G. E. Moore and Scepticism: ‘Internal’ and ‘External’
G. E. Moore and Scepticism: ‘Internal’ and ‘External’
Abstract and Keywords
In this chapter, Stroud analyses the response to scepticism given by G. E. Moore in his famous ‘Proof of an External World’.
Moore seeks to prove that the proposition that there are no external things is in fact false. His proof consists of two premisses and a conclusion: ‘Here is one hand’, ‘And here is another’, therefore ‘Two human hands exist’; the conclusion follows from the premisses, and establishes conclusively that there are external things.
Contrasting his interpretation with those of N. Malcolm and A. Ambrose, Stroud argues that Moore in his proof fails to refute philosophical scepticism, but nevertheless makes a significant and legitimate use of the words ‘I know’ in formulating it and succeeds in establishing his conclusion from premisses he knows. Stroud argues that Moore, although he himself does not speak in those terms, can be understood to be giving a conclusive internal answer to the question whether he knows that there are external objects: a response from within his current knowledge to a question about knowledge that merely asks whether a piece of knowledge is already included among all the things he knows, or can be included among them by finding good reason to accept that thing on the basis of other things he already knows; but while this is a perfectly common and legitimate use of the expression ‘know’, it does nothing to answer the external question about knowledge that purports to cast doubt on all knowledge claims at once, and it is this question that must be answered if the sceptic is to be refuted.
Someone who doubts or denies that he knows a certain thing—the name of Lear's second daughter or the date of the battle of Waterloo, for example—can sometimes be reminded that he does know it after all. And the best response to someone who insists that he doesn't know a certain thing or that it is not known by anyone or perhaps is not even true would be to prove that very thing to him. G. E. Moore's approach to the problem of the external world embodies both of these forthright reactions. He seeks to remind philosophers that they do indeed know many things they say they doubt, and he thinks he can prove to those who doubt or deny it that there are indeed external things. We are perhaps familiar enough with the problem of the external world to be doubtful about the prospects of such a straight‐forward approach. I nevertheless want to look at what Moore does in his celebrated ‘proof of an external world’ and what he achieves by it. I think he can be seen to achieve a great deal even if it turns out not to be what he thinks he achieves.
He starts, characteristically, with some words he finds in Kant's Critique of Pure Reason.1 He takes them to express the complaint that a proof of the existence of things outside us has never been given. Whether that is what Kant complains of in that passage is open to question, but Moore thinks there is no doubt he can meet the challenge he finds expressed there. He begins by explaining in considerable detail precisely what he is going to prove and what it would take to prove it—clearing off the table and rolling up his sleeves, as it were, while describing exactly what he is going to do before getting down to performing the remarkable feat itself.
(p.84) He explains that by ‘things outside us’ he means things to be met with in space, and he carefully distinguishes that class of things from pains, after‐images, double‐images and the like, which are not to be met with in space. From the statement that such ‘inner’ things exist it follows that someone is having or has had some experience or other, but from the existence of things outside us no such inference can be drawn. Things outside us can in that sense be said to be independent of us—not dependent for their existence on being perceived or experienced. Examples of things outside us, so defined, are sheets of paper, shoes and socks, human hands, and soap bubbles. From the fact that things of that kind exist it does not follow that anyone is perceiving or experiencing anything. If at least two things of that kind could be proved to exist, the existence of things outside us would have been proved. That is precisely what Moore then tries to do.
The proof is short. It starts with his holding up his two hands and making a certain gesture with the right hand while saying ‘Here is one hand’ and a certain gesture with the left while saying ‘And here is another’. He thereby proves that two human hands exist. But it was just explained that that would be sufficient for proving the existence of things outside us; anyone who proved what Moore claims to have proved has ‘proved ipso facto the existence of external things’ (PP, 146). So here is a ‘perfectly rigorous’ proof of the existence of external things; Moore thinks it is probably impossible to give a better or more rigorous proof of anything (PP, 146).
It certainly seems true, as he points out, that:
He gives an everyday example of proving that there are at least three misprints on a certain page. If there has been some dispute on the question—or even if there hasn't, for that matter—the issue can be conclusively settled in the affirmative by finding that here is one misprint and here is another and here is another. That does prove it. Our daily experience—not (p.85) to mention what goes on in scientific laboratories or in courts of law—is full of similar examples every day. The best proof we could possibly have of something's existence would be to find it right before our eyes.
we all of us do constantly take proofs of this sort as absolutely conclusive proofs of certain conclusions—as finally settling certain questions, as to which we were previously in doubt. (PP, 147.)
That we do often prove things and come to know them in this way in everyday and scientific life seems to me undeniable; it must be kept clearly in mind in any discussion of Moore's proof. It is what I think makes his curious performance so important for an understanding of the philosophical problem of our knowledge of the external world. If Moore's proof is just like the proofs we give and accept of similar matters in everyday life, then by studying his proof and asking how it bears on the philosophical problem we can hope to clarify the relation between the philosophical problem and our ordinary procedures and claims to know things in everyday life. If Moore really does prove that there are external things, doesn't that settle the question of whether we know that such things exist? And if he does answer the question, isn't it also answered in the affirmative a thousand times a day by proofs we give in everyday life that are just as rigorous and conclusive as Moore's? Anyone trying to explain the philosophical problem of the external world would have to have convincing answers to these questions. If, on the other hand, we think Moore does not really establish what he sets out to prove, doesn't it follow that our ordinary attempts to know and prove things in everyday life are deficient in the same way? It would follow that nobody ever does establish that there are as many as three misprints on a certain page. But where exactly does Moore go wrong? And what mistake are all the rest of us making in everyday life when we give and accept proofs we regard as conclusive?
I considered one sort of answer to those questions in Chapter Two. On that account we would have to accept the sceptical conclusion that we never know anything about the world around us in everyday and scientific life. A consideration of Moore's proof gives us one way of testing the plausibility, or perhaps even the intelligibility, of that answer.
Moore points out that his proof satisfies three necessary conditions of a successful proof. (1) Its premiss is different from the conclusion it is used to prove. The proposition ‘Two (p.86) human hands exist’ is different from ‘Here is a human hand’ and ‘Here is another human hand’; that conclusion could be true even if both of those premisses were false. (2) The premisses are something known to be true. It would be quite absurd, Moore says, to suggest that he does not know that here is one hand and here is another, or that he only believes it to be true but is not or cannot be certain of it. ‘You might as well suggest that I do not know that I am now standing up and talking—that perhaps after all I'm not, and that it's not quite certain that I am!’, he replies (PP, 146–7). (3) The conclusion follows from the premisses. If it is true that here is one hand and here is another it could not possibly be false that two human hands exist at this moment. So Moore's proof is like other proofs in which a conclusion is validly drawn from premisses that differ from it and are known to be true.
Once we are familiar with the philosophical problem of our knowledge of the external world, I think we immediately feel that Moore's proof is inadequate. We are then most strongly inclined to object that he does not really satisfy the second of his three conditions for a successful proof—he does not really know that here is one hand and here is another. It seems hopeless to protest that the conclusion does not follow from the premisses—that even if it were true that here is a hand and here is another it still might not be true that two human hands exist—so his ‘knowledge’ of the premisses strikes us as the most questionable claim Moore makes for his proof.
It is worth looking into the source of this very natural reaction. I think it is more complicated than it might seem. We can distinguish two different questions that are not usually asked separately—probably because a certain answer to one of them is taken for granted. We can ask whether Moore's proof is a good one—whether he knows what he claims to know and legitimately establishes his conclusion on that basis. If so, he has proved that there are external things. We can also ask whether Moore refutes philosophical scepticism and answers affirmatively the philosophical problem of the external world. I think we do immediately feel that the answer to this second question is ‘No’. But must (p.87) we then conclude that the answer to the first question must be ‘No’ also; that there must be something wrong with Moore's proof? From the fact that Moore's performance does not answer a certain philosophical problem (assuming for the moment it does not) does it follow that there must be something wrong with that performance, that it does not succeed in doing what Moore intended it to do? I think it does not immediately follow. Explaining why that is so might help to illuminate the special character of the philosophical problem.
For one thing, if that conclusion did follow, it would be because what Moore says is inconsistent with what philosophical scepticism says, and because his proof was intended to refute that very philosophical thesis. It is of course extremely natural to assume that that is so. Moore says ‘I know that there are at least two external things’, and philosophical scepticism says ‘No one knows whether there are any external things’; it is difficult to imagine how they could fail to be inconsistent with each other. And Moore does take himself to be refuting the very thing he thinks sceptical philosophers are saying. But—and here I only introduce the possibility—if there were in fact no incompatibility between them, and if what Moore claims to refute turned out not to be the thesis of philosophical scepticism at all, we would not be forced on the usual grounds to say that there is something wrong with his proof as a proof. Perhaps we could even say that he really does know that there are external things, and that it really can be proved that there are as many as three misprints on a certain page, as long as we do not imply in either case that it is thereby demonstrated that philosophical scepticism is not true. What philosophical scepticism could possibly amount to might then become much more difficult to understand, but at least we would have a less distorted picture of how we prove and know things in everyday life against which to try to illuminate it.
To begin to explore the question of the compatibility or incompatibility between everyday knowledge and the thesis of philosophical scepticism I want to look at some of the ways Moore's proof can get misinterpreted when it is judged too much in the light of a certain understanding of the philosophical problem. I take as illustrations two (p.88) sympathetically‐motivated but, it seems to me, ultimately unacceptable accounts of Moore's proof. Something can be learned from each of them.
Norman Malcolm has tried to identify the great force and philosophical importance of Moore's work despite the fact that, taken at face value, his proof seems simply to ‘beg the question’ against philosophical scepticism, and his reply to philosophical paradox ‘does not seem to be a fruitful one’ or ‘one which ought to convince the philosopher that what he said was false’.2 Moore says that he does know that there are hands before him, and so any arguments to the contrary must be wrong, but according to Malcolm, ‘he does not say how they are wrong; so is he not begging the question?’3 If Moore in his ‘proof’ were simply asserting that there are two human hands, Malcolm thinks he would not disprove what a sceptical philosopher says—‘at least it seems a poor sort of refutation’ (S, 348–9). In response to a philosopher who says we can never know whether there is a tree before us, for example, Malcolm thinks it would be ‘pointless’ for Moore to say ‘I know there is a tree there because I have a clear, unobstructed view of it’.4 But that seems to be just the sort of thing Moore does say.
