Chapter 5

Wittgenstein on privacy

Wittgenstein’s Remarks on Private Language and the Myth of the Given

And hence also ‘obeying a rule’ is a practice.  And to think one is obeying a rule is not to obey a rule.  Hence it is not possible to obey a rule ‘privately’: otherwise thinking one was obeying a rule would be the same as obeying it.[1]

 

Wittgenstein draws this conclusion immediately after his rejection of the sceptical paradox and the corresponding model of meaning that it gives rise to.  If instead meaning is use, if it requires a regular practice, then private meaning is not possible.  For merely thinking that one is following a rule is not to follow a rule.

But it is not yet clear why a private language is not possible.  Kripke thinks that Wittgenstein has argued that meaning (insofar as there is anything called ‘meaning’ left on Kripke’s reading) requires a community.  It requires the agreement of other language users.  Hence simply thinking one is following a rule is not enough for rule following to be actually taking place.  Where there can be no criteria by which other members of the community can check if a word is being applied correctly, there can be no such thing as a language.  Kripke thinks that this is about all there is to the “private language argument”.[2]  The various comments about privacy that follow are then presumably a defence of Wittgenstein’s account of meaning in that they defend it against an obvious counter example: that the language of sensations is inherently private in character.  Wittgenstein must defeat this idea if he is to defend his ‘new account of meaning’ in terms of agreement within a community[3].

But it is not at all obvious that the statement above follows so easily from the discussion of meaning in the first 200 sections.  We have already seen that Kripke’s interpretation of Wittgenstein’s remarks on rule following has missed the mark.  And there are further good reasons to doubt the community interpretation of Wittgenstein’s remarks on private language.  The most decisive point is this: even if something like the community interpretation were correct, it remains unclear why this in itself should rule out an alternative form of language that is based on private objects; just because public language works in such-and-such a way, why should a private language?  Surely the whole point about it being private is that it has a different nature to ordinary language.

I think Wittgenstein’s swift conclusion in section 202 is the product of more thorough analysis.  After all, Wittgenstein rearranged his remarks obsessively, and there is little reason to expect them to follow the standard form of argument[4].  There is much yet to be explained, both by section 258, which is most often taken to be the definitive statement of “the private language argument”, and by the remarks before and after it.  Many of the remarks derive from previous work on sense data and private objects.  Wittgenstein had been working for a long time against the prevalent ‘Myth of the Given’.  So I suggest we look for an interpretation that forms a natural continuation with this previous work.

One way to approach Wittgenstein’s remarks on private language is suggested by the conclusion of the previous chapter.  If Wittgenstein’s remarks on meaning amount to a rejection of conceptual realism, then perhaps we can understand his rejection of the given as being motivated by considerations similar to those that lead to conceptual idealism.  This is not yet to suggest that Wittgenstein was a conceptual idealist; that question will be raised and examined in detail in the following chapter.  But we might consider the remarks on privacy with the working hypothesis that Wittgenstein does indeed share some assumptions and aims with the conceptual idealist, and that he is similarly motivated to reject the dualism of scheme and content.  The notion of the private object is just the notion of an ineffable, pre-conceptual object.  For something to be private to the individual would mean that it does not fall under the restriction of a regular, specifiable use.  But according to Wittgenstein we cannot talk of any class of objects without assuming that a grammar is involved in individuating those objects.  And if there is a grammar we can specify it, and this shows that the supposed private object is covered by a public concept after all.

Of course, a great deal more needs to be said to hammer out and make clear this interpretation.  So this chapter will proceed as follows.  First I will consider the shortcomings of the community interpretation and suggest an alternative.  Then I will try to make sense of the remarks on privacy using the working hypothesis as our guide.  I will conclude, by way of preparation for the next chapter, by drawing some distinctions between Wittgenstein’s reasoning and that of the conceptual idealist.

5.1 Customs and The Community

We can distinguish two important interpretations of why private use is ruled out on Wittgenstein’s account.  The ‘community interpretation’ (to which Kripke subscribes) emphasises the communal nature of language use.  Since language is a custom, because it is from the beginning an integral part of a communal form of life, there can be no private language.  There are two problems such an interpretation must face.  Firstly, this by itself does not seem to rule out the possibility of a private language.  Even if it is true that language as we know it is embedded in a community of language users, this does not by itself rule out the possibility of a language that works within a different context.  It remains to be shown that the possibility of rule following requires the community.  The second problem with this interpretation, suggested by Malcolm Budd[5], is that it is weakened by allowing for the possibility of a ‘public’ language in isolation from a community.  It is not an actual community of speakers that is required, but the possibility of speakers that could, in principle, understand the language user that makes his utterances or thoughts public in the required sense.

