The point that emerges is that awareness is not the home or origin of intentions or volitions. In fact we have only limited and fallible access to the mechanisms that direct our behaviour. Nothing that goes on ‘in awareness’ can be construed as an act of will or volition, and nothing that is subconscious would fit the ordinary connotations of these words. Once again, getting rid of the little man in the brain, this time in the guise of ‘conscious agent and source of volitions’, also involves getting rid of the tools of his trade.
XXIII THE IMPORTANCE OF INTENTIONAL ACTIONS
The class of intentional actions has now been characterized as the class of motions under particular descriptions of which the actor has practical knowledge and for which he is prepared to offer reasons. It is not at all clear why these conditions should make intentional actions so special. The concept of intentional action is critical in our conceptual scheme, for our bestowing and withholding of praise and blame is generally tied to the decisions we reach regarding intentions. Quite literally a lot hangs on our ascriptions of intentions, and we cannot answer the question of their importance by pointing out that a lot hangs on them; it is just this that is in need of explanation and eventually moral justification. Unless there is some important, efficacious difference between those motions that are intentional actions and those that are not, the distinction is pernicious.
A brief look at the range of bodily motions and actions shows that there is no clear-cut line between the intentional and unintentional or the voluntary and involuntary from the point of view of causal determination. The circuitry that causes these motions varies only in complexity and degree of mediation. The only ‘indeterminacy’ that can be held out for the causal sequences governing intentional or voluntary action is due to our lack of knowledge of the nervous system, not to any random effects.9 Moreover it is now widely recognized that causal indeterminacy is not the ‘freedom’ we should look for to account for ‘freedom of the will’. Nor does mere complexity of causal antecedents promise to be the important distinguishing characteristic. As we saw in Chapter 8, even in cases of casual acts where there is no conscious reasoning and no awareness, such as picking up and biting into an apple, the complexity of the causal antecedents, characterized either extensionally or Intentionally, will be roughly equal to their complexity in cases of intentional action. These are considerations of the wrong sort; assigning responsibility for actions cannot hinge on the complexity or determinacy of causal sequences, for when one speaks of responsibility, one is already firmly in the personal realm of our conceptual scheme, where such mechanical questions cannot even significantly be raised, let alone be relevant.
I do not intend to present a ‘solution’ to the problems of responsibility and free will here, but certainly a first step in any such solution must be finding the crucial difference between intentional and unintentional actions. It has been a recurring theme in this book that awareness and the control of behaviour are only circumstantially related, and yet the distinction that has so far been drawn between intentional and unintentional actions is one of awareness: one is aware1 of the efferent commands in cases of intentional action. The awareness line is no centre of personal control, so it can hardly be that the importance of intentions has to do with a person’s control over his own motions. It could, on the other hand, be that the importance of intentions has to do with the control or influence another person can have over a person’s actions. The concept of intentional action hinges on the effect on people of verbal stimulation. Verbal stimulation – talking to someone – contributes to the control of behaviour in much the same way non-verbal stimulation does. Efferent signals have been likened to orders or commands, but a verbal order, telling someone to do something, does not have the same function as such an efferent signal. It is a bit of information the contribution of which depends on the pre-existing neural organizations and states; the order may be obeyed or disobeyed. In extremely docile or dependent people, or in the face of over-powering authority or when one is caught off guard, verbal stimulation may in fact contribute so strongly to the determination of behaviour that it is like pushing a button. In such a case it would be tempting to say the verbal order causes the action, just as the efferent order causes the action, but there is a difference. The verbal order’s content is determined by its linguistic parts, and it may or may not have the effect it is ‘designed for’, depending on conditions in the recipient. The effects of suggestions, requests, reports of information and criticisms are similarly dependent on these conditions.
In order for verbal stimulation, as for non-verbal stimulation, to contribute to behaviour control, the behaviour in question must be amenable to influence and the stimulation must be relevant. Verbal stimulation must strike at the actual controls of behaviour if it is to have its effect, and in some sorts of behaviour, verbal stimulation is simply not effective. There is little or nothing one can say to stop a person from crying, and nothing one says will stop a person from shivering. When actions are intentional, on the other hand, the actor can report – fallibly, but normally reliably – some part of the controls, and these reports allow others to aim verbal stimulation with some assurance of accuracy and efficacy. Suppose, for example, the lady next door is yelling at the top of her lungs and I wish to change her behaviour. Setting aside the sub-human alternative of physically muffling her, my first move is to find out why she is yelling. Such verbal stimulation as suggesting she install a telephone will miss the mark if the description under which her noisy behaviour is an intentional action is ‘rehearsing “Vissi d’arte” ’. Having learned the description of her intentional action, a number of alternatives are open to me, depending on my tact and subtlety, and if I learn that an anvil has just fallen on her foot, I will abandon verbal stimulation and set about finding medical aid.