Alice Ambrose also finds Moore's proof unsatisfactory when taken at face value. She thinks it would never convince a philosophical sceptic because he would require in addition that Moore prove his premisses that here is one hand and here is another. It is not simply that the philosopher has higher standards than Moore or that Moore is more easily convinced of things than he ought to be. Rather, it is the very possibility of knowing such things as Moore's premisses that Ambrose thinks the philosopher questions; that is what is involved in questioning the possibility of knowing whether any external things exist. Since the sceptical philosopher would agree with Moore that being a hand entails being an (p.89) external object, Ambrose thinks Moore's proof ‘will not be to the point’ if it demonstrates only that the existence of external things follows from the existence of human hands.5 She concludes that if ‘There are external things’ is taken as a straightforward empirical proposition deduced from another proposition established by the evidence of the sense, it could never serve as a refutation of scepticism (S, 399).
Both responses start from the reaction that Moore's proof, taken as it stands, is unsuccessful as a refutation of philosophical scepticism about the external world. But both nevertheless regard the proof as extremely valuable and indeed correct when properly understood. For Malcolm it is a philosophical advance of great importance which does refute scepticism (S, 349); for Ambrose it is successful in making one see that scepticism is unacceptable (S, 418). They therefore conclude that it cannot be taken at face value; Moore must be doing something different from what he seems to be doing. His proof cannot be the ordinary empirical demonstration it appears to be.
For both Malcolm and Ambrose the real force of Moore's proof cannot be appreciated without understanding the peculiar nature of philosophical scepticism. They both think it is clear on reflection that for the philosophical sceptic nothing could possibly count as knowing by means of the senses that there are external things. It is not just that the evidence Moore gives for the existence of his hands is incomplete or inadequate in specifiable ways, but that all the empirical evidence there could possibly be would still not be enough. Ambrose thinks the philosophical sceptic cannot even describe what sort of thing could make one's evidence for ‘There are external things’ complete; there are for her no describable circumstances in which anyone could be said to know that there are external things. She concludes that the sceptic's claim ‘No one knows external things exist’ cannot be falsified and so cannot be an ‘empirical assertion’ about our inability to know (S, 402). She thinks this in turn shows that ‘the sceptic is arguing for the logical impossibility of knowledge and not for any empirical fact’ (S, 402). For him, every statement such as ‘I do not know there is a dollar in (p.90) my purse’ is ‘necessarily true’ (S, 402). Malcolm also holds that philosophical scepticism is the view that it is ‘logically impossible’ for anyone to know that that is a tree or that here is a human hand; for the sceptic there is a ‘contradiction’ involved in supposing that such things are known (S, 353). Malcolm and Ambrose agree about the philosophical sceptic's invulnerability to straightforward empirical refutation, and therefore about the source of the weakness or irrelevance of Moore's proof taken simply as a claim to know that there are hands before him and that therefore at least two external things exist. They also think the proof must therefore be understood to be something different from what on first sight it appears to be. Only then would it refute philosophical scepticism correctly understood. Both Malcolm and Ambrose assume that that is the only way it could have the profound philosophical significance it so obviously has for them.
What, then, is Moore really doing in his proof, and how does he succeed in refuting philosophical scepticism? According to Malcolm he is really pointing out that there is no contradiction involved in asserting that someone knows that there is a hand, and therefore an external object, before him. He does so in the only way Malcolm seems to allow it is possible to show that something is non‐contradictory—by showing that it is ‘perfectly correct language’ to say on certain occasions things like ‘I see a tree’, or that ‘it is a proper way of speaking to say that we know for certain that there are several chairs in this room’ (S, 354). Moore appeals to ‘our language‐sense’ (S, 354). By insisting that he knows the things he does Moore ‘reminds us that there is an ordinary use of the phrase “know for certain” in which it is applied to empirical statements’ (S, 355). That in itself ‘constitutes a refutation of the philosophical statement that we can never have certain knowledge of material‐thing statements’ (S, 355). Moore's proof really makes a point about what is correct language, and from that in turn a conclusion about the possibility of a certain state of affairs is said to follow.
Ambrose's account agrees with most of what Malcolm says but goes one step further. She thinks the philosophical sceptic could not fail to be aware of the facts about how we (p.91) speak. He knows that language is in fact so used that the sentence ‘I know there is a dollar in my purse’ describes something that could be the case. He would admit that it is not a necessary falsehood as language is now used to say that such a thing is known. That is why ‘it will not settle the dispute for Moore to show the sceptic he is using language incorrectly’ (S, 410). But philosophical scepticism according to Ambrose is really an insistence that such a sentence should be used to express a necessary falsehood. The sceptic argues as if the sentence ‘No one knows that hands exist’ expressed a necessary proposition, so for Ambrose he is really recommending or proposing that certain expressions of our language should be deprived of what he acknowledges is their current use. Moore's insistence that he knows that there are hands is therefore to be seen as working against scepticism because it ‘constitutes an insistence on retaining present usage’ (S, 411); it is a recommendation opposed to the recommendation of the philosophical sceptic. ‘It is the sceptic's recommendation which makes Moore's insistence relevant’ (S, 411).
These interpretations of Moore's proof are offered to explain how something which on the surface seems so inadequate as a reply to philosophical scepticism can nevertheless be of great philosophical force and importance. Moore certainly does not seem to be doing what they claim he is doing, so in order to increase the plausibility of their interpretations both Malcolm and Ambrose go on to argue directly against what Moore appears to be doing. Not only would his proof be ineffectual against scepticism if he were doing only what he seems to be doing, they argue, but (for different reasons in each case) Moore simply could not be doing what he appears to be doing. What he superficially seems to be doing simply cannot be done. To see why I think both these criticisms fail will take us a long way towards understanding Moore's proof as he understands it. That will leave us with the problem of its relation to philosophical scepticism.
Ambrose admits that Moore does seem to be trying to ‘establish the proposition that there exist things external to our minds’ (S, 397) by ‘an ordinary empirical argument’ (p.92) (S, 405) of a very common form. Just as one can establish an existential proposition such as ‘There is a coin in the collection plate’ by pointing out a specific instance—say, a particular dime in the collection plate—so Moore appears to be pointing out his hands as a way of establishing that there are external things. But for Ambrose what Moore does could not be of that form because pointing in the ordinary case ‘calls attention to a thing with features differentiating this thing from things of other kinds’ (S, 405), but one cannot point out an ‘external thing’ to someone in that way. It is impossible to point to something that is not an ‘external thing’. There is therefore nothing from which ‘external things’ could be distinguished, no contrasting class of ‘non‐external things’, and no features differentiating ‘external things’ from things of other kinds. Ambrose concludes that the term ‘external thing’ is not ‘a general name for some kind of thing, designating features distinguishing that kind of thing from some other kind’ (S, 406). It therefore is not simply a term that is more general than ‘dime’ or ‘coin’ or ‘piece of money’, all of which do serve to pick out things of certain sorts. But then one cannot establish the existence of external things by pointing to a human hand in the way one can establish the existence of a coin by pointing to a dime. So Moore's proof, whatever it is, cannot be a straightforward empirical demonstration of an empirical proposition. It cannot be an empirical refutation of a philosophical sceptic who denies that external things can be known. This objection to what Moore appears to be doing is meant to support Ambrose's claim that he is really doing something else in his proof—in particular that he is recommending a certain familiar linguistic usage, or resisting the sceptic's radical recommendation that certain words be used in new ways.
Fortunately we have Moore's reply to this interpretation, and it is not surprising to find that he repudiates it entirely. He insists that his assertion that there are external things is ‘empirical’ and was meant to be ‘empirical’, and that in proving it he meant to be proving that the proposition ‘There are no external things’ is actually false (S, 672). Consistent with that conception of his proof, he also says that he took the term ‘external object’ to be an ‘empirical’ (p.93) term: producing or pointing to a dime can prove that at least one external object exists just as it can prove that at least one coin exists (S, 671). Moore admits that there are differences between the term ‘external object’ and the term ‘coin’, but he thinks the terms do not differ with respect to the possibility of pointing to instances that fall under them. It might be true that one cannot literally point with one's finger at something that is not an external object, but one can certainly draw someone's attention to, and in that sense point out to him, a sense‐datum, an after‐image, or some other object which is not ‘external’ in the sense Moore specified before giving his proof. So for Moore the term ‘external object’ has a significant empirical contrast within our experience; it denotes things which can be pointed out and distinguished from other things that do not fall into that class. That is why the proposition ‘There are external things’ follows directly and obviously from ‘There are coins’ or ‘There are human hands’ just as ‘There are coins’ follows directly and obviously from ‘There are dimes’. The term ‘external thing’ as Moore understands it is just a more general empirical term than ‘dime’, ‘coin’, and ‘piece of money’, but not everything that exists falls under it.
Given this conception of his proof it is no wonder Moore thinks the only objection one could possibly make to it is that he has not proved his premisses that here is one hand and here is another. To object that the argument is not valid would be as silly as refusing to agree that there are coins while conceding that there are dimes. If the terms in which the conclusion is expressed are just more general than those in the premisses, the only possible objection would seem to be that the premisses are not really known. That is perhaps what Wittgenstein is conceding at the beginning of On Certainty when he says ‘If you do know that here is one hand, we'll grant you all the rest’.6
Because Moore understands his proof and the proposition established by it to be ‘empirical’, he does not hesitate to dismiss Ambrose's interpretation of him as making a certain recommendation about the use of words. He sees himself (p.94) in his proof as appealing to one fact—that here is a hand and here is another—in order to prove another—that there are external objects. He thinks the fact he appeals to proves just what he wanted to prove, but ‘I could not have supposed’, Moore says, ‘that the fact that I had a hand proved anything as to how the expression “external objects” ought to be used’ (S, 674). When he finds that here is one hand and here is another, he proves that there are external things, just as one might find that here is one misprint and here is another and here is another and thereby prove that there are three misprints. In neither case is anything proved or even said about the expression ‘hand’, ‘external thing’, ‘misprint’, or even the expression ‘I know that here is a hand’ or ‘I know that there are three misprints’. Moore takes himself to be appealing to a certain fact in order to prove something which itself is nothing linguistic. His proof as he understands it would be completely ineffectual against someone who was making a linguistic recommendation. Nothing about how words should be used follows one way or the other from his premisses.
