The Value of Consciousness

It is widely assumed that consciousness matters. It matters in a way, and to a degree, that many of the other topics on which philosophers outside the explicitly normative areas of the discipline spend their time and energy do not. It matters in the kind of way that central topics in ethics and political philosophy matter. Indeed, it is, or should, be a central topic in ethics, since it is intrinsically valuable and/or confers value on those who possess it. Accordingly, consciousness, or its lack, is often invoked in discussions in applied and normative ethics. Ethical issues that arise at the beginning and the end of life are often held to turn, in part at any rate, on whether the organism is conscious. If the fetus is conscious, it is widely held, its moral value is greater than if it is not yet conscious, and this extra value must be taken into account in discussions of the permissibility of abortion. The presence or absence of consciousness is held to make a difference in a parallel way when the question concerns the permissibility of switching off life support, or the pursuit of further treatments for those near the end of life.

Consider, for illustration, the fraught case of the persistent vegetative state. After traumatic brain injury or anoxia, people sometimes awaken from a coma to a state in which consciousness seems to be absent, at least so far as behavioral measures are able to glean. Patients like this fail to respond to stimuli in a way that indicates awareness of their surroundings or of themselves. They do not respond to words, including their own name, and they do not track movement with their eyes, despite being awake and despite having no physical impairment that would prevent them from doing these things. If this state persists for more than three months, it is said to be a persistent vegetative state, and chances of further meaningful recovery are low.

The persistent vegetative state is difficult to diagnose and misdiagnosis is not uncommon (37-43%; Schnakers et al., 2009). PVS may be diagnosed when the persons is really in the minimally conscious state, in which response to external stimuli is inconsistent and fluctuating, or in the locked in state, in which the person is fully conscious but incapable of voluntary movements (beyond, at most, eye blinks). Moreover, since patients correctly diagnosed as PVS may behave in ways that may be taken to indicate consciousness (they may laugh, cry, move their heads and so on), families sometimes refuse to accept the diagnosis. Opponents of withdrawing nutrition and hydration from these patients sometimes base their opposition on claims about the sanctity of life, but they may also, or instead, insist that the person may be conscious. Again, we see the way in which a major ethical question is thought to turn, at least in very significant part, on the presence or absence of consciousness.

Recent advances in technologies for the detection of consciousness in patients who are unresponsive to external stimuli (Owen et al., 2006; Monti et al., 2010) promise to alter the terms of this debate, insofar as they will allow us diagnose the presence of consciousness more accurately, even in cases when the person is incapable of overt behaviour. In this paper, however, I do not aim to assess how we ought to respond to these new technologies, nor whether it is permissible to withdraw nutrition and hydration from patients correctly diagnosed as PVS. Rather, my aim is to address a prior question, which has been little discussed previously, but which appears to be relevant to these debates within ethics, as well as, indirectly, to the significance of work in philosophy of mind. That question concerns the value of consciousness.

All sides appear to have assumed that the presence or absence of consciousness is central to the question whether the person has a life worth preserving. It is the plausibility of this assumption that I aim to examine, by asking whether consciousness is itself valuable, or makes the life of a conscious being more valuable. My focus is on phenomenal consciousness, since it is this kind of consciousness that is usually held to be valuable. I shall argue that a great deal, if not all, of what we value in life and makes beings valuable can be explained by other kinds of consciousness. In conclusion, I shall briefly comment on the implications of my claims for debates in applied ethics.

‘Consciousness’ is not a univocal term. Philosophers distinguish a number of different senses of the word. The most important distinction here seems to be the distinction between phenomenal consciousness and access consciousness (Block, 1995). Unfortunately, it seems to be impossible to define phenomenal consciousness (Chalmers, 1996). The best we can do is to point to instances of it. Phenomenal consciousness is the kind of conscious such that there is something it is like to be phenomenally conscious (Nagel, 1974). There is something it is like to taste a glass of pinot noir, to hear the opening notes of Tristan und Isolde, to feel the warmth of the sun on your face or to feel a pain in your left knee. Each of these experiences has a distinctive phenomenal quality to it. That quality seems inexpressible. We often use metaphorical language to communicate that quality (“it is a dull throbbing ache”; “a sharp stabbing pain”; “a rich lush red”), but when we do so we seem to rely on our shared experience with such phenomenal qualities to calibrate our talk.

