In a person, if a signal (an event given content) does not cross the awareness line, the person cannot express its content; he is not aware1 of its content. Such a signal, however, could contribute to some very useful circuit in the nervous system. A simple reflex like blinking or pulling one’s finger away from something hot is apparently controlled by relatively short neural arcs that one would not expect to involve speech centre activity. Our experience shows us that by the time we become aware1 that our finger has touched something hot the reflex has already occurred. One can become aware1 of one’s blinking, but one almost never is, and awareness1 is certainly not a necessary step between stimulus and response in this case.
It is not only simple reflexes that can apparently be controlled without the intervention of awareness1. An accomplished pianist can play difficult music beautifully ‘with his mind on something else’, and need not be aware1 of the notes on the page, the sounds of his playing or the motions of his hands and fingers. He must, of course, be aware2 of these. There is nothing particularly remarkable about this, for could one not build a machine that read music and then played it (a sophisticated player-piano)? There is no temptation to suppose that it would be anything more than aware2 of what was going on. Experience suggests that although we can only be aware1 of one thing at a time, the brain can control a number of complex activities at the same time. As we say, we do many things without thinking about them, but surely we do not do these things without the brain’s controlling them. It would be rare for a man to drive long distances without occasionally being aware1 of his driving or the landmarks, and similarly the pianist would not long remain unaware1 of the notes, the sounds or his finger motions. In particular, if he made a mistake, some sort of ‘negative feedback’ would no doubt shift him to awareness1 of what he was doing.
This suggests that awareness1 does have some efficacy in behaviour control. What is the point of paying attention if not to control one’s behaviour better, and does not paying attention involve awareness1 of what one is doing? But at the same time it is clear that simply bringing a signal across the awareness line, ‘into speaking position’ so to speak, could have in itself no beneficial effect on behaviour control. There can be no logical relation between being aware1 of something and improving one’s control of related behaviour, but there could be a contingent and coincidental relation.
We bring activities into awareness1 to correct them or improve them. The pianist who keeps fumbling a trill starts paying attention to the particular motions of his fingers when trilling. When learning to drive one is very much aware1 of raising the clutch, shifting gears, looking in the mirror and so forth, although these activities eventually become ‘automatic’. It is also clear that we are inevitably aware1 of the sights, noises and other sensations (to use the word in its most ordinary sense) that are particularly bright, sudden, acute, bizarre, unexpected or otherwise outstanding. It would be appropriate for us to be aware1 of these if our awareness1 contributed to better coping with the environment, since the outstanding sensations are usually the ones that make the most difference to the person’s well-being. Could it be that becoming aware1 of these crucial events is a contingent but perfectly natural by-product of some shift in controls that occurs on the far (inner) side of the awareness line? There seem to be two levels from which we direct our behaviour. At the ‘high’ level (apparently in the cortex) we correlate information from a variety of sources, the behaviour controlled is versatile and changeable – and not particularly coordinated. Once under control, the behaviour is often made into a routine and the control is packed off into a more automatic and specialized system. (Apparently the site of these controls is the cerebellum.) If ‘paying attention’ is a matter of dealing with the relevant parts of the environment at the high level, it might also happen to be a matter of bringing certain high-level signals across the awareness line, just because that is the way the brain is wired. We can even suppose that such connections would have survival value in that they essentially contributed to the human activity of (verbal) teaching and learning.
By viewing the relation between concentration or paying attention and awareness1 as not only contingent but exceptioned, we can account for phenomena not otherwise describable without confusion. We can say, for example, that dumb animals can pay attention (e.g., the first step in training a dog is to teach him to pay attention to his master) without sliding from an acknowledgment of this obvious fact into the supposition that in that case dumb animals are aware just the way people are. Or consider a trained seal balancing on a ball while balancing another on its nose. A seal has a cerebellum and perhaps the seal, like a man, puts part of its trick on automatic pilot in the cerebellum, while concentrating on the other. We can investigate such possibilities without imagining that the seal, in concentrating, is rehearsing thoughts in seal language in his head. Lacking a speech centre the seal cannot be aware1 of anything, cannot introspect. This does not mean the seal is doomed to an unhappy life of uncomprehending darkness, always wondering what was going on both inside and out. Only a being that can be aware1 of something can be sadly unaware1 of anything. People struck blind are depressed by the loss; blindness does not bother stones.
