Your complimentary articles
You’ve read one of your four complimentary articles for this month.
You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please
Articles
The Problem With Zombies
Rebecca Hanrahan says that just because you can imagine zombies doesn’t mean they’re possible, or that they can tell us anything about consciousness.
We know all sorts of things. We know whether it is raining, whether the computer is on or off, and whether the kettle is boiling. And putting aside the worries of the skeptic, we understand how we came to have such knowledge. Our senses informed us of these facts. But not only do we know these sorts of empirical facts, we also have knowledge of modal facts: facts about possibilities. We know that even if it is raining, it is possible for it to have been a sunny day. We know that even if the computer is on, it is possible for it to have been off.
But how do we come to have knowledge of modal facts? It isn’t through empirical investigation. The scope of empirical investigation is the here and now, the actual world, not possible worlds. So, here is a dilemma. We clearly have some knowledge about what is possible, yet how we usually acquire knowledge – empirical investigation – won’t tell us about these possibilities. How then do we know what is possible?
Many philosophers hold that conceivability tells us what is possible. For example, right now I am typing away at my computer with my cat sitting on my lap. My computer is working just fine, but I know it is possible for it to crash. How do I know t his? Philosopher’s answer: because I can imagine a series of events that would lead to a systems meltdown. For example, I can imagine my cat being spooked by my dogs, jumping onto my keyboard, and in his frenzy striking just the right combination of keys to overload my antiquated system. If we endorse this line of reasoning, then we are endorsing the conceivability principle: the principle that conceivability is a guide to possibility.
The conceivability principle is a foundational principle in analytic philosophy. It has been employed in every field, to support disparate and controversial positions. My interest here is in exploring the way it has been used to resuscitate mind/body dualism. I maintain that this resurrection of dualism comes at a great price. Specifically, I will argue that dualism can be salvaged via the conceivability principle only if we are willing to embrace solipsism. If we want to reject solipsism, and instead hold that we are justified in our belief that each of us isn’t the only conscious life that has ever existed, we will have to concede that conceivability is limited in such a way that it can’t provide us with a demonstration of dualism.
First, I will explain how the conceivability principle has been used to argue that not all mental properties are physical properties. This is where zombies are relevant. Next, I will consider the notion of conceivability used in this discussion of dualism and argue that it doesn’t provide us with a guide to possibility. Then I will show that even if conceivability does provide us with a guide to possibility, it is limited by our acceptance of inferences to the best explanation in multiple alternative contexts. Finally, I will argue that because we are so limited, we will only be able to endorse dualism if we are also willing to endorse solipsism.
Chalmers, Zombies and Dualism
In the last few years dualism has enjoyed a revival, due to David Chalmers’ 1996 book The Conscious Mind: In Search of a Fundamental Theory. Chalmers argues for dualism and against materialism. Materialism is the position that there is only one kind of stuff and/or property in this world, and it is physical, that is, material, in nature. According to the materialists, everything that exists exists in space and time. All there is can be measured or weighed. If the materialists are right, then everything in this world, including us, can be accounted for by science, at least in principle. Our scientists will explain not only how plants change light into energy, but will also unlock the nature of consciousness.
In its most basic form dualism is the position that this world is constituted by more than the physical. According to dualism there is another kind of stuff or property: hence no purely physical explanation of this world could fully account for all that is in this world. Specifically, a physicalist explanation could never account for us, for we are special kinds of creatures: we can’t be reduced to our biological, chemical, or other physical properties. For there is something about us that enables us to feel, love and experience the world. To explain it, we need to go beyond the physical, beyond what can be measured or weighed, and posit another kind of stuff or property – which we call the mind or the mental, or even the soul.
