[ previous | contents | next ] |
Wittgenstein on meaning and rule following
“I set the brake up by connecting up the rod and lever.”—Yes, given the whole of the rest of the mechanism. Only in conjunction with that is it a brake-lever, and separated from its support it is not even a lever; it may be anything, or nothing.
Philosophical Investigations, §6.
The initial problem investigated in the Tractatus was the question of meaning: how can our words signify anything? On the basis of the assumption that, in order to say anything at all, our words must have a determinate meaning[1], Wittgenstein constructed a metaphysical theory that had far reaching consequences. The result was not just a treatise on the central question of meaning, but a work that made concise pronouncements on a wide range of metaphysical disputes: on the self and subjectivity, ethics and religion. Furthermore, it was intended to act as a kind of ‘Prolegomenon to Any Future Metaphysics’, albeit one that was more radical and restrictive than most other philosophers would want as a solution to the ‘problem of metaphysics’. Metaphysical truths, Wittgenstein concluded, cannot be expressed in meaningful discourse, which is restricted to the statements of natural science[2]. That Wittgenstein went on to draw such conclusions from the original question of ‘what is meaning?’ shows how central he took this question to be. The answer to this question, he assumed, must surely provide the key to all metaphysical problems. Under the influence of Russell, Wittgenstein took the question of meaning to be the metaphysical investigation par excellence. And the solution to the problem of meaning was to provide the key for a general solution to the problems of metaphysics.
Philosophical Investigations begins with ‘a particular picture of the essence of language’[3] provided by a passage from Augustine. ‘The individual words in language name objects—sentences are combinations of such names’[4]. This picture, whether or not it was one held by Augustine, was certainly held by Wittgenstein himself during the period of the Tractatus. It is presented at the beginning of Philosophical Investigations as a target to be attacked, which Wittgenstein does by pointing out inadequacies in this picture: it is too narrow a characterisation, for instance, for not everything we call language can be so simply described. The assumption that words must have a determinate and precise meaning is also brought into question, as is the idea that language must consist of simple names that stand in a pictorial relation to simple objects.
Some commentators have wondered why so much fuss is made of this obviously inadequate picture of language, one that was not even really held by Augustine. In particular, why did Wittgenstein choose to begin his work in such a way? The standard view is that the extent and rigour of the attack on the ‘Augustinian picture’ reflects the extent of Wittgenstein’s break with his earlier thought. It seems clear to me, however, that the picture of meaning in question, and the emphasis Wittgenstein places on its rejection, has far more significance in his later work than this simple view suggests. It does show a rejection of the Tractarian notion of meaning, but the importance it is given also shows a salient continuity with his previous work. Meaning is still seen as central to philosophical theorising, and since Wittgenstein’s concern is still to show how this enterprise is misguided, he again begins with the concept of meaning. One difference between the two approaches is this: in the earlier work he tried to show that the philosophical concept of meaning necessarily excluded itself, along with all other metaphysical notions, from the realm of meaningful discourse. This is presented as an inevitable consequence of correct metaphysical investigations. In the later work, however, metaphysical theorising is attacked at a more fundamental level. His aim is to question the intuitions and motivations that bring us to suggest philosophical theories in the first place. He examines the tendency to produce theories, and the result is a series of remarks on where we go wrong in philosophy. It is only natural then, that Wittgenstein should take the tendency to theorise about meaning as his departure point. Just as in the Tractatus metaphysics begins with a theory of meaning, so Philosophical Investigations takes as its starting point the tendency to provide metaphysical answers to questions about meaning. These metaphysical misconceptions all have one thing in common: they take the meaning of a word or proposition as something independent of its use. The model of meaning that Wittgenstein rejects places too much emphasis on the things that we talk about, and treats them as if they were all objects, with essences that are independent of our concepts. Wittgenstein’s remarks amount to nothing less than a rejection of conceptual realism.
At the beginning of the Philosophical Investigations we are introduced to the notion of a primitive language-game in which we can see clearly the functions of the words and phrases used by those involved. In particular, Wittgenstein pays special attention to the use of the words that function as names in these language games, emphasising the contexts and conditions under which it is appropriate to apply names to objects, and what role their application might play. In section 43 he makes the following observation from his reflections:
For a large class of cases–though not for all–in which we employ the word “meaning” it can be defined thus: the meaning of a word is its use in the language.
And the meaning of
a name is sometimes explained by pointing to its bearer.