An alternative account stresses Wittgenstein’s emphasis of the fact that an instance of rule following could not occur in isolation of other applications of the rule, or at least not in isolation of the context of rule following behaviour in general.  As Budd points out, Wittgenstein’s concept of following a rule as a practice or custom is illustrated, not by a contrast between a single individual in isolation and a community acting in agreement, but by the contrast of a single occasion and a practice spanning a series of occasions.

The reason that the community is so often proffered as the seat of meaning by commentators of Wittgenstein is that the problem is taken to be that nothing grounds my decision to take a rule one way rather than another.  Whatever seems right is going to be right.  And that means we cannot talk about right here.  But, to repeat a point made by McDowell[6], what good will it do us to look to the community?  What ever seems right to the community is going to be right, so we cannot talk about right here either.  This paradox only disappears on the second interpretation.  A judgement can be considered right when it is in agreement with judgements that are part of a regular practice.  That is not to say that the notion of a community is irrelevant.  We have learned our notion of rule following in the context of a community.  Could I follow a rule just once in my life?  We could say so if I acted in the context of a rule-following community.

So what is the connection between my decision to interpret a rule in a certain way, and the communities’ sanction of that decision?  Well, imagine that I did reinterpret the rules of chess, so as to allow a certain move that was considered illegal by other players of chess (say, that the king can move two squares whenever the player is in check).  The other chess players will insist that I have got the rules wrong.  But suppose I insisted on playing this way, and always played this way.  I find someone who has not previously played, and convince them that this is a superior way to play.  (I train them my way).  I play this way regularly.  It is no good to say that there is no justification for this new move.  There is no justification for the old way of playing.  Rather, people would say, “I don’t know what you’re playing, but it’s not chess”.  I have invented a new game.

One might say that I have a ‘right’ to play my way, but this is not simply because I have found someone to play my way with me.  I could invent a new way of playing patience, which only I ever play.  I will have still invented a new game.  I might even never play the game I invent, but only consider the new rules in my imagination.  Of course, if someone asks me the rules, I should be able to explain them.  (The response “I knew them a while ago, but now I have forgotten” would rightly be regarded with suspicion.  But it is possible that I am telling the truth.)  But what about the following situation: I cut a deck of cards, and declare, “I won!  I found the 8 of spades.”  If, on being asked if I could win again, I say “but the rules of the next game are not decided yet”, you will not accept that there was any winning or losing going on at all.

5.2 Wittgenstein on Private Language

Before going on to present what is widely regarded as “the private language argument” in section 258, Wittgenstein continues his discussion of rule following.  During this discussion he touches on the interdependence of such concepts as ‘agreement’ and ‘same’ with the concept of a rule:

One might say to the person one was training [to follow a rule]: “Look, I always do the same thing: I…..”

224.  The word “agreement” and the word “rule” are related to one another, they are cousins.  If I teach anyone the use of the one word, he learns the use of the other with it.

225.  The use of the word “rule” and the use of the word “same” are interwoven.  (As are the use of “proposition” and the use of “true”.)[7]

Thus, if we are to apply a rule there must be some sense in which it can agree with other applications of the rule (made by ourselves or by someone else).  There must be some sense in which what we are doing is the same thing.  It is this possibility of agreement or disagreement that is essential for an act of rule following.  Now, what happens when I try to name a sensation by the act of concentrating on that sensation – as it were, giving a “kind of ostensive definition” of a word by pointing to its referent inwardly.  What is the purpose of this ceremony?  To impress on myself the connection between the sign and the sensation?  But that is only so that “I remember the connection right in the future.”  And Wittgenstein goes on:

But in the present case I have no criterion of correctness.  One would like to say: whatever is going to seem right to me is right.  And that only means that here we can’t talk about ‘right’.[8]

Now some commentators have taken this to be a case where verification is required for the act of applying a symbol to a sensation.  If this were the case we could reject the argument so long as we reject this kind of Verificationism.  But there is no indication here, or in the surrounding text, that what is missing for the private linguist is merely the test of agreement with other cases.  Rather, there is no room here for the concept of ‘agreement’ or ‘same’ at all. 

