AA308

TMA06

Are Zombies Conceivable?

A philosophical zombie is a hypothetical being, physically indistinguishable from a normal human but with no conscious experiences. It forms the basis for the Conceivability Argument, which attempts to show that consciousness is a non-physical property. Paraphrased from Frankish [2005] the argument is thus:

1. Zombies are conceivable.
2. Conceivability implies metaphysical possibility.
3. Therefore it is metaphysically possible for two entities to be physically identical, yet differ in their conscious states.
4. Therefore conscious states are independent of physical properties.

In this paper I shall be focussing on premise 1, and asking whether zombies are conceivable or logically incoherent.

So what exactly is a zombie? To illustrate, it is useful to imagine I have a zombie twin. My zombie twin is physically identical to me, right down to the cellular and subatomic levels. Assuming a materialist universe, physically identical means functionally identical, and thus my zombie twin behaves identically to me. However, he has no consciousness, no first-person point of view. There is nothing it is like to be a zombie. He has mental states (as we shall discuss later) but those states have no qualia.

There is nothing immediately incoherent about this idea. Imagining a being physically identical to ourselves is easy enough, and imagining something with no consciousness (e.g. a rock) is easy enough. However, if we start considering zombies in more depth, the conception becomes less than trivial.

Conceiving zombies is difficult, because conceiving consciousness is difficult. We can certainly conceive of our own consciousness but, as discussed in Lewis [1998], consciousness is inherently private, and intrinsically bound up with a point of view. We know what it is like to be us, but we cannot know what it is like to be someone else except by assuming it is comparable to our own experiences. Our language lacks the vocabulary for people to describe their conscious experiences directly. The best they can do is describe them indirectly, with reference to experiences we are acquainted with. What cannot be described in words is extremely hard to coherently conceive.

Conceiving a lack of consciousness is even more difficult, as to be able to conceive the absence of something, one must first have a coherent idea of its presence. If you only drink white coffee you might be able to coherently conceive of coffee without milk. However, if you only ever drink black coffee it is harder, as you must first conceive milk's presence before you can then conceive its absence. Can you coherently conceive of coffee without curry powder? To take another example, I have always been an atheist, so when Christians ask me what it is like to live without God, I cannot give them any sensible answer as I personally cannot conceive of a life with God, so I have nothing to contrast my life with.

Thus, as it is extremely difficult to conceive what it is like to be another person, it is even harder to conceive that there is nothing that it is like to be that person. This doesn't necessarily make zombies logically incoherent though.

Let's take a different approach, and examine whether we can conceive of mental states without qualia. We can certainly conceive of qualitatively different mental states. Consider the idea of inverted qualia, where two people look at the same colour yet one sees it as red while the other sees it as green. It is simple enough to conceive of the idea that animals with different visual apparatus to us could see infra-red as a distinct colour, even though we cannot imagine what that colour would look like. We can conceive absence of qualia, as in someone who has been born blind, but can we conceive of absent qualia while the person is functionally unimpaired?

There is such a thing as "blind-sight" where someone has only partial vision, but they can still identify objects that are held within their blind zone (see Crawford [2005] or Frankish [2005]). Many philosophers have extended this idea to a hypothetical perfect blind-sight, where a person has no vision at all but somehow 'just knows' what is being held in front of them. Compare the blind mystic, as featured in many a kung-fu movie. He has no vision, but with near-psychic precision he 'just knows' when the hero is going to try and snatch the pebbles out of his hand.

We can extend this idea further and imagine a person who has no sensations at all, but who 'just knows' where they are, who is speaking to them, how close they are to the open fire etc. They navigate through life using the same indescribable gut-feelings you sometimes get when you 'just know' you're being watched, or when you 'just know' your twin brother has just died.

However, in an echo of Descartes' meditations, no matter how many qualia we strip away, ultimately we are still left with the notion that there is something within the zombie's mind that is doing the 'just knowing'. My zombie-twin thinks, therefore he is. This hints at logical incoherency. However, there may yet be something that assists in the conceiving.

Chalmers [1996] maintains that the conceivability of zombies is intuitively clear, but to bolster the case he refers to the China Brain. In this thought experiment the people of a vast population - such as that of China - all take a role as a neuron and through a complex communications network precisely model the functioning of a human brain. This network is connected up to a humanoid simulacrum which will then behave identically to a normal human. While many philosophers (including Chalmers in fact) believe that such a system could produce consciousness, it is if anything easier to imagine that it does not. This system, if modelled on me, would be functionally identical to me in just the same way as my zombie twin would be. If we're happy with the idea that the China Brain is not conscious, then we should have no problem conceiving the zombie. However, as we shall see later when we re-examine the China Brain, it's not that easy.