It is worth emphasising that Wittgenstein did not see these observations as constituting some sort of theory of meaning. They are simply meant to be observations. But describing the meaning of a word as its use does bring it into conflict with a certain prevalent philosophical notion of meaning: that the meaning of a word is given solely by the object it refers to. On this philosophical conception of meaning, how a sign should be used (it rule for use) is provided by the nature of the object it refers to. But then the use is not its meaning, but something derived from it meaning. Wittgenstein thinks this false conception of meaning comes about when we make too much of the fact that the meaning of a name is sometimes explained by pointing to its bearer. This thought tempts us to confuse the meaning of the word with its bearer, and to generalise this misleading model of ‘object and designation’ for the whole of language.
This conflict between meaning as use and the rejected model can be stated more generally using the notion of rule following: one applies a word correctly if one applies it in accordance with a rule, and this implies the possibility of using it incorrectly, in conflict with the rule. But what is this rule, and what counts as acting in accordance with it or against it? The answer that tempts us supposes that the rule is always determined by something external to its use (as if it were latent in the ‘object’ that a sign refers to). The thought is that unless the rule is determined by something independent of my actual application of a sign, that rule could not justify that application. (It must provide an independent standard of correctness).
Consider the way this question was dealt with in the Tractatus. The central question there was ‘what (metaphysical) facts determine the meaning of a sentence?’ The answer was that names refer to objects, and these objects are the meanings of the names. So the meaning of a name (the object to which it refers) is something external to it. But then how can we guarantee that the name has the meaning we suppose it to have? Well, Wittgenstein supposed that the logical form shared by a fact and a corresponding proposition constituted a set of isomorphic relations between names and objects. A name refers to an object so long as that name is part of a logical picture that corresponds to the state of affair to which the object belongs[5]. More generally, a sign has meaning in virtue of having a logical syntax (in virtue of being used in accordance with a rule) that matches the internal properties of the object that it refers to. But the meaning itself is the object referred to[6].
At one time after writing the Tractatus Wittgenstein talked about the possibility of a certain ‘phenomenology’[7]. The logical form of elementary propositions is not revealed by ordinary language use, which uses the same forms for grammatically quite different propositions (“This paper is boring” and “The weather is fine” have the same superficial subject-predicate form). Since he held that the logical grammar of analysed language reflects the nature of the objects it refers to, he also held that the analysis of phenomena could provide an insight into the grammar that ordinary language disguises. So Wittgenstein held that the grammatical form of words could be derived from the phenomena (objects) that they refer to. In this paper he expresses a change of opinion from a certain tractarian position (that the elementary propositions are independent of each other), but in the main the model of meaning from the Tractatus is still presupposed. Though more complex in its details, it is the same model of language that is presented at the beginning of Philosophical Investigations. Given a proper analysis of the way we use words, the underlying grammar must accord with the objects that they signify. So when we learn to use words, we do so on the basis of examining the objects they refer to. One might say that Wittgenstein assumed that the objects that our words refer to present us with rules for the use of those words: that the rules we use in language must somehow be answerable to the natures of objects. By examining the nature of objects directly – that is their logical form – one learns how to use a name such that it will have the same logical form and hence refer to that object. At the time of writing, however, Wittgenstein gave little in the way of detail of how this derivation might take place. He took for granted our ability to grasp essences that are somehow intimated to us by external objects. But how do we know that we have grasped a logical form (and thus discerned a rule for the use of a name) correctly? The Tractatus left unanswered the question of how we know we are using a sign correctly; how we know we are applying a rule in accordance with the object we take it to signify.
He reassesses this question in more general terms in Philosophical Investigations by introducing the concept of interpreting a rule. How are we to know how we are to interpret a given expression of a rule? Philosophical Investigations offers a new description of meaning—that it can often be defined as its use. But if the meaning of a word is its use, what does understanding the meaning of a word consist in? The answer that tempts us drags us back to the tractarian conception of meaning: that what we grasp is something external to our use of a sign. When we are taught the meaning of a sign (or asked to guess it from examples of its use) we often have the feeling that we have caught hold of something that ‘lies behind’ the examples that we have been shown.
The notion of ‘grasping’ a rule is problematic for Wittgenstein’s later conception of meaning as use, for clearly we can grasp the meaning of a word ‘in a flash’. “We understand the meaning of a word when we hear or say it.”[8] What kind of strange mental state is it that constitutes grasping a whole use of a word, or more generally, the whole application of a rule, which is extended in time? What does one grasp then? The obvious answer is that what one grasps is an interpretation of some kind. But any interpretation of a rule that we have in mind is apparently nothing more than a rule for how the first rule should be followed. It too must be grasped, and this seems to only deepen the regress.