Anscombe says, ‘Roughly speaking, it establishes something as a reason if one argues against it.’10 This is the basis of importance of intentional actions: they are actions one can argue against. We exculpate the insane on the grounds that rationally directed verbal stimulation fails to have its proper effect: ‘It’s no use talking to him – he’s mad’, ‘He won’t listen to reason’, ‘Arguing will get you nowhere’. To argue with an entity is to treat that entity as a person, a rational agent. Thus personal responsibility – and only people are responsible – is founded on the general assessment of the limits of the contributions one can make to another’s behavioural control by means of rational discourse. Anscombe argues that ‘the concept of voluntary or intentional action would not exist, if the question “Why?”, with answers that give reasons for acting, did not.’11 The ordinary concept of intention, which is dependent on the concepts of awareness and rational control, and on which depends the concept of responsibility, is important because for the average man the best hope he has of contributing to the control of another’s behaviour lies in aiming verbal stimulation accurately at those controls that can be altered. The ascendancy of this method has until recently been supported by results, and is entrenched in our conceptual scheme in such concepts as rational agent, conscious act, and ultimately, person. A shift to other methods, including non-rational verbal stimulation (hypnotism, psychoanalytic therapy, brainwashing) and possibly chemical and electrical methods, could result in a shift in this part of our conceptual scheme, and of course this eventuality has already been dimly discerned by the users of such neologisms as ‘depersonalization’ and ‘dehumanization’ and in the prophecies of a disappearance of the concept of responsibility under the new wave of ‘causal’ explanations of human behaviour in the social sciences. The distinction between manipulation and persuasion is thus fundamental to our conceptual scheme, since on it rests ultimately the concept of a person. This is the ‘different light’ in which we view motions when we view them Intentionally as actions.
10
LANGUAGE AND UNDERSTANDING
XXIV KNOWING AND UNDERSTANDING
Viewing a person as an Intentional system i
s viewing a person as working with information, knowing facts, believing statements. In Chapter 4 a number of obstacles were placed in the way of the centralist programme of associating verbal formulae (reports, statements of fact, commands, etc.) with events and states of such a system, and it might seem to be a corollary of this that the centralist, working his way up from the sub-personal, physical account to his merely approximate ascriptions of content, must necessarily fail to achieve the precision with which we speak at the personal, purely Intentional level of people’s beliefs and knowledge. The precision we find in our ordinary talk of beliefs and knowledge is an illusion, however, for the obstacles that face the centralist have their counterparts on the purely personal, Intentional level.
We talk about what a person knows as if we could make a list of the things he knows, or at least specify quite precisely a few of the things he knows, but such specifications as we can make are always open-ended and depend on an indefinite number of assumptions. To bring out this systematic impossibility of precisely determining things known we must first set aside a difficulty that infects our ordinary concept of knowledge. The ordinary use of ‘know’ carries with it the claim that what is known is true. If I claim to know that p, and p turns out to be false, my claim to knowledge is disallowed. It will be said that I only believed that p, but did not know that p. Yet at the same time we suppose that what is known by a person can occupy a special psychological position, so that a person can tell the things he knows from the things he merely believes, can follow such maxims as ‘Don’t commit yourself until you know for sure’, can tell us what he knows about a particular subject. These two notions about knowledge, the truth condition and the ability of the person to tell knowledge from belief, are incompatible. No one could ‘intuit’ or ‘introspect’ a difference between those things merely believed and those things believed and actually true. There can be degrees of belief; a person can order a group of statements according to how willing he would be to stake something on them – money, or his life or reputation – and he could even decide to draw a line somewhere dividing what he claims to be knowledge from what he claims to be mere belief, but such a line must be arbitrary. When called upon to produce one’s knowledge one can do no better than to produce what one believes to be true, and whether or not what one believes to be true is true does not affect its being one of those things one will produce as knowledge when asked, or will otherwise act on as if one knew them. If we suppose for the moment that it is safe to think in this way of things (perhaps facts or propositions) one will act on as if one knew them, then it is clear that the class of these things for a person need not (and probably never does) coincide with the class of things known by the person (which, however else it is characterized, will include the truth condition). A thing (a fact or proposition or whatever) could not occupy a special psychological position (e.g., have a special functional potential in the direction of behaviour) in virtue of its truth, so knowing something cannot be purely a matter of being in a particular psychological state. When someone claims to know something and is proved wrong, it would be absurd for him to suppose that he had misidentified the state he was in, had mistaken the marks of belief for the marks of knowledge. It is easy to confuse these two ‘classes of things’, the psychologically characterized class of things that are as if known by a person with the class of things actually known. For example, characterizations of knowledge often specify a condition of ‘adequate evidence’ or ‘justification’. What is known must pass certain tests, but which tests? If what is known is to be distinguished from what is merely believed (even believed on good evidence or with good reason) then these tests that must be passed must be foolproof tests – but of course there are no such tests. What is as if known, on the other hand, must merely pass whatever tests a person sets, whether these are good or bad, wise or foolish; it is what has arrived at a certain functional position regardless of the rigours of its journey. Stipulating conditions of adequate evidence or reason is a mistake, for only ‘perfect’ evidence would ensure knowledge; such stipulations should be construed as normative: one should test one’s beliefs rigorously if one wants to achieve knowledge.