This insistence of Moore's is important because it shows that if he is doing what he here claims to have been doing in his proof, not just Ambrose's interpretation, but Malcolm's interpretation as well, must be wrong. Just as the fact that here is a hand does not prove anything about how certain expressions ought to be used, so it does not prove anything about how certain expressions are used, or are correctly used, either. The point of Moore's proof could not be to show that such‐and‐such is ‘perfectly correct language’ or ‘a proper way of speaking’, if Moore is right about what he himself is doing in his proof. Malcolm believes that Moore never repudiated Malcolm's interpretation of the proof—he even suggests that Moore actually accepted it7—but it seems to me that that cannot be so if Moore was doing what he says he was doing. Of course Malcolm does have an additional argument to show that Moore could not have been doing what he says (p.95) he was doing. I will consider that additional argument in a moment. But on Moore's own understanding of his proof there is nothing to be said in favour of the idea that its premisses or conclusion ‘may be interpreted as meaning “It is correct language to say . . . ” ’ (S, 350).
It might still be true, as Malcolm says, that the proof or Moore's other typical assertions against philosophers serve to ‘remind’ us that ‘situations constantly occur which ordinary language allows us to describe by uttering sentences of the sort “I see my pen” ’ (S, 351), or that he ‘reminds us that there is an ordinary use of the phrase “know for certain” in which it is applied to empirical statements’ (S, 355). Moore's remarks might serve to remind us of many things, but that does not show that those reminders are the very point or conclusion of his proof of an external world, or that they are what it is meant to achieve. If I ask whether there is anything to eat in the kitchen and am told that there is spaghetti and broccoli, I might be reminded that some English words for food are taken over from Italian. But that is not the point of the reply, nor would I have found out what I wanted to know if I had been told only the facts of language that the reply served to remind me of. Even if I am reminded of those facts of language, it would not be correct to say that the reply ‘There is spaghetti and broccoli’ may be interpreted as meaning ‘Some English words for food are taken over from Italian’.
Behind both Malcolm's and Ambrose's interpretation of Moore's proof is the idea that the philosophical sceptic is not putting forward an empirical statement when he says that no one knows whether any external things exist. That is why they think Moore cannot be understood to be giving a straightforward empirical argument. Not only are their interpretations repudiated by Moore, as we have seen, but the inference on which they are based seems to be mistaken. Moore himself makes, or half makes, the point in his reply to Ambrose.
The philosophical sceptic might think he had a priori reasons for denying that there are external things or that anyone knows that there are. But even if he argues for his conclusion in that way, it does not follow that his conclusion (p.96) cannot be refuted empirically. If someone claims to have established on a priori grounds that there are no Xs, or even that there could not possibly be, it would be sufficient refutation of his view to point to the presence of Xs right before our eyes. The statement that we know a priori that there are no Xs, or the statement that there could not possibly be any Xs, both do imply, after all, that there are no Xs. And if it is obviously or even discoverably true that there are Xs, the original claim will have been refuted, whatever the arguments for it might have been. Moore believes that there is no difficulty in refuting scepticism empirically even if the sceptic's reasons are thought to be, and actually are, a priori or non‐empirical (S, 672–3). But with that non sequitur out of the way, Malcolm's and Ambrose's reasons for saying that Moore must be doing something different from what he seems to be doing in his proof reduce to nothing more than the reaction that the proof as it stands seems ineffective against philosophical scepticism. I suggest that that reaction, which I think we all share, is compatible with Moore's doing just what he appears to be doing in the proof.
Malcolm's direct argument to show that what Moore superficially appears to be doing simply cannot be done tries to explain Moore's repeated insistence to the contrary as due to confusion. He thinks Moore fails to see that if he had been simply asserting that he knows that here is a human hand and here is another in the circumstances in which he found himself, he would have been misusing the word ‘know’; it is simply not possible in those circumstances to use the word ‘know’ as Moore says he was using it. To understand and evaluate this line of criticism will take us a long way towards understanding both Moore's proof and the philosophical problem of the external world. But it must be kept in mind throughout that this diagnosis of how and why Moore could not succeed in doing what he says he is doing is to be assessed quite independently of any feelings we might have about the ineffectiveness of his proof against philosophical scepticism. That will lead us closer to an examination of what Moore actually does that is uncoloured for the moment by expectations (p.97) about what he should be doing if he wants to refute philosophical scepticism.
Malcolm's criticism of Moore's proof and other assertions against philosophers is that remarks like ‘I know that here is a human hand’ or ‘I know that that is a tree’ cannot be intelligibly made in just any situation, at any time, but require for their sense certain special conditions. In his paper ‘Defending Common Sense’ he says that for the proper use of the word ‘know’ there must be some question at issue or some doubt to be removed, the person saying he knows something must be able to give some reason for his assertion, and there must be some investigation which, if carried out, would settle the question.8 Malcolm holds that Moore violates all three conditions in his proof and in his typical responses to sceptical philosophers. The philosopher who denies that anyone knows that there are external things is not in fact in doubt about the existence of external things, there is no question at issue, Moore cannot give anything that counts as a reason for what he claims to know, and there is no investigation that could ever settle the question. Malcolm concludes that Moore misuses the word ‘know’. We cannot even ask whether it is true that Moore knows that here is a human hand, since he does not succeed in using the word ‘know’ correctly when he says that. His ‘proof’, understood straightforwardly, cannot get off the ground.
As a way of proving that Moore misuses the word ‘know’ this argument is obviously only as good as Malcolm's contention that the conditions he lists are necessary for the proper use of that word in an assertion of knowledge. It seems clear that they are not. The use of the word ‘know’ is more complicated than Malcolm's three examples and three conditions would indicate. In a more recent paper, he gives not three but twelve examples of ordinary uses of ‘I know’. He admits that the twelve do not yield a complete account of the proper use of ‘I know’; he now thinks ‘there is no such thing as a “complete account” ’ (MW, 179). That makes it difficult to give a proof, in the old way, that Moore (p.98) must be misusing ‘know’. Nevertheless Malcolm goes on from his twelve examples to conclude that ‘It is clear that Moore was not giving any everyday employment to the words “I know” ’ (MW, 185).
Unfortunately Malcolm does not try to show precisely how and why Moore fails to say anything or misuses ‘I know’ in typical utterances against sceptical philosophers. He is more interested in a ‘picture’ of knowledge he thinks Moore must have been operating with and which must have led him astray. But that ‘picture’ can be shown to have led Moore astray only if it can be shown that Moore was led astray—that he was in fact misusing the expression ‘I know’ or not giving it any everyday employment. And that could be shown only by a more careful examination of what Moore was actually doing or trying to do.
In a letter quoted in Malcolm's paper, Moore points out that the only reason Malcolm gives for saying he misused ‘I know’ is that he did not use it in circumstances in which it would normally be used; for example, no doubt or uncertainty had been expressed which was then cleared up by some new‐found knowledge. Moore concedes that it might serve no useful purpose to say a certain thing or to utter certain words on a certain occasion, but he insists that ‘this is an entirely different thing from saying that the words in question don't, on that occasion “make sense” ’ (MW, 174). He thinks ‘it is perfectly possible that a person who uses [certain words] senselessly, in the sense that he uses them where no sensible person would use them because, under those circumstances, they serve no useful purpose, should be using them in their normal sense, and that what he asserts by so using them should be true’ (MW, 174). Moore here seems to be invoking something like the distinction between the conditions for the appropriate or useful application of expressions and the conditions of their truth that I introduced in the sceptic's defence in Chapter Two. It might seem odd to find the anti‐sceptical Moore insisting on a distinction that I argued makes scepticism invulnerable to the charge that it misuses or distorts the meanings of its terms. But that charge is never part of Moore's attack on philosophical scepticism. In his typical assertions against philosophers he (p.99) is not making a point about the actual or proper use of expressions.
In this published letter to Malcolm he is of course making a point about the use of expressions; he is denying that he misused ‘I know’ in his proof or in his assertion that he knew that that was a tree before him. But he was not making a point about the use of expressions in making those assertions themselves. He does not even admit that the words he used against philosophers were ‘senseless’ in the sense of serving no useful purpose.
This repeats Moore's insistence that in his proof he meant to be doing exactly what he appears to be doing—proving the truth of a certain proposition. We now need to ask whether he succeeds and if so, whether what he succeeds in disproving is ‘a general proposition which many philosophers have made’.
Of course, in my case, I was using them with a purpose—the purpose of disproving a general proposition which many philosophers have made; so that I was not only using them in their usual sense, but also under circumstances where they might possibly serve a useful purpose, though not a purpose for which they would be commonly used. (MW, 174.)
I think there is a way of understanding Moore's assertions in which they are perfectly intelligible and legitimate and involve no misuse of ‘I know’ or any other expressions. Whether he thereby settles affirmatively the philosophical problem of the external world depends in part on what that problem amounts to and what the negative sceptical answer to it means. But if Moore does prove or know that there are external things, there must be some general proposition to the effect that there are no external things which he proves or knows to be false. We know that some philosophers have said or implied that no one knows whether there are external things. But if there is a way of understanding Moore's assertions as fully legitimate we are now faced with the possibility that what those philosophers meant to assert is not the same thing Moore proves to be false. This is precisely why I think G. E. Moore's proof of an external world is so important; he better than anyone else opens up this possibility for us. He of course would never explain the significance of (p.100) his work in this way. He thinks he is refuting the very thing sceptical philosophers said or implied. But if there is nothing wrong with his saying what he says, he could be unwittingly presenting us with the possibility that what he and all the rest of us say and do in everyday life could be perfectly true and legitimate without thereby answering one way or the other the philosophical problem of our knowledge of the external world. If that were so, the philosophical problem and its sceptical answer would perhaps be seen to stand in a much more complicated and puzzling relation to what we say and do in everyday life than the traditional conception outlined in Chapter Two would imply. That in itself could be a philosophical advance of great importance.