Consider what we mean when we say, ‘I now understand how to go on,’ in the case of being given a segment of a series of numbers. Perhaps a formula comes to mind. We are given, say, the numbers ‘1 3 5 7’ and the formula ‘xn+1 = 2n + 1’ occurs to us. Does this constitute our understanding how to go on? The formula is a symbolic expression, and as such can be interpreted in an indefinite number of ways. For now we need rules for how we should interpret the signs of this expression. Any rule, or expression of that rule, can be interpreted in any number of ways. Even an arrow, which points in one direction, could be interpreted as meaning, ‘walk in the opposite direction’, and this interpretation in turn could be given a further interpretation. Eventually we must just act, and the meaning of the last interpretation can only be given in terms of our following it or going against it. But what counts as following it or going against it here?
In the Tractatus Wittgenstein asked the question ‘what is meaning?’ In Philosophical Investigations he reassessed the assumptions that brought him to his earlier conclusions. He had assumed that for a proposition to have meaning it must have a determinate sense. From this he concluded that certain (inexpressible) metaphysical facts must determine unambiguously what our words mean. But the later Wittgenstein realised that this requirement amounted to a kind of scepticism. And once one expresses this scepticism explicitly, it seems to have no answer. For if we consider the theory offered in the Tractatus more carefully, we realise it will not solve the problem that it was designed to. The picture theory of meaning demonstrably fails in its task, for any arrangement of (mental) elements can be taken in any number of ways. We still have the problem of how the arrangement of (mental) elements should be taken as one picture rather than another. Any image or picture can always be applied in a different way[9].
In Philosophical Investigations the question ‘what is meaning?’ has been replaced by the more general question ‘what is it to follow a rule?’ The scepticism that brought him to his earlier (flawed) solution can then be made explicit in the following way: what is it that justifies me in thinking that, in using a sign on a particular occasion in a particular way, I am using it in accordance with a rule that endows it with meaning? But note that this way of reformulating the question presupposes a notion of meaning analogous to the tractarian model of ‘object and designation’. It demands that meaning has the form of a ‘justification’ of the use of the word in the sense that it must both (a) be something (some ‘object’) that is independent of that use, and (b) provide me with reasons for using it (it must guide my use, or determine how I ought to use it). So it seems that an independent standard of correctness[10] must be compared to any particular application of a word. But whatever this standard is, it must also fill the role of the thing that guides my application of a rule. So the search is on for this thing that we grasp when we understand a rule. But every time we think of a way of grasping the rule, it seems to slip away from us as we think of new ways of interpreting what we have grasped.
As well as rejecting the picture theory of meaning, Wittgenstein also rejects a series of other natural proposals. The application of a rule cannot be fully determined by some kind of look up table, for it could always be read in another way[11]. Neither can meaning be fully determined by a mental process, or some kind of formula or logically pure expression of a rule. Any formulation could always be interpreted differently. None of the proposed solutions are up to the job of determining how a rule should be applied, for the scepticism that was first raised can be reformulated for any proposed solution. Nothing in my act of understanding a rule seems to justify my application of it.
The strangeness of the act of understanding a meaning is also reflected in the case of intending. If I intend to play a game of chess, I surely know what I intend to do. But the meaning of the word ‘chess’ is wrapped up in the fact that chess is defined by its rules. Does the whole set of rules then flash before my mind? Surely not. Then do I not really know what I intend until I have played the game? Surely I do. So what is the connection between my act of intending to play chess and the rules of the game? Wittgenstein answers, “Well, in the list of the rules of the game, in the teaching of it, in the day-to-day practice of playing.”[12] That is to say, in the actual rules, teaching and practice of the game, not in a mental shadow of them. Once I have learned to play chess, I am in the position to intend to play chess (as opposed to merely intending to play the game that everyone refers to as ‘chess’, however that is played).
The significance of this simple answer is easily overlooked. Indeed, it may not strike one as an answer to the sceptical problem of meaning at all. The problem was how a rule can show me how to proceed at any particular point, and whatever one does is in accordance with the rule on some interpretation. What this shows, says Wittgenstein, is that interpretations do not support rules. The connection between a rule and following it in the correct way lies in the training one receives, and the actual instances of following it. This is not to merely give a causal connection, he insists, for a use exists for a rule only in so far as there exists a regular use for the rule – a custom. Thus language is characterised as an activity, which in turn is characterised in terms of an ability to act. And, he adds, there is no such thing as a single isolated act. There is only action in the context of a common custom. This is the only way we can give meaning to a particular application of a rule, and we should neither look for a more independent measure of correct application, nor think we can give meaning to a word without the possibility of a regular use.