Dualism and materialism are in opposition to each other. If materialism is true, dualism is false, and vice versa. Again, Chalmers’ goal is to undermine materialism and support dualism. To that end, he offers the following argument:
1. We can conceive of a world populated by our zombie twins.
2. Conceivability provides us with a guide to possibility.
3. It is possible that there is a world populated by our zombie twins.
4. If this is possible, materialism is false.
5. If materialism is false, then dualism is true.
Conclusion: There is more to this world than the physical. There is in addition some other kind of stuff or property that accounts for our ability to have conscious experiences.
How can the possible existence of zombies prove that dualism is true? To see how this argument works, note that the kind of zombies at issue here aren’t the George Romero, Hollywood type of zombies. They aren’t the undead. Their flesh isn’t rotting, nor do they particularly enjoy the taste of human meat. Rather, they are phil-zombies.
Phil-zombies are just like us in almost every way. They have our physiology. They look like us on the outside as well as on the inside. Their brains and bodies are structured just like ours, and are composed of the same kind of material. Because phil-zombies share our physiology, they will also share our psychology: Their brain processes will be just like ours, hence they will believe, think, and act just like us What they lack is our phenomenology. They lack all conscious experiences of any kind. (Note that we usually speak of psychology in such a way that it includes the phenomenological aspects of our mental lives. Chalmers, though, separates the phenomenological (how stuff feels) from the psychological (how information is processed and behavior controlled).)
Consider me in comparison to my phil-zombie twin. My twin is a molecule for molecule replica of me who functions just like me in every way. Thus, if you ask either me or my twin whether we like chocolate, we will both answer in the affirmative. If you put before either of us a plate of chocolate chip cookies, we will each immediately eat two and then hide two for later. If either of us is asked to wax poetic on the wonders of chocolate, we will offer up the same verse – and that verse will begin with disparagements directed at white chocolate. Again, my phil-zombie twin and I will look the same, behave the same, and we will think the same things. How then do we differ from each other?
Upon consuming a piece of chocolate, I have a series of sweet, smooth, and rich sensations. But when my twin eats chocolate she does not have these luxurious sensations, for she never experiences anything at all. It isn’t that the chocolate tastes or feels differently to my twin compared to me. Rather, eating chocolate does not produce any experiences of any kind whatsoever in my twin: but she will behave as if it does. Her eyes will close seemingly in delight as she takes the first bite, even though she won’t be experiencing anything, let alone delight. In addition, even though eating this chocolate won’t produce any experiences in my twin, she will believe that it does. She will believe that she enjoyed its rich taste. She will believe that she is now more relaxed as a consequence of her having this tasty treat. But she will be wrong. This lack of experience isn’t just associated with chocolate. It is associated with every aspect of my twin’s life. My phil-zombie twin never feels, experiences, or senses anything.
So we now know what kind of zombies are at issue here. The question is, how would the possible existence of phil-zombies establish that dualism is true? Whatever my doctor, neurologist, or therapist discovers about me, is true of my twin. But yet there are certain propositions whose truth applies to me and not to my twin. Consider this proposition:
‘The coffee tasted bitter this morning’
My twin’s brain works just like mine does when processing information. For this reason, we share the belief that the coffee tasted bitter. But yet when my twin drank the coffee she had no experience of bitterness, but I did. I drank the two-day-old coffee and I had a profound experience of bitterness. Thus, this proposition is true of me, but not of my twin. Hence, we must differ from each other in some way. But in what does this difference lie?
By definition this difference can’t rest in the physical. And because it can’t rest in the physical, it can’t rest in the psychological. Thus, this difference must lie in something that has nothing to do with the physical. But if this is so, then there is more to this world than the physical. There must be, in addition, something non-physical that our doctors, therapists, and neurologists can’t get at. But if this is the case, then materialism is false, for materialists hold that there is only the physical.
So we now know why phil-zombies are so important. If there could be phil-zombies, the dualists are right, and we are more than merely bones, flesh, synapses and electricity.
With that, let’s return to Chalmers. His argument needs his third premise. If Chalmers can establish that there could be phil-zombies, creatures that shared our physiology and our psychology but lacked our phenomenology, then materialism is false and dualism is true. But can he get us to concede to premise three? Here is where he has problems.