Malcolm thinks ‘it is evident that Moore was not giving any everyday employment to the words “I know . . . ” ’ in his typical assertions against philosophers, even though ‘he was saying something of deep philosophical interest’ (MW, 185). He was giving what Malcolm calls ‘a philosophical employment to the words “I know . . . ” ’ (MW, 185). I agree that Moore does something of deep philosophical interest, but I would like to suggest that it might be possible to do something of deep philosophical interest without giving a ‘philosophical’ employment to one's words. Perhaps a steadfast refusal or inability to speak or think in a ‘philosophical’ or non‐everyday way could reveal something of the greatest philosophical significance. Exploring that possibility will involve looking closely, and if possible without philosophical preconceptions, at the sort of thing Moore actually does and says in his typical assertions against philosophers.
Is it possible to use ‘I know’ as Moore does without misusing it? Could one then really fail to answer the philosophical problem of the external world? I think the answer to both questions is ‘Yes’. Malcolm apparently thinks the answer to the second question is ‘No’, and perhaps that, along with his understanding of philosophical scepticism, is part of what leads him to conclude that it is not possible to use ‘I know’ in the way Moore tries to use it. To see that it is possible to use Moore's very words in contexts that appear nowhere in Malcolm's list of correct uses of ‘know’, we can recall Thompson Clarke's example of the physiologist lecturing (p.101) on mental abnormalities. Near the beginning of his lecture he might say:
Here the lecturer uses the same words often used by philosophers who make or question general statements about the world and our knowledge of it. When he says that each of us knows that there is a public world of three‐dimensional bodies, he is stating what can only be regarded as a straight‐forward empirical fact. Most of us do know the things he mentions, and those with the abnormalities he has in mind presumably do not. That is a real difference between people that can be observed or ascertained.
Each of us who is normal knows that he is now awake, not dreaming or hallucinating, that there is a real public world outside his mind which he is now perceiving, that in this world there are three‐dimensional animate and inanimate bodies of many shapes and sizes . . . . In contrast, individuals suffering from certain mental abnormalities each believes that what we know to be the real world is his imaginative creation.9
I think we do not regard the lecturer in this context as having settled affirmatively the philosophical problem of our knowledge of the external world. If we had once raised the philosophical question of whether anyone knows that there is a public world of enduring, three‐dimensional bodies, it would be ludicrous to reply, ‘The answer is “Yes”; we do know of the existence of the external world. That physiologist says we do, and he is a reputable scientist who knows what he is talking about’. It is difficult to say precisely why that is absurd—after all, the lecturer did say that we know it, and (we can suppose) he does know what he is talking about—but I think there is no doubt that that would be our reaction. Certainly the physiologist in his lecture is not responding to any challenge he sees coming his way from philosophy. No philosophical thoughts need ever have entered his head; he could say and mean exactly what he says even if there had never been any such thing as philosophy. He is simply distinguishing two groups of people on the basis of what one group knows and the other does not; he is stating what he and all the rest of us know to be a fact.
Whatever we might think about the relation between what (p.102) the lecturer says and the philosophical problem of the external world, it is clear that he cannot be accused of misusing the word ‘know’. He makes a perfectly legitimate and intelligible application of the word, and once we are reminded of examples like this we see that similar remarks can be made in such relatively ordinary circumstances every day. But none of Malcolm's original three conditions are satisfied, nor does an example anything like this appear in his augmented list. When the physiologist was giving his lecture there was no question at issue about the external world and no doubt to be removed. He gave no reason for his assertion and there was no investigation in the offing that would settle the question of the existence of the external world. But violating Malcolm's conditions or not appearing in his augmented list is no proof of misuse. There are more legitimate uses of the word ‘know’ than the fifteen or so that are dreamt up in Malcolm's philosophy.
Moore, of course, was not lecturing on mental abormalities, and he was addressing his remarks to philosophers, so the example of the lecturing physiologist does not automatically settle the question of how we are to understand Moore. But it does show that it is possible to say, legitimately and undogmatically, what at least looks like the very same thing the philosophical sceptic doubts or denies, without settling or perhaps even touching the philosophical problem of our knowledge of the external world. That should make us more suspicious of a direct step from the understandable reaction that Moore does not really refute philosophical scepticism to the conclusion that he misuses ‘know’ or does not give it an everyday employment.
Consider another everyday use of ‘know’ not considered by Malcolm which is closer than that of the lecturing physiologist to Moore's own use. Suppose a murder has just been committed in a country house during a weekend party. The young duke is found stabbed on the far side of the large table in the hall, although the butler was with him the whole time except for a few seconds when he left to answer the telephone in the foyer where there were many people. An experienced detective and his younger assistant are among the guests and are trying to determine how it could have (p.103) happened. After considerable reflection the eager assistant announces that someone must have dashed into the room and stabbed the victim and dashed out again before the butler returned from answering the telephone. ‘That's the only way it could have happened’ he says, ‘the only thing we don't know is who did it’. ‘No,’ says the master detective at the scene of the crime, ‘we know this table is here and is so large that no one could have come through that door and got around to this side of the table and stabbed the victim and got back out again before the butler returned’.
The master detective is not misusing the word ‘know’. But when he says they know the table is there and was there a few minutes ago, there is no question at issue about the table's presence and no doubt about it to be removed. He gives no reason for his assertion, and there is no investigation that would settle the question. Nor is he doing any of the other things Malcolm lists as possible jobs for ‘I know’. He is simply reminding his colleague of something he knows and appears to have overlooked or denied in his attempted explanation of the murder. That is often a valuable procedure in trying to determine what is true or what to believe. The detective knows that the reflections of his younger colleague must be wrong, since they conflict with something both of them already know to be true. He does not even need to know what thoughts led the assistant to that conclusion. Even without finding some specific flaw in his colleague's thinking, he knows it is wrong since it could be true only if the table were not there. The presence of the table is something that is known and cannot be denied in their reflections, and the detective is quite right to remind his apprentice of it. It brings the enthusiastic but misguided speculations of his colleague back down to earth.
Could Moore have been using ‘know’ in some such way? The technique was certainly familiar to him. In an address to the Aristotelian Society on judgements of perception like ‘That is an inkstand’ or ‘This is a finger’, for example, he acknowledges that some philosophers seem to have denied that we ever know such things to be true, and even that they are ever true, but Moore replies: (p.104)
Here Moore is clearly reminding his audience of something the philosophers he refers to seemed to deny. Imagine him saying it, as he did, and not just writing it11 (‘This, after all, you know, really is a finger: there is no doubt about it: I know it, and you all know it’). He thinks what he says is enough in itself to show that such views must be wrong; it is a ‘sufficient refutation’ as it stands. He tries to bring the philosophers back down to earth.
It seems to me a sufficient refutation of such views as these, simply to point to cases in which we do know such things. This, after all, you know, really is a finger: there is no doubt about it: I know it and you all know it. And I think we may safely challenge any philosopher to bring forward any argument in favour either of the proposition that we do not know it, or of the proposition that it is not true, which does not at some point, rest upon some premiss which is, beyond comparison, less certain than is the proposition which it is designed to attack. The questions whether we ever do know such things as these, and whether there are any material things, seem to me, therefore, to be questions which there is no need to take seriously; they are questions which it is quite easy to answer, with certainty, in the affirmative.10
Moore thinks he can safely challenge philosophers in this way because he thinks nothing is more certain than that this is a finger. That is why he is confident that any argument (p.105) against it would have to rest at some point on some premiss that is less certain. This capacity to remain unruffled by apparently disturbing philosophical reasoning, and never to cast a second glance at his certainty, is characteristic of Moore's confrontations with other philosophers. Philosophers tend to regard him as dogmatic or stubborn in that respect, and to think he should take more seriously the possibility that his certainty might be ill‐founded. Do we think the master detective is dogmatic or hasty in his reply to the apprentice? He knows the table is there, and so does the apprentice, and that knowledge is what assures him that any explanation of the murder must acknowledge that the table is there. He too could be confident that anyone who tried to explain the murder by denying that the table is there would have to rely at some point on something that is less certain than that is. There can be no objection to the detective's assessing the apprentice's hypothesis on the basis of how it fits in with what is already known. It is difficult to think of any other way to judge the truth or plausibility of something. The detective's remaining unruffled by the apprentice's suggestion, and his having no second thoughts about his certainty that the table is there, is not dogmatism. He would be hopeless as a detective if he could be led to deny obvious facts simply in order to have some explanation or other. Dismissing without further investigation something that conflicts with what is already known is the very heart of rationality. It is one of the things that makes him the master detective that he is.
I think Moore sees himself as following the same eminently rational procedure. He thinks that what he says conflicts with what sceptical philosophers say, and he thinks it is a ‘sufficient refutation’ of philosophical scepticism to point as he does to some particular thing that is known.
In lectures delivered in 1910, for example, he picks out two ‘principles’ accepted by Hume which he thinks together imply12 that it is impossible for anyone to know about (p.106) anything external to his mind. To prove that those ‘principles’ are false, Moore says:
It seems to me that, in fact, there really is no stronger and better argument than the following. I do know that this pencil exists; but I could not know this, if Hume's principles were true; therefore, Hume's principles, one or both of them, are false. I think this argument really is as strong and good a one as any that could be used: and I think it really is conclusive. In other words, I think that the fact that, if Hume's principles were true, I could not know of the existence of this pencil, is a reductio ad absurdum of those principles.13
He acknowledges that a defender of Hume's views would accept that same conditional proposition and would argue from the truth of those principles to the conclusion that Moore does not know that his pencil exists. Each argument is simply the reverse of the other. Both are valid and they share a common premiss. For Moore the question of which conclusion to accept therefore comes down to the question of whether it is more certain that he knows that his pencil exists or that Hume's two ‘principles’ are true. Moore thinks it is obvious that it is more certain that he knows that his pencil exists. His aim is pretty clearly to refute Hume's philosophy by relying on the procedure of retaining what is known or is more certain when it conflicts with what is less certain.