But how does this answer the sceptic? The short answer is: it does not. Wittgenstein did not intend to answer scepticism, but to reject it. The real question is ‘what is it to follow a rule?’ and this is what he has answered. It was the sceptic who demanded that this question must be taken to mean ‘what justifies my using a sign this way?’ This sceptical question has been rejected. We have looked for justification behind justification, and finally the “spade is turned” and we are inclined to say, “This is simply what I do”[13].
Before I give a longer answer, it will help to examine the somewhat different interpretation of Wittgenstein given by Saul Kripke. Kripke has characterised Wittgenstein as offering “a sceptical solution to a sceptical problem”. A ‘straight solution’ would be an argument to the effect that the sceptic has overlooked some justification. A ‘sceptic solution’, on the other hand, starts by admitting that there is no answer that will meet the sceptic’s challenge. Kripke reads Wittgenstein as reaching a sceptical conclusion: nothing determines meaning, and there is no fact of the matter as to what I mean with my words[14]. Kripke’s interpretation of Wittgenstein is wrong, but wrong in an interesting respect.
Kripke has endeavoured to reformulate Wittgenstein’s problem of meaning and the sceptical paradox that it gives rise to in a more precise form. Take ‘quus’ to be an operation defined as follows: x quus y is the same as x plus y, so long as x and y are both less than 57, otherwise x quus y is 5. So for numbers less than 57, the results of ‘quaddition’ are identical to addition, and thereafter they are radically different. Assume for the sake of argument that I have not previously added numbers greater than 57 (or move the point at which plus and quus diverge to a point greater than any actual addition I have so far performed or considered). The ‘sceptical problem of meaning’ can be raised with the following question: How do I know that in the past I meant plus and not quus? And if there is nothing that distinguishes these two cases, what makes it the case that I am carrying on as before? For we feel that some fact about my past intentions to use the plus function must justify my current responses to certain arithmetic question in that it must determine what I ought to say.
Saul Kripke claims that his version of the ‘sceptical paradox’ is somewhat more carefully formulated than Wittgenstein’s problem. He has indeed defined a more precise sceptical problem. But in doing so he has lost sight of Wittgenstein’s motivations for raising the problem in the first place. Kripke is one of many commentators who find a strong sense of tension in Wittgenstein[15]. On the one hand Wittgenstein is explicitly and ardently opposed to any kind of philosophical theorising. On the other, these commentators find substantial philosophical projects that are (a least partly) undertaken in his work, but because of Wittgenstein’s anti-philosophical attitude, they claim, he is unable to acknowledge them as such. Thus Kripke talks enthusiastically about the ‘new form of philosophical scepticism’[16] that Wittgenstein has invented, but concedes that ‘Wittgenstein never avows, and almost surely would not avow, the label “sceptic”…’[17]. Thus Kripke sees Wittgenstein as setting up a ‘real’ philosophical problem, and offering a sceptical solution it. But it is well known that the later Wittgenstein explicitly rejects the method of philosophy that offers solutions on the basis of their explanatory power[18]. So I think we should look for another interpretation.
Fortunately we do not have to look far. There is a sceptical element to Wittgenstein’s remarks, and he is accepting that nothing will answer the sceptic. No metaphysical or meta-logical facts determine that I mean plus rather than quus. There are no metaphysical facts that justify language use. This is also an anti-sceptical result, however, for Wittgenstein takes it to show that, since nothing could possibly satisfy the sceptic’s desire for justification, the sceptical question must have been misguided in the first place. It is the sceptic after all who requires such metaphysical justification, and this is just what Wittgenstein concludes cannot be given:
This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule. The answer was: if everything can be made out to accord with the rule, then it can also be made out to conflict with it. And so there would be neither accord nor conflict here.