Another case on which light is shed is that of pain. A sore foot may be so sore that I cannot ‘get my mind off it’, and am almost continuously aware1 of the pain. Or I may be only intermittently aware1 of the pain. In the latter case I may or may not be able to concentrate usefully on other things during the intervals between awareness1 of the pain. Philosophers are fond of asking whether there are unfelt pains. The answer that suggests itself is that being in pain, for a person, is a dual situation, involving both awareness1 and awareness2 of the pain. I might not at every moment ‘be conscious of’ the pain (be aware1 of the pain), and yet the continuing neural excitation might disrupt the high-level operation of the brain and hence indirectly bother me, regardless of what I happened at any moment to be aware1 of. Not being always aware1 of the pain, I would not always have unimpeachable authority as to whether the pain was bothering me (saying that it was the pain that was preventing me from working would be a hypothesis for which my only evidence would be the intermittent awareness1 of the pain). When I ceased to be even aware2 of the pain we would say the pain had ceased altogether.
I think we should resist the temptation to choose one of the two new senses of ‘aware’ as the sense of the term; neither one can claim a clear majority of supporting intuitions drawn from our ordinary language. If what is held to be essential to awareness is its relation to behavioural control (‘he must have been aware of the tree, for he swerved’) then it must only be an exceptioned coincidence that one can infallibly introspect what one is aware of; if what is held to be essential is this infallibility of expression (‘only I can tell for sure what I am aware of’) then it is only usually true that we are aware of the information of foremost importance to the control of current behaviour. It is easy to confuse awareness1 with awareness2 and extrapolate from the fact that one must be aware of something in order to say it (express it, report it) to the untruth that one must be aware of something in the same sense in order to do anything with it. For example, Sayre says that ‘we would not say under any ordinary circumstances that we recognized an apple, or some other object, but were aware of no such object’.3 Of course I could not say that I recognized an apple and yet was not aware of an apple, since in order to say anything I must be aware1 of it, but this should not lead to the conclusion that I cannot say he recognizes an apple without implying that he is aware of the apple in the same sense. This illicit move is obscured by the use of ‘we’ in the quotation above. What of sorting machines or animals that cannot say anything? They can fulfil all the functions of recognition short of saying they recognize; does this bar them from recognizing? It is not that one must be aware1 in order to recognize, but that one must be aware1 in order to say that one recognizes.4 To the extent that our ordinary concept of awareness leads to such confusions
it is a poor concept in spite of its ordinariness.
XVI CONSCIOUSNESS
In § XIV, ‘conscious’ was found to have both Intentional and non-Intentional uses, and the Intentional uses were subsumed under the Intentional uses of ‘aware’. We distinguished the question of what it is to be conscious of something from the question of what it is to be conscious, and although this initial distinction allowed an account of the Intentional sense of ‘conscious’, we are not through making distinctions, for even in its non-Intentional uses ‘conscious’ is ambiguous. As Scriven for one points out, ‘conscious’ can mean something like (1) ‘awake’ or ‘aware’ (in the non-Intentional sense of ‘alert’), or (2) having the capacity to be awake or aware.5 That is, in sense (2) ‘conscious’ is used to distinguish beings capable of consciousness in sense (1) from inanimate objects. It is not paradoxical to say, then, that only conscious beings can be unconscious. This is not an unusual ambiguity; only rational creatures can be irrational, only seeing beings can truly be called blind. Since the question of what it is to be conscious in sense (2) is entirely dependent on what it is to be conscious in sense (1), we can skirt this ambiguity by restricting our use of the word to sense (1). Once we know about it, we will be able to say with no further ado what it is to be conscious in sense (2).
