrace the bold position that all sensations are true.
The position that all sensations are true appears to be itself obviously false. But Epicurus is willing to embrace its apparently absurd consequences, asserting that even the figments experienced by dreamers and madmen are true (DL X 32). The first move the Epicureans make in order to render this position at all plausible is to distinguish sharply between our sensations and the judgements about objects made on the basis of those sensations. When there is error, it enters in always in the “added opinion”, that is, in the judgements we make about the world on the basis of our sensations (Ep. Hdt. 50). When we see a “bent” oar in the water, the sensation does not tell us that the oar itself is bent. The bent-shaped patch in our visual field is just the impression we are receiving from the oar, and we make a mistake when we infer that the oar itself is bent because of that impression. (In fact, with experience we learn that straight objects normally appear bent when partially stuck under the water, so we would be more likely to go wrong if the straight oar for some reason did not appear bent.) If the bent-oar example is unconvincing, we should conclude that the senses deceive us when they report that people far away are really tiny, but, of course, the senses report no such thing. As Lucretius puts it, it is the business of the mind to make such judgements (e.g. what size the people are) on the basis of the information the senses furnish, and so we should not blame the eyes for the mind’s shortcomings (DRN IV 379–86).
Sensation does not make any judgements about the world. It just apprehends what is present to it, for example colour in the case of sight (Sext. Emp. Math. VII 210). One sense cannot refute another, because each has a different sort of object – colour for sight, sounds for hearing and so on – so their reports cannot conflict (DL X 32; DRN IV 486–96). This insulates sensations from the possibility of error, but at the apparent cost of their not having any propositional content at all, of not saying that anything is the case. So it becomes quite murky the sense in which they are all true, as opposed to neither true nor false.
One way to make all sensations true is to restrict their propositional content to what is immediately given in the sensation itself, that is, that I am having a certain sort of visual, auditory, olfactory, gustatory or tactile experience. I may not know whether the wall is white, or the orange juice sweet. But that I am having a certain sort of sense-experience is obvious and cannot be denied, and this is all that the sensation itself is really saying. The Cyrenaics (a group of rival hedonists active shortly before and around the time of Epicurus) give a theory of this sort. They coin neologisms to make it clear that they are restricting themselves to just the sensation itself, saying that we can apprehend that “I am being whitened” when I (apparently) see the white wall, or that “I am being sweetened” when I (apparently) drink the sweet orange juice, because this content is self-evident and knowledge of it incorrigible (Adv. Col. 1120e; Sext. Emp. Math. VII 191).
But the Epicureans reject this position, even though they say that sensation as such is concerned only with what is immediately present to it, for example colo
ur in the case of sight. The Epicurean Colotes mocks the Cyrenaics’ neologisms, saying that the Cyrenaics refuse to admit that a horse exists, stating instead that they are horsed. But more seriously, if sensations inform us only of themselves, and we cannot use them to gain knowledge of the world, then we would be unable to act (Adv. Col. 1120d). The Epicureans have little interest in the supposedly immediate knowledge of one’s own mental states; sensations are supposed to be one of the criteria whereby we can gain knowledge of the world and are thereby able to act effectively.
At this point, it is worth noting an ambiguity in the Epicurean slogan “all sensations are true”. The Greek term I have been translating as “true”, alēthēs, can also mean “real”. And, in fact, many of the Epicurean discussions of the “truth” of all sensations point towards their using the term to mean that all sensations are real, not that they are true. Diogenes Laertius reports that our awareness proves that every sensation is alēthēs because it is as much a fact that we see and hear as that we feel pain. Our awareness of pain, colours and sounds may show that pain and sensations of colour and sound are real, but not that they are all true.
Another proof the Epicureans give for every sensation being alēthēs – including even the figments of madmen and dreamers – is that they all cause movement, whereas what does not exist does not move anything (DL X 32). Once again, this is a fine proof that sensations are real, but a bad one that they are true, and the obvious contradictory of “what does not exist” is “what does exist”, not “what is true”. So the Epicureans are not trying to say that the propositional contents of sensations are all true (as they have none), but simply that all sensations are real things. Sextus reports that even the sensations of Orestes, who saw the Furies pursuing him, were alētheis in so far as the images existed and moved his sensation; the error was in his mind when he concluded falsely that the Furies were solid bodies instead of phantasms (Sext. Emp. Math. XIII 63, LS 16F).
But this defence of the Epicurean position leads to a very different problem. Whereas the slogan “all sensations are true” seems absurd, the slogan “all sensations are real”, that is, they exist, seems trivial and unhelpful. The Epicurean doctrine that every sensation is alēthēs is supposed to be a contribution to a debate about how we gain knowledge of the world and a response to scepticism. Responding to sceptical worries about the reliability of the senses by asserting “Yes, but sensations all exist” seems radically beside the point. The Epicureans need some way of avoiding both the Scylla of absurdity and the Charybdis of triviality.
The Epicureans do have such a way, as they do not think of sensations as purely private mental phenomena. Instead, they are effects of external causes. Objects in the environment throw off a continuous flow of “images” (eidōla) from their surfaces, and visual sensations result when these images bang into our eyeballs. Other sensations are also analysed as the result of atomic interactions between external objects and our sense-organs, for example the bitter taste of coffee is the result of barbed atoms tearing at our tongue. As such, sensations may not have propositional content in the way a statement or belief does, but they can still be informative. As Sextus Empiricus puts it, in reporting the Epicurean doctrine that every sensation is alēthēs, “every impression is the product of something existent and is like the thing which moves the sense” (Math. XIII 63). That is, not only are sensations effects, but we can reason from these effects about what sorts of things caused them.
