Saying, Showing, and Knowing: Epistemological Dimensions of Wittgenstein’s Response to Scepticism in On Certainty

Abstract

This paper evaluates a certain kind of response to philosophical scepticism that is broadly inspired by Wittgenstein’s discussion in On Certainty (OC). Special attention is paid to the idea that we show our certainty and belief in particular assumptions through our actions, practices, and language-games, and that due to their global, impractical, and largely context-free nature, sceptical doubts are not genuinely meaningful doubts. How successful is such a response? Wittgenstein’s own philosophy helps us to evaluate this type of response to scepticism. It is shown that Wittgenstein was well aware that a mere appeal to actions or practices, unaccompanied by a contextualist approach to meaning, would be insufficient as a satisfying response to scepticism. Moreover, the main thrust of Wittgenstein’s response in OC largely departs from foundationalism in that it is not an attempt to provide a ‘justification’ of our beliefs as with traditional forms of foundationalism. For this reason, this rationally reconstructed Wittgensteinian response to scepticism does not fall prey to the ‘myth of the given’ with which foundationalists are sometimes accused. Thus, this Wittgensteinian response to scepticism is a robust and successful one.

1. Introduction

Attending to actions and practices in their proper limited contexts can have an anti-sceptical effect1. Being practically oriented and engaged in coping with everyday life tends to limit the extent of one’s scepticism insofar as it is difficult, though not obviously logically impossible, to doubt the existence of the very things with which one is immediately engaged and interacting. To doubt that, say, a door exists, is to step back from merely using the door (opening it) and to reflect on it in a detached, theoretical way. To be sure, we do sometimes need to doubt the existence of particulars in everyday life. For example: ‘Is there an apple in the fruit bowl?’, ‘Is someone in the next room?’, ‘Is that a child in the back of that car?’, one might ask. We raise particular doubts in particular contexts, and these doubts are easily settled on the whole. The stakes can be high, but the means of settling these doubts are not in question. Philosophical sceptical doubts are not like these everyday doubts: they are general, all encompassing, largely theoretical doubts, without an easy means of resolution.

It is hard to live with philosophical doubt while functioning as a reasonable person.

• Sceptical doubts—such as ‘Is this really a door?’, ‘Am I really walking?’—require a withdrawal in thought from the situation at hand2. Perhaps maintaining sceptical doubt while coping with everyday life requires a split consciousness and a kind of bad faith, with one part of consciousness doubting the existence of things that the other part at some level takes for granted. It appears to involve a kind of cognitive dissonance. For this reason, a sustained lived sceptical doubt is sometimes thought to be impossible. George Santayana expressed a view like this one and concluded that our natural predicament is one of belief, not doubt. It is one of ‘animal faith’ [4]3. Philosophical scepticism is something that we acquire through education and reflection; it is not natural.

Not only does a focus on action, practice, and context help to postpone or bracket sceptical doubts, but acting and enacting practices are also the way in which we show what we take to be certain. In acting, we manifest those beliefs and assumptions on the basis of which we live our lives, assumptions which Wittgenstein thinks could hardly be called into question in most cases without our being regarded as mad. Thus, the appeal to actions and practices serves a complex, multifaceted purpose in responding to scepticism. Much of this ‘practical’ anti-sceptical strategy can be seen in Wittgenstein’s discussion in On Certainty.

A quasi-pragmatist strand in Wittgenstein’s response to scepticism has been the topic of sustained scholarship and forms the pivot of studies of the so-called ‘third Wittgenstein’, the philosopher as author of works coming after the Philosophical Investigations (see [5,6,7]). Much of the focus of this scholarship is on interpretation and textual exegesis, and documenting that this is actually the position that Wittgenstein held. Given the unfinished nature of Wittgenstein’s comments in On Certainty, this is a difficult task. In any case, the search for “the position” of Wittgenstein in OC perhaps should be resisted in light of the unfinished, fragmentary nature of Wittgenstein’s notes in OC and his views about philosophical activity. The focus of this paper is on the analytical exposition and evaluation of the philosophical ideas behind a broadly Wittgensteinian response to scepticism, both in its contextualist and pragmatist elements. The aim is to evaluate the success of the kind of reply to scepticism consisting of these elements, not to conclusively demonstrate that the resulting position actually belonged to Wittgenstein. As it so happens, I do think it represents the main thrust of Wittgenstein’s response to scepticism in On Certainty4. For the sake of simplicity, I will sometimes speak of Wittgenstein’s view, but it should be understood as a reconstruction broadly inspired by his view.

Wittgenstein does not seem to endorse pragmatism completely in OC, but he is clearly tempted by it. There is, as we shall see, an element correct in this idea: Wittgenstein—like the pragmatists —does in some way regard the appeal to actions and practices as somehow, at least partially, unravelling, loosening and untying the knots of sceptical doubts. He considers our actions as lying at the basis of our beliefs. We can be as certain about our actions and practices as we can be about anything. However, Wittgenstein is under no illusion that an appeal to actions in a way provides a justification (in the traditional sense) for our beliefs. For this reason, Wittgenstein’s overall response to scepticism in OC is not—as sometimes suggested—foundationalist. At best, one can say that there is a foundationalist strand in his writing in OC (Grayling [8]), but in reconstructing a successful Wittgensteinian response from the materials of OC, traditional foundationalism falls away. Furthermore, I will show that the Wittgensteinian response also avoids the great weakness of foundationalist views in assuming that the meaning of our experiences is given absolutely in a way that transcends particular contexts. Thus, contextualism is an integral part of a successful Wittgensteinian response to scepticism.