Consciousness seems itself to be a capacity that comes and goes in beings that have the capacity to be conscious, but what capacity is it? At first glance it seems to be the capacity to be conscious of, or aware of things, but is this awareness1 or awareness2? This question is implicit in the alternatives, ‘awake’ or ‘aware’. When a dumb animal is awake, it can be aware2 of things; is it then conscious? We have few qualms about saying a dumb animal in a coma is unconscious, but what is its state when it is not asleep or in a coma? Is our reluctance to call animals conscious just a matter of confusing aware1 with aware2? Where do we draw the line between conscious and unconscious in any case? Is a person unconscious when he is asleep or only when he has ‘passed out’ or is in a coma?
So long as we fix our eyes doggedly on the ordinary word ‘conscious’ and consider all that has been said in the past by philosophers and psychologists about consciousness, it is easy to convince ourselves that ‘conscious’ and ‘unconscious’ sunder the universe in a very fundamental way, and from this it is easy to arrive at the conviction that being conscious must be an all-or-nothing matter. Capacities can be partial, however, and admit of degrees (consider the common but ill-defined term ‘semi-conscious’). If being conscious is held to be having the present capacity to be aware2 of things, then when one is in a coma, one is unconscious to a very great degree, and when one is asleep one is unconscious to a lesser degree, since reflexes still work and one can wake up, which is itself a behavioural response to incoming information. If this is what we should mean by ‘conscious’, then animals without the power of speech can be conscious and people can even be aware of things while (relatively) unconscious. Consider a person who said, in describing a dream, ‘Suddenly I was aware of a man with a knife, and that scared me so much I woke up.’ The dreamer could be aware2 of the man in the dream; if he talked in his sleep he might also be aware1 of the man in his dream – and all this while asleep or only partially conscious.
If, on the other hand, consciousness is seen more restrictedly as the capacity to be aware1, of things, dumb animals can never be conscious, and dreamers and people who talk while in hypnotic trances are conscious. I do not believe that an analysis of our ordinary language here will reveal that we mean one of these rather than the other. Ordinary language does not determine which of these alternatives is right, because it mixes them together. A decision on ‘conscious’, therefore, cannot be a decision about what the word actually means, but only what it ought to mean, and this can only be relative to our purposes. Most of the interesting theoretical questions about consciousness seem to tie it to the notion of awareness1. This is particularly clear in the case of the terms ‘subconscious’ and the Freudian ‘Unconscious’.
The control of reflexes in man is subconscious, as are the stages of perceptual analysis, and in fact all information processing. We are not aware1 of the processes at all (as one might, with suitable incisions and mirrors, be aware1 of one’s digestive processes). What is subconscious, clearly, is everything that happens in the brain except what crosses the awareness line. As Lashley says, ‘No activity of mind is ever conscious.’6 He gives an example: responding to the request to think a thought in dactylic hexameter. We are conscious of the thought we produce, but not of its production, or of how it was produced. The consciousness Lashley is concerned with is clearly the capacity for awareness1; we cannot say how we produce this thought, and have no introspective access to the activity or process. Similarly, the Freudian Unconscious, if it is anything at all, is a region inaccessible to awareness1, and has nothing directly to do with comas and sleep.
This proposed treatment of ‘conscious’ – and its brethren, ‘unconscious’, ‘subconscious’ and ‘Unconscious’ – completes the fragmentation of the ordinary words of the chapter title. The philosophically trained reader is apt to feel uncomfortable about this analysis for two reasons. First, because of his recognition of the privileged position of ordinary language, he is apt to conclude that I have not so much analysed the ordinary words as tampered with them, an error that puts me in danger of speaking nonsense or at least irrelevancies. Second, he has been taught, by Ryle and others, to avoid the ‘bogy of mechanism’, and my analysis is accompanied by the spinning out of some unabashedly mechanical and quasi-mechanical speculation. The two philosophers’ rules, ‘Tamper not with ordinary words’ and ‘Avoid mechanism’, are good rules, but their sound application is not universal. I would like to defend my particular transgressions of these rules.