The form of anti‐sceptical argument described here in 1910 is precisely what Moore follows in his more famous ‘Proof of an External World’ twenty‐nine years later. He never abandoned the idea of bringing forward particular things that are known to refute denials of knowledge that appear to conflict with their being known.15
That is why I say that the strongest argument to prove that Hume's principles are false is the argument from a particular case, like this in which we do know of the existence of some material object. And similarly, if the object is to prove in general that we do know of the existence of material things, no argument which is really stronger can, I think, be brought forward to prove this than particular instances in which we do in fact know of the existence of such an object.14
In ‘Four Forms of Scepticism’ he follows the same strategy. After identifying four ‘assumptions’ he claims are behind several sceptical arguments of Russell's, he ends by confessing that he ‘can't help asking’ himself whether it is as certain that those four assumptions are true as it is that he knows that this is a pencil or that his audience is conscious:
(p.108) Even Moore's acceptance of three of Russell's assumptions is not enough to persuade him that there might be something in Russell's argument. Not only is he less certain of their truth than he is that he knows that this is a pencil, he does not think it would be rational to believe otherwise.
I cannot help answering: It seems to me more certain that I do know that this is a pencil and that you are conscious, than that any single one of these four assumptions is true, let alone all four . . . I agree with Russell that (1), (2) and (3) are true; yet of no one even of these three do I feel as certain as that I do know for certain that this is a pencil. Nay more: I do not think it is rational to be as certain of any one of these four propositions, as of the proposition that I do know that this is a pencil. (PP, 226.)
The detective in his reply to the apprentice followed the eminently rational policy of rejecting the less certain because it conflicts with the more certain, or rejecting a hypothesis that conflicts with what is already known. Is Moore right in thinking that his arguments work against sceptical philosophers in the same way? It depends on the source of the philosophical conclusion. It is not always possible to reject a denial of knowledge by simply appealing to some particular thing that is known. Imagine a slightly later stage in the investigation of the murder. The apprentice, properly chastened, tries to be thorough and systematic and decides to consider everyone who could possibly have committed the murder and to eliminate them one by one. He gets from the duke's secretary a list of all those who were in the house at the time and with careful research shows conclusively and, let us suppose, correctly that the only one on the list who could possibly have done it is the butler. He then announces to the detective that he now knows that the butler did it. ‘No,’ the master replies, ‘that list was simply given to you by the secretary; it could be that someone whose name is not on the list was in the house at the time and committed the murder. We still don't know who did it.’
This is a successful objection to the apprentice's claim to know. If he has not checked the completeness of the list, we recognize that he has been hasty and does not yet know who committed the murder. It would obviously be absurd at this point for him to try to reject what the detective said by appealing to his ‘knowledge’ that the butler did it. The detective said that even after all the apprentice's valuable work they still do not know who committed the murder, and the apprentice cannot reply by saying ‘No. You're wrong because I know the butler did it’. In the earlier exchange the detective was quite right to reject the apprentice's claim by saying, ‘No. You're wrong because we know that this table is here and was right here a few minutes ago’. That did (p.109) refute the apprentice. But what might look like a formally similar reply to the detective in this later exchange would be ludicrous. It would not be a ‘sufficient refutation’ of the detective's denial of knowledge or a ‘good and conclusive argument’ against him. If what the apprentice says (‘I know the butler did it’) is true, what the detective says (‘We still don't know who did it’) is not true, but that does not provide the apprentice with a reductio ad absurdum of what the detective said. He cannot argue from the truth of that conditional to the conclusion that what the detective said is not true.
That is not to say that one can never appeal to one's own knowledge in just that way to refute someone who denies one's knowledge. Suppose that later still the detective and his assistant have established beyond doubt that the butler did in fact do it—they have just found a hidden camera that recorded the whole event and the film they have just watched clearly shows the butler in action. If a newspaper reporter who knows nothing about the discovery of the camera is saying into a telephone in the foyer ‘It is still not known here who committed the murder’ the apprentice overhearing him can easily refute him by saying ‘No. I know that the butler did it’. In this case he could ‘argue’ as follows: what the reporter said implies that I do not know that the butler did it, but I do know that the butler did it, so what the reporter said is not true. That is a conclusive argument. It is ‘as strong and good a one as any that could be used’ to show that the uninformed reporter is wrong. But those same words in an argument of the same form do not work against the detective who denies the apprentice's knowledge.
The difference between the two cases is obviously that the detective, unlike the reporter, is denying the apprentice's knowledge by pointing out a deficiency in the way the apprentice's conclusion was reached. A certain possibility is raised which is compatible with all the apprentice's evidence for his claim and, if realized, would mean that he does not know that the butler did it. I think we recognize that even if that possibility is not in fact realized—no one whose name is not on the list was in fact in the house at that time—the apprentice still does not know in that way that the butler (p.110) did it unless he has also established that the list is complete. He cannot meet the detective's challenge simply by asking whether it is more certain that the butler did it or that there is someone whose name is not on the list. The detective is not to be understood as putting forward a competing hypothesis about who committed the murder that he regards as more certain than that the butler did it. Nor is it a question of which ‘hypothesis’ it is rational to be more certain of. If the apprentice does not know that the list is complete his certainty that the butler did it is unwarranted. The detective's challenge obviously must be met in some other way, and that would require some understanding of the source of the conflict between what he says and what the apprentice says; the mere conflict itself is not enough to determine what a successful counter‐argument will be. What succeeds against the reporter's ‘It is still not known who committed the murder’ does not succeed against the detective's denial expressed in those same words. What matters are the reasons for the denial in each case.
How then are we to understand Moore's typical responses to sceptical philosophers? When we see his arguments as ineffective against scepticism I think it is because we see them as parallel to the apprentice's ludicrous response to the detective's verdict that it is still not known who committed the murder. That same assertion made by the uninformed reporter is refuted by appealing to the knowledge that the butler did it, but no such appeal works against the detective who challenges the basis of that putative knowledge. I have explained the philosophical sceptic's denial of our knowledge as the outcome of an investigation into the basis of all the knowledge or certainty we think we have about the world around us. That is why I think we feel it is not a ‘sufficient refutation’ of that scepticism simply to bring forward ‘a particular case . . . in which we do know of the existence of some material object’. The philosopher's assessment of all of our knowledge of the world around us is meant to apply to every particular case in which we do think we know of some material object, so no case that could be brought forward would escape that scrutiny.
The two Humean ‘principles’ that Moore tries to refute (p.111) (along with the sense‐datum thesis that Moore accepts) do indeed imply that Moore does not know that this pencil exists. But whether he can argue that since he does know that this pencil exists, those ‘principles’ (along with that thesis) must be false, will depend on the source of those ‘principles’ (and that thesis). Descartes reached his general negative conclusion from an assessment of all of our knowledge of the world—from asking how we know what we do, and taking seriously certain general features of the senses as a source of knowledge. Hume shared that conception, and hence that conclusion, with Descartes. Perhaps that negative conclusion is not correct, or perhaps we are not driven to it by a general assessment of our sensory knowledge, but so far that has not been shown to be so. If there can be a general assessment of our knowledge of the sort the philosopher engages in, and if the most careful execution of that assessment leads to the conclusion that we never know of the existence of material objects, Moore's attempt to argue against that conclusion by appealing simply to his knowledge that this pencil exists would indeed be like the apprentice's ludicrous response to the detective. He would be trying to deny the correctness of the assessment by appealing to one of the pieces of ‘knowledge’ that had been called into question by that very assessment.
From the ‘assumptions’ said to be behind Russell's sceptical conclusion it does indeed follow that Moore does not know that this is a pencil. But if those ‘assumptions’ are nothing more than truths unavoidably involved in any general assessment of our knowledge of the world, Moore does not successfully refute them any more than the apprentice refutes the detective. The detective in his objection might be said to be ‘assuming’ (1) that it is possible that someone whose name is not on the list committed the murder, (2) that that possibility has not yet been ruled out, and (3) that that possibility must be ruled out if the apprentice is to know by his eliminative reasoning that the butler did it. From those three ‘assumptions’ it follows that the apprentice does not know by eliminating all the other listed subjects that the butler did it. But that does not enable the apprentice to refute those ‘assumptions’ simply on the grounds that they have (p.112) that implication. Those ‘assumptions’ amount to an objection to the apprentice's claim to know. Whether Russell's ‘assumptions’ can be refuted in Moore's way will similarly depend on whether they are part of a negative assessment of the grounds for, among other things, Moore's assertion that he knows that this is a pencil.
It certainly seems as if Hume's ‘principles’ and Russell's ‘assumptions’ and Descartes's ‘requirement’ are all meant in just that way, so it is difficult to resist the conclusion that Moore's attempted refutations fail because he does not recognize that fact about the philosophers' denials of knowledge. That would be a serious deficiency in what Moore says and does. The reluctance of Ambrose and Malcolm and others to attribute such an apparent lack of philosophical understanding to Moore is what leads them to believe that he must be doing something different in his proof from what he seems to be doing.
I suggested earlier that we should not infer directly from the felt ineffectiveness of Moore's proof against philosophical scepticism that it involves some misunderstanding or misuse of words on his part. Nor should we infer on that basis alone that his assertions are dogmatic or hasty or ill‐supported either. We do not regard the apprentice as dogmatic or hasty in his refutation of what the reporter said. Nor was the detective dogmatic in his original denial of the apprentice's hypothesis. There was no misuse of words in those fully effective replies, they were perfectly reasonable appeals to something that is known, and they did not require careful consideration of the reasons behind the assertions they rejected. The detective knew immediately that the apprentice's hypothesis was incorrect, however it was arrived at, and the apprentice knew without further ado that the reporter was wrong. If Moore saw his attack on philosophers' remarks in that way, he too could not be faulted for not going carefully into their reasoning.