Chalmers’ Intuitions
Chalmers’ method of conceiving depends on intuitions. He claims in The Conscious Mind: In Search of a Fundamental Theory, p.96 that he has a ‘brute intuition’ of a world which is populated by phil-zombies, and this is enough to establish that such a world is possible. But a method of conceiving based on intuitions could never authoritatively tell us what is possible.
Intuitions are spooky things. They are only accessible to the person who has them. I can’t have Chalmers’ intuitions, and he can’t have mine. But this means that if our intuitions point us in different directions, there will be no way of adjudicating between us. Say for example I claim that I intuit that phil-zombie world is an impossibility. The disagreement between Chalmers and myself might be because one of us misread our intuitions, or because one of our intuitors isn’t working. But how are we to tell which?
One way we might have been able to diagnose and treat the problem would be if simple empirical investigation informed us about possibility; for then we would have independent means of establishing whose modal intuitions are reliable. If more of my modal intuitions were confirmed by empirical investigation than Chalmers’, then we should trust my intuitions about phil-zombie world over Chalmers’. But because empirical investigation doesn’t provide us with a guide to possibility, we have no way of determining whose modal intuitions are reliable.
The problem isn’t just that we can’t settle our modal disputes: the deeper problem is that we have no way of confirming our modal intuitions. We can’t test them against the intuitions of others, nor can we test them against the reports of our empirical investigations. But if we can do neither of these things even when our modal intuitions agree, we can never be sure that we are intuiting correctly, or that what we are intuiting is a reflection of the possible. Our intuitions might instead be reflections of our prejudices. If this possibility can’t be ruled out, then intuitions can’t be the basis for a system of knowledge.
Chalmers claims that he can conceive of a world populated by phil-zombies, and hence he concludes that it is possible for there to be such a world. But what it means for him to conceive of this world is just that he intuits its possibility. But intuitions can’t be at the foundation of any justification of what is possible, so Chalmers isn’t justified in concluding that the phil-zombie world is possible.
Thus, Chalmers hasn’t yet established premise three of his argument. But might there be another way to establish this premise? Next I will argue that the only way we could justifiably claim that there could be a world populated by phil-zombies is if we endorse solipsism with regards to the actual world. But if instead we reject solipsism – that is, if we hold that we are justified in our belief that other conscious minds surround us – we must concede that Chalmers’ premise three can never be justified. To see why, instead of considering phil-zombies, let us turn our attention to Hollywood zombies.
The Eating of Flesh
Imagine you were confronted by a humanoid who feasted on the living flesh of human beings. Would you be tempted to think that it was a phil-zombie? Would you think it lacked phenomenological experiences? No! Instead you would believe that, like you, it experienced the world; but unlike you, it found the taste of human flesh to be very pleasing, and hence it felt the overwhelming compulsion to consume such meat.
In this belief, Chalmers concurs. Chalmers says of Hollywood-style zombies that “it is reasonable to suppose that there is something it tastes like when they eat their victims.” (The Conscious Mind, p.95.) But why is such a belief about these creatures reasonable? My position is that this belief is deemed reasonable because it is a product of our best explanation.
There are many different conditions under which we deem a belief reasonable. One of them is when the belief is generated via an inference to our best explanation. Each of us has constructed an explanation as to how the world works, and we continually modify it to better explain our world. My best explanation (as well as Chalmers’) includes the idea that mammals consume nutrients because they are compelled by their desires to do so. Thus, if I see some animal eating victuals, I infer via my best explanation that it is because it is feeling either hunger or pleasure. Therefore it is reasonable to conclude that Hollywood zombies enjoy the taste of human flesh, because such a belief is licensed by an inference to our best explanation.
Note two points here:
1.) Our best explanation endorses the claim that we are surrounded by other conscious beings who enjoy phenomenological experiences. This best explanation further endorses the claim that the behaviour of others is good evidence as to the nature of those experiences.