In what follows, Section 2 of the paper sets the scene for an investigation of philosophical scepticism by explaining the contextualist diagnosis of the nature of radical sceptical doubts. Section 3 focuses specifically on Wittgenstein’s response to scepticism, considering: his distinction between certainty and knowledge; his appeal to actions and practices; and his departure from traditional foundationalism. Section 4 concludes with an assessment of the Wittgensteinian idea that it is the degenerate context of traditional philosophy (and epistemology in particular) in the mould of Cartesian meditations—abstracted from the rich individual practical contexts of specific inquiries as in everyday life—that is itself to blame for the creation of epistemological problems such as philosophical scepticism.

2. The All-Encompassing Nature of Sceptical Doubt and the Contextualist Diagnosis of Scepticism

The contextualist diagnosis of scepticism involves noticing that it may be appropriate for someone to claim knowledge in some contexts and not others. It then involves noticing that the standards of justification one has to meet to claim knowledge vary contextually. When the sceptic then claims that one does not know that there is a chair there on the grounds that one cannot discriminate perceptually between a veridical perception of a chair and a hallucination of one, contextualists say that sceptic is insensitive to a switching of context. By drawing attention to hallucinations, the context has changed, and one can no longer take perceptual evidence at face value. One is now in the context of a philosophical reflection requiring a higher degree of certainty than required in everyday contexts. Such is, in a nutshell, the contextual diagnosis of scepticism. Wittgenstein’s On Certainty inspires much of the contemporary contextualist diagnosis of scepticism, and crucially makes the claim that philosophical sceptical doubts are actually meaningless.

In stating that scepticism does not arise in everyday life, it is necessary to appreciate the global, all-encompassing nature of philosophical scepticism. Philosophical scepticism is not restricted to doubts about our capacity to know particular truths. Philosophical scepticism raises doubts about our ability to know of the existence of anything, because it raises questions about our very general methods of gaining knowledge. The very general nature of sceptical doubt can be illustrated by considering briefly Descartes’ two great sceptical arguments: the dreaming argument and the evil demon argument in Meditations I. I assume that the readers are familiar with Descartes’ arguments and simply present the arguments in schematic form. This presentation has the advantage of making it easier to concentrate on the peculiar form of sceptical doubts, an issue that Wittgenstein pays great attention to in On Certainty, and thus an issue that can be carried over into an evaluation of a Wittgensteinian diagnosis of sceptical arguments.

Descartes’ dreaming argument has the following schematic structure:

(1)

If I am to know some particular empirical proposition p, then I must know that I am not dreaming that p.

(2)

I do not know that I am not dreaming that p.

(3)

I do not know that p.

On this reconstruction, knowing that one is not dreaming is a necessary condition for knowledge. But knowledge that one is not dreaming is not obtainable, since any apparent empirically justified belief that one is not dreaming could itself be dreamt up and therefore (assuming (1)) not knowledge. Scepticism embracing all empirical knowledge appears to follow as the argument is applied to various empirical propositions.

The Cartesian “evil demon argument for scepticism” has a similar structure:

• (ED 1) It seems to me that p.
• (ED 2) In order to infer from its seeming that p to the conclusion that actually p,
• I must know that I am not being deceived by an evil demon who brings it about that it merely looks to me as though p, but actually not-p.
• (ED 3) I don’t know that I am not being deceived by an evil demon.
• (ED 4) Therefore, I don’t know that p.

The greater scope of the argument is reflected in the fact that the proposition that p may be virtually any kind of proposition, not just a truth learned through sensory experience, but also one known through pure reason such as a truth of mathematics and logic. As with the dreaming argument, exclusion of a sceptical hypothesis is a necessary condition on knowledge. In the evil demon argument, however, the sceptical hypothesis is all-embracing. Even if one thought that one was not being deceived by an evil demon, one might be deceived in that thought. So, radical scepticism concerning the possibility of all knowledge seems to follow5.

In each argument, the sceptical hypothesis is extremely general and immune from refutation. The possibility that one is dreaming, once admitted, is a possibility that cannot be certainly excluded and will undermine one’s claim to know the necessary condition on knowledge (that one is not dreaming). Similarly, the possibility that one is being deceived by an evil demon cannot be certainly excluded once raised, and will undermine one’s claim to know the necessary condition (that one is not being deceived by an evil demon). It is this feature of sceptical doubt—its all-encompassing extreme and theoretical generality—that sets it apart from everyday doubts. It is also part of what, in the minds of contextualists, makes sceptical doubts not genuinely meaningful and not truly intelligible. For meaning is dependent on context, and the degenerate context of sceptical doubts deprives it of real meaning. This feature sets the contextualist Wittgensteinian approach apart from those responses to scepticism—including that found in American pragmatist Charles Sanders Peirce—who claimed that although the sceptic’s doubts are intelligible and meaningful, they are just implausible, artificial ‘paper doubts’ [10,11].

One of the problems with contextualism as a response to scepticism is that sceptical doubts appear intelligible and thus do seem to have meaning. Moreover, it is somewhat incredible that philosophers should grapple with the problem of scepticism perennially if the problem were merely one of noticing shifting contexts and different standards for what counts as knowledge. The problem of scepticism as a threat to claims to knowledge appears more robust than the contextualist analysis suggests6.