Access consciousness is not a state that has a phenomenal feel essentially associated with it, though it may be that access conscious states typically (or even always) have phenomenal qualities associated with them. Access consciousness is a kind of availability of informational contents: a state is access conscious if the person is able to access the contents. Paradigmatically, if contents are access conscious, we can report them; if they are not access conscious, we can’t report them. This isn’t definitive of access consciousness, though: sometimes we become conscious of the content of our minds by reporting them (Carruthers, 2009) and sometimes we can’t report what we are conscious of (because we have lost the use of our voice, say).

Sometimes, philosophers say that phenomenal consciousness is the only kind of consciousness there is. According to these philosophers, what ordinary people mean by ‘consciousness’ is phenomenal consciousness, and access consciousness, which is not essentially phenomenal, is simply some other kind of phenomenon. We might buttress this line of thought by noting that because access consciousness is an information processing concept, there is no reason why a computer can’t be conscious in that sense; that is, there is no reason why a computer should not be set up so that its central processing mechanisms have access to the contents of any of its other systems. But since people commonly doubt whether a computer can be conscious, they can’t mean access consciousness by ‘consciousness’.

However, there are good reasons to think that ordinary people and theorists sometimes do mean access consciousness by ‘consciousness’. In all his writings, Freud seems to have had this kind of notion of consciousness in mind. When he claims that neurosis is caused by unconscious thoughts, or something along those lines, he clearly means thoughts with contents that can’t be accessed, not thoughts which fail to have a phenomenal feel associated with them. Many people have taken issue with Freud, for many different reasons, but no one has ever objected that his postulates aren’t really unconscious, because we mean phenomenal consciousness by ‘consciousness’. Similarly, in contemporary debates over whether consciousness is a necessary condition of moral responsibility (Arpaly, 2002; Smith, 2005; Sher, 2009), it is clearly access consciousness that is at issue (Levy, forthcoming). The sense of ‘consciousness’ may vary from context to context.

Nevertheless, in this paper I shall focus on phenomenal consciousness, which is certainly the central concept of consciousness in contemporary philosophy of mind. In debates about the ‘hard problem’ (Chalmers, 1996), for instance, it is phenomenal consciousness that is at issue. The hard problem is supposed to be hard precisely because it is so difficult to explain consciousness in information processing or functionalist terms and because it is so difficult to see how the physical systems that implement cognitive processes could give rise to consciousness. There is no in principle difficulty with understanding how physical systems could implement access consciousness, so clearly it is phenomenal consciousness on which these debates turn (as the disputants explicitly recognize).

It is widely assumed that it is phenomenal consciousness on which the ethical disputes mentioned above turn (Kahane and Savulescu, 2009). It is not obvious to me that this assumption is correct. It may be that disputants have no clear notion of consciousness in mind at all, and that clarifying the notion of consciousness at work would all by itself advance the ethical debates. Settling what sense of consciousness those on different sides of debates have in mind would take exegetical work, and perhaps empirical work as well (perhaps it would be useful to conduct some experimental philosophy, asking people who think that, for instance, it would be wrong to withdraw nutrition from patients because they may be conscious what sense of ‘consciousness’ they have in mind). Here, however, I will simply accept the common assumption and ask whether phenomenal consciousness can make the kind of difference to moral status apparently invoked in debates in applied and normative ethics. Does phenomenal consciousness make a significant difference to the value of the lives of organisms?

How might we go about attempting to answer this question? Siewert (1998), in the most sustained consideration of the value of consciousness, argues that the best way to approach this question is via thought experiment. Imagine you are offered the opportunity to live your life without phenomenal consciousness. Imagine, that is, you are offered the chance to become a zombie, in the sense of the word employed in philosophy of mind. A zombie in this sense is a functional (perhaps even a physical) duplicate of an ordinary person. A zombie twin of me would do all the things I do (eat, sleep, write philosophy papers), would be indistinguishable from me to my friends and family, and would have mental states with contents just like mine. My zombie twin would believe what I believe, hope for what I hope for and fear what I fear. But these mental states would be unaccompanied by any phenomenal states. When he eats, he is able to detect the qualities of the food (if asked, he can tell you if it is salty or bitter, say), but he does this in a way that does not involve the distinctive qualitative experience of these tastes (just as a machine might be designed to detect saltiness; such a machine might have salt detectors similar to ours, but it would not experience the saltiness). Similarly, when it fears what I fear, its fear would be a functional state (it would withdraw from what it fears; be motivated to do things that make it less likely that it will encounter things it fears, and so on) and would be accompanied by the physical states that accompany fear in me (heart pounding, sweaty palms, dry throat, and so on). But again, there would be no qualitative feel associated with its fear.