Determining precisely what a person actually knows would involve two tasks: first determining what a person ‘knows’ (what he will offer or exhibit as knowledge), and then determining which of this is true. The second task does not concern us here, for the issue is whether there is any way of describing or expressing or determining precisely what a person ‘knows’. We are inclined to think of a person as having a store of information and misinformation, and the question is whether we can specify the contents of this store with any precision.
Storage of information (including misinformation) does not by itself constitute ‘knowing’ (I shall drop the scare-quotes from this non-ordinary term in what follows); dictionaries and encyclopaedias and libraries are stores of information, but they do not know the information stored. Knowing requires understanding, and here we must be careful to distinguish different sorts of understanding. One can understand each word in a sentence without understanding the sentence (‘Sufficient unto the day is the evil thereof’ is one that puzzled me as a child even though I understood each word); one can understand a sentence without understanding what a person is saying or stating by uttering the sentence; one can understand a person’s statement without understanding the person; one can understand a subject, a state of affairs, a problem.
If a person utters a sentence and we take this as an indication that the person knows (or believes) what the sentence states, we assume that the person understands the statement he has made in uttering the sentence. What is involved in understanding a statement? It is not enough that the person be able to produce paraphrases of his sentence. A computer programmed to translate English into Russian might be capable of producing passable Russian paraphrases of English sentences, but this would not suffice to show that the computer understood statements. It might even produce English paraphrases of English sentences. Suppose we feed in ‘I just murdered my uncle’. Even if it responded with ‘You have recently slain the brother of one of your parents’ it would not be said to have understood the statement. If, however, the computer immediately made a discreet telephone call to police headquarters, one would be tempted to say it had understood the statement, but only if it also had the capacity to do other quite different things with different input. If it is merely the local ADIAC computer (Apparatus of Dubious Intelligence for Acknowledging Confessions) no one will grant it understanding. Only a being that is non-verbally active in the world could meet our requirements for understanding. A computer whose only input and output was verbal would always be blind to the meaning of what was written. It might ‘grasp’ all the verbal connections, but it would lack ‘acquaintance’ with the things the words are about. Suppose we fed a computer a description of the Taj Mahal. It might paraphrase this or even respond with an output like ‘The Taj Mahal must be very beautiful’, but one wants the computer also to produce outputs like ‘Take me there; I want to see for myself’, and such outputs would be a hoax if the computer did not have some perceptual apparatus and many other sophisticated capacities.
The tests for understanding in each particular case involve a ‘family’ of behavioural capacities, some of which must be demonstrated. If Jones says ‘Smith is here’ no one will allow that Jones understands this, and hence no one will allow that he knows or believes this unless Jones can also say and do a variety of other things ‘with his knowledge’. He must be able to assert, for instance, ‘Smith is not in Siam’, ‘Smith is that friend of Black’s’, or ‘By “here” I mean “in town”, not “in this room” ’. If Jones knows Smith is here he must be able to point him out or at least direct the search party. If no such corroborating behaviour is in the offing, Jones may be in this instance no more than a parrot or tape recorder. The particular tests that must be passed in any one case are not entirely determined by the information in question (the candidate for
what is known). Rather, what tests must be passed depend largely on what else the person knows and understands, and whereas a great deal of the corroborating behaviour can be verbal – explaining, asserting related statements, paraphrasing, and expanding on the subject – if there are available non-verbal tests and they are failed, the verbal testimony will be shaken. Much information, of course, is so intimately verbal in being about verbal states of affairs that no strictly non-verbal behaviour could tend to corroborate the claim to knowledge: e.g., the information that yesterday was (called) Friday, this place is (called) New York.
What are the conditions that would suffice to show that a child understood his own statement: ‘Daddy is a doctor’? Must the child be able to produce paraphrases, or expand on the subject by saying his father cures sick people? Or is it enough if the child knows that Daddy’s being a doctor precludes his being a butcher, a baker, a candlestick maker? Does the child know what a doctor is if he lacks the concept of a fake doctor, a quack, an unlicensed practitioner? Surely the child’s understanding of what it is to be a doctor (as well as what it is to be a father, etc.) will grow through the years, and hence his understanding of the sentence ‘Daddy is a doctor’ will grow. Can we specify what the child knows when he tells us his Daddy is a doctor? It may seem simple: what the child knows is that his Daddy is a doctor – that is, the object or content of his knowledge in this case is the proposition, ‘that Daddy is a doctor’. But does the child really know this? One is inclined to say that he only ‘sort of’ knows this, or ‘half’ knows this. If the proposition is to be the thing known, we have to allow for quasi-knowledge of propositions. Yet one might argue that when the child only half knows the proposition there is still something – something somehow ‘less’ – that he fully or really knows.
Content and Consciousness Page 22