If the contextualist response to scepticism is to be satisfactory, it must explain why the appearance of the intelligibility and legitimacy of sceptical doubts is in fact misleading. I take it to be the case that Wittgenstein prepares us to meet this challenge in On Certainty as argued in detail below. Part of what renders sceptical doubts meaningless is that they are not raised in the context of everyday life and action. They are raised in the artificial context of doing philosophy, where the ordinary use of words is forgotten, and yet there is often not enough agreement in practice to settle on an acceptable philosophical use of words. Once philosophers set out to ‘discover’ by epistemological analysis ‘the meaning of knowledge’, they may lose a grip on how we use the term ‘knowledge’ in everyday life. But if one accepts a broadly Wittgensteinian approach to meaning, there is no other sure guide to the meaning of ‘knowledge’ other than actual everyday usage. If epistemologists ignore everyday usage, then epistemologists run the risk that the target concept of their inquiry—knowledge—becomes either arbitrary, empty of content, and of limited relevance.

3. Saying, Showing, and Knowing

It will be argued that one strand of Wittgenstein’s response in On Certainty to scepticism is to suggest that the proper response to sceptical claims involves showing rather than saying that one knows that sceptical claims cannot be taken seriously. When Wittgenstein’s response is seen in this light, the circle is complete: the contrast between what can be said and what can be shown was present already in the Tractatus, and returns in Wittgenstein’s last work as an important part of his response to scepticism7. Moreover, the notion of ‘showing’ takes on a new light in the context of the pragmatist response to scepticism. For a key aspect of pragmatist responses to scepticism is that the pragmatist attempts to show that scepticism is mistaken by performing some action. Most famously, Dr Johnson is alleged to have ‘refuted’ Berkeleian idealism by kicking a stone. Johnson’s ‘refutation’ contains the essence of a germ of the pragmatist response to idealist scepticism. Dr Johnson’s playful, irreverent response based on ‘common sense’ and ‘action’ is of the same sort, I think, as part of G.E. Moore’s common-sense reply to the sceptic. Dr Johnson acts and perceives that he acts on objects, and does not doubt that these objects exist while acting. Similarly, G.E. Moore perceives that he has hands and infers that the external world exists.

As contemporary interpretations of Wittgenstein’s response to scepticism in OC make clear, it is not really possible to understand what Wittgenstein is up to without the Moorean background. G.E. Moore’s notorious ‘Proof of the External World’ purports to be a refutation of external world scepticism8. Moore holds up each of his hands and says ‘Here’s a hand and here’s another hand’. These are the premises of the proof. (The fact that Moore gestures and holds up his hands as he says this is an essential part of the proof.) The conclusion is ‘There are (at least two) external things (at this moment)’. In bare outline, Moore’s proof consists in these claims:

(A)

Here’s a hand, and here’s another hand.

(B)

If there are hands, there’s a world.

(C)

So, there’s a world.

• As Moore is at pains to establish, his proof appears to possess the attributes of a sound and valid argument. He claims that the premises are known and different from the conclusion and indeed follow from the conclusion9. Moore is right that his proof is of a deductively valid form, and that the premises differ from the conclusion. Nonetheless, most people think that Moore’s proof is a bad one and that it begs the question of the existence of the world at all. One way of putting this point is to object to Moore’s claim that the premises are something he knows. For it seems that Moore could only know (A) if he already knew (C). Moore’s basis for claiming to know (A) is his perceptual experience of apparently seeing his hands. Such perceptual evidence, however, can only be trusted if one already knows that perception functions to give knowledge of the world, rather than just to create a visual experience like a dream that is not actually representative of objects in one’s environment10. Not having defeated Descartes’ dreaming doubt, it does not seem that Moore can claim knowledge on the basis of perception.

Moore, however, is aware of the objection based on Descartes’ dreaming doubt. Moore’s concept of knowledge appears sufficiently different from Descartes’ in that he is willing to claim knowledge despite not having defeated the dreaming doubt. Moore plainly acknowledges that he cannot prove the premises of his proof:

How am I to prove now that here’s a hand and here’s another? I do not believe that I can do it. In order to do it, I should need to prove for one thing, as Descartes pointed out, that I am not now dreaming. But how can I prove that I am not? I have, no doubt, conclusive evidence reasons for asserting that I am not now dreaming; I have conclusive evidence that I am awake: but that is a very different thing from being able to prove it. I could not tell you what all my evidence is; and I should require to do this at least, in order to give you a proof [of these premises]11.

• Nonetheless, Moore claims that he is entitled to claim knowledge in the absence of proof or a method of verification. He says, ‘I can know things, which I cannot prove: and among the things which I certainly did know, even if (as I think) I could not prove them, were the premises of my two proofs’12.

Moore thinks it is evident to him that he is holding up a hand and that he knows it: but he can only show his knowledge, he cannot prove—cannot give you exhaustive verbal reasons why—he knows. That is why the gesture of holding up one’s hands is an essential part of the demonstration or proof: the gesture is part of what ‘shows off’ Moore’s knowledge that he is holding up his hands. A proof must proceed from known premises or axioms, but some of these premises function like axioms and do not need further proof, especially if they seem self-evident. That is a charitable interpretation of Moore’s ‘proof’.

Wittgenstein’s general reaction to Moore’s proof can be divided into a positive aspect and a negative aspect. On the positive side, Wittgenstein thinks that what is right about Moore’s proof is that we do not demand a proof of statements like ‘I have hands’, because these statements cannot be sensibly doubted except under special circumstances. ‘I have hands’ is such that no one doubts it who has a body in the normal course of human development after about three months of age: one can of course imagine circumstances that falsify the claim: e.g., if an amputee without hands makes the claim while suffering from delusions13.