It can be seen that there is a misunderstanding here from the mere fact that in the course of our argument we give one interpretation after another; as if each one contented us just for a moment, until we thought of yet another standing behind it. What this shows is that there is a way of grasping a rule which is not an interpretation, but which is exhibited in what we call “obeying the rule” and “going against it” in actual cases…[19]
The ‘answer’ that Wittgenstein offers is not a solution to, but a rejection of the sceptical paradox. (A fortiori, he is not offering a sceptical solution). And the ‘misunderstanding’ that Wittgenstein goes on to talk about is not just the mistaken nature of the answers, but also the sceptic’s insistence that some such answer is needed to support meaning. The sceptic began with a misconstrued concept of meaning in the first place, so it should not worry us that nothing answers to that concept. In asking for a justification for a particular application of a word, the sceptic had a particular model in mind. The justification must be in the form of something that is independent of the use I make of a word (indeed, independent of the use that the community I am part of make of it). He wanted something (some object) that was external to the use of a word, which nevertheless gave the meaning of that word. Nothing will fill that role because the meaning of the word (for a large class of cases) simply is its use.
The result is not that there is no fact of the matter about whether I mean plus or quus. I mean plus. I have been trained at school in the practice of adding numbers together and I quite rightly call my acquired skill ‘the ability to add.’ If you asked me what 73 plus 91 is, I would answer (in the majority of cases), 164. Furthermore, when I learned to add, a new fact became true of me: It became true that I could add. I gained an ability. (I mastered a technique). But does this not miss the sceptic’s point? In a way it does, but Wittgenstein is saying that the sceptic does not have a point worth answering. The question of whether I mean plus or quus only arises if we imagine that these functions exist (already, as it were) independently of our application of the plus sign.
The confusion about the demand for a justification can be made clearer with the use of some Wittgensteinian analogies. Imagine you are playing chess, and a sceptical observer demands that you offer a justification for moving you king in such-and-such a way. He is not quieted by an explanation of your strategy within the game. He wants to know what the rule is that you thought you were being guided by (why you moved the King one square and not two). How do you know, he says with a rye smile, that you are interpreting this rule as you have done in the past? Let us assume, he says, that for all the many times you have previously played chess, the board has not been arranged precisely like this, with this configuration of pieces, at this stage (number of moves) into the game. He insists that at this point in the game, the rule demands that you move the king two squares instead of one. You describe the rule to him carefully, but however precisely you try to define it, he comes up with an interpretation that is inconsistent with the move you have made.
If the sceptic persists with this line of reasoning (and it seems it has no end) you will perhaps become annoyed and at best consider his questions as a distraction from the game. “This is just how chess is played!” you shout. To continue with the retort, “But how do you know?” will now seem so weary that even the sceptic might suspect that there was something wrong with it.
Wittgenstein is suggesting that what is wrong with it is that it demands a kind of justification that is as unnecessary as it is impossible. Moreover, its impossibility demonstrates that it is unnecessary. The sceptic wants a justification that is independent, not only of your decision to move your king in such-and-such a way, but independent of your taking the rules to be such as to legitimise such a move. Nevertheless, he wants the justification to be the thing that guided your decision to move the piece as you did. The supposed decision is not that you moved it to this rather than that legal square, but which squares you took to be legal moves in the first place. But this confuses the matter, for you did not decide that the rule for the king piece is that it can move only one square at a time. You followed that rule blindly. To ask for the justification of that rule following would be to ask for a justification from outside the way your opponent and you take to be the way to play chess. And the way you take chess to be played is not something independent of the activity you and your community engage in. It makes itself manifest in the way we play chess.
Kripke may have developed Wittgenstein’s paradox into a clearer and more precise formulation, but in doing so he lost sight of the nature of Wittgenstein’s enquiry. Wittgenstein was not discovering a ‘new kind of scepticism’, but unearthing the foundations on which we build philosophical theories. He was digging for the hidden questions that had previously caused him to find a particular theory of meaning inevitable. Once those questions are made explicit, they loose their power, for we see that nothing could answer them. This could indeed be characterised as a ‘sceptical solution to a sceptical paradox’ — or better a sceptical response, for it does not solve the sceptical problem, but rejects it[20]. But if the sceptical questions were not explicitly formulated and considered in the Tractatus, the response was:
Scepticism is not irrefutable, but obviously nonsensical, when it tries to raise doubts where no questions can be asked.
For doubt exist only where a question exists, a question only where an answer exists, and an answer only where something can be said.[21]
Now it may be argued that the mere impossibility of
answering a question does not obviously make the question nonsensical. Certainly this is a controversial idea to
put forward without justification.