To defend the propriety and legitimacy of Moore's assertions in this way will no doubt seem only to postpone the difficulty. It will now be puzzling how Moore could ever have come to understand philosophers' remarks in the way he does. In insisting that he knows that this (p.113) pencil exists or that here is a human hand, how could he have thought he was responding to the sceptical philosopher in the way the apprentice responded to the uninformed reporter or the detective reminded his colleague of something right before their eyes? How could he have missed the fact that philosophical scepticism is not to be refuted in that way because it comes from a general challenge to all our knowledge of the world? How could he miss the parallel between the sceptical denial of our knowledge and the detective's successful undermining of the apprentice's claim to know that the butler did it? How could Moore have failed to entertain the possibility that the philosophers' denials of knowledge might be based on general considerations designed to cast doubt on the adequacy of the very reasons Moore or anyone else thinks he has got for claiming to know such things?
I would like to know the answers to those questions. I think they are all genuine questions—although largely questions about the thoughts or perceptions of G. E. Moore. It is very natural to take them as simply rhetorical questions—as exposing the absurdity of the idea that Moore could be doing exactly what he seems to be doing in his proof. But that natural reaction does rest on a certain acceptance of at least the intelligibility, if not the feasibility, of a general philosophical assessment of our knowledge of the world. If we think that Moore fails to answer a certain question or to refute a certain thesis, and that he can be expected to recognize that fact, we must also believe that there is some definite question he avoids or some intelligible thesis he fails to refute. We know that philosophers have certainly intended to scrutinize the grounds of all our knowledge of the world, including those particular pieces of knowledge that Moore would cite, and they have certainly thought they reached general sceptical conclusions. But intention alone is no guarantee of success, or even of a coherent project or thesis. Only if there is an intelligible general question about knowledge which, once asked, makes it impossible for Moore to answer it in the way he does will there be some deficiency in Moore's remarks against philosophers. If we find his assertions inadequate it is because we are taking it for granted (p.114) that there is such a question and that we understand what it is. But if that turned out to be an illusion, if it were not really possible to subject all of our knowledge of the world all at once to the kind of assessment that would render Moore's assertions ineffective responses to it, we would be in no position to accuse Moore of having missed something or of not succeeding in doing what he seems to be doing in his ‘proof’ or his other claims to know things.
Even if the philosophical question and the sceptical answer to it did turn out to make perfect sense, and even if Moore does not answer it or refute philosophical scepticism, it does not follow that if he were doing and saying precisely what he appears to be doing and saying there would be nothing right in what he says or that his saying it would be of no philosophical significance. He does not have to answer a philosophical question or make what Malcolm calls a ‘philosophical’ employment of his words in order to reveal something of great philosophical importance. Moore says things like ‘I know that here is a human hand’ or ‘I know that there are external things’, and I think it cannot be denied that there are questions to which those assertions are answers and that there are statements about human knowledge which must be false if what Moore says is true. We cannot deny that he says something that answers some question or implies the falsity of some proposition.
Moore says his proof was meant to prove the falsity of ‘There are no external things’, and it must be granted that there is a way in which it does that. If there are some apples in the cupboard it is false that there are no apples in the cupboard, and the answer to the question whether there are any apples in the cupboard is ‘Yes’. Apples are pieces of fruit, so it is also false that there are no pieces of fruit in the cupboard. Apples are also external things—from the existence of an apple it does not follow that someone is having or has had some experience or other—so it is also false that there are no external things. That is just how Moore sees his proof, and aside from the question of why such a ‘proof’ should be thought to be needed, can it be said that there is actually something wrong with it? Similarly, if Moore knows that there are external things, it is false that nobody (p.115) knows that there are external things, just as it is false that nobody knows that there are external things if what the physiologist said at the beginning of his lecture is true. He said that almost everyone knows it, except those suffering from certain mental abnormalities. G. E. Moore does not suffer from those abnormalities; he is one of us who know that there are external things. But if that is so, what prevents G. E. Moore from saying it? We need not expect him to be answering the philosophical problem of the external world in saying it; we did not suppose the lecturing physiologist to be answering that question when he said what he did, but his assertion was none the worse for that.
Suppose it occurs to me to ask whether there were any apples in Sicily in the fourth century B.C. I do not know the answer to that question, but I have a good idea of how to find out. Suppose it had occurred to me instead to ask whether it is known whether there were any apples in Sicily in the fourth century B.C. If I found that historians familiar with the place and period had established that there were plenty of apples there I would have found out that it is known that there were apples in Sicily then. If some historians tell me ‘We know there were apples in Sicily then’ they are simply reporting on the state of historical knowledge; they are telling me one of the things that is known about the past. If someone then asks me whether it is known that there were apples in Sicily, I can say ‘Yes, that is known’. Similarly, if I am asked whether anything is known about Sicily in the fourth century B.C., I can reply that among other things, it is known that there were apples there then. All these are answers to questions about our knowledge, and they are answered in the most straightforward way. It might be thought that no one could be so ignorant as not to know whether anything at all is known about Sicily in the fourth century B.C. Even if that is so, the answer I gave does imply that something is known about Sicily then, and therefore that the answer to that general question is ‘Yes’, whether anyone ever asks it or not. Moore thinks the questions whether we ever do know such things as that this is a finger, or whether there are any material things, are questions ‘there is no need to take seriously’ because ‘it (p.116) is quite easy to answer [them] with certainty, in the affirmative’. It is perhaps even easier to answer them than it is for the historian to answer my questions about the past and about what is known about the past.
The point is that there are general truths about human knowledge which simply follow from the fact that this or that or the other thing is known. A general question about knowledge therefore could be answered simply by appeal to one or more of those particular pieces of knowledge. That appears to be the way Moore understands general questions about what is known. He says astronomy gives us information about the size and motions of various planets and other heavenly bodies; geology tells us about the present state and past history of different layers of rock and soil; physics and chemistry provide knowledge about the composition of different kinds of physical things. From what has been known for a long time in these and other sciences it simply follows that there are material things. If there are nine planets, there are material things (at least nine of them). That is why Moore thinks anyone who says it is not known that there are material things is simply flying in the face of what science already knows to be true. The implication is that there is no reason to take that person's denial seriously. But we do not need science to show his denial to be false; it is contradicted by what all of us know and observe in the most ordinary circumstances every day. If there are human hands, there are material things, and if someone knows there are two of them in front of him, it is known that there are material things.
That Moore understands general questions of knowledge in just this way is shown by another of his responses to Russell. Russell reports that he came to philosophy through the wish to find some reason to believe in the truth of mathematics; he thought the best chance of finding indubitable truth lay in that domain.16 Moore takes Russell to be saying that the question whether any of the propositions of pure mathematics are true is a question for philosophers to answer. But, Moore replies:
The same is true for sciences other than mathematics. Whether we know anything of a certain sort is answered in the affirmative by the fact that this or that or the other thing is known in that area.
Surely it's the business of the mathematicians to decide whether (p.117) particular mathematical propositions are true? And if so what's the use of the philosopher discussing whether any mathematical propositions are true? Suppose he decides they are, can he give better reasons than the mathematicians give? Suppose he decides they aren't. He's contradicting the mathematicians. And aren't they better judges? It's admitted not to be the business of philosophers to discuss whether particular theorems are true. But if he insists on discussing whether any are, he's bound either to contradict the mathematicians, or to be doing something which seems superfluous.17
This is what might be called an ‘internal’ reaction to the question what is known or whether anything is known in a certain area. By that I mean that the question is answered by actually establishing some truths in that area or by finding out what has been established by others. It is to react, in one's own case, to the question ‘Do I know it?’ by asking oneself ‘Is it true?’ or ‘Should I believe it?’ The answer is to be found by trying to establish the thing in question, to see whether, given what one already knows, that thing also is or must be true. I call the reaction ‘internal’ because it is a response from ‘within’ one's current knowledge; the question whether one knows a certain thing is just the question whether that thing is already included among all the things one knows, or can be included among them by finding good reason to accept that thing on the basis of other things one already knows. Given that conception of the question ‘Is it known that p?’, it seems to me that there is no good reason for denying that what Moore says about our knowledge of external things is perfectly correct. That we know of external things follows trivially from our knowing many of the things we already know.
If we have the feeling that Moore nevertheless fails to answer the philosophical question about our knowledge of external things, as we do, it is because we understand that question as requiring a certain withdrawal or detachment from the whole body of our knowledge of the world. We recognize that when I ask in that detached philosophical (p.118) way whether I know that there are external things, I am not supposed to be allowed to appeal to other things I think I know about external things in order to help me settle the question. All of my knowledge of the external world is supposed to have been brought into question at one fell swoop; no particular piece of it is to be available as unquestioned knowledge to help me decide whether or not another particular candidate is true. I am to focus on my relation to the whole body of beliefs which I take to be knowledge of the external world and to ask, from ‘outside’ it as it were, not simply whether it is true but whether and how I know it even if it is in fact true. It is no longer simply a question about what to believe, but whether and how any of the things I admittedly do believe are things that I know or can have any reason to believe. That might be called an ‘external’ reaction to the question whether anything is known about the external world.
The terms ‘internal’ and ‘external’ are so far nothing more than labels; they do not serve to describe unambiguously the difference between two ways of understanding questions about our knowledge. Although I think there is a difference to be captured, those terms alone do not explain what it is. It is easy to think we understand it when we do not. I have emphasized the complete generality of the philosophical question of our knowledge of the external world. Descartes was not interested in whether we know this or that particular thing about the world around us, but whether we know anything at all about it. To answer that philosophical question we cannot appeal to one thing known about the external world in order to support another; all of it is meant to be in question all at once. But whatever the special feature of the ‘external’ philosophical question might be, and whatever might be the explanation of Moore's failure to answer it, it cannot be simply that the philosophical question is more general than any question Moore answers or addresses himself to. When Moore says he knows there are human hands and therefore that there are external things he is giving an affirmative answer to a completely general question about whether anyone at all knows anything at all about the external world. It will not do, therefore, to try to characterize (p.119) the philosophical or ‘external’ understanding of the question simply by asking ‘Does anything at all that is believed about the external world amount to knowledge of it?’ or ‘Do we know, not this or that fact about the world around us, but even whether there are any external things at all?’ All such general questions can be answered in Moore's way. No ordinary form of words alone can be guaranteed to express only the philosophical question or assertion. There will always be a way for a Moore to take it in which it does not have what we feel is its special ‘philosophical’ significance.