2.) Even though we construct our best explanation with respect to this world, we can employ it in alternative contexts. There aren’t any zombies in this world. But we can imagine a world populated by flesh-eating humanoids and then employ our best explanation in the context of this world to generate the reasonable belief that such creatures enjoy the taste of human flesh.
But together this means that there could be no context in which we would be justified in concluding that the creature before us was a phil-zombie.
Why? Imagine that you are given the power to travel to any world to determine if it contained phil-zombies. Imagine that during your interdimensional travels you do find creatures that look like us and act like us. But when you do, you will employ your best explanation to account for their behaviour. So, when you see someone who looks and acts like me eating a chocolate truffle, reason will demand that you conclude that she is very much enjoying some chocolatey sensations. You might be wrong: but reason cannot reveal your mistake. Thus, no matter how many worlds you visit, you could never conclude that the creature before you was a phil-zombie – even if in fact it was.
In sum, we think it is reasonable to conclude that Hollywood zombies enjoy the taste of human flesh. Even Chalmers concedes this. But this demonstrates that the conclusions we draw even in the most outlandish situations are limited by our best explanations. Thus, even if conceivability could provide us with a guide to possibility, our being limited by our best explanation will preclude us from ever concluding that we have conceived of a phil-zombie. We can imagine a creature that looks like us and acts like us; but once we have, our best explanation demands we attribute our phenomenology to that creature. So in fact Chalmers’ argument fails at premise one: he doesn’t even get to three. We can’t conceive of a world populated by phil-zombies; hence we aren’t justified in believing that such a world is possible.
Some might raise an objection here: best explanations change. Might we not have call to modify our best explanation so that it could license the conclusion that the creatures before us are phil-zombies? Note, though, how our best explanation would have to change to justify this conclusion. We would have to change it so that it endorses the claim that some creatures who look and act just like us lack conscious experiences. But to modify our best explanation in this way is to endorse solipsism.
To see this, imagine that in some world you find creatures you suspect are phil-zombies. These creatures have our physiology and our psychology. Hence, in every way, they look and act like us. You are now wondering whether they also have our phenomenology.
Say you have discovered something about them which forces you to deny that their behaviour is evidence that they possess conscious experiences. What could you have discovered? Given that we have no direct access to another being’s phenomenology, the only thing that you could have discovered is some physical or behavioural fact about these creatures. But this evidence, whatever it may be, will be equally true of us as of them, by definition. So, you could only deny the link between mind and behaviour with respect to these creatures if you were equally willing to deny this link with respect to us real world humans. But to deny this link with respect to us is to embrace solipsism. It is to embrace the position that how others act isn’t evidence as to whether they feel anything, and hence, there is no reason to think that there are any conscious beings other than oneself.
The End Is Nigh
To close, let’s return to the issue of dualism. At most, what Chalmers has shown is that if we could justify our belief in the possible existence of phil-zombies we would have reason to reject materialism and endorse dualism. But I have shown that we can only justify our belief in the possibility of phil-zombies by embracing solipsism. If we reject solipsism, we have to concede that we would never be justified in believing that phil-zombies are possible. Thus Chalmers’ argument for dualism fails. Of course, this doesn’t mean that dualism is false: it only means that Chalmers’ zombie argument gives us no reason to endorse it.
This is a negative conclusion. Is there something positive that can be taken away from this discussion? I began this article by declaring that empirical investigation doesn’t provide us with a guide to possibility. But one thing we’ve learnt is that what empirical investigation does produce, ie our best explanation, is relevant when it comes to justifying our beliefs about what is possible. Via our best explanation, we can generate reasonable beliefs even in contexts outside the actual world. This, as I argue elsewhere, is an insight on which an entire epistemology of possibility can be built.
© Rebecca Hanrahan 2008
Rebecca Hanrahan is an Assistant Professor of Philosophy at Whitman College in Walla Walla, Washington, U.S.A.