Abstract
Wittgenstein suggests that we can understand ascriptions of thoughts that we have no means of verifying: thoughts that not only are not but could not be manifested in behaviour. For example, we can understand claims about what Queen Victoria was thinking about as she lay dying. But how do we understand such claims? The paper explores Wittgenstein’s answer to that question. Three possible accounts are examined. It is argued that there are traces of each account in Wittgenstein; that Wittgenstein himself did not clearly favour one of these accounts over the others; but that one of these accounts is philosophically preferable to the others.
Table of contents
In a number of passages, Wittgenstein suggests that we can make perfectly good sense of ascriptions of thoughts that we have no means of verifying: thoughts that not only are not but could not be manifested in behaviour. For example:
Lytton Strachey writes that as Queen Victoria lay dying she ‘may have thought of’, say, her mother’s youth, her own youth, Prince Albert in a Grenadier’s uniform (LPP 274. See also LPP 32-3, 99, 152, 229; RPP i 366).
We clearly understand Strachey’s speculation. But it seems perfectly possible not only that Queen Victoria did not report her dying thoughts but that she could not have done so. And in that case, we cannot make sense of claims about her dying thoughts in terms of what she was disposed to report thinking; she had no such disposition. So how do we understand what Strachey says?
One idea would be to appeal to counterfactuals: if Queen Victoria had been able and willing to report what she was thinking, shewould have reported thinking such-and-such. But Wittgenstein takes a different line. We learn ‘She thought X’, he thinks, in cases where people say what they thought, and where the question what they thought has some practical importance. But with our understanding secure in those basic cases, we can go on to apply the same words to cases where there is no possibility of verification, and where no practical consequences attach to someone’s having thought one thing or another. Thus:
We understand ‘He thought X but would not admit it’, but we get the use of ‘He thought X’ from ‘He admits X’, i.e. says X, writes in his diary X, acts in an X-like way . . . Thinking and not admitting comes in only after thinking and admitting. It’s an exception-concept. You’d have to explain to someone who did not know what ‘thinking and not admitting’ was in terms of thinking and admitting (LPP 329).
In Wittgenstein’s view, then, a central role is played, in determining the content of the concept of thought, by cases in which someone’s thoughts are manifest in their words or actions. That is a particular case of a more general principle: that a central role is played in determining the content of a concept by cases in which the concept is manifestly instantiated. That principle does not apply to every concept. The content of a highly theoretical concept, for instance, is determined by the theory in which it appears, not by cases where it is manifestly instantiated. Similarly for concepts that can be analyzed in terms of descriptive conditions. But it is very plausible that there are some cases where the principle does apply.
Colour concepts are an obvious example. Cases where something is manifestly red, where it is observed to be red, have a crucial role in determining the content of the concept red. But the concept red also applies to things that are not observed to be red, and to things that in some reasonably strong sense could not be observed to be red: things that can only exist in conditions where human life is impossible, and so on. How should we understand the application of the concept in those cases? One idea is to appeal to counterfactuals: for an unobserved object to be red is for it to be true that, if it were observed by a suitable observer in suitable conditions, it would look red. That proposal might work in explaining how we understand applications of the concept red to objects that merely are not observed. But it is hard to see how it could work for the case of an object that could not be observed to be red. Yet we do seem able to make sense of the thought that such an object is red. So we need a different idea. An obvious proposal is this: cases in which objects are manifestly red play an essential role in determining the content of the concept red. What it is for an unobserved object to be red is then explained by relation to what it is for an observed object to be red: an unobserved object is red just in case it is the same colour as an object that is observed to be red.1
Now Wittgenstein might complain that such a view would be question-begging. If we are trying to explain what it is for an unobserved object to be red, we cannot simply help ourselves to the idea of the object’s being the same colour as an observed red object. For (adapting what he says about a different case): I know well enough that one can call an observed red thing and an unobserved red thing ‘the same colour’, but what I do not know is in what cases one is to speak of an observed and an unobserved thing being the same colour’ (cf. PI §350). But how far would Wittgenstein push this objection? He would certainly insist that what it takes for one thing to be the same colour as another cannot just be taken for granted: it must be understood by reference to a humanly-created concept of colour; and the existence of the concept depends on a whole practice of sorting and classifying things according to their colours, of agreeing and disagreeing about which things are the same colours, and so on. But once that point is accepted, does Wittgenstein think there is a further problem about extending the concept red from things that are observed to things that are not, and could not be, observed? It seems plausible that, for the case of objects that are unobserved but could be observed, he would accept the dispositional view mentioned in the previous paragraph: what it is for an unobserved table to be brown is for it to be disposed to appear brown to the normal sighted under certain circumstances (see RC §97). But how would he understand the application of the concept red to things that could not be observed? I know no passage where Wittgenstein explicitly considers that question.2 Perhaps he would regard such an application as unintelligible. But if that is his view, it needs further argument. For, on the face of it, there is no obvious reason why the concept of colour that we develop in connection with practices involving observed things should not be straightforwardly applicable to things whose colours we could not observe.