Because a zombie twin of you can do all the things you can do, have the same experiences (in some sense of ‘experience’, at any rate) and achieve the same ends, if you do not value phenomenal consciousness, you should be indifferent with regard to the offer to become a zombie. If you are offered even a small incentive (a small amount of money, or a slight increase in cognitive capacity, or what you), you should be happy to choose zombification over your current state. But Siewert thinks that a moment’s reflection reveals that all of us would be very reluctant to accept the offer. It would take a very large incentive to induce most of us to accept it (say a cure for the terminal illness of a loved one). This shows, Siewert claims, that we think phenomenal consciousness is intrinsically valuable: valuable not for what it allows us to achieve (since by hypothesis our zombie twin can achieve anything we can achieve) but just for itself.

This thought experiment is designed to show us that we do value phenomenal consciousness. But how much do we value it? Siewert argues that we place so much value on phenomenal consciousness that we would view a life lacking in conscious experiences “as little or no better than death” (1998, 329). Careful consideration shows, therefore, that we value phenomenal experiences intrinsically, and that we value it a great deal. Indeed, Siewert thinks that there is a case to be made for thinking that we would not survive zombification: if I were transformed so that I no longer had qualitative experiences, I would no longer exist. Obviously, these claims are directly relevant to ethical theorizing. Their truth would underwrite the truth of the intuitive thought that a conscious being has more value than a nonconscious, and therefore make the appeal to consciousness in ethical decision-making a rational and appropriate one.

However, we need to be very careful in conducting the thought experiment Siewert recommends. It is very hard to know when we have succeeded in imagining what we lose when we lose phenomenal consciousness. Lack of phenomenal consciousness is not lack of consciousness; a zombie is still conscious in some sense. My zombie twin is self-conscious and access conscious. Perhaps the kinds of consciousness it continues to have underwrite more of the value of its life (to itself, or objectively) than Siewert thinks. Even if these other kinds of consciousness cannot underwrite all the values that Siewert attributes to phenomenal consciousness alone, they can underwrite some kind of value, and this kind of value seems more significant than Siewert seems to credit. I will begin with this claim, prior to a direct consideration of how much we lose when we lose phenomenal consciousness.

It is sometimes claimed that phenomenal consciousness is a necessary condition of anything mattering to an organism. Kahane and Savulescu (2009) claim, for instance, that only a phenomenally conscious being has “a point of view, a subjective take on things” (13). Perhaps it is this kind of consideration that Siewert has in mind in claiming that I would not survive the permanent loss of phenomenal consciousness. More strongly, Seager (2001) suggests that there can be nothing of any value without valuings, and that valuings have to be conscious. But even if it is true that only a conscious being has a point of view, and that there can only be values if there are conscious acts of valuing, I see no reason at all to believe that the consciousness in question has to be phenomenal consciousness. My zombie twin has a point of view. He sees the world from a particular perspective, in an attitude-infused way. Indeed, his idiosyncratic take on things is identical to mine. He also values things. He believes that his life will go better if certain things happen, and therefore he desires that those things happen. He is motivated to do things that make it more likely that those things happen. There are states of affairs that he does not believe he can influence, but toward which he takes attitudes too (he hopes for peace in the Middle East, for instance). He is inclined to say that he values these things, and I am inclined to agree with him.

To assess how well or badly his life is going, it might be worth measuring his well-being against the three most influential accounts of happiness available. Hedonists believe that happiness consists, roughly, in having the pleasant experiences in one’s life outweigh the unpleasant (by some margin). Desire-satisfaction theorists hold that a happy life is one in which enough of one’s desires are satisfied. Objective list theorists hold that there is some set of goods that a life must instantiate (the goods of friendship, artistic fulfillment, intellectual challenge... just what should feature on the list is controversial) for that life to be a happy one, regardless of what the individual wants or thinks she wants.