What aspect of the sensation is going to be used as a criterion of identity?  In virtue of what does one say, “this is another instance of S”?  Just that it seems to me to be so.  “Whatever seems right to me is right”, and there is no room for the possibility of disagreement.  But if there is no room for disagreement, we cannot talk of agreement either, and thus we cannot talk of ‘right’, or of the application of a rule.

The proponent of a private language will not accept this so easily, however.  What makes Wittgenstein so sure that there is no room for the notion of agreement here?  Perhaps I do get my application of the word ‘S’ wrong sometimes, and I just don’t know it.  But Wittgenstein is insisting that if that is possible, then we should be able to articulate what it is that we are getting right or wrong.  There must be a game of getting it right or wrong set up, and that is precisely what the private linguist has not done.  Otherwise what is it that we are supposed to be pointing at?

Consider Wittgenstein’s remarks on ostensive definition early on in Philosophical Investigations.  Wittgenstein considers (in section 28) how someone might be taught the meaning of the number two by pointing at two nuts.  On the face of it, ostensive definitions look hopelessly inadequate:

But how can ‘two’ be defined like that?  The person one gives the definition to doesn’t know what one wants to call “two”; he will suppose that “two” is the name given to this group of nuts!

And Wittgenstein concedes, “an ostensive definition can be variously interpreted in every case.”  This is, in fact, just a special application of the ‘sceptical paradox’ that Kripke has made so much of.  We want the ‘object’ that is the referent of our word to determine the rule for its use, but what we are trying to point to can be variously interpreted, and so it seems we have not provided any guide to the meaning of the word at all.  Yet the sentence that precedes the passage quoted runs: ‘The definition of the number two, “That is called ‘two’”––pointing to two nuts––is perfectly exact.’[9]  The answer to the problem of ostensive definition is that the person you are teaching must know or guess that you are referring to a number, and what a number is.  That is, he must already understand the use of the word.  And he will demonstrate that he has correctly understood if he goes on to use the word ‘two’ correctly.  If he does not then we must offer further examples or instructions.

The point that is relevant to the discussion of private language is that the ostensive definition only makes sense in the context of a use.  This is the point that Wittgenstein makes in section 257, immediately preceding the passage widely regarded as “the private language argument”:

When one says “He has given a name to his sensation” one forgets that a great deal of stage setting in the language is presupposed if the mere act of naming is to make sense.  And when we speak of someone’s having given a name to pain, what is presupposed is the existence of the grammar of the word “pain”; it shews the post where the new word is stationed.

So the point could be put this way: you can very well be defining something when you say you are ‘pointing inwardly’, but you are only doing so insofar as you have a regularity of application to go by.  Otherwise your pointing remains hopelessly ambiguous.  But if there is some regularity of use that will disambiguate your pointing gesture, then you should be able (at least in principle) to share it with us.  Once we imagine that the private linguist has something before him, we have assumed that he has something that has conceptual shape.  The basic idea has much in common with the conceptual idealist’s rejection of the given.  If it is ‘a something’ then it can be described.  Hence the assumption that a private linguist has something before him, but we cannot say what, is mistaken:

If you say he sees a private picture before him, which he is describing, you have still made an assumption about what he has before him.  And that means that you can describe it or do describe it more closely.  If you admit that you haven’t any notion what kind of thing it might be that he has before him—then what leads you into to saying, in spite of that, that he has something before him?[10]

Do these remarks constitute a cogent argument against the possibility of a private language?  The private linguist will argue that as an argument, this line of reasoning begs the question against his position – for it is precisely his contention that some concepts are known from a private sample that cannot be publicly articulated.  Put in this way, Wittgenstein’s demand that the private linguist further describe what he thinks he has before him seems nothing more than a refusal to take his opponent seriously.  But I think this unsatisfactory result comes about through misinterpreting the later Wittgenstein’s method.  There is no “private language argument”.  What Wittgenstein’s collected remarks offer us is a series of approaches to a set of more or less related philosophical theses.  Wittgenstein’s method is to uncover the intuitions that underlie them.  Once these intuitions are laid bare, we can see how ill founded and muddled they are.  While any single remark may not prove beyond possible doubt that the private linguist is mistaken, taken together they strip away the façade that there is good reason to believe the story that the private linguist tells.  The private linguist is made to retreat to a position that invites us to say: “But isn’t that ridiculous?  Why should we believe that?”  And of course the private linguist (in particular, the younger Wittgenstein) put forward his thesis as part of an explanation of language.  But an explanation that turns on something inherently mysterious, it seems to me, is no explanation at all.  If the private linguist wants to deny that, so be it, but it leaves his position without the intuitive appeal it once seemed to have.  Wittgenstein does not offer a cogent argument against private language, because he does not offer an argument in the traditional sense.  His remarks, however, are convincing.

