1I am indebted to J. O. Urmson (1952) for some of these distinctions. Even in this very simple sort of explanation in terms of a manâs reason for doing something there are, as a matter of fact, concealed assumptions. We assume, for instance, that walking across the street is an efficient way of getting to the tobacconist. This counts as an explanation not simply because Jones envisaged walking across the street as a means to getting the tobacco but because it really is a means to getting it. We assume, too, that a man who has this information will act on it if he wants some tobacco. We assume that men are rational in that they will take means which lead to ends if they have the information and want the ends. âHis reasonâ is an explanation in terms of what Popper (1945) calls âthe logic of the situationâ.
But it is not only norms of efficiency and consistency that are implicit in the concept of âhis reasonâ. There are also norms or standards of social appropriateness. After all Jones might have crawled or run across the road. But âto get some tobaccoâ would be a very odd answer to the question âWhy did Jones run across the road?â Yet running would be quite an efficient way of getting across the road. It would, however, be socially odd as a way of crossing the road to get some tobacco. Man is a rule-following animal. His actions are not simply directed towards ends; they also conform to social standards and conventions, and unlike a calculating machine he acts because of his knowledge of rules and objectives. For instance, we ascribe to people traits of character like honesty, punctuality, considerateness and meanness. Such terms do not, like ambition or hunger or sexual desire, indicate the sorts of goals that a man tends to pursue; rather they indicate the type of regulation that he imposes on his conduct whatever his goals may be. A man who is ruthless, selfish, punctual, considerate, persistent, and honest, does not have any particular goals; rather he pursues whatever goals he has in particular sorts of ways.
This simple purposive model of a man taking means to bring about an end is further complicated by the fact that norms enter into and often entirely define the end. Ends like passing an examination, getting married, becoming a professor, and reading a paper, explain quite adequately a great deal of the goings on in the precincts of a university; yet they are defined almost entirely by social convention. It is a gross over-simplification to think of ends merely as terminating points of activity. Actually even a rat, after eating or achieving some other end, will continue being active in a variety of waysâsniffing, preening, and so on. If eating can be regarded as an end this is not because it is a definite terminating point of activity but because activity previous to it varies concomitantly with changes in the conditions necessary to define it as an end. The concept of means is just as necessary to bring out what is meant by an end as the concept of end is to bring out what is meant by a means. Ends are not given as natural terminating points like a chain of oases distributed across a desert. And, to a large extent, what counts as falling within a means-to-end explanatory framework is determined by convention. Even those ends, like eating and sexual intercourse, which are universal and which have an obvious biological basis, can scarcely be specified without recourse to norms. For there are countless ways of performing the acts which can be regarded as ends and in every culture a few particular ways are stamped with the hallmark of conventionality. Eating is not just getting food into the stomach. Jonesâ movements across the road are classifiable as means to the end of buying tobacco because of a vast system of norms defining âbuying tobaccoâ as an end as well as a system of norms regulating what is an efficient and socially appropriate way of attaining it.
My reasons for stressing this rule-following purposive pattern into which we fit our common-sense explanations are twofold. In the first place I want to insist that most of our explanations are couched in terms of this model and our predictions of peopleâs behaviour presuppose it. We know what the parson will do when he begins to walk towards the pulpit because we know the conventions regulating church services. And we can make such predictions without knowing anything about the causes of peopleâs behaviour unless we include under âcausesâ things like the parsonâs training and grasp of the rules, which are things of a different order from âcausesâ in the sense of antecedent movements. Man in society is like a chess-player writ large. Requests for explanation are usually reflections of our ignorance about the particular rule or goal which is relevant to the behaviour in question. We usually know the general pattern but are unsure which part of it is relevant. Sometimes, of course, we are in the position of a free-thinker at a Roman Catholic mass. The question âWhy did X do that?â is then usually a request for an elucidation of the whole pattern of conventions. In explaining human actions we, like anthropologists, must all, in the first place be structuralists. Indeed I would go so far as to say that anthropology or sociology must be the basic sciences of human action in that they exhibit the systematic framework of norms and goals which are necessary to classify actions as being of a certain sort. They bothâlike classical economicsâpresuppose the purposive, rule-following model; in this respect they are quite unlike sciences which imply a mechanical model of explanation.
In the second place this rule-following purposive pattern of explanation must be sketched in some detail because a proper understanding of what is meant by a human action has very important logical consequences. It shows, for instance, as I shall argue, that human actions cannot be sufficiently explained in terms of causal concepts like âcolourless movementsâ. Indeed to claim that we are confronted with an action is ipso facto to rule out such mechanical explanations, as being sufficient.