It is precisely Moore's refusal or inability to take his own or anyone else's words in that increasingly elusive ‘external’ or ‘philosophical’ way that seems to me to constitute the philosophical importance of his remarks. He steadfastly remains within the familiar, unproblematic understanding of those general questions and assertions with which the philosopher would attempt to bring all of our knowledge of the world into question. He resists, or more probably does not even feel, the pressure towards the philosophical project as it is understood by the philosophers he discusses. For Ambrose, Moore is like an ordinary man who dismisses the sceptical conclusion by simply denying it without bothering to counter the argument for it, thereby making one feel that there is something ridiculous about the sceptical conclusion. He shocks us into recognizing the contrast between what the philosopher says and ordinary life. ‘Because he is himself a great philosopher,’ Ambrose says, ‘Moore can succeed in this, whereas the ordinary man's remarks would have no influence. For the ordinary man can so easily be lured into talking in the same way’ (S, 416). That is true. It is borne out by the ease with which we feel we must go along with Descartes in his sceptical reasoning. We are ‘lured’ to his conclusion because it seems to be the only answer to his questions as he understands them. If Moore in his responses represents the ordinary man, he is a most extraordinary ordinary man in not being ‘lured’ into the traditional philosopher's understanding of his questions.
Moore does not resist the lure by simply shutting his ears to sceptical philosophers and refusing to get involved in their disputes. On the contrary. He listens to what they (p.120) say, he understands the words they use, and he then answers the questions expressed in just those words on the basis of what we and they have known all along. I think what Moore says, understood as he means it, is perfectly acceptable. If it nevertheless seems completely irrelevant to the philosophical questions and does not refute the paradoxical conclusions philosophers reach, that is a very important fact about those philosophical questions and conclusions. It will now need to be explained more carefully why the philosophical questions are not answered if everything Moore says is correct. That would focus attention on what I think is the right issue: precisely how the questions and assertions of the traditional philosopher are related to the questions and assertions we express in the very same words every day without managing to raise or answer philosophical questions.
But even Homer nods, and from time to time Moore is lured further towards seeing things in the philosopher's way than I think is consistent with his total immersion in the non‐‘philosophical’ or everyday understanding of the remarks philosophers make. That no doubt testifies even more strongly to the power of the philosophical project. No one, however firmly his feet are planted on the ground, seems able to resist it entirely. The philosopher asks not only whether it is known that there are external things but also how it is known, and Moore thinks he can answer the question. In everyday life, under normal circumstances, we often can say how we know a certain thing, so it would seem that there is no special difficulty in answering the question satisfactorily in Moore's way. But in this case he does not seem to me to stick closely enough to the straightforward, everyday response.
He is aware that philosophers will object to his claim to know that here is a human hand by raising the possibility that he might be dreaming, and he thinks he can meet that objection. In his lecture ‘Certainty’ Moore grants that if he does not know that he is not dreaming he does not know that he is standing up,18 but he is undaunted because, as he puts it, it is ‘a consideration which cuts both ways’. (p.121)
For, if it is true, it follows that it is also true that if I do know that I'm standing up, then I do know that I'm not dreaming. I can therefore just as well argue: since I do know that I'm standing up, it follows that I do know that I'm not dreaming; as my opponent can argue: since you don't know that you're not dreaming, it follows that you don't know that you're standing up. The one argument is just as good as the other, unless my opponent can give better reasons for asserting that I don't know that I'm not dreaming, than I can give for asserting that I do know that I am standing up. (PP. 247.)
Is Moore justified in his comfortable acceptance of what looks like a strong condition on our knowledge of the world? If the possibility of his dreaming is put forward by the philosopher as a criticism of Moore's claim to know that he is standing up (as it certainly is), that philosophical criticism would be parallel to the detective's criticism of his apprentice's announcement that he knows the butler did it because he has eliminated everyone else whose name is on the list. If the apprentice does not know that the list is complete he does not know in that way that the butler did it. But he could not comfortably accept that as ‘a consideration which cuts both ways’. He could not say:
If the apprentice did not even check the list, then for all he knows there could have been people in the house whose names are not on the list; he has to show how he knows that that possibility does not obtain. In the same way Moore would have to (p.122) show how he knows that the dream‐possibility does not obtain in his case. He cannot simply deflate the objection by reversing the philosopher's argument in the way he does.
I can just as well argue: since I do know that the butler did it, it follows that I do know that the list is complete; as the detective can argue: since you don't know that the list is complete, it follows that you don't know that the butler did it. The one argument is just as good as the other, unless the detective can give better reasons for asserting that I don't know that the list is complete, than I can give for asserting that I do know that the butler did it.
Of course, in the earlier example, the detective might have been wrong—and in any case he could eventually be given an answer. When he pointed out that the apprentice does not know that the list is complete, the apprentice might have been in a position to answer ‘No. I checked it. I also examined all the doors and windows, none of the guests reports seeing anyone else, the trustworthy doorman admitted only those on the list, the social secretary was a reliable, devoted servant of the duke . . . ’, and so on. He might have very good reasons for believing that the list is complete. He would thereby meet the detective's challenge and fulfil the condition for knowing by his eliminative reasoning that the butler did it. There is nothing in the detective's objection which by itself implies that it cannot be met. What is Moore to say in a similar vein about how he knows that he is not dreaming?
In his ‘Proof of an External World’ he thinks he has ‘conclusive reasons’ for asserting that he is not dreaming, ‘conclusive evidence’ that he is awake, although he admits that he cannot say what all that evidence is (PP, 149). But in ‘Certainty’ he goes so far as to admit that he would have ‘the evidence of his senses’ for the proposition that he is standing up only if he were not dreaming; if he were dreaming he would only be having ‘an experience which is very like having the evidence of my senses that I am standing up’ (PP, 248). He therefore could not be said simply to have ‘the evidence of his senses’ that he is awake, however much his present experiences resemble those he has when he is awake. By now he seems already a step or two down the slope to scepticism. Moore was apparently never satisfied with this part of his lecture.19 He makes a feeble objection to what he believes is one of the philosopher's arguments for the conclusion that he does not know that he is not dreaming. He even concedes that if it is ‘logically possible’ for all his ‘sensory experiences’ at a certain moment to be dream‐images, and to be the only experiences he is having at that (p.123) moment, he could not then know that he is not dreaming (PP, 250). He pins his remaining hopes on the possibility of remembering some things about the recent past that would enable him to know that he is not dreaming, but he is forced to admit that if that remembering itself could occur while he is dreaming he could never know that he was not dreaming and so could never know that he is standing up. But even a successful objection to an argument of the philosopher would not have been enough. Once he accepts the possibility of his dreaming as an objection to his claim to know that he is standing up, Moore must show how he knows that the possibility does not obtain. Not surprisingly, that is something he fails to do.
M. F. Burnyeat also notices and laments Moore's failure to extricate himself completely from the traditional epistemological predicament.20 He thinks the promise of Moore's philosophy was that it would simply avoid the traditional route to scepticism by continually insisting on the certainty expressed in particular everyday instances. But any such reductio ad absurdum of scepticism would work, Burnyeat thinks, only if Moore could ‘explain’ the certainty of his examples and give ‘a general rationale’ that ‘explains and justifies his belief that examples of knowledge . . . are the primary thing to which a philosopher should respond’.21 Moore never does that.
My explanation of Moore's relation to the philosophical problem is different. If he had never deviated even slightly from the attitudes and assertions of the plain man, if he had always put on the philosopher's words the interpretation that can be put on those same words in everyday, non‐philosophical life, he would never have taken a step down the ‘philosophical’ path. Although there would be no philosophical question he answered and no paradoxical philosophical assertion he managed to refute, his remarks would be none the worse for that. He would thereby represent more of a challenge to traditional philosophy, it seems to me, than if he tried to follow Burnyeat's suggestion. In fact, I do not see how trying to give a general explanation and justification (p.124) of the idea that particular examples ‘are the primary thing to which a philosopher should respond’ could keep Moore or anyone else out of ‘the traditional maze of epistemological argument’.22
On Burnyeat's suggestion Moore would have to know very well what the sceptical philosopher is really up to and explain to him why he cannot do it and why particular examples are the primary thing to which he should respond. On my suggestion Moore takes general questions and assertions expressed in the same words as those of the philosopher and answers or refutes them by appeal to particular examples. There is nothing wrong with that procedure as such. How else are general questions to be answered or general assertions tested? He does not go on to diagnose the philosopher's assertions or try to explain why they cannot be made; he simply denies them. On his way of understanding those words—which is a way of understanding them—they are simply false. Moore gives the impression of having no idea what the sceptical philosopher really wants to say or do. We feel he constantly construes the epistemologist's words only in a non‐‘philosophical’, everyday, and therefore completely uninteresting way.
J. L. Austin, by contrast, has a quite definite view about the source of certain philosophical problems. He makes a detailed scrutiny of the expressions philosophers use in formulating their questions and doctrines about perception and the external world and tries to demonstrate that those expressions are not actually used in that way. Austin's work rests on a shrewd grasp of the traditional epistemological project; he knows only too well what the sceptical philosopher is trying to do, and he thinks he can show that it cannot be done.
It is just possible, I suppose, that Moore has that same shrewd understanding of traditional epistemology. If one thought or knew that what the philosopher wants to say is really incoherent and the result only of some identifiable confusion, one might adopt the clever policy of never speaking or responding to his remarks in the ‘philosophical’ way. One could deliberately strive to avoid everything but (p.125) the straightforward questions and assertions of everyday life and always reply only as a plain man, and in that way refuse to be drawn into (what one knew to be confused) philosophical dispute. That ironic policy could even extend to announcing that one had refuted the philosopher's views, since that would give the impression that one did not even understand anything other than the straightforward, everyday assertion made by the words the philosopher uses. If it turned out that the philosopher's views really were incoherent, one would have been correct all along in behaving as if his words cannot be understood in any other way. Behind this plan might lie the hope that philosophers would eventually catch on and realize that what they were trying to say makes no sense or cannot be said with the significance they want to give to it.
