Content and Consciousness

by Daniel C. Dennett

  Suppose a man with a hammer decides to finish driving a nail into a door. Suppose that he happens to see the nail sticking out, and this visual input has the effect, after running through various cerebral organizations, of giving rise to an efferent state or event with the content ‘drive in the nail’. (Note that no appeal is made to anything like ‘conscious decision’.) Then suppose this efferent state or event sends a signal across the awareness line; the man will then be aware1 of what he is about. He may say to himself or out loud ‘about time I finished driving in that nail’ – or he may only be aware1 of what he is about and never go through the longer temporal process of formulating an utterance aloud or to himself. To vary the case, suppose the man is prompted to make a dent in the wood next to the nail – for some arcane reason, or just to be silly. A different efferent state would direct this activity and the man would be aware1 of this different action were the efferent state to send a signal across the awareness line. Then suppose in the former case the man accidentally misses the nail and makes a dent; the muscular and skeletal motions might be quite indistinguishable in the two cases, but they would be controlled by different efferent states, and the man would be able to tell the difference – not because he recognized a qualitative difference in efferent states (he would have no inkling of what his efferent states were) but because these different efferent states produced in him different dispositions to express certain things, produced in him different contents of awareness.

  How do we know when we have made a mistake or through accident failed to achieve what we are trying to achieve? Anscombe gives an example where the utterance of an action description and the action do not match:

  …I say to myself ‘Now I press Button A’ – pressing Button B – a thing which can certainly happen … And here, to use Theophrastus’ expression again, the mistake is not one of judgment but of performance. This is, we do not say: What you said was a mistake, because it was supposed to describe what you did and did not describe it, but: What you did was a mistake, because it was not in accordance with what you said.4

  As she points out, this is just like obeying an order wrong, which is not a case of disobedience, but of malfunction. There are a number of places where the malfunction can occur, however. It can occur in the implementing of the button-pressing action or of the verbal utterance – I may have made a slip of the tongue, and have meant to say ‘Now I press Button B’. In the latter case I can correct my utterance via ‘feedback loops’. There is still one more place for malfunction to occur, however. Malfunction could occur between the directing efferent state and the awareness line, so that I would be aware1 of ‘ordering myself’ to push Button A while the efferent state actually directing my motions had the content ‘press Button B’. This would be a case similar to those discussed in § XX, where my sincere report of what I am about is not veridical, where my awareness1 is not a reliable manifestation of my inner direction. In this case I cannot be mistaken about that of which I am aware1 – for only correctible verbal slips are possible there – but only mistaken in supposing that I have practical knowledge of what I am doing; i.e., in supposing that the content of my awareness1 is a veridical report of my inner direction.

  The possibility of this distinct source of knowledge is easy to overlook in the course of a casual examination of one’s own experiences just because ‘practical’ knowledge of our overt actions never exists unmixed. There is a contingent, functional interdependence of ‘practical’ and proprioceptive knowledge, so that we are constitutionally unable to write words on a blackboard while blindfolded (to use one of Anscombe’s examples) without being informed proprioceptively, and so we can become aware1 of the proprioceptive information. Or, without becoming aware1 of the proprioceptive information, we can still be informed by it; i.e., it can contribute to our (inferential) knowledge that not only are we ‘ordering’ a certain action, but it is being correctly carried out. For some actions, such as drawing a cow, visual information is required in addition to proprioceptive information if I am to know what I am doing (unlike the action of, say, writing one’s signature with one’s finger in the air), and in these cases if I am blindfolded I will only know what I am trying to do, not what I am doing. It is often seen as a puzzle that a man can quite directly (without conscious inference or observation) know what he is doing, and not merely what he is trying to do, but this puzzle dissolves when one recognizes that the experiential directness (lack of conscious inference) masks an epistemic mediacy, in which information from two or more sources combines to give us the contents of awareness. For overt actions, involving skeletal and muscular motions, it is contingently impossible to isolate cases in our normal experience, in which we have only practical knowledge – knowledge of the efferent ‘commands’ given. We are all, however, familiar with the experience of trying to move an arm that is ‘asleep’, and in these cases our knowledge of what we are trying to do is pure practical knowledge, purely a matter of being aware1 that this is the efferent command being given.