On the negative side, Wittgenstein thinks that what is wrong about Moore’s proof is that he should not have countered the sceptic’s claim that we cannot know anything by saying anything. In particular, to say ‘I do know that there’s a world’ is just to engage in the same kind of overblown, not obviously meaningful statements that the sceptic engages in when he purports to deny knowledge of the existence of the world.

There are a number of reasons why Moore should not have given the sceptic a direct response. First, it is not clear that the sceptics’ doubts really constitute meaningful propositions. Moore simply assumed that the sceptics’ propositions are meaningful; Wittgenstein disputes this. He disputed it from his earliest philosophical work to his last. In the Tractatus, he says:

Scepticism is not irrefutable, but palpably senseless [unsinnig], if it would doubt where a question cannot be asked.

For doubt can only exist where there is a question; a question only where there is an answer, and this only where something can be said.

([22], Tractatus 6.51)

• In On Certainty, he similarly says that the sceptic’s doubts are not real doubts:

The statement ‘I know that here is a hand’ may then be continued: “For it’s my hand I am looking at’. Then a reasonable man will not doubt that I know.—

Nor will the idealist [sceptic]; rather he will say that he was not dealing with the practical doubt which is being dismissed, but there is a further doubt behind that one.—That this is an illusion has to be shewn in a different way.

([20], OC 19)

• The sceptic’s doubt—is there a world?—is something with very general content. It is not a meaningful proposition unless we could know what it would be to verify that there is a world or is not one. But since we cannot simply experience there being a world or not, there is no way of knowing such a general claim. The claim is then in a certain sense meaningless. In fact it is arguably nonsense: it is neither true nor false14. Prefixing the general claim ‘there is a world’ with ‘I know’ or ‘I don’t know’ can only make matters worse. If p is not a meaningful proposition, then ‘I know that p’ and ‘I do not know that p’ are not meaningful either. So, Wittgenstein thinks Moore should have refrained from crediting the sceptic with meaningful propositions.

Furthermore, Wittgenstein held that Moore misused the word ‘knows’. The proper use of ‘knows’ is usually to claim that one has grounds for one’s belief:

“I know” often means: I have proper grounds for my statement. So, if the other person is acquainted with the language-game, he would admit I know. The other if he is acquainted with the language game, must be able to imagine how one may know something of the kind.

([20], OC 18)

• Moore should not say ‘I know there’s a world’, because his grounds for this knowledge are not something he can completely communicate, and whatever those grounds—perhaps perceptual experiences as of a world—these are not regarded as settling the issue by the sceptic. Moore assumes that the sceptic will acknowledge certain moves, as a person in ordinary life would—pointing to the data of perception—as conclusive. But the sceptic undercuts this strategy and does not regard these moves as beyond doubt. Moore’s moves can only play into the hands of the sceptic.

Finally, Wittgenstein thinks that Moore is wrong to react to the sceptic by claiming knowledge. That reaction can only encourage a sceptical search for further justifications for regarding those propositions as known. Hinge propositions (framework propositions, background assumptions) can only be justified from within a framework, and so do not have grounds. Therefore, Wittgenstein says it is better to think of ourselves as certain of the hinge propositions, rather than as knowing them. He thinks it is a perversion of the ordinary use of ‘knows’ to extend it to cases where no evidence is possible. Instead, the category of certainty is used to explain our relation to assumption that ‘stand fast for us’ and are exempt from doubt. Certainty is not a matter of exhibiting grounds for a claim; it is an attitude and a way of acting that shows one has no doubt about the claim in question; one simply accepts the claim. As Wittgenstein says, ‘My life consists in being content to accept many things.’ ([20], OC 344).

This distinction between certainty and knowledge ([20], OC 308) is one of Wittgenstein’s main innovations in rejecting Cartesian epistemology. Wittgenstein is saying that we can retain certainty in our way of life, habitual beliefs and background assumptions, without claiming knowledge, which would suggest that these beliefs are open to doubt when they are not. This is one way of insulating our background assumptions from sceptical doubt. But, of course, it comes at the price that these certain beliefs are now not rationally scrutable. Reason’s pretensions to govern all of our mental lives must be deflated.

4. The Appeal to Action and Practice

A recurring theme in On Certainty is the idea that actions and practices are the final court of appeal. The appeal to actions is so central that there is now discussion of Wittgenstein as a kind of pragmatist [16]. Setting aside that exegetical question, though, we must consider the role that the appeal to action plays in a broadly Wittgensteinian response to scepticism. Can actions ever really show something to be the case? That is the implication of the pragmatist response to scepticism, and it is philosophically problematic. I contend that the appeal to action is just one part of Wittgenstein’s response to scepticism in OC. The appeal to action cannot stand on its own as an adequate response to scepticism. Moreover, I contend that the appeal to action is neither a kind of foundationalism nor pragmatism, but wholly transformed into something new by his contextualism about meaning.

The essence of the response based on action is the idea that we show our certainty through action; that acting temporarily dispels sceptical doubts; that the proper therapy for sceptics is simply to engage in action and pay attention to when people claim knowledge. It might seem surprising to attribute this kind of response to Wittgenstein. Yet there is strong textual evidence for the claim that at the very least, Wittgenstein considers the response at length, even if he does not endorse it outright. The key passage for this interpretation of Wittgenstein’s response to scepticism is OC 20415. It says:

Giving grounds, however, justifying the evidence, comes to an end—but the end is not certain propositions striking us immediately as true, i.e., it is not a kind of seeing on our part, it is our acting, which lies at the bottom of the language game.