In Chapter 1, I proposed as a modus operandi that sentences using the difficult ‘mental’ words be treated as significant, true and false, but not automatically subject to the sort of semantic analysis that generates ontology. The question of whether ‘thought’, for example, refers to anything was to be left for the time being up in the air. In Chapter 4 this hands-off policy was invoked when the questions were considered: how do we tell pains from non-pains, and how do we locate pains? These questions, framed in the language of people and pains, not bodies and neural events, have no answer, or only the brusque answer: we just do, that’s all. Different questions may be asked and answered, but these are not strictly speaking about pains. If one is not confused by the brusque answer to the original questions, then the concept of pain is not confusing, is not to be criticized or discarded. Similarly, in Chapter 5 the temptation to identify thoughts with certain brain processes or their contents was resisted since we do say that we report our thoughts, and do not run into any ordinary confusions in saying this. No reason was seen for abandoning the idiom ‘report one’s thoughts’ in favour of something clearer. In these instances the requisite correlations between the mental-language sentences and physical-language sentences could apparently be made without first doctoring up the mental-language sentences.
With ‘conscious’ and ‘aware’, however, the situation is different. Ordinary usage of these words is not remotely consistent. We do say that both people and animals are aware of things, that they are conscious or unconscious, that one can say what one is aware of or conscious of, and that one must be aware of something in order to recognize it. We do say these things, but we say them, even ordinarily (when not engaged in philosophical discussion) with misgivings. Our intuitions conflict when we are confronted with the crucial test cases. It is not merely that philosophers can generate confusions by misusing these words, but that the words in their most time-honoured uses are confused. Order can be brought out of this chaos, but only by abandoning the conviction that ordinary usage here is conceptually sound, that it meets the standards that are met by most words as a matter of course. And as soon as we abandon this conviction we can no longer rely on the totality of usage to tell u
s what lies behind our notions of awareness and consciousness, because nothing consistent lies behind the totality of usage. When we look for distinctions that will serve to mark off consistent, separable senses of these words all we can find are distinctions of function, and pinning down these distinctions of function involves drawing a plausible mechanistic picture. There is an alternative – a ‘para-mechanical’ picture, to use Ryle’s term, replete with non-physical, non-causal ‘bits of not-clockwork’. Avoiding this alternative, one is still forced to do psychology rather than ‘pure’ philosophy.
The psychological hypotheses, however, do not miss the mark when applied to the philosophical questions. Can a machine be conscious? This question cannot be answered until we arrive at a conclusion about what it is to be conscious, and ordinary language does not tell us. Consider the philosophical puzzle of whether or not animals can be conscious and the related puzzle of whether or not animals perform intentional actions. Intentional actions are characterized Intentionally and it is commonly accepted by philosophers that Intentionality presupposes consciousness.7 What, indeed, could be more obvious? One must be conscious in order to believe, or in order to want, or in order to perform any other ‘mental act’. But what is this consciousness that is presupposed? Is it something only people are capable of, or can animals be conscious? If we decide – and this is fairly common – that only language-users can be conscious, does this suffice to demonstrate that it would be wrong to ascribe Intentionally characterized actions to dumb animals? Anything that suffices to demonstrate this must itself be suspect, for we do describe animal behaviour in Intentional terms quite successfully.8 The way out of this impasse must be more than a solution to a conceptual problem via analysis of language, for language is deficient in this area. The way out is an analysis of phenomena at the sub-personal level, and although this leads one into areas many philosophers would prefer to avoid, the alternative is the perpetuation of traditional confusions.
Content and Consciousness Page 16