Let’s leave hedonism to one side for the moment. It should be obvious that my zombie-twin has no less a chance of satisfying his desires than I do. He wants precisely the same things I do, and he does precisely the same things I do to achieve them. We are evenly matched on desire-satisfaction; if his life doesn’t go well on that score, its not because he’s a zombie: it’s because he didn’t desire the right things or doesn’t or can’t achieve them. For most of the items on the objective list, we are evenly matched too. Friendship, health, engagement with nature, intimate relationships – all of these are as much within his grasp as mine. But some items on the list might seem out of his grasp. Consider aesthetic experience. He might be able to understand the harmonic relations between the chords in a Beethoven sonata, but isn’t appreciation of the rich timbre of the piano denied to him in principle? He is blind, is he not, to the vibrant colors of a Mondrian painting? These experiences require phenomenal consciousness, it seems. Likewise, some of the experiences associated with intimate relationships are phenomenal: love without its qualitative character would be a pallid imitation.

I think that it is actually far from obvious that aesthetic experience is even partially inaccessible to my zombie twin. Because he is my functional twin, he has some kind of grasp of color and timbre; since I respond to these features of the world, so does he. Clearly these properties make a functional difference to me: I might be disposed to prefer one recording of a sonata to another because of the timbre of the piano, and that preference might cause me, inter alia, to listen to that recording more often. So will he be, for precisely the same reason. If I am disposed to comment on the richness of the colors in a painting, so will he be, and he will take himself to respond to them due to the way he experiences them. In some sense, he does experience the colors. He also experiences emotions: given that my emotions make a difference to my behavior, we can be sure of that.

If there are differences in the extent to which his life can go well by having his desires satisfied, or by the extent to which he achieves those goods that actually make up a flourishing life, and the extent to which my life can go well for the same reasons, these differences are subtle. Let’s turn now to hedonism: surely experiences of pleasure and pain are barred to him, and for that reason his life is less good than mine? This is Seager’s (2001) view. He claims that pleasure, as invoked by Bentham, and happiness, as it figured in Mill’s utilitarianism, are phenomenally conscious states. Again, I think we should hesitate before we accept the claim that my zombie twin can’t experience happiness or pleasure.

To a great extent, happiness and pleasure might be devoid of proprietary phenomenology. Many of the moments subjects report as most happy are experienced in the state called ‘flow’ (Csikszentmihalyi, 1990). Flow is experienced as absorption in an activity; the person is aware of the activity and not themselves. A fortiori, they are not aware of any experiential state, phenomenal or not (on some views, they might nevertheless have phenomenal experiences to which they lack access – see Block, 1995; 2007 – but it is difficult to see how phenomenal states of which agents are unaware contribute to their well-being). Moreover, even when we report feeling joyful (say), it is not clear to what extent the reported experience overflows the functional; joy, too, plays a role in my behavior. The same goes for less florid states: I experience a quiet satisfaction when my desires are satisfied, but to at least some extent that satisfaction is cognitive – consisting in the knowledge that I have achieved what I sought – and some of any remaining feeling beyond that can be captured functionally. There may still be a remainder, but a lot of the value of the state to me has been accounted for by these functional considerations.

Even a state like pain, the badness of which seems to consist essentially in the way it feels (pace Carruthers, 2004), might be as well captured by a functionalist analysis as a purely phenomenal. Though it is intuitive that pain is essentially aversive, the aversiveness of pain and its phenomenal feel may actually dissociate. In pain asymbolia, produced by prefrontal lobotomy, cingulotomy and some analgesics the person reports that the pain feels just like it always has, but that they no longer mind it. The aversiveness has vanished yet the experience remains the same. Pain asymbolia seems to demonstrate that the badness of pain does not lie in its phenomenal character (Clark, 2005 goes further, arguing that it shows that pain is not a quale at all; I think it might be better to say that pain has two faces, one of which is not a quale). If the badness of pain does not consist in the way it feels, though, then the goodness of pleasure seems unlikely to consist in the way it feels.