Scepticism is not obviously nonsensical – a point the later
Wittgenstein would have conceded. While
his later treatment of scepticism does indeed aim to show it does not make
sense, he does not simply attempt to enforce some principle such as the one
pronounced in the Tractatus. The
later Wittgenstein diagnosed each piece of nonsense carefully and
individually. But it is clear that in
the case of meaning scepticism, little more work need be done than to show that
it is unanswerable, for, pace Kripke, meaning scepticism is intolerable, if not
self-contradictory. If “there is no fact of the matter as to whether I
mean plus or quus”, there is no fact of the matter as to whether the sceptic
means by this that “there is no fact of the matter as to whether I mean plus or
quus”. Meaning scepticism is compatible with its inverse (a point made,
for somewhat different reasons, in §201 of the Investigations[22]). So if the sceptics demand for justification
of meaning cannot be answered, this indeed shows that a justification of this
kind is not required, and Wittgenstein has shown that the motivation for a
whole line of enquiry, that he himself had once found to be central to
philosophy, rests on false foundations.
The idea that a word corresponds to a meaning, given in advance by the world, is referred to in Philosophical Grammar as a ‘mythology of symbolism’[23]. This myth in mathematics, for example, suggests that numbers are mind independent[24] objects to which our use of numerals must correspond. We assume that the number is a pure form of the numeral that is immune to misuse. The problem with this notion of numbers is that it leaves mysterious just how our minds get hold of these objects, prompting the idea that they are ‘perceived’ by some faculty of intuition. This is precisely the kind of mistaken notion of meaning that Wittgenstein is trying to expose. In §186, having dealt with the feeling that there was something that I must grasp (once and for all) in order to guide all my moves in future, the suggestion is that I require a new ‘insight’ or ‘intuition’ at every stage. Wittgenstein responds with ‘It would almost be more correct to say, not that a new intuition was needed at every stage, but that a new decision was needed at every stage.’
The point is not, of course, that I can simply decide to use a word or a symbol how I like (though, in a sense, I can). Should I decide to respond with ‘five’ to the question ‘what is 57 + 68?’ I would be deciding to use the symbol ‘+’ as (say) the quus function. The point is that the meanings of my words are not independent of the way that I, and other members of my community, decide to use them. (And, to anticipate a further discussion, if there is no regularity in my decisions, we cannot talk about an ‘application’ at all.) The rule is not an abstract object or Platonic form that one requires a special intuition to grasp. We feel that a rule must be something quite different and independent from the way we use it in particular cases. But the difference in kind between a rule and an instance of following it lies not in some difference in their mode of existence. The difference between rule and rule following lies in the different roles they play in the language games we use them in[25].
One of the reasons that Wittgenstein thinks we are tempted to look for the mysterious abstract object is that we have the feeling that we grasp the meaning of a sign ‘in a flash’. How does Wittgenstein deal with this phenomenon? Is he claiming that there is nothing that I grasp? When we talk about ‘grasping’ meanings, we must think about the circumstances in which we use such expressions, and that will stop us from sliding back to metaphysical notions of meaning. When do we say, ‘I have grasped the meaning of this expression’? In what circumstances is saying this appropriate? And in such circumstances, need there be something that I have got hold of – something that determines how I should use the expression on every occasion in future? Finding no metaphysical concept of meaning that justifies the use we make of a word, we feel uneasy. There must be something that determines unambiguously how I should carry on. We still cannot rid ourselves of the tendency to offer explanations, even when we have accepted that nothing will satisfy this desire to say something more. What have we acted with or against? But the reflections on how we are to interpret a rule show that we cannot drive a wedge between our use of a word and its meaning (and this, in retrospect, should be clear from §43). It quiets the feeling of discomfort to look at the situations where we talk about acting in accordance and acting against a rule in actual cases. Talk of acting in accordance with a rule or against a rule only makes sense in the context of a regularity of use. We say, ‘that is not how that word is normally used’. (But it would be a great mistake to think that the ‘solution’ to the sceptical paradox lay in the community[26], as if that was where meaning was. After all, could I not follow a rule alone, on a dessert island, say? The problems with the community interpretation will be considered in greater detail in the following chapter).