5.3 McDowell on Wittgenstein’s Rejection of the Given

This connection with conceptual idealism (and so far it is only a connection—I am not claiming that Wittgenstein was a conceptual idealist) can be put like this: The possibility of a regularity of use can also be called a public concept.  To possess a concept, on this account, would be to be a master of the technique of the application of a word.  Wittgenstein’s claim then amounts to this: language can only be used to refer to that which falls under some public concept.  Hence we cannot use language to talk about a private object that does not fall under a public concept.  It follows that talk of the ‘given’ as a pre-conceptual something is mistaken.

The idea of Wittgenstein’s remarks on private language as a rejection of ‘the Myth of the given’ has been taken up by John McDowell in ‘One Strand in the Private Language Argument’.  McDowell’s reasons for so interpreting Wittgenstein seem to stem from a general approval of the project that rejects this “dualism of conceptual scheme and pre-conceptual given”[11]:

[W]e ought to look with favour on a thesis on these lines: nothing can count as an episode in a stream of consciousness unless it has (already, we might say) a conceptual shape, an articulable experiential content.[12]

But McDowell concedes that this may not be precisely the way Wittgenstein intended his remarks:

I do not mean to suggest that Wittgenstein sees his polemic in precisely these terms, as an application of a general rejection of the dualism.  Opposition to the dualism makes good sense of some of what he says.  I think it leaves some of what he says unexplained, and some looking positively mistaken.[13]

But if such an interpretation leaves some of Wittgenstein’s remarks unexplained, we have good reason to look for another interpretation.  I would like to argue, however, that the problem lies in the way that McDowell conceives of Wittgenstein’s rejection of the dualism.  Once we have corrected that, we can make good sense of all of Wittgenstein’s remark on this issue.

The rejection of the given that McDowell has in mind is the one exemplified by Wilfred Sellers argument in Empiricism and the Philosophy of Mind.  This argument centres on an inconsistency in the notion of the given as it is supposed to play a role in empiricist epistemology.  The given is supposed to be pre-conceptual, and as such cannot stand in rational relations to anything.  It therefore falls outside of ‘the logical space of reasons’.  But it must ‘impinge’ on that space, so that it can act as a ground or justification for our lowest level conceptualisations.  The traditional rejection of the given argues that nothing could play both these roles.  Note that this argument leaves open whether there is anything that corresponds to each of these roles individually.  It merely argues that nothing could play both roles, and that is what the proponent of the given wants.

McDowell quotes Rorty’s attempt to expound a “Wittgensteinian point” by identifying two different forms of knowledge.  The first is “the way in which the pre-linguistic infant knows that it has a pain”, which “is the way in which the record-changer knows the spindle is empty, the plant the direction of the sun, and the amoeba the temperature of the water.”[14]  This is to be sharply distinguished from “what a language-user knows when he knows what pain is”.  Rorty argues that “the mistake that Wittgenstein exposed” is “the notion that knowledge in the first sense–the sort manifested by behavioural discrimination–is the ‘foundation’ (rather than simply one possible causal antecedent) of knowledge in the second sense”.

While McDowell has serious misgivings about talking about the pre-linguistic infant as ‘knowing’ that he is in pain, rather than simply ‘feeling’ it, and the behaviouristic remarks this way of talking gives rise to, he is otherwise in broad agreement with Rorty’s interpretation of Wittgenstein.  McDowell wants to take the point about knowing one is in pain and transpose it into a point about having the concept of pain:

[T]hen his point can be put as one about the relation between pre-conceptually felt pain and episodes of pain that belong in full-fledged streams of consciousness, conceived as necessarily in conceptual shape.  The fundamental point is the distinction between foundations and (mere) causal antecedents: non-conceptual pain (in pre-linguistic infants) is a causal antecedent of the ability to have conceptually structured pain episodes, not a continuing ingredient in them that grounds the conceptual structure involved.  Put like this, Rorty’s point perfectly fits the reading of Wittgenstein I am recommending.[15]

