Grice's intentions and uncertainties

by JLS
for the GC

After discussing the 'problem of free will', briefly, Hampshire and Hart summarise their conclusions on 'intend'. It is this section that is most influential in Grice's neo-Prichardian alternative.

Hampshire and Hart write in the subsection "Intention", with which they close their "Mind" essay after discussing the problem of free will:

"Usually a person engaged in doing something knows, and is able to specify, what action he is doing, and often, though not always, a person knows, and is able to specify, some action which he will later try to do. Yet in neither of these cases is the agent's knowledge of his own present or future action normally derived from observation nor is it a conclusion from evidence. Other people are often able to say what action someone is doing or is going to do, when they know this either from observation of his move- ments or as a conclusion from evidence available to them. They make use of criteria or evidence in determining what he is doing or going to do; but the agent himself does not. The segregation of this form of knowledge which the agent has of his own present or future actions, and the identification of corresponding uses of expressions such as "not knowing what one is doing ", " being certain or uncertain ", "being mistaken ", is an essential element in the elucidation of the concept of intention. We shall consider the dependence upon this form of knowledge (practical know- ledge) of certain distinct but related applications of the notion of intention. There are two principal ways in which intention is connected with action. First, when a person has done something, e.g. struck another person, the question of whether he did it intention- ally or unintentionally may arise, and this is equivalent, except in certain trivial respects, to the question whether he intended, or did not intend, to do what he has in fact done. Secondly, the question may arise as to whether a person intends, or does not intend, to do some action in the future; and a similar question may arise about his past intention to do some action, even if he did not in fact do it. These two applications of the notion of intention do not of course exhaust the notion. Besides them there is the special application of intention in cases where it has close connexion with the notions of meaning and reference. What did he intend by those words ? To whom did he intend to refer by that name ? There is also the use of intention in conjunction with the actions of other people: we may intend other people to do certain things or even to have certain experiences. 'I intended you to take that book ' I intended him to suffer'. Although there are these and other applications of the concept of intention, the first two are of special relevance to the present subject. Acting intentionally. We must first consider the type of context required if the question whether a person has done some- thing intentionally or not is to have any point. In any ordinary narrative describing ordinary actions done in normal circum- stances, it would be pointless to say that a person did these things intentionally; for normally there is a flfilled presumption that if a person does something, he does it intentionally. This is a feature of the whole conceptual scheme involved in our descrip- tion of persons in terms of actions. If I am telling you simply what someone did, e.g. took off his hat or sat down, it would normally be redundant, and hence misleading, though not false, to say that he sat down intentionally. The primary point of saying that someone acted intentionally is to rebut a prima facie suggestion that he was in some way ignorant of, or mistaken about, some element involved in this action. Usually the sugges- tion arises from the fact that what he has done is abnormal or wrong in some way or that it is something which ordinary people would not do except unintentionally. Part of the force of 'He did it intentionally' is just to rule out the suggestion that he did it unintentionally, where 'unintentionally' means that the agent did what he did through some accident or by mistake. This suggests that the whole meaning of 'intention- ally' simply lies in its negation of accident or mistake, and that once these two ideas are elucidated, as they easily can be, the analysis of the notion of intentionally doing something is within our grasp. On this view the analysis will simply be (1) a descrip- tion of the appropriate context for the use of the expression ' intentionally ' as rebutting accident or mistake, and (2) the elucidation of the ideas of accident and mistake. But this would only be the first step towards the analysis of the notion of intentionally doing something; for the assertion that someone has done something intentionally, or that he intended to do what he has done, is not merely the equivalent of the assertion that he did not do it unintentionally, or that he did not do it acci- dentally or by mistake. Accident and mistake are certainly incompatible with the agent doing what he did intentionally; but the assertion that someone did what he did intentionally does not merely exclude these cases of unintentional action. This may be seen from the following example. A man fires and shoots another: if asked what he was doing, and if he was prepared to give an honest answer, he would identify his action as shooting at someone. But in a perfectly ordinary sense of 'know ', he would know that shooting at someone involved making the loud noise which in fact the shot had made. In such a case it is clear that he did not make the noise by mistake or by accident; it is therefore clear that he did not make it unintentionally. But ori these facts it would be misleading to say he intentionally made the noise, or that he intended to make the noise; for this would suggest that this is what the agent would say that he was doing, if asked. Always the expression 'He intended to do it ' means more than, though it also includes, 'He did not do it unintentionally '. If an action is to be intentional, or to be what the agent intended to do, two different kinds of requirement must be satisfied. First, the agent must have ordinary empirical know- ledge of certain features of his environment and of the nature and characteristics of certain