• This passage contains a critique of the Cartesian foundationalist response to the sceptic’s endless requests for justification, and recommends an alternative. So long as we seek to anchor knowledge in mental states which are supposed to be self-evident, the spectre of scepticism will haunt us. The sceptic can always press the Cartesian thinker to justify his knowledge of his own mental states, since even these mental states could be misinterpreted by the thinker if he is being deceived by an evil demon. In particular, Descartes seems not to have realised that a very deep scepticism about meaning might set in if the evil demon doubt is allowed to run wild, so that the Cartesian thinker cannot be sure even of the content of his thoughts (including the cogito)16. What we need to do to escape the sceptic’s argument is to reject his request for further justification in terms of giving reasons or stating propositions. We need simply to act, and in acting, we show our certainty. We refuse to give the sceptic what he wants, because what he wants is not something that can ever be given. Moreover, in accepting some beliefs as certain and primitive, we will then have enough of a basis to render the meaning of terms in our framework determinate. Our practice will fix the meaning of our terms. Moreover, our practice is something that can only be shown, not fully described or stated. References to the idea that we show our grasp of meaning and our beliefs, rather than stating it, are numerous: cf. OC 7, 28, 29, 431, 618.

In his commentary on Wittgenstein, Robert Fogelin raises the question of whether Wittgenstein has succumbed to a new kind of foundationalism, this time based on immediate and evident actions rather than mental states. Fogelin asks whether ‘Wittgenstein…departs from the neo-Pyrrhonian by treating acting (doing) as a new kind of foundation, or alternative foundation, to the foundations [for knowledge] favoured by philosophers’ (Fogelin [23], p. 218). Grayling also thinks that he discerns a strand in OC which is ‘a foundationalist refutation of scepticism’ (Grayling [8], p. 306). Both commentators have in mind the passage quoted above, among others.

The attribution of any sort of foundationalism to Wittgenstein has to be tempered, however, with the reflection that, unlike the basic beliefs of traditional foundationalism, the basic ‘foundational’ actions and practices that underlie our language-games are not in themselves true or false, justified or unjustified ([20], OC, 205). That is why it is not correct to read Wittgenstein in OC as a foundationalist (Williams [24]). Wittgenstein is at pains to stress ‘the groundlessness of our believing’ ([20], OC 166). So, the sense in which some actions and practices are ‘foundational’ is just that they inform and underlie our more theoretical practices of inquiry and are not themselves called into question. They are fixed points. Is it propositions or practices that are fixed points? This question has been debated, because the textual evidence points both ways. Wittgenstein often describes the fixed points of inquiry as hinge propositions: ‘some propositions are exempt from doubt, are as it were like hinges on which those turn’ ([20], OC 341) and speaks of ‘propositions exempt from doubt’ that ‘lie apart from the route travelled by enquiry’ ([20], OC 88)17. However, there is a good philosophical reason to view practices as primary for Wittgenstein (Rudd [3]). Taking contextualism about meaning seriously, a proposition has content only in the context of a practice. So, we cannot extract propositions without also taking into account the practices that underlie these propositions. For this reason, the dichotomy between proposition and practice is unhelpful and contrary to the spirit of Wittgenstein’s views on meaning. Furthermore, there are many statements to the effect that what is really largely fixed are practices, as, for example: ‘The end is not ungrounded propositions, it is an ungrounded way of acting’ ([20], OC 110). For example, one such fixed assumption is that in measuring, we use a straight, not a bent, ruler. These fixed points have this status not because of some peculiar property of indubitability or necessity, but because of how our practices took shape. The fixed points and the practices in which they are embedded are ‘not reasonable (or unreasonable)’ but ‘there—like our life’ ([20], OC 559).

Practices, or forms of life, are the bedrock on which our beliefs rest. Does this mean that Wittgenstein was a pragmatist? Certainly some aspects of Wittgenstein’s treatment of sceptical doubts look pragmatist. Both Peirce and Wittgenstein agree that sceptical doubts are not real doubts ([10], p. 14; [11], p. 5). For both, the fact that the sceptic’s doubts are not serious is shown by the fact that the sceptic cannot live his doubts. The sceptics’ actions do not betray the doubts he voices. But Wittgenstein goes further than the pragmatist and suggests that the denial of a certainty like ‘These are hands’ goes further than merely being a pretend doubt, but would actually be mad—that is, unintelligible to us18. Such a sceptic’s doubts so depart from our normal framework of belief that we really would have difficulty in even understanding someone with these doubts. Much contemporary philosophical work on psychopathology has fruitfully developed this insight (See, for example, [27,28]). Between madness and artificiality, there is a wide gap, so Wittgenstein’s treatment of sceptical doubts as mad would seem to go further than the pragmatist dismissal of sceptical doubts as not genuine.

4.1. The Insufficiency of the Appeal to Action

What kind of response to scepticism is the response based on acting? How does acting answer the sceptic at all? For example, suppose that one encounters a sceptic who doubts the existence of the external world. How could acting help him not to doubt the world’s existence at all? After all, for each action, there is an alternative sceptical interpretation on which the action does not involve a real causal interaction with an external object, but rather only a simulation of such an interaction. So, suppose one attempts to show one’s certainty that there is a world by continuing to live in the world: slamming doors, sitting on chairs, and tending to the sick, for example ([20], OC 7). These actions take for granted the existence of doors and chairs and sick people, but how do they satisfy the sceptic that there really are such things?