So McDowell finds a story in Wittgenstein that has the form of Sellar’s rejection of the given.  Note that the non-conceptual pain still has the role of a (mere) causal antecedent to the ability to have conceptual pains.  The distinction between the two is that a pain must have conceptual shape if it is to feature in the stream of consciousness.  This leaves McDowell in the uncomfortable position of having to defend the claim that pre-linguistic infants and animals do not have pains in the same sense we do.  Surely this is not a claim that we should be happy about saddling Wittgenstein with.  Furthermore, by his own admission, the interpretation McDowell is advocating leaves some of what Wittgenstein says unexplained.  The problem is that in transposing Rorty’s point, McDowell conflates two separate Wittgensteinian points.

The point about how we know we are in pain is indeed related to Wittgenstein’s remarks on private language, though perhaps not as closely as McDowell proposes.  The standard view is that what I mean by the word ‘pain’ can only really be known to me, for I am referring to what can only be known to me, namely, my pains.  We are tempted to express this thought by saying, “I only know what pain is from my own case.”  The remarks on private language are a direct attack on this view of pain language.  Given that this picture is rejected, Wittgenstein must offer some indication of how pain language might be learned.

This is where Rorty’s ‘Wittgensteinian point’ has a its source in Philosophical Investigations.  Against the rejected picture of pain language, Wittgenstein offers the following alternative as one possibility: “words are connected with the primitive, the natural, expressions of sensation and used in their place.”[16]  Thus the verbal expression of pain is taught as a replacement of crying behaviour.  That is not to say that the word ‘pain’ describes crying, but replaces crying and inherits from it the connection with pain as the public expression of pain.

This ‘connection’ requires careful reflection.  It is not that we say ‘I am in pain’ instead of crying out — suppressing the tendency to scream, as it were — or as a new involuntary reaction to pain stimulus.  Wittgenstein describes the way pain language is learned.  He is suggesting one way that the word pain enters our language games.  We could put the point like this: having unconceptualised pains (in pre-linguistic infants) is a causal antecedent of the ability to know one is in pain in the sense of being able to say one is in pain[17].

But note that I used the word ‘unconceptualised’ (rather than ‘non-conceptual’) to describe the pains that infants have.  They are not conceptualised, that is, by the infant.  Yet they are the very same pains that we attribute to linguistically competent adults when they wince and say, “I am in pain!”  The knowledge that the infant, or an articulate adult, or anything else for that matter, is in pain requires the public concept of pain[18].

I want to use the expression ‘non-conceptual’ in a somewhat more fundamental sense than it is used by McDowell.  For him, having a ‘non-conceptual pain’ is having something that has not been conceptualised as part of a stream of consciousness.  Only the linguistically competent can, on McDowell’s account, have fully fledged streams of consciousness.  The ‘pre-conceptual’ pains of infants act as mere causal antecedents of the things we commonly refer to as pains[19].  But I want to use the term ‘conceptual’ to refer to anything that has a nature that can be expressed as a public concept.  Something is therefore non-conceptual if it does not admit of being captured by a public concept[20].  On the traditional picture of pain-language that Wittgenstein attacks, pains are just such ‘non-conceptual somethings’.  But Wittgenstein argues that language can only refer at all insofar as we have a grammar set up for our referring expressions[21].  Since it makes no sense to talk about an essentially private use of a word, a ‘something’ which makes contact with language must fall under a public concept (use).  It follows that nothing referred to by language is non-conceptual in the sense I am using the term.

Not only does this way of looking at Wittgenstein’s remarks avoid saddling him with the unworthy view that infants do not have real pains in the sense that we do, but it makes sense of those passages that McDowell finds obscure or in error.  The two offending passages that he quotes are to be found in section 304 of Philosophical Investigations:

“And yet you again and again reach the conclusion that the sensation itself is a nothing.”–Not at all.  It is not a something, but not a nothing either!  The conclusion was only that a nothing would serve just as well as a something about which nothing could be said.  We have only rejected the grammar which tries to force itself on us here.

And section 293:

The thing in the box has no place in the language-game at all; not even as a something … That is to say: if we construe the grammar of the expression of sensation on the model of ‘object and designation’ the object drops out of consideration as irrelevant.