The idea that we can grasp a rule in a way that reveals its correct application, and the accompanying idea that this is somehow necessary in order to explain language, results in the tendency ‘to invent a mythology of symbolism or of psychology’[27]. The mythology of symbolism was something that Wittgenstein was partly aware of when he wrote the Tractatus, but he failed to see that it had infected his system. He had already realised that the logical constants were not names[28]. The logical symbol ‘not’ does not stand for a logical object that one must grasp in order to understand its meaning. He expressed the temptation to make this mistake in Philosophical Grammar in the following way:
It looks as if one could infer from the meaning of negation that “~~p” means p. As if the rules for the negation sign follow from the nature of negation. So that in a certain sense there is first of all negation, and then the rules of grammar.[29]
But by the time of writing this he had realised that this kind of mistake applied to meaning in general: we look for the meaning of a word in addition to its use, either as an abstraction from the symbol, or as a mental entity. Both of these are chimeras, suggested to us by a false picture of the way meaning is bestowed upon words. G. E. Moore reports Wittgenstein as saying, that “the mere fact that we have the expression ‘the meaning’ of a word is bound to lead us wrong: we are led to think that rules are responsible to something not a rule, whereas they are only responsible to rules.”[30]
One could make this point by saying that the mythologies of symbolism and psychology are ways of making a ‘category mistake’ with respect to meaning[31]. Having had the various uses of a word explained my means of examples, the philosopher goes on to ask, ‘But what is the meaning of the word?’ The tendency to make this mistake is compounded by the fact that often, no matter how many examples are given, they will never exhaust the concept involved. But then, neither will a definition or a rule.
The theories of meaning that Wittgenstein’s remarks are aimed at come about from an over-generalisation of the picture of language that Wittgenstein criticises from section 1 of Philosophical Investigations. It is ‘the model of object and designation’ that he finds in the passage from Augustine: that the way words get meaning is by signifying objects. In the Tractatus, the idea that a proposition must have a determinate meaning, which supposedly left no room for scepticism about meaning, brought Wittgenstein to construe the whole of language on the model of ‘object and designation’. The only way to give words the determinate meaning he felt they must have was in terms of the naming relation. The logical form of a word corresponds to the nature of the object it stands for[32]. But this produces the mythology of symbolism or psychology: that the meaning of a word is always something in addition to the use of the word.
The model of object and designation supposes that first we have a world populated with various kinds of objects (physical objects, colours, numbers, and so on), and then we attach words to those things. The way we should use each word is then determined by the self-intimating nature of the object that corresponds to it. In rejecting this picture, Wittgenstein is not rejecting the notion of a mind-independent world. He is not rejecting the idea that there really are colours, numbers, or other objects. He is merely rejecting the use they are supposedly put to in philosophical conceptions of meaning. He is rejecting their use in a philosophical explanation of the way we use language. He rejects the over-generalisation of the idea of defining the meaning of words as a kind of simple ostensive definition that does not take into account the way the grammar of a word contributes to setting up the link with the thing named. ‘The definition of the number two, “That is called ‘two’”—pointing to two nuts—is perfectly exact’, but what makes this definition possible is the grammar of numerals. In order to set up a link between language and the world, we must bring a grammar (concepts, if you like, that are embedded in our practices) to the world[33].
The mythologies of symbolism or psychology are theories of meaning naturally held by the conceptual realist[34]. Conceptual realism maintains that the nature of reality is independent of our ability to think about it or grasp it. But it also claims that there can be interesting metaphysical theories or explanations. Traditionally, metaphysical theorising for the conceptual realist has assumed a distinction between the way things appear and the way things really are. The apparent structure of the world is explained in terms of, or reduced to, an underlying real structure. So the conceptual realist must hold that, although the connection between our concepts and the true nature of reality is at best contingent, there nonetheless is (hopefully) some way of discovering metaphysical truths by way of analysis of our ordinary concepts.
Now the question arises, on what basis does this analysis proceed? For if reality is independent of our concepts, there is a great deal of room for scepticism about how much of the true nature of reality is in fact revealed to our epistemic point of view. (Indeed the conceptual realist has the problem of explaining how our words and concepts can mean anything if they are divorced from the true nature of things.)
The intuition that saves the conceptual realist’s faith in metaphysics is that, however affected and distorted our view of reality is, it is nevertheless one possible view of reality. In a theory of meaning this intuition is expressed as the claim that our concepts are somehow derived from some aspect of reality. Though the idiosyncrasies of our perceptual systems and mental abilities may confuse and diminish our picture of reality, some of its true nature must provide our thoughts with content[35]. In order to make good this claim, the conceptual realist must provide a theory according to which meaning is derived from an independent reality.