It would be very difficult to keep up this act without slipping from time to time. The ordinary man with no awareness of philosophy can easily be lured towards the ‘external’ perspective. A lifelong performance would take great vigilance and care. But to adopt such a clever policy would require understanding at least the aims or intentions of the philosophical project. Moore in his writing shows few signs of that. Perhaps that is because he is extremely clever and consistent in his performance, and his mask almost never slips. But the whole idea of such deception seems incompatible with the child‐like honesty, directness, and lack of guile described by so many admirers of Moore. If that is so, we are left with the conclusion that Moore really did not understand the philosopher's assertions in any way other than the everyday ‘internal’ way he seems to have understood them.
This brings us back to the question how he could ever have come to give only that everyday interpretation to the philosopher's remarks. I have suggested that his way of taking them involves no misuse of words and is perfectly acceptable even if it does not refute philosophical scepticism. I have even conceded that there might be nothing intelligible that Moore missed; perhaps there is no comprehensible ‘philosophical’ way of taking the philosopher's questions and assertions. But how could Moore show no signs of acknowledging that they are even intended to be taken in a special (p.126) ‘external’ way derived from the Cartesian project of assessing all our knowledge of the external world all at once? That is the question about the mind of G. E. Moore that I cannot answer. Moore is an extremely puzzling philosophical phenomenon.
For all my efforts to separate what Moore says and does from the sceptical philosopher's own understanding of his questions and assertions, there remains a disturbing question that I have not answered or squarely faced. I have suggested that Moore does not provide a ‘sufficient refutation’ of philosophical scepticism, but the possibility seems to remain that what he says is nevertheless incompatible with philosophical scepticism. When the detective objected that the list was not known to be complete and so the apprentice did not know who committed the murder, the apprentice did not refute him by saying ‘No, you're wrong because I know that the butler did it’. But it is difficult to deny that that remark by the apprentice nevertheless contradicts what the detective said; they cannot both be true. It seems equally difficult to escape the idea that even if the sceptical philosopher's ‘No one knows whether there are external things’ is not refuted by Moore's ‘I know there are external things’, what Moore says nevertheless contradicts what the philosopher says; they cannot both be true. I have argued that there is a statement expressed in the philosopher's words which is incompatible with the truth of what Moore says. The present suggestion is that that statement must be the thesis of philosophical scepticism about the external world. If that is so, the relation between philosophical scepticism and the assertions Moore and all the rest of us make every day would be in one important respect as direct and straightforward as it seemed at the beginning of Descartes's argument.
That is the way Moore understands what the philosopher says. He thinks his own assertions of knowledge are true and that they obviously contradict what the sceptical philosopher says, and he concludes on that basis that the philosopher is wrong. The sceptical philosopher holds that what Moore says is no refutation of philosophical scepticism. I think the sceptical philosopher is right on that point. We (p.127) seem forced to conclude either that Moore's assertions of knowledge are not true or that they do not even contradict philosophical scepticism. Descartes and other sceptical philosophers take the first alternative—no one, including Moore, knows anything about the external world. But if, as on the other alternative, the two did not conflict, it might be possible to hold that Moore does not refute philosophical scepticism even though his assertions of knowledge are nevertheless true. The price of conceding the truth of Moore's assertions, as it were, would be their lack of logical connection with the thesis of philosophical scepticism. But on that alternative philosophical scepticism would no longer imply the falsity of the knowledge‐claims made by Moore and all the rest of us in everyday life. The price of philosophical scepticism's immunity, as it were, would be the corresponding immunity of all our ordinary assertions to philosophical attack. That is a price many would be eager to pay. It would mean that, however cogent and convincing the arguments for philosophical scepticism might be, they could not cast any aspersions on the knowledge we possess and seek in science and in everyday life.
Could philosophical scepticism be compatible with the truth of what we say and believe in ordinary life? I confess it is difficult for me to see how it could be so. Once one grasps the traditional epistemological project it is difficult to see the claims of everyday life as anything other than restricted in the way outlined in Chapter Two. It would then be difficult to see how philosophical scepticism could fail to be true. The chapters that follow explore other ways of trying to understand the relation between philosophical theories of knowledge and the everyday claims to knowledge which are presumably their subject‐matter. Only something other than that traditional conception would enable us to avoid or defuse philosophical scepticism.
(1) G. E. Moore, ‘Proof of an External World’, in his Philosophical Papers (London, 1959), p. 127. (This volume is hereafter cited as PP.)
(2) N. Malcolm, ‘Moore and Ordinary Language’, in P. A. Schilpp (ed.), The Philosophy of G. E. Moore (New York, 1952), p. 48. (This volume is hereafter cited as S.)
(3) N. Malcolm, ‘George Edward Moore’, in his Knowledge and Certainty (Englewood Cliffs, N.J., 1963), p. 177.
(4) N. Malcolm, ‘Defending Common Sense’, The Philosophical Review, 1949, p. 209.
(5) A. Ambrose, ‘Moore's “Proof of an External World” ’ (S, 399).
(6) L. Wittgenstein, On Certainty (Oxford, 1969), p. §1.
(7) N. Malcolm, ‘Moore and Wittgenstein on the Sense of “I Know” ’, in his Thought and Knowledge (Ithaca, N.Y., 1977), p. 171. (Hereafter cited as MW.)
(8) N. Malcolm, ‘Defending Common Sense’, The Philosophical Review, 1949, pp. 203 ff. It is not completely clear whether Malcolm regards each of the three conditions as necessary, or only their disjunction.
(9) Clarke, ‘The Legacy of Skepticism’, p. 756.
(10) G. E. Moore, Philosophical Studies (London, 1958), p. 228. I see no conflict, as M. Lazerowitz does, between Moore's saying that such questions need not be taken seriously and his continuing to write about them and even to offer a proof of an external world. Nor do I think it shows any ambivalence on Moore's part about the force of his ‘refutations’. Saying over and over again that a certain thing is already known, especially when others appear to persist in denying it, does not show that one even half thinks there is a serious question about whether it is known. And the ‘Proof of an External World’ concentrates on the further question whether it can be proved that there are external things. That does not show that whether there are material things, or whether we know there are, are in any way serious questions for Moore. See M. Lazerowitz, ‘Moore's Paradox’ (S, 376).
(11) I think it is significant that almost all Moore's papers were invited public addresses. They were meant to be heard by an audience, and they were written specifically for particular occasions, not simply published by Moore. Even the apparent exception, ‘A Defence of Common Sense’, was an invited contribution to a volume in which authors were given ‘an opportunity of stating authentically what they regard as the main problem of philosophy and what they have endeavoured to make central in their own speculation upon it’ (J. H. Muirhead (ed.), Contemporary British Philosophy, First and Second Series (London, 1925), p. 10). The ‘defence of common sense’ was solicited, and not spontaneously offered to the philosophical world by Moore.
(12) The two ‘principles’ are (1) nobody can know that something he has not directly apprehended exists unless he knows that something he has directly apprehended is a sign of its existence, and (2) nobody can know that one thing is a sign of another unless he has directly apprehended things of both kinds. Obviously it will follow from (1) and (2) that no one can know of the existence of material things only if material things cannot be directly apprehended. Moore does believe that ‘nobody ever does know, by direct apprehension, of the existence of anything whatever except his own acts of consciousness and the sense‐data and images he directly apprehends’ (G. E. Moore, Some Main Problems of Philosophy (London, 1958), p. 111). I do not understand why Moore accepts that thesis. Sceptical reasoning like that of Descartes can lead one to accept it, but Moore seems immune to the force of such reasoning. Nor do I understand how Moore fails to see the sceptical consequences of that sense‐datum thesis. Here my puzzlement extends well beyond Moore. Many philosophers appear to hold some such thesis while also believing that they know things about the world around them.
(13) Some Main Problems, pp. 119–20.
(14) Some Main Problems, pp. 125–6.
(15) Moore here speaks of bringing forward particular instances in which we know of the existence of a material thing to prove that ‘we know of the existence of material things’. In his reply to Ambrose he says he never intended his proof of an external world to be a refutation of ‘Nobody knows that there are external things’. He takes Ambrose to task for not making the distinction, and adds ‘I do not think I have ever implied that [‘Nobody knows for certain that there are any material things’] could be proved to be false in any such simple way; e.g., by holding up one of your hands and saying “I know that this hand is a material thing; therefore at least one person knows that there is at least one material thing” ' (S, 668). But that seems to be precisely what he does say in his 1910 lecture. Either Moore is mistaken in this later disclaimer, or else he does not regard holding up his hand and saying ‘I know that this hand is a material thing’ as bringing forward a particular instance in which one knows of the existence of a material thing. I do not see why not. I therefore continue to treat his proof of an external world as also implying that we know there are external things.
(16) B. Russell, ‘Logical Atomism’, in R. Marsh (ed.), Logic and Knowledge (London, 1956), p. 323.
(17) G. E. Moore, Lectures on Philosophy (London, 1966), p. 185.
(18) This is the assumption or requirement which in Descartes's reasoning seemed to lead inevitably to scepticism. I do not understand why Moore accepts it so uncritically. He never explains why. After a page and a half explaining that if he were dreaming he would not then know that he is standing up (what I called an ‘undeniable fact about dreams’), he immediately concludes: ‘I agree, therefore, with that part of the argument which asserts that if I don't know that I'm not dreaming, it follows that I don't know that I'm standing up . . . ’ (PP, 247). Moore's ‘therefore’ suggests that he thinks that the epistemic requirement he here accepts follows from the ‘undeniable fact about dreams’ stated earlier. I do not see that it does. I argued in Chapter One that if it does follow, scepticism about the external world must be correct, since that epistemic requirement leads directly to scepticism, and if it follows from an ‘undeniable fact’, it must be true.
(19) See Moore's preface and also the note by C. Lewy on p. 251.
(20) M. F. Burnyeat, ‘Examples in Epistemology: Socrates, Theaetetus and G. E. Moore’, Philosophy, 1977, pp. 396–7.
(21) Burnyeat, p. 397.
(22) Burnyeat, p. 396.