things affected by his movements. Precisely what knowledge of this sort he must have will depend upon precisely what action is ascribed to him. If, for example, he is said intentionally to have shot at a bird, he must know, in the ordinary sense of' know', that what he has in his hands is a gun, and that there is a bird in the line of fire. He must also have certain types of general knowledge, for example, of the consequences of pulling the trigger of a loaded gun. Without such knowledge as this, his action in shooting the bird would be accidental or done by mistake, and hence would be ranked as unintentional. Second, and more important, if his action is intentional (what he intended to do), the agent must know what he was doing in some sense which would differentiate his shooting at the bird from other non-accidental actions performed at the same time, such as making the cartridge explode. This is the action which he himself would specify, if he were prepared to give an honest reply to the question ' What are you doing ? ' The special kind of knowledge involved in intentional action may emerge from a comparison of the agent's own declaration or description of what he is doing ('I am shooting at X') and state- ments made about his action by others (' He is shooting at X '). The latter statements are generally made on the strength of observa- tion of the agent's movements; and these movements provide logically sufficient grounds for such statements ('He is shooting at X '), though they do not exhaust the meaning of the statement, which is liable to qualification, though not withdrawal, if it later appears that certain abnormalities were present in the situation. In the case of statements made by others, the question 'How do you know that he is shooting at X ? ' is one which could be properly answered by referring to his observed movements. But the agent does not himself use these, or any other criteria, in declaring what he is doing; there is a corres- ponding pointlessness in the question, 'How do you know you are shooting at X ? ' or ' What grounds have you for saying that you were shooting at X ? '. The suggestion that, before answer- ing the question, 'What are you doing ? ', we first look and see how our body is disposed is absurd. The ascription to ourselves of some action (the declaration of what we are currently doing) is not a report of our bodily movements nor a conclusion from the observation of movements. Perhaps the absurdity is most evident in the case of speech action; if someone to whom we are talking asks us what we are saying, we do not have to listen or recall the sound of our own words before answering, and our answer is not a report Qf those sounds nor a conclusion from evidence provided by them. Secondly, a knowledge of the position or movements of our own body would not be sufficient to enable us to answer the question 'What are you doing ? ' In the relevant sense of knowing what we are doing, it is perfectly possible that we should recognize, and be able to describe, such features of the situation as that a knife is in our hands, and yet we may have forgotten what we were engaged in doing. Except in abnormal cases, memory returns and we may then say that we were sharpening a pencil. This illustrates the irrelevance of prior observation by the agent of his bodily movements to his declara- tions of what he is currently engaged in doing. Thirdly, there is a sense in which our own declarations about our current actions may be mistaken. We may say that we are sharpening a pencil when in fact, owing to inadvertence, we were cutting away at a pen. In these cases the natural comment would be, not that the agent's statement was false (though there would also be occasions for that comment), but that he was doing something unintentionally or by mistake. But it is important to notice that, whereas statements made by others concerning a person's action characteristically leave open the question whether he has done what he has done unintentionally, it is a distinguishing feature of the agent's own statement about his actions that an answer of the form ' I am doing this but I am not doing it intentionally ' would be absurd. It would be the virtual equivalent of 'I am doing this but I do not know what I am doing'. The normal ability of an agent to say what he is doing without prior observation suggests two explanations, both of which, though tempting, are mistaken. It may seem that the agent's ability to specify what he is doing in this way is explained by the fact that, prior to acting, he must always have considered and decided what to do; in saying what he is actually doing he is simply recalling the previous decision. The objections to this are, first, that very frequenitly we know what we are doing, and so are doing it intentionally, although this has never been pre- ceded by any stage of prior deliberation or doubt. I may just break off drawing in order to sharpen a pencil without prior deliberation, and yet I can still answer correctly the question 'What are you doing ? ' without first observing my own move- ments. And, even where there has been prior deliberation and decision, the question 'Wlhat are you doing U is answered without recall of this earlier stage. The second explanation, suggested by certain passages in Wittgenstein's Philosophicat Investigations, is that, since the agent in declaring what he is doing makes no use of any criteria, and in particular does not use the criteria upon which other persons rely, the agent's own declaration should be treated, not as a statement, but as a ' signal', to be assimilated to the behaviour-criteria which are the basis of third-person statements about his actions. But this explanation, though it may serve to correct some mistakes (notably that the agent's own state- ments about his action are made on the basis of observation), surely distorts the facts and is open to now familiar objections. The agent's own statements about his actions may be true or false; we contrast an agent's telling us what he is doing with other forms of words which he might use and which might be our evidence that he is doing some action. His statement about his own action stands in recognizable logical relations with state- ments made by other persons. If the agent says, ' I am shooting a bird ', and