I take the forgoing considerations to show that we lose a lot less than might be thought in losing phenomenal consciousness. I think they also provide us with reason to think that the thought experiment Siewert recommends to us is one that is difficult to conduct. When we try to imagine that we have lost phenomenal consciousness, we must be careful not to allow ourselves to be misled into imagining the loss of subjectivity or experience (assuming it is possible to imagine such a thing). Siewert compares the loss of phenomenal consciousness to the loss of color vision; it is akin to moving from a colored world to a black-and-white world (1998, 323). Perhaps the comparison is an apt one, but it is very hard to be sure. We don’t lose color vision when we lose phenomenal consciousness. Nor do we lose the ability to experience pleasure and pain, emotions or sounds. We lose the ability to experience these things phenomenally, but to what extent that is a significant loss is very hard to judge.

Some theorists of consciousness might go further, invoking the claim that much of what we value in conscious experience does not overflow the functional (or the representational), to argue that in fact it is not possible to lose phenomenal consciousness without losing these non-phenomenal qualities, because phenomenality reduces to or supervenes on these qualities. These philosophers take the zombie thought experiment to be not merely metaphysically impossible but actually incoherent: I cannot have a zombie twin, since a functional duplicate of me must be conscious, to whatever extent I am. These theorists hold that qualia, the qualities that make up phenomenal experiences, are functional properties, or (more commonly) they take them to be the representational contents of experiences, or the properties represented by experiences. If functionalism about qualia is true, then a functional duplicate of me is a phenomenal duplicate of me. Matters are trickier if representationalism is true: the properties I represent may be externally individuated, to some degree, and a functional duplicate of me might represent different properties. But even on a representationalist view with externalist commitments, my representational duplicate would be a phenomenal duplicate, though a microphysical duplicate might differ in what it represents due to differences in its environment. Since my zombie twin can represent states of affairs, or the properties or contents of experiences, it is just as phenomenally conscious as I am.

If any of these proposals succeed, we need not ask what kind of consciousness is at issue: tests for access consciousness (which is far more empirically tractable than phenomenal consciousness) are, ipso facto, tests for phenomenal consciousness as well. These proposals are, of course, highly contentious. However, it is worth emphasizing how much these proposals capture of what can plausibly matter, so far as the value of consciousness is concerned. Even if those philosophers who deny that phenomenal consciousness can be reduced to or explained by functional and representational qualities are right, these proposals show that much of what we value in consciousness is not irreducibly phenomenal.

What follows from these considerations, with regard to the invocation of consciousness in debates in applied and normative ethics? In one way, these considerations make the debate more tractable: phenomenal consciousness seems mysterious and inaccessible, but other notions of consciousness are far more easily tested for. There are extant tests for self-awareness and meta-cognitive abilities even in animals, and no reason why they cannot be applied to humans, too.

One way in which debates might change in the light of the arguments given above is that we might become more inclined to resist the temptation to think that moral value increases discontinuously. Some philosophers believe that phenomenal consciousness cannot be a matter of degree (Kahane and Savulescu, 2009). But clearly psychological complexity is a matter of degree, and psychological complexity seems to correlate well with access consciousness. To the extent to which other kinds of consciousness underwrite moral value, it seems likely that the moral value of a fetus will gradually develop through pregnancy and perhaps even after birth. It may often gradually diminish at the end of life, too: though sometimes a person may abruptly go from a state in which their life is valuable to them to one in which it is not as a result of brain injury (for instance), often their psychological complexity will decrease more gradually. Finally, the temptation to think there is a gulf between our moral value and the value of non-human animals, because we alone are conscious, diminishes once we recognize that many of them are conscious, too, in a way that apparently underwrites moral value.

These kinds of considerations are merely preliminary; developing them further requires detailed work in moral philosophy, in conjunction with work in the philosophy of mind. Here I have aimed at a project that is a necessary condition of that work: to argue that there are good reasons to think that though phenomenal consciousness may be valuable, other kinds of consciousness underwrite a great deal of what we value and is valuable in our lives. As well as asking ‘is the patient conscious?’ we need to ask ‘how is the patient conscious’, and have our ethical decision-making guided by the answers to both questions.