What about the concept of thinking? Two points about Wittgenstein’s view seem clear. First that, as I have said, a central role is played in determining the content of the concept by cases in which what someone is thinking is manifest because she says or otherwise manifests what she is thinking. Second, that our grasp of what it is for someone to think so-and-so in a case where her thoughts cannot be manifested is dependent on our grasp of what it is for someone to think so-and-so in a case where her thoughts are manifested. But exactly what is the relation between the content of the concept in the two kinds of case? We can distinguish three quite different models, each of which is consistent with the two points just made.
On the first model, the relation between the case where someone says what she is thinking and the Queen Victoria case is like the relation between the cases of observed colour and unobservable colour suggested above. The concept of thinking cannot be explained without making use of examples of thinking; we acquire the concept of thought, in part, in connection with cases where we can tell what someone is thinking. But, having explained the concept of thinking as it applies in cases where we can tell what someone is thinking, we can apply the same concept without further explanation to cases where people’s thoughts are not and could not be manifested. At one point, Wittgenstein presses the question, ‘what we can do with’ a sentence about Queen Victoria’s dying thoughts – ‘how we use it’ (RPP i 366). On the current model, that question has a straightforward answer. We use the sentence ‘Queen Victoria saw so-and-so before her mind’s eye’ to speculate about Queen Victoria’s dying thoughts. We engage in such speculation because we are interested in what she was thinking about immediately before her death. And we are interested in that question for its own sake – not because we think it has any practical implications.
Maybe Wittgenstein would accept that answer. But some of what he says suggests a quite different model. On this second model, the content of the concept of thought as applied in the Queen Victoria case cannot simply be read off the content of the concept in the more basic cases; it must be understood by giving a direct account of the nature and point of the practice of describing and speculating about thoughts whose ascription cannot possibly be verified. We find it natural to take the word ‘thought’ from the basic cases, where we can tell what someone is thinking, and apply it in Queen Victoria cases. The meaning of the word in these new applications is parasitic on its meaning in the basic cases, but it is not fully determined by that use; it depends also on the actual use of the word in the new applications. And that use is a matter of our shared interest in developing narratives about the inner lives of others: narratives that have no practical purpose, and for which there is no standard of correctness other than what people agree in regarding as plausible or appropriate. On this view, the practice of discussing Queen Victoria’s dying thoughts comes closer to the practice of discussing fiction than to that of ascribing thoughts in more basic cases.
A third model is suggested by the following passage:
What is the purpose of a sentence saying: perhaps N had the experience E but never gave any sign of it? Well, it is at any rate possible to think of an application for the sentence. Suppose, for example, that a trace of the experience were to be found in the brain, and then we say it has turned out that before his death he had thought or seen such and such etc. Such an application might be held to be artificial or far-fetched; but it is important that it is possible (RPP i 157).
On this view, the sentence ‘perhaps N had the experience E but never gave any sign of it’ has an application, a meaning, because there is in principle some way of verifying whether or not N did have the experience E. If we apply this line to the Queen Victoria case, we will say that we understand the ascription of thoughts in such a case by supposing that there is, after all, a method of verifying such ascriptions, albeit a method that looks not to the subject’s actual or potential words and actions, but to physical traces of her thoughts.
If Wittgenstein accepts the first model of our understanding of the ascription of thoughts in the Queen Victoria case, his treatment will be decisively non-verificationist. If he accepts the second model, his account of the meanings of such ascriptions will, again, avoid verificationism; but it will nonetheless be a form of anti-realism. For it will explain the meanings of such ascriptions in a way that gives up the idea that there is any independent fact of the matter about what Queen Victoria was thinking in her dying moments. If he accepts the third model, his account of the Queen Victoria case will, after all, be a form of verificationism. For on this view, the meaningfulness of ascriptions of thought in the Queen Victoria case depends on the supposition that those ascriptions are not, after all, inaccessible to every form of verification.
Which of the three models would Wittgenstein accept? I think his position is unclear. The first model is consistent with much that he wants to say. But there is some evidence that he would reject that model; that he would insist that an account of the meaning of the word ‘think’ as applied in Queen Victoria cases must say something more substantive about our practice of using the word in such cases. The very fact that he presses the question, what we do with the sentence ‘Queen Victoria may have thought . . .’ suggests that, even when we have explained the meaning of ascriptions of thought in cases where a subject’s thoughts are manifested, there is a further question, how we understand ascriptions of thoughts that lie beyond our normal methods of verification. That, in turn, suggests that when we apply the concept of thought in Queen Victoria cases, we are in some way developing or extending the concept, or using it in a secondary sense. A view of that sort seems right for the application of the adjectives ‘fat’ and ‘lean’ to days of the week. Perhaps it is right for the application of the concept calculating to cases in which there is no overt process of calculation. But it is hard to believe that it is right for the application of the concept thinking to Queen Victoria cases. If Wittgenstein was tempted by such a view, it is a temptation he should have resisted.
Literature
- Peacocke, Christopher 2008 Truly Understood, Oxford: Oxford University Press.
- Wittgenstein, L. LPP Wittgenstein’s Lectures on Philosophical Psychology 1946-47, London: Harvester, 1988.
- Wittgenstein, L. PI Philosophical Investigations, 2nd edition, Oxford: Blackwell, 1958.
- Wittgenstein, L. RC Remarks on Colour, Oxford: Blackwell, 1977.
- Wittgenstein, L. RPP i Remarks on the Philosophy of Psychology vol i, Oxford: Blackwell, 1980.
Refbacks
- There are currently no refbacks.