One story that gets told goes something like this. The true nature of reality is given by certain abstract ideas or ‘forms’[36] that may or may not be graspable by human minds. It is these forms that embody the true nature of reality, and they themselves are real in the sense of being mind and language independent. Nevertheless, language does afford us some insight into reality, since these abstract forms are the standards to which our concepts must aspire. It is only in virtue of using a word in accordance with such objects that we manage to give our words meaning. (Or to put the point in terms of rule following, the rules themselves are such language independent objects.) The job of the metaphysician is ultimately to discern those concepts that both have application and are independent of any particular point of view – independent, that is, of any particular set of interests or perceptual mechanisms.
But this involves the mythology of symbolism that Wittgenstein goes to such lengths to criticise and reject. Conceptual realism involves the demand for an independent standard of correctness against which we should compare our application of a word. Wittgenstein argues that nothing could fulfil this role, including an abstract, mind-independent object. For how is the comparison supposed to be made? And if the demand cannot possibly be met, that shows that it was a misguided requirement in the first place.
Note that it will not help the conceptual realist to turn to some kind of naturalism about abstract objects. The meaning of our words is not derived solely from the (set of) objects they refer to in the world. We do indeed learn the use of words by being shown samples as a kind of ostensive definition. But such definitions can only be useful if the grammar of a word has already been set up; when the rules of the game to be played with the word has been laid down as a foundation[37]. Even in the case of a proper name, the meaning is not independent of the use. The game of naming objects must first be set-up.
The story the conceptual realist tells about meaning may not go exactly as I have outlined. An alternative account is given by Empiricism. A conceptual form is imposed on, or derived from, the pre-conceptual ‘given’ – that which grounds our lowest level conceptualisations. But this is just one form of the mythology of psychology. We will see in the next chapter that Wittgenstein’s rejection of the given is a special case of his arguments against metaphysical theories of meaning. Empiricism is just another way of trying to specify how our concepts are derived from something independent of those concepts.
The philosophical mistakes about meaning that Wittgenstein criticises are all attempts to give a justification of a use of a word in terms of something independent of that use. It becomes clear how this amounts to a critique of conceptual realism if we put the same point in terms of concepts: Wittgenstein criticises the attempt to give a justification of our concepts in terms of something independent of those concepts. We cannot speak of anything without using concepts, so a justification of those concepts will never be complete. No matter what kind of metaphysical reduction we perform, we will always have to assume a grammar that remains unjustified in which to couch that reduction. Conceptual realism assumes that we can speak of objects (or their forms, or some mental representation of them) independently of the concepts that pick those objects out. And this we cannot do.
Conceptual realism comes about from scepticism about whether (human) language is an adequate tool for describing the world. How can the way we choose to use a word at any moment be an adequate measuring rod to place against reality? Surely what we are aspiring to is an independent gauge given by the nature of reality itself. But Wittgenstein argues that nothing corresponds to this independent gauge. We just have the way we use words within the context of a custom or practice. The tendency to search for an independent measure of reality is one that is brought about by unnecessarily searching for a justification of these practices where there is none. We simply use a word in a particular way, and the use itself will determine whether an object, property or fact falls under its application.
Now, given that Wittgenstein is rejecting these philosophical conceptions of meaning, and therewith any plausible form of conceptual realism, does he thereby embrace its converse, conceptual idealism? One thing that we can note in passing is that the transcendental idealism of the Tractatus was put forward in part as a solution to meaning scepticism as applied to his early realism[38]: how is the connection between a language independent world and language ensured? He answered that both language and the world share a logical form. Anything thinkable must have a logical (pictorial) form since all thoughts are facts, and thus have a logical form of the objects that constitute them, as well as the states of affairs that they depict. But he later realised that there were an indefinite number of ways of ‘projecting’ the thoughts onto states of affairs in the world. This realisation brought him to reject the mythology of symbolism that formed one part of the realist-idealist hybrid of the Tractatus. Does that mean that what we are left with is still (or even more so) a kind of conceptual idealism? We will answer this question in chapter 6. But we can already note that it throws up the same problem of privacy that conceptual idealism faces. A word has meaning so long as it has a use. We cannot simply look to the world or our private experience and name something, hoping that that object will give us the criterion for applying the name. The name must have a certain use already before it can be used as a name. In other words, only a public concept can be used to refer to anything, so we cannot refer to something that is inherently private, thinking that the object itself will bestow a use on the name we give it. Hence we cannot talk about private objects. But Wittgenstein does not leave the matter there. He deals at length with the intuitions that suggest to us that we have sensations that are logically private. Whether his treatment is successful will be the topic of the next chapter.
[ previous | contents | next ] |
[ previous | contents | next ] |
| © July 2001 |