Note that Wittgenstein here stresses, as he was fond of stressing, that what he is doing is making a grammatical remark.  He has not denied that we have sensations, but “only rejected the grammar which tries to force itself on us”.  This point has remained obscure to most commentators on Wittgenstein, but I think a proper understanding of it is essential to a proper understanding of Wittgenstein’s later philosophy in general.  It is particularly relevant to understanding the rather odd expression “not a something, but not a nothing either”.  But first, let us consider the way in which McDowell objects to this locution.

According to McDowell, given the framework of the rejection of the given that he has outlined (namely the Sellars-Rorty rejection of something that impinges on the space of possible reasons), Wittgenstein should have said something more like the following:

The sensation (the pain, say) is a perfectly good something—an object, if you like, of concept-involving awareness.  What is a nothing (and this is simply a nothing, not “not a something, but not a nothing either”) is the supposed pre-conceptual this that is supposed to ground our conceptualizations.[22]

So, according to McDowell, Wittgenstein should have denied something that he did not (the pre-conceptual this).  Furthermore, Wittgenstein should have accepted something that he apparently denies, namely the sensation that occurs in ‘concept-involving awareness’.  McDowell thinks that Wittgenstein is motivated to deny sensations as objects, because of the following thought:

The idea of encountering a particular is in place here only because the experience involves a concept (pain, say, or toothache): the particular has no status except as what is experienced as instantiating the concept.  So the idea of encountering a particular in this application lacks a kind of independent robustness that we can credit it with in other applications.[23]

Given this interpretation of the main force of “the private language argument”, McDowell concludes that we should not say that the sensation is not a something, but rather that at best it is a limiting case of the model of object and designation.  But this simplifies Wittgenstein’s remarks on private language considerably.  For a start, what Wittgenstein is objecting to is not the idea that we call our pains ‘objects’, but that in doing so we should realise that we have not yet said anything revealing about their nature.  The danger of calling them objects is that the picture of sensation language that this engenders: namely, the idea that I know what the word ‘pain’ means from my own acquaintance with my pains.  The passage that is commonly called ‘the private language argument’ (namely, section 258) has the following form: Wittgenstein starts with the assumption that we only know pains from are own case.  And Wittgenstein asks what it would be to introduce a term that refers to such private objects into a language.  The private linguist thought that he could first point inwardly at his pain, and then derive from the object designated a (private) use for the label that he has supposedly attached to it.  No one else will be privy to this use because no one else will be privy to that object.  This is the ‘model of object and designation’ that Wittgenstein objects to.  He objects that such a method of introducing a term into language would not work.  The object that the private linguist thought he was pointing to, which is not a pre-conceptual this, but a ‘this that cannot be conceptualised’, ends up playing no role in the language-game that results.  Thus Wittgenstein claims the sensation as it is conceived in this story is ‘not a something’.  It has not yet been given the status of a something in the story told by the private linguist.  Only with a language-game already set up can we make sense of attaching a label to a thing.  So the first thing to note is that Wittgenstein is not denying that our sensations are ‘objects of concept involving awareness’ (for this expression can, of course, be given a use).  He is denying that in thinking of it as a ‘limiting case of object of designation’ we have settled its nature prior to examining how pain language is used.  He is denying that the private linguist can help himself to the something prior to this examination.

But even this account of Wittgenstein’s remarks on private language is too simple, for Wittgenstein does not present a single “private language argument”, but a battery of attacks on a group of related philosophical theses.  The common theme is to start from the assumption that ‘I only know such-and-such a sensation from my own case’, and to show how this would not enable us to talk about the sensation in question at all.

The second passage quoted above that McDowell objects to gives only Wittgenstein’s conclusion to section 293.  McDowell has neglected the idea that Wittgenstein was considering, which is the idea that “it is only from my own case that I know what the word ‘pain’ means”.  McDowell misses the following from the beginning of section 293:

If I say of myself that it only from my own case that I know what the word ‘pain’ means–must I not say the same of other people too?  And how can I generalize the one case so irresponsibly?