My zombie twin is physically, functionally and behaviourally identical to me. This means he is completely indistinguishable from me. We are physically identical, therefore any physical test would return identical results for both of us. We are behaviourally identical, therefore any behavioural test would return identical results.

In fact, any test we can conceive would return identical results for us both. This includes the most sophisticated Turing tests and psychological profiles. It includes hypnotic regression, and interrogation by Derren Brown. It includes attaching us both to an infallible lie-detector and asking "are you conscious?".

This indistinguishableness is an essential part of the zombie definition, but it is often watered down or misunderstood. Moody [1994] goes so far as to write a paper on how we could distinguish philosophical zombies using behavioural differences in their manner of speech. Dennett [1995] attempts to highlight this misunderstanding by introducing the idea of a "zimboe" - a zombie that, by hypothesis, is physically and functionally indistinguishable from a normal conscious human - and showing how zimboes in fact differ from most people's conceptions of zombies when they should be identical.

This indistinguishableness makes conceiving zombies trickier than at first appears. It is easy enough to conceive of something without consciousness, such as a rock. However, a rock does not have passionate discussions about the nature of consciousness, whereas my zombie twin would. Zombies fall in love, plead for their life, grumble about politics and generally act as if they were conscious. Rocks do not.

Let's look again at the China Brain scenario. As Cottrell [1999] points out, there is a danger that this is merely appealing to our biases. Instead of focussing on the vast population of people pretending to be neurons, focus instead on the simulacrum. It acts identically to a human. If you produce a gun and threaten to shoot it in the head, it will fall to its knees and beg for its life with tears of terror flooding down its cheeks. You may have no qualms about unplugging the China Brain, but could you pull the trigger on something so obviously afraid to die? Maybe you could, but if you accept it would be even slightly harder, you are accepting that the human appearance and behaviour is affecting your judgement.

You might counter this by arguing that, while zombies and simulacrums might act as if they are afraid to die, they are not really, as they do not really have thoughts, emotions or beliefs. However, there is certainly something going on. If I believe there is coffee in my mug, I raise it to my lips and drink. My zombie twin is behaviourally identical, so would do the same thing in the same circumstances. There must be something that is serving the same functional purpose as a belief, even if it's not a belief as such. One might call it a zombie-belief.

However, if we accept Functionalism as a theory of mental states, we can go further. Based on Crawford [2005] Functionalism says that mental states are generally i) caused by sensory input, ii) cause behavioural output, and iii) cause and are caused by other mental states. My zombie has the same sensory inputs as me, displays the same behavioural outputs as me, and his equivalent zombie-states are caused and are caused by other zombie-states (his zombie-desire for coffee combines with his zombie-belief that coffee remains in the mug, and causes the zombie-intention to drink from the mug). Therefore zombie states are mental states. Thus, if you threaten to kill my zombie-twin, he really will be afraid to die. Can we conceive of him being afraid to die, if there is nothing that it is like for him to live? I suggest that this again hints at logical incoherence.

I have examined a number of difficulties in the conceiving of zombies. While I don't pretend that any are knock-down arguments, I think that they suggest significant worries about their logical coherence and cast significant doubts on conceivability.

I therefore conclude that no, zombies are not conceivable.

Bibliography

• Chalmers, D.J. (1995), "Facing up to the problem of consciousness", in Journal of Consciousness Studies, 2 (3) pp 200-19
• Chalmers, D.J (1996), The conscious mind: In search of a fundamental theory, reprinted in Frankish (2005).
• Cottrell, A (1999), "Sniffing the Camembert: on the conceivability of zombies", in Journal of Consciousness Studies, 6 (1), pp. 4-12.
• Crawford, S (2005), Aspects of Mind, (Open University)
• Dennett, D.C. (1995), "The unimagined preposterousness of zombies", in Journal of Consciousness Studies, 2 (4), pp. 322-6.
• Frankish, K (2005), Consciousness, (Open University)
• Lewis, H.A. (1998), "Consciousness: Inexplicable and useless too?", in Journal of Consciousness Studies, 5 (1), pp 59-66
• Moody, T.C. (1994), "Conversations with zombies", in Journal of Consciousness Studies, 1 (2), pp. 196.200.