  There are thus a variety of ways in which we can be informed about what we are about. We can see what we are doing just the way we see what others are doing; we can feel, proprioceptively, what we are doing; and we can have practical knowledge of what we are doing, that is, we can be aware1 of our efferent commands. The latter mode of knowledge is a necessary condition of intentional action, but it is still not a sufficient condition. One may have practical knowledge of what one is doing and still not be doing it intentionally if one’s answer to the question ‘Why are you x-ing?’ is ‘no particular reason’ or ‘I don’t know; I was just doodling’. As Anscombe points out, such an answer is not a rejection of the question, as ‘I was not aware I was doing that’ is. ‘The question is not refused application because the answer to it says that there is no reason, any more than the question how much money I have in my pocket is refused application by the answer “None”.’5 The final requirement for an action to be intentional is that there must be a reason that can be given by the actor for the action. Where there is no reason in the offing, as in doodling, Anscombe would call the action voluntary, but not intentional, and this seems harmonious with our ordinary usage. Other actions may qualify as voluntary, but at least all actions of which one has practical knowledge but can offer no reasons for doing are voluntary. In the case of doodling, for example, the psychoanalyst may claim – even correctly – that there is in fact an unavowed deeper reason for the doodling, but this does not make the action intentional. This point will become important in § XXIII, when we consider the significance of intentional actions, for in finding a reason for the apparently inadvertent and unintentional, is the psychoanalyst not claiming to put the action on a par with intentional actions? That is, should we not treat the action as if it were intentional?

  The giving of reasons by the actor is not a foolproof activity, as we saw in § XX. Even where conscious reasoning occurs, so that we are aware1 of the apparent input and output of the reasoning process – though not of the process itself – we cannot know with certainty that what we offer is reasoning and not rationalization. When conscious reasoning has not occurred our reason giving is simply conjecture, although often highly reliable conjecture. The ‘Why?’ routine brings this out: ‘Why are you sawing the plank?’, ‘I’m making a table’, ‘Why are you making a table?’, ‘To put our food on’, ‘Why put your food on a table?’, ‘Just because, that’s why’. After the initial answer or two, what follows is conjecture or fabrication, not based on any conscious reasoning one remembers having performed. It is true or false, however, since either the information cited has contributed indirectly to one’s behaviour or not. The repetition of the ‘Why?’ question is supposed to have the effect of probing deeper and deeper into the beliefs and methods of the actor, but once the responder has reported whatever thoughts he was initially aware1 of that might have contributed to the direction of his behaviour, the subsequent answers are merely parts of his own personal theory of motivation. I
t still makes sense to ask the question, though, for the responder has his memories of past behaviour and the thinking that accompanied it, and this is certainly relevant information not held by the questioner. The actor is simply empirically better acquainted with his own style of behaviour than anyone else is, although he may not be particularly perceptive or critical about his own behaviour.

  The distinction between awareness1 of efferent commands and awareness1 of one’s further reasons is not a sharp one. An intentional action is a motion under a particular description, and we saw that sawing the plank and sawing one of Smith’s oak planks are two different actions. For an action to be intentional, moreover, we saw that the actor must be aware1 of the action under that description, via efferent commands. Yet is it reasonable to suppose that different motor commands would have the different contents ‘saw the plank’ and ‘saw Smith’s plank’? The differentiation here must come at the level of awareness of reasons, not awareness of motor commands, but it is not always clear where to draw this line. Does one push the pen in order to sign the document or in signing the document does one push the pen; does one sign the document in order to close the deal, or in closing the deal does one sign the document? One can say that one must be aware of one’s further reasons for performing an intentional action, or one can say that one must be aware of a wider description of what one is doing when one performs an intentional action. The demand for reasons for intentional actions is not a demand with fixed limits, since there is no fixed length for the nested reasons one must give, and no fixed description from which to start. Sawing the plank is intentional if the question ‘Why are you sawing the plank?’ is answered ‘I’m making a table’, but if someone a bit more observant asks, ‘Why are you making a table?’, this requires something in the way of further reasons if it is to be considered intentional.6

  XXII WILLING

  The account of intention that has been given includes no talk about volitions or willing. That is because, as Anscombe argues, the verb ‘to will’ is a hoax. There are no such things as acts of will or volitions.

  People sometimes say that one can get one’s arm to move by an act of will but not a matchbox; but if they mean ‘Will a matchbox to move and it won’t’, the answer is ‘If I will my arm to move in that way, it won’t’, and if they mean ‘I can move my arm but not the matchbox’ the answer is that I can move the matchbox – nothing easier.7

  The idea that willing is some sort of radiation generated by gritting the teeth and saying, ‘move, move, move’ is hopeless. It arises, no doubt, from such experiences as lying in bed and saying to oneself ‘I must get up, I must get up; it’s late. On the count of three: one, two, three …’ until finally one gets up. The causal link in these cases has been debated at great length, for on the one hand thinking these thoughts often seems to help or even cause the action, and yet on the other hand very often thinking the thoughts has no effect at all.

  It is supposed, perhaps, that when thinking these thoughts does not work, one is just not thinking hard enough or with enough conviction, but these explanations are obvious dead ends. The ‘tone of voice’ with which one says these things to oneself clearly does not make any difference, and what else can one do to simulate or bring about conviction? The facts of the matter – that sometimes the thoughts seem to help and sometimes not – suggest that thinking to oneself is merely an accompaniment or by-product of the actual business of determining action. It seems most likely, that is, that if I have the conviction that I must get up, I can say the words and will thereupon arise from bed – but I will arise equally well if I do not say the words to myself. And if I do not have the conviction that I must get up, I will not get up, whether or not I mutter exhortations to myself with great vehemence.