Now someone tells me that he knows what pain is only from his own case!—…

This idea suggests that pain language is set up in a way parallel to the following:

Suppose everyone had a box with something it: we call it a “beetle”.  No one can look into anyone else’s box, and everyone says he knows what a beetle is only by looking at his beetle.—Here it would be quite possible for everyone to have something different in his box.  One might even imagine such a thing constantly changing.—But suppose the word “beetle” had a use in these people’s languages?—If so it would not be used as the name of a thing.  The thing in the box has no place in the language-game at all; not even as a something: for the box might even be empty.—No, one can ‘divide through’ by the thing in the box; it cancels out, whatever it is.

So again we have Wittgenstein wonder what it would be like if the ‘object of designation’ model were correct.  That model has it that we have our own sample of the thing to be designated, and we know what the word means by examining that object.  But then whatever use we make of such a ‘designating’ word would not allow us to talk about the very thing that we thought we were supposed to be talking about: “the object drops out as irrelevant”.  This is not a rejection of the idea that the sensation is a something, but a reductio on the notion of construing the sensation on the model of ‘object and designation’.

The first passage quoted above that McDowell objects to (from section 304) is also clearly a reductio on the object of designation model.  “The conclusion was only that a nothing would serve just as well as a something about which nothing could be said.”  Wittgenstein is rejecting the sensation as it is construed as a private object.  But Wittgenstein does not want to deny sensations.  The passage continues:

We have only rejected the grammar which tries to force itself on us here.

The paradox disappears only if we make a radical break with the idea that language always functions in one way, always serves the same purpose: to convey thoughts—which may be about houses, pains, good and evil, or anything else you please.

The paradox is that, if we construe the sensation on the model of object and designation, we end up not being able to talk about sensations at all: they fall outside of language.  The paradox is dissolved if we do not bring a notion of an object to the understanding of, say, pain, but look at how pain language is actually used.  The grammar of the word pain shows us the place that the word ‘pain’ has in our language.

5.4 Wittgenstein, Idealism and the Limits of Language

But then why does Wittgenstein not say that what the private linguist wants is simply a nothing?  The short answer is that the private linguist has not made enough sense to give us anything to deny.  If to assert nonsense does not make sense, then its denial will equally lack sense.  Although this may seem a rather flippant answer, I believe that, properly understood, this holds the key to understanding many of Wittgenstein’s more perplexing passages.

Wittgenstein repeatedly urged that he “was not trying to deny anything”, and that he was only making “grammatical remarks”.  These two ideas are clearly related in Wittgenstein’s mind, as is clear from section 307:

“Are you not really a behaviourist in disguise?  Aren’t you at bottom really saying that everything except human behaviour is a fiction?”—If I do speak of a fiction, then it is of a grammatical fiction.

I think it is significant that, after a further comment on “the philosophical problem about mental processes and states and about behaviourism”, which ends with “And naturally we don’t want to deny them [mental processes]”, Wittgenstein makes one of his elliptical comments about  “the aim of philosophy”.  The connection is this: the aim of philosophy is not about asserting the existence of this, or denying the existence of that.  It is simply a matter of getting the grammar of our language clear.  Wittgenstein is not in the business of doing ontology.  On the contrary, he wants to demonstrate the philosophical practice of ontology is mistaken.  All we can do in philosophy is examine the way language works, and this does not provide the philosopher with any special insight into reality.  Both the private linguist who wants to make an inarticulate sound, and the Platonist who wants to ground our language use in a theory of meaning, share a common mistake.  They both assume that philosophy is about reaching out beyond the ordinary workings of language (to what is ‘really there’, as it were).  But in so doing they leave the limits of language.

The full answer to this question can only be appreciated with a better understanding of Wittgenstein’s philosophy.  We are tempted to think (along with the conceptual idealist) that if we have shown that sensations as the private linguist construes them result in a paradox, then we have shown that private objects do no exist—that they are nothings.  But according to the later Wittgenstein, this misconstrues the nature of language.  Language does not have the complete contact with reality that the conceptual idealist takes it to have.  So we cannot say that beyond the limits of language there is ‘nothing’.  Nor are the things we talk about connected in a contingent manner with language, such that beyond the limits of language there are things that cannot be expressed in our language.  There is not, as the conceptual realist would have it, ‘something’ beyond the limits of language.  Both the words ‘nothing’ and ‘something’ belong to language, and beyond its limits, we can say nothing.  Understanding this thought provides the key to understanding Wittgenstein’s remarks on philosophy, and his relationship to idealism.  So it is to these matters that I will turn in the next chapter.