First of all, the response based on acting is not obviously an evasion of the sceptic’s question, ‘How do you know that there are tables, chairs, and things that exist when you are not perceiving them?’. An evasion of the question does not engage with the question at all. But actions can be specifically targeted at practical doubts. Is there a chair in the room? Yes, sure enough, one can discover that there is and satisfy oneself by looking at it or sitting down in it. Sitting in a chair answers the first level of doubt as to whether there’s a chair in the room. To be sure, it does not answer the second-level doubt about whether this experience is a hallucination, of course. That doubt can be answered by a transcendental argument to the effect that we must trust our sensory experience on the whole—unless specifically given reasons not to—if we are to gain empirical knowledge at all. So, the problem is that while actions can settle practical (first level) doubts, actions cannot settle hyperbolical (reflective, second-level) doubts raised by the philosophical sceptic.

The problem is that the sceptic’s very general doubt about the entire world leaves nothing untouched: therefore, it cannot be targeted by an action. Nothing one does could show the world as a whole to either exist or not exist, because the world is not an object of experience. The world is not an object with which one can causally interact at once. This is why Wittgenstein thinks that the propositions being debated between sceptic and realist such as ‘There’s a world’ are, strictly speaking, meaningless. But at OC 37, Wittgenstein expresses a doubt about the adequacy of dismissing the sceptic’s doubts as meaningless. Rather the sceptic’s assertion ‘You don’t know there are physical objects’ and the assertion ‘There are physical objects’ are ‘a misfiring attempt to express what can’t be expressed like that’. Perhaps the assertion or denial of hinge propositions may serve a purpose: to delineate the scaffolding of our thought.

Nonetheless, perhaps a pragmatist might say, surely one could build up a reply to the sceptic over time simply by continuing to act? In acting, one shows oneself to have a certain kind of justified true belief. For example, suppose one is hunting for one’s keys and searches a drawer for them based on a memory of where one last left them. One then takes them out and says ‘Here they are’. One’s searching behaviour expresses the belief that something wanted is in the drawer, and one’s behaviour on taking out the keys confirms that one was looking for keys. One’s belief that the keys are in the drawer turned out to be true, as well as justified. But such justified true beliefs may well fall short of knowledge. For one’s justification—seeming to remember leaving them there—cannot be trusted unless one can trust one’s perceptions and memories, which the sceptic doubts. So, obviously mere action will never answer the sceptic satisfactorily.

The sceptic, Wittgenstein says, is like someone who has not learned to search properly—nothing satisfies him and brings his search to the close (OC 315). Finding an object cannot satisfy the sceptic that objects exist—he cannot even recognise his finding as the finding of an object. For he must first doubt whether what seems to be the case—that he has found an object—is true. Nothing could count as a satisfactory reply to the sceptic. For just this reason, Wittgenstein concludes that the sceptic’s questions are meaningless. Satisfying ourselves that our assertions are correct is an essential part of the language-game of claiming knowledge, just as there are rules for calculating correctly in arithmetic that allow us to claim ‘the right answer’. But these rules are realised and exemplified in practice rather than theory, even if we could codify them (OC 39). Without such standards for correctness, we would lose a sense of the meaning of our words (OC 114).

So, if the reply based on acting does constitute a kind of answer to the sceptic, is it an adequate answer? And, did Wittgenstein think it was adequate? The answer in both cases is, of course, ‘No’. First, I will show that the response based on action succumbs to the dilemma that it is either (a) inadequate or else (b) question-begging. That is, the response is only adequate—in the sense of answering the question—if it is also question-begging. Second, I will demonstrate that Wittgenstein was aware of the temptation of appealing to actions as if they settled the matter.

A response is inadequate if it fails to answer the question by having nothing to say. For example, my action of sitting in a chair, can only satisfy the ‘chair-sceptic’ if it is interpreted as an action of sitting in a chair, (rather than, say, a hallucination of sitting in a chair, caused by an evil demon). So there must be a correct description (or range of correct descriptions) of my action. But how do we know which description is correct? Why should we trust one’s perception that one is sitting down? Is it obvious that this perception is of sitting down, rather than a simulation of sitting down, when in fact one is a brain in a vat? What could show that one was in the situation s1 (in which one really is sitting in a chair) rather than the sceptical situation s2 (in which one is deceived by a simulation)? The sceptic, of course, accepts the principle that knowledge requires discrimination. That is,

(KD) IF S knows p, then S can tell that S is in a world where p holds rather than not-p.

Notice that KD is a completely general principle. It differs from the view of relevant alternative theorists, who are not sceptics, in admitting any other world where not-p holds as a potential threat to knowledge that p if S cannot discriminate such a world from a p-world. Now, it is generally agreed that one cannot say why one can discriminate a p-situation from one in which everything looks similar except not-p holds. As Moore admits in his ‘Proof of the External World’, he says he knows he has two hands, but cannot prove it. Proof requires stating how one proposition follows from another according to the rules or axioms that license transitions between propositions in that system. In proof, each transition must be made explicit and legitimated by appeal to an axiom. Even though he has ‘conclusive evidence’ that he is awake and not dreaming, he cannot articulate this evidence in words. If attempts to justify one’s beliefs by statements are hopeless, what about attempting to justify those beliefs by further actions? No appeal to action is of the right form to demonstrate knowledge. The fact that a subject S performs an action ϕ does not entail that S knows that p (where p is some proposition). At best, S’s ϕ-ing shows that S probably believes that p.