(b) âthe reasonâ explanations. But, of course, as psychologists will be the first to point out, people often invent reasons for doing things or delude themselves into thinking that the reasons they offer for their actions are operative reasons. We therefore often say of a man that his reason may have been x but the reason why he acted like that was y. For instance we might say that Jones said that he crossed the road in order to buy some tobacco but the reason why he did it was not really his desire for tobacco; it was sex. There was a pretty girl looking in the window of the tobacconist. This explanation may of course be erroneous. For instance a psychologist once told me that I delayed crossing the road to College because of an aversion to getting down to work. I replied, and I think more convincingly, that I stayed on the other side in order to look at the row of glistening cars drawn up opposite. But whether the explanation in question is correct or incorrect does not much matter; the point is that to speak of the reason why a person does something is different in that it is a way of calling attention to the law or assumed law that a given case actually falls under. His reason may coincide with the reason. The reason why Jones crossed the road might in fact be his desire for tobacco. He might also be aware that he wanted to inspect the girl at close quarters, but was concealing this by the camouflage of buying tobacco. This would then be his real reason. But whereas his reasonâwhether real or notâentails that a man is conscious of his objective, the reason why he did it does not.1 The reason why he did it might well be sex or aversion to work; yet the individual might be quite unaware of pursuing or avoiding the relevant goals. And whereas to say that he had a reason for doing something is more or less to rule out a causal explanation, to give the reason why he did it is sometimes to subsume it under a law-like proposition of a causal kind. This is not necessarily so. For we can say that sex or aversion to work was the reason why he did it and simply be insisting that a different directive disposition is being exercised. But the reason why he did it might also be that he was pushed or assailed by an attack of giddiness. These would be causal explanations which would rebut the suggestion that he had a reason for crossing the road. Causal explanations, in other words, can count as the reason why a person does something; but they are only one type of answer to the question âWhat was the reason why he did it?â
1 Hamlyn has pointed out to me the use of âthe reason for his actionâ as well as âthe reason why he did itâ. âThe reason forâ seems to be similar to âhis reasonâ but to imply a coincidence between âhis reasonâ and âthe reason why he did itâ. I am not here concerned with the use of âreasonâ in the context of justification as when we say that a reason for giving up smoking is that it causes lung cancer. âHis reasonâ and âthe reason forâ can be used in contexts both of justification and of explanation. Needless to say âthe reason why he did itâ is reserved for contexts of explanation with which I am here concerned. (c) causal explanations. There are, however, other questions about particular goings onâI omit to say actions on purposeâto which answers in causal terms are appropriate. Instead of the omnibus question âWhy did Jones do that?â we often ask what made, drove, or possessed him to do it. These are usually cases of lapses from action or failure to actâwhen there is some kind of deviation from the purposive rule-following model, when people, as it were, get it wrong. This may be in respect of an efficiency normâfor example, when a person refuses to take the only quick route to his destination by underground train, or when he canât remember a well-known name when he is performing an introduction. Or the behaviour may go wrong in respect of a norm governing social appropriatenessâas with a business man who runs to work when he is not late or a tutor who crawls round the room sniffing while listening to an undergraduate essay. Or behaviour may go wrong by being deflected towards a peculiar goal as with a married man who suddenly makes an advance to a choir boy. In such cases it is as if the man suffers something rather than does something. It is because things seem to be happening to him that it is appropriate to ask what made, drove, or possessed him to do that. The appropriate answer in such cases may be in terms of a causal theory
These cases of particular goings on which look like breakdowns of action are very similar to a whole class of general activities which seem to have no point or a very odd pointâdreams, hallucinations, obsessions, anxieties and perversions. In such cases the Greeks suggested that the gods intervene and take possession of the individualâs mind. Very often recourse is made to crude physiological explanations. It was not till the advent of Freud that any systematic explanation of such goings-on was offered in psychological terms. Indeed Freud claimed in 1913 that the main contribution of psycho-analysis to general psychology was to link together and to give psychological explanations for happenings which had previously been left to physiology or to folk-lore. Many have claimed that Freud, by reclaiming these phenomena for psychology, was in fact extending the model of purposive rule-following behaviour to cover the unconscious. He showed, it is argued, that we have reasons for acts which were previously only explained in terms of causes. I shall argue later that this thesis is mistaken. Freud showed, perhaps, that the concept of âwishâ has a wider application than was previously thought. But his account of the working of the primary processes creaks with causality. In maintaining that in the unconscious there is no sense of causal or logical connexion he was ipso facto denying that the model of âhis reasonâ, implying norms of efficiency and social appropriateness, was relevant. Freud, I shall argue, provides the classic case of giving quasi-causal explanations where causal explanations seem prima facie appropriate.
I shall also argue that Freud in fact only intended to explain by reference to unconscious mental processes cases where the purposive rule-following model breaks down or is inappropriate. He did not thinkâand often explicitly deniedâthat this sort of explanation can be appropriately given for everythingâfor cases where a man acts as well as for cases where something happens to a man. In this respect Freud was, from the point of view of my argument, on the side of the angels. For my case is not simply that causal explanations are otiose when we know the point of a personâs action in that, life being short and time limited, we no longer feel inclined to ask âwhyâ once we have accommodated a piece of behaviour within the rule-following purposive model. It is also that if we are in fact confronted with a case of a genuine action (i.e. an act of doing something as opposed to suffering something), then causal explanations are ipso facto inappropriate as sufficient explanations. Indeed they may rule out rule-following purposive explanations. To ask what made Jones do something is at least to suggest that he had no good reason for doing it. Similarly to asc...