a third party says, ' He is not shooting a bird ', these statements are contradictories. If the third party says ' He is shooting a bird ' and the agent says, ' I am shooting a bird the latter statement confirms the first. These relation- ships would be impossible if first-person statements were to be treated as signals, and therefore as having a different meaning from third-person statements about action. Only a doctrinaire identification of the meaning of the statement with the means of its verification entails this result. Action is a concept which, like many concepts involving reference to states of con- sciousness (I expect, he expects; I believe, he believes), exhibits this asymmetry between first-person and third-person statements. Intention to do a future act. In those cases where the agent announces his intention to do an action in the future, there is a similar -need to distinguish a belief which the agent may have formed as to the course of his future actions as a result of observation, or as a conclusion from evidence, from a belief which he has formed independently of observation or evidence; for the second is essentially involved in this application of the notion of intention. It is clear that a,person's announcement of his intention to do some action in the future is not a prediction that he will do this action, although others may base their predictions upon such announcements by the agent. That such statements are not predictions is evident from the fact that if the agent does not act as he says that he intends to act, this exposes him, not to the criticism that what he said was false, or that he-was mistaken, but to the charge that he has changed his mind. He may be accused, if he does not do what he says that he intends to do, of having lied about his intentions. But it is possible for him to exculpate himself from this charge by convincing others that he had changed his mind. The obviously mistaken analysis of announcements of inten- tion to do an act in the future as a prediction must be dis- tinguished from an analysis of such an announcement as a statement by the agent of his present belief as to his future action. On this view, ' I intend to do X ' will at least entail, ' I believe that I shall do X '. Certainly this needs qualification: for it is plain that ' I intend to do X ' is compatible with ' I believe I shall do X unless prevented or unless I fail by reason of circumstances outside my control'. The statement, 'I intend to do X ' is also compatible with 'I believe I shall do X unless a change in circumstances leads me to change my mind', though there is at least a suggestion that if the agent thus leaves open the possibility of a change of mind, he does not yet really intend to do the future action. If we take these qualifcations into account, the minimum force of ' I intend to do X ' is ' I believe that I will try to do X'. Hence it would be a contra- diction to say, 'I intend to do X though I do not believe I will even try to do X when an opportunity arises '. But again the salient characteristic of this form of belief is that it is not a con- clusion from evidence, and that it is neither proper to ask, nor necessary to answer, the question: 'Why do you believe that you will try to do X ?' The contrast with ordinary professions of belief is plain. The contrast between first and third-person statements is instructive. An observer who says ' He intends to do X ' makes his statement on the basis of observation or evidence, and he could be asked to support his statement by evidence, which may range from remote circumstantial evidence to a report of X's statement of his own future intentions. On the other hand, the third person who says that X intends to do a future action is committed to the statement that X believes that he will at least try to do X if the occasion arises. He has evidence that X does s6 believe, but it would be absurd to suggest that X himself has evidence that he does so believe or has evidence in favour of the truth of his belief. In many cases the agent will have formed the intention to do a future action as a result of considering alternatives and deciding between them, and perhaps the characteristics of intention to do a future act best emerge from a study of such cases. But not all intention is formed as a result of prior decision. I intend to go home after writing this essay, but I have never decided to do this. If we consider those intentions which emerge as a result of deliberation and decision, we can trace certain parallels with the formation of theoretical certainty about the future, as well as the major contrast we have noticed with regard to the inde- pendence of evidence. In deliberation we consider whether to do, or not to do, something, and we oscillate between these alternatives: we attend to reasons for or against the proposed action, and we attribute more or less weight to these reasons: we then decide what to do. In the theoretical case we consider whether something is or is not be the case, and we attend to the evidence in favour of one or other alternative: we find the evidence in favour of one alternative convincing and then decide that it is or is not be the case. In both cases we could sub- stitute for the expression 'decision' expressions such as 'being certain 'or' making up our mindc'. But neither the deliberative process concerning future action nor its theoretical counterpart need issue in decision. We may remain undecided, on the one hand whether to do something or not, and on the other hand whether something is the case or not. We may be unable to decide, and may give up both the practical or theoretical problem as too difficult, leaving others to tell us what to do or what is to be the case. The characteristic termination of the practical inquiry is the settled frame of mind when we are no longer undecided what to do. We have made up our mind and are both certain what to do and certain what we will try to do. In describing this ter- mination of deliberation, we cannot separate the temporal reference to the future from the solution of the practical question. We have decided what to do, and that we shall at least try to do it. We canot have this form of confident belief about our future voluntary action without this form of practical certainty about what to do."