However, even though I might be certain that I am about to sit in a chair, I can only show my subjective certainty through action. I could not show that a particular fact—say, that that is a chair—was true. Knowledge is factive; it requires that a fact obtain, not just that one thinks that a fact obtains. But all my action shows is that I think certain facts obtain, not that these are the facts. So, my action can never demonstrate knowledge.

Wittgenstein is aware of the problem. Tempting as it is to appeal to actions or experiences to justify our beliefs, neither provides states with content that can be simply ‘read off’ without interpretation. To think that actions have their meanings writ large is no less to succumb to the Myth of the Given than to think that inner mental states or sensations have self-evident meaning and indubitable content. At OC 145, Wittgenstein says:

One wants to say “All my experiences shew that it is so”. But how do they do that? For that proposition to which they point itself belongs to a particular interpretation of them….

• This remark shows an awareness that experiences—be they of acting or having sensations—cannot be taken as having a meaning prior to interpretation. There are no more ‘raw actions’ than there are ‘raw sensations’, unadulterated by concepts. Actions appear to settle doubts and to express certain beliefs. But the content of these beliefs and actions is still in question.

Though action tends to dissipate doubt, appealing to actions does not provide a justification of any sort for the propositions that the sceptic says we do not know. That is in part why it is, strictly speaking, incorrect to see a new kind of foundationalism in On Certainty19. Foundationalism requires that at the bottom of our knowledge are certain basic propositions, where these basic propositions are themselves true, certain, and justified (self-warranting). The foundationalist holds that the basic propositions may be justified in a way different from other propositions, but they are justified nonetheless. On Wittgenstein’s picture, what lies at the bottom of our knowledge is a set of ungrounded beliefs, practices, and actions that make up ‘form of life’. These basic certainties are neither true nor false nor justified (or unjustified). There is no external framework against which these certainties can be objectively assessed. One could view this as a very good critique of traditional foundationalism: that it is not entitled to the idea that mere experience justifies basic propositions.

Therefore, Wittgenstein is aware that the appeal to action is not a refutation of scepticism. It constitutes a kind of therapeutic tonic that temporarily makes scepticism go away when we are not reflecting. The appeal to action, then, does not enable one to say anything to the sceptic. Even the idea that it may show something is itself subject to scepticism about meaning. Appealing to action cannot answer scepticism.

4.2. The Appeal to Context

Given the ultimate inadequacy of the appeal to action, Wittgenstein needs to provide something else in his response to scepticism. He provides an additional diagnosis of how scepticism arises through the misuse of language in epistemological debate. This closes the gap left open by the appeal to action. As we saw, appeals to action, like appeals to mental states, do not by themselves deliver indubitable mental representations with a certain content, because one can always doubt that one has correctly identified the content of those representations20. It is clear that scepticism about meaning is the larger issue that must be tackled before any appeal to actions or experiences can carry significant weight in responding to scepticism.

I take it that the burden of remarks scattered in the 400s of OC (334,348,374,433,460) is to provide such a contextualist account of meaning. The main theme is that expressions make sense only in a given context. This is a general aspect of Wittgenstein’s view of meaning, but it also has specific application to knowledge and doubting. Both claims to knowledge and claims to have doubt are significant only in certain contexts. At OC 348, Wittgenstein compares ‘I know that’s a tree’ to ‘I am here’, as a statement that has no definite meaning (content) until a context is assigned. It could only make sense to assert knowledge that something is a tree if the circumstances were such as to call it in question. For example, if one is walking in an amusement park full of fake trees, then there would be a special point to asking ‘That’s a tree, isn’t it?’ and someone’s answering, ‘I know that’s a tree’. The proper way to continue the claim to knowledge would be by presenting mutually acceptable evidence: ‘This has bark, and roots, so it must be a tree’. Thus, there will be some special circumstances in which it does make sense to doubt a proposition that p or claim knowledge that p. But these propositions will not be hinge propositions—things that it does not make sense to doubt. For example, one cannot sensibly point at a recognisably, clearly green leaf and say ‘I doubt that’s green’. This kind of doubt just shows that one does not understand the word ‘green’. It is true that one can have doubts about something’s colour in some contexts. For example, one might look at a laser through tinted goggles and sensibly wonder, ‘Is that a green laser, or does it just look like it now?’.

The sceptic’s doubts lack meaning precisely because they are not grounded in practice; they arise from prying words such as ‘know’ and ‘exist’ apart from their usual context of use. In everyday life, one might ask, ‘Do you know if we have any butter?’ and get the answer, ‘Yes, it’s in the pantry’. Getting the answer ‘No, there’s no butter in the pantry’ would hardly generate butter scepticism. But the sceptic does something similar with his words. He asks, ‘Do you know if there’s anything outside of your mind?’. But what would be the verification procedure for determining the existence of something outside of one’s mind? By definition, we cannot step outside of our own minds and see how things are from that perspective. So, when someone concedes to the sceptic, ‘No, I really don’t know for certain that there is anything outside of my mind’, what he is conceding is that he has no way of knowing. But if there is no way of knowing, then the word ‘know’ is being misused. For the ordinary normative use of ‘know’ is to say that one knows something if one has a way of finding it out ([20], OC 18, 550).

Contextualism is a very important part of Wittgenstein’s response to scepticism. It is important because it complements the appeal to action in providing Wittgenstein’s adequate response to scepticism. He diagnoses scepticism as an illness resulting from the abstraction of words from their ordinary context. The therapy for scepticism then consists in bringing the sceptic back to the ordinary use of words, by making him use them in ordinary contexts.