  The view of neural activity so far developed provides a plausible if sketchy explanation of this phenomenon. Roughly, in order for the brain to initiate the activity of getting up, its input must be such that it outweighs, say, the pleasure of just lying in bed, the influence of stored information to the effect that getting out of bed is unpleasant, the input to the effect that the body is still tired, and so forth. As soon as the balance is tipped, the brain initiates the activity and one gets up. There is no need to suppose that once the balance is tipped something must recognize that the balance is tipped and then proceed to will the act of getting up, any more than a spinning top must recognize that its gyrostatic force is no longer strong enough to balance it in order to ‘decide’ to fall over.

  When the balance has been tipped or is being tipped in the brain, messages that cross the awareness line to the effect that it is time to get up, there is much to be done, and so forth, are merely accompaniments to the ‘decision-making’ of the brain. But when the balance has not been tipped, no amount of repetition of these messages, if in themselves they are not enough to tip the balance, will bring about the action.

  The way, presumably, to tip the balance is to increase and facilitate the sort of input that would outweigh one’s inertia. The input that would accomplish this would depend on the dominant organizations at the time. The information: ‘the British are coming!’ would seldom serve to tip the balance. The information ‘it’s time to get up’ can tip the balance if the person has some reason to get up on time or some natural tendency to regularity and punctuality, which amounts to a ‘weakness’ for this sort of input. But if the information ‘it’s time to get up’ is not sufficient to outweigh the inertia, no amount of feeding this information repeatedly into the relevant parts of the brain will tip the balance. It is not awareness or consciousness, however, that is producing this information, but the rest of the brain; awareness is simply the verbal outlet for the information.

  Just as one person can devise input information to stimulate another person – a stubborn or lethargic person – to do one thing and not another, so perhaps the brain, when more or less stalled, resorts to self-stimulation by the production or retrieval of information. And the success in both cases, of course, would depend on the relevance and abundance of the information produced. Something like this may well go on, but if it does, awareness of the information, or saying the information to oneself with feeling, would add nothing to the process. As suggested in Chapter 6, it may be that any information that is boosted into the higher levels of neural organization must come to awareness as a matter of physical fact, but then it is still the boosting and not the awareness itself that is necessary. The notion that must be avoided is that awareness is in any way a centre from which efficacious signals, volitions, or any sort of psychic radiation emanates.

  Anscombe touches on this, but in quite different terms.

  We can imagine an intention which is a purely interior matter nevertheless changing the whole character of certain things. A contemptuous thought might enter a man’s mind so that he meant his polite and affectionate behaviour to someone on a particular occasion only ironically, without there being any outward sign of this (for perhaps he did not venture to give any outward sign) … Let us suppose that the thought in his mind is ‘you silly little twit!’ Now here too, it is not enough that these words should occur to him. He has to mean them. This shows, once more, that you cannot take any performance (even an interior performance) as itself an act of intention.8

  There is an internal difference, quite clearly, between just saying ‘you silly little twit!’ and meaning it, but this difference is not itself a performance. The difference must depend on what the function is of the part of the brain that produces this message in awareness. If the message is produced in the course of the brain’s maintaining a particular antagonistic state, if the production of this message is caused by some neural activity that, say, brings into play stored information on the shortcomings of the ‘silly little twit’, then the ‘thought is meant’. If, on the other hand, the message is produced in the course of, say, mere experimentation, such as seeing if it is in fact possible to be polite while thinking ‘you silly little twit’ and not meaning it, then th
e ‘thought is not meant’. Suppose the thinker of this phrase has been trying to think of a four-word phrase with internal assonance; suppose in other words someone has just said ‘give me a four-word phrase’ with as much assonance as ‘Philip spilt the milk’, or for one reason or another this task has just occurred to the person. The neural mechanism that produces the message ‘you silly little twit’, like the neural mechanism that produced in me ‘Philip spilt the milk’, has, or can have, virtually no other effect on behaviour or neural state than the production of words. The activity involved does not influence or mesh with any other activity.

  How does one know whether one means it or not? One knows this simply because one knows what one is about; and one knows this only by knowing what messages preceded and followed the message in question. Imagine a person all of a sudden finding himself saying to himself ‘you silly little twit’. What if no other rancorous thoughts had been going through his head; what if there was no obvious candidate for the epithet; what if further the thinker had not just been aware of thinking he would try this little experiment? Could there be anything intrinsic in the mere unheralded, unaccompanied phrase occurring in his awareness that would tell him whether or not he meant it? Strange, isolated thoughts do spring to people’s minds occasionally, and they can be totally baffled as to the meaning or importance of these thoughts.