So think of the images thrown off the surface of a square tower. A continuous series of them impacting our eyes gives us a visual sensation, like the impressions created on the stills of an old-fashioned movie camera as they absorb the light waves reflected off of the tower. The eye passively receives the impact of the images and records it. This sensation is “true” of the images in so far as the images are what (directly) move the sense, so that the sensation accurately reflects its shape. This shape will not be the same as the shape of the tower: the corners of the images are blunted as they pass through a great distance and are buffeted by the intervening air, so it looks roundish. But the image can still give us information regarding the tower, just as a photo can, even though the shape of the image on the film is not the same as the shape of the object itself.1 After all, an object seen from a distance should not look the same as one seen close up.
Preconceptions
The second criterion of truth is “preconception”, or prolēpsis. Prolēpsis is a technical term coined by Epicurus, and can also be translated “basic grasp”. Lēpsis comes from the verb “to grasp”, but it often is used to form words that have overtones of a cognitive grasp, so “basic cognition” would serve as a translation too.
Epicurus uses preconceptions to solve the celebrated paradox of enquiry, which Socrates puts forward in Plato’s dialogue the Meno (Men. 80d–e). The paradox is supposed to show that enquiry is impossible. Let us suppose you are trying to learn something, such as what virtue is. Either you know what you are looking for, or you do not. If you already know it, there is no point in enquiring after it. But if you do not already know what you are looking for, enquiry is impossible, as you would not be able to recognize the correct answer even if you were to come across it. In response to this paradox, Socrates develops the theory that all “learning” is really recollection. You already know what you are enquiring after, but only implicitly. When you discover truth you recognize it, fully remembering what you had half-forgotten. This innate knowledge that makes enquiry possible, Socrates asserts, must have existed in our souls in a prenatal existence, with the shock of being embodied causing our souls to have amnesia.
Epicurus opts for a more economical solution. He agrees that enquiry requires previous knowledge; for example, I cannot ask “Is that thing over there a horse or a cow?” unless I already know what a horse and a cow are (DL X 33). And when it comes to the definitions of words, not all words must require definition in terms of other words, on pain of an infinite regress. Instead, the meanings of some words we simply grasp without need of additional proof (Ep. Hdt. 37–8). When we have repeated sense-experiences of the same sort of thing, this gives us a memory of what has often appeared, and this memory is the universal idea or preconception. When the word associated with that concept is uttered, we call up the memory; for example, when somebody says “human being”, I immediately have a general outline of a human being and think “that sort of a thing is a human being” (DL X 33). We may enquire further about the features human beings have in common that make them human beings, but that is not the meaning of “human being”. Instead, “human being” simply picks out those sorts of things over there, those things I have seen around all over the place. As with the instinctive animal utterances that form the basis of human language (see Ch. 6, § “Language”), the meanings of preconceptions are set by the sorts of items that cause them.
Once we have preconceptions, further ideas are formed by analogy or by similarity or by compounding these basic ideas. But since preconceptions themselves are formed by our experiences, ultimately all of our ideas are based on our experiences. Whatever questions may be raised about this, the basic idea seems fairly straightforward for many of our preconceptions, such as “human being” and “body”. But included among our preconceptions are usefulness (DRN IV 853–7), truth (DRN IV 473–9), god (which we shall look at in Chapter 16) and our own agency, where the supposed empirical basis is much less obvious. Unfortunately, we do not have explanations of how most of these preconceptions arise.
But a discussion of how we come to have the preconception of our own agency has been preserved. Some things happen of necessity, whereas others depend on us and are our own responsibility. Our preconceptions of necessity and of our own responsibility arise by observing ourselves in action, as we deliberate among options, advise one another and admonish one another. We see that sometimes we can do things we do not want to (e.g. submitting to a root canal treatment now to avoid greater pain later), and that we can dissuade
others from doing something they are considering only because they are being threatened (e.g. convincing someone not to betray his friend despite the prospect of the rack). It is from these observations that we get the idea that some things are of necessity while others depend on us. We show our awareness of the distinction in our interactions: we try to dissuade others from actions that “depend on us”, those that are under our rational control, which would be pointless for those that are of necessity. So our idea of our own agency does have an empirical basis, albeit one that is not as straightforward as the empirical basis of the idea “cow”.2
Confirmation and multiple explanations
Sensations form the first and primary part of the canon, the yardstick whereby we decide what is the case. Every one of them gives us some information about the world. But this obviously leaves open the question of how we go from this information to make judgements about objects, as sensations themselves do not do this. Once repeated sense-experiences have formed preconceptions in us (and we develop further ideas formed via operations on these basic ideas), we can make judgements about the world, framed in terms of these concepts, based on the information we receive through our senses. We do not just get some colour and shape in our visual field, we think “Oh, maybe there is a round tower over there”. As we have noted, this stage is where error arises.
The Epicurean account of how we can make accurate judgements about the world regarding day-to-day empirical matters is surprisingly humdrum. We form a preliminary opinion about something based on our sensation: for example, “Plato is over there” or “that tower is round”. This conjectural opinion still awaits confirmation, and whether it turns out to be true or false depends on further testimony from the senses, to see whether the object of opinion is as I believed it to be: for example, I see Plato close up and it is obvious that it is Plato; or I approach the tower, and close up it is clear that I was wrong, as the tower is actually square (DL X 34; Sext. Emp. Math. VII 211–12). So these clear, obvious observations, which allow us to confirm and disconfirm our opinions, are the basis of everything (Sext. Emp. Math. VII 216).