Wittgenstein thinks the sceptic has to be taught that words derive their core meaning from usage in everyday life. This goes not just for words for common objects, but also for epistemic words, like ‘know’. Wittgenstein says his purpose is to return the usage of ‘know’ to its everyday context. The statement inat Philosophical Investigations (PI) 116 serves as a general maxim or guide to Wittgensteinian method:

When philosophers use a word—“knowledge”, “being”, “object”, “I”, “proposition”, “name”—and try to grasp the essence of the thing: one must always ask oneself: is the word ever actually used in this way in the language-game which is its original home?—What we do is to bring words back from their metaphysical to their everyday use.

([21], PI 116)

A common use of ‘know’ in asking a question is to request further grounds for someone’s statement (OC 18), and a common use of ‘know’ in answering a question is to say that one has particularly compelling grounds (OC 243). (e.g., She’ll win, alright, I am sure of it. How do you know?). ‘Know’ gets misused when one asks, as the sceptic does, ‘How do you know there’s a world at all?’ since no grounds for this belief can be offered that are themselves more certain or more entrenched than the belief itself.

Philosophical scepticism discards the certainties of everyday life in search of absolute certainty, but never finds it. Wittgenstein explains the barrenness of philosophical scepticism by showing that knowledge claims are emptied of meaning when abstracted from their everyday use. OC §§552,553 illustrates the process of giving up everyday certainties in philosophy and ending with empty statements:

552. Do I know I am now sitting in a chair?—Don’t I know it?! In the present circumstances no one is going to say that I know this; but no more will he say, for example, that I am conscious. Nor will one normally say this of passerbys in the street.

But now, even if one doesn’t say it, does that make it untrue?

553. It is queer: if I say, without any special occasion, “I know”—for example, “I know I am now sitting in a chair”, this statement seems to me unjustified and presumptuous. But if I make the statement where there is some need for it, then, although I am not one jot more certain of its truth, it seems to me to be perfectly justified and everyday.

554. In its language-game, it is not presumptuous. There, it has no higher position than, simply, the human language-game. For there it has restricted application.

But as soon as I say this sentence outside its context, it appears in a false light. For then it is as if I wanted to insist that there are things that I can know. God himself can’t say anything to me about them.

So, epistemological problems like scepticism derive from trying to take the verb ‘knows’ outside of its usual restricted contexts. The very generality of sceptical doubt and the epistemologist’s claim to knowledge mean that they do not arise in everyday life. Thus diagnosed, the recommended treatment for scepticism is to reacquaint the sceptic with the ordinary use of our words, such as ‘knows’. Wittgenstein’s view appears to be that it would be better if we did not use words such as ‘knowledge’ in philosophical investigations at all. ‘It is as if ‘I know’ did not tolerate a metaphysical emphasis.’ ([20], OC 482). We should stick to the everyday meaning of words. This raises some disturbing questions about the value of doing academic epistemology. It also raises a question that contemporary epistemologists have ducked: what sort of context is doing philosophy?

5. Conclusions: The Context(s) of Philosophy—Some Further Questions

What kind of context is doing philosophy in a particular situation? Though it is divorced from the demands of daily life, it is a rich context with a rich set of traditions and ways of proceeding. On closer inspection, philosophy (at least of a certain analytic stripe in some places) has its method and fixed assumptions as much as any other ordinary context. Can we isolate the ‘fixed points’ of the context of philosophy? One of the fixed points of the practice of traditional philosophy is

(Rationalism) There’s a reason for everything! Do not assert any claim without a reason.

• It is this assumption that gives scepticism its bite. Wittgenstein’s response to scepticism relinquishes this assumption. To point to and describe claims held without reason (without further grounds) largely brings the traditional enterprise of philosophy to an end. For this reason, we should expect resistance from traditional philosophy to Wittgenstein’s ‘quietistic’ response to scepticism. Yet, though Wittgenstein’s response does not fit within the traditional approaches, there is something valuable about its redirection.

If a collection of epistemologists decide to use the word ‘know’ in a special sense totally unconnected with its everyday usage, it would be of limited interest. If everyday usage forms the basis for the meaning of words, as Wittgenstein suggested in his Philosophical Investigations, then the epistemologists’ use of ‘know’ must be parasitic on that everyday use. If it is not parasitic on ordinary language, then academic epistemologists are playing a game of limited interest, like a parlour game. Given that epistemology, on some conception, should be an investigation of how knowledge—understood as something valuable that we can possess in everyday life—is possible, there should not be a total disconnection between the use of ‘know’ in epistemology and everyday life. But to grant this insight is already to start pursuing philosophy on a non-sceptical basis. It prompts one to look for explanations and descriptions of human knowledge with the aim of understanding it in all its variety21. The goal of non-sceptical epistemology is to seek understanding, and understanding comes with attending to practices. Moreover, this understanding, like the actions that show our certainties, may be at first non-propositional in nature. It involves having an oversight of the whole process of giving and asking for reasons, claiming and abandoning the claim to knowledge. One contemporary epistemologist, Linda Zagzebski (not especially concerned with Wittgenstein), writes:

Understanding deepens our own cognitive grasp of that which is already known. So, a person can know the individual propositions that make up a body of knowledge without understanding them. Understanding the propositions requires seeing how the parts of that body of knowledge fit together, where the fitting together is not itself propositional in form.

[29] (p. 244)

• I think that Zagzebski’s insight into the nature of understanding suggests a way of reading the Wittgensteinian project in On Certainty. Wittgenstein’s ultimate unravelling of sceptical doubt depends on acquiring an understanding of the human practices of producing, gathering, and claiming knowledge.