Photo by Angus Mordant/Bloomberg/Getty


The philosopher’s zombie

The infamous thought experiment, flawed as it is, does demonstrate one thing: physics alone can’t explain consciousness

by Dan Falk + BIO

Photo by Angus Mordant/Bloomberg/Getty

In his book Until the End of Time (2020), the physicist Brian Greene sums up the standard physicalist view of reality: ‘Particles and fields. Physical laws and initial conditions. To the depth of reality we have so far plumbed, there is no evidence for anything else.’ This physicalist approach has a heck of a track record. For some 400 years – roughly from the time of Galileo – scientists have had great success in figuring out how the Universe works by breaking up big, messy problems into smaller ones that could be tackled quantitatively through physics, with the help of mathematics. But there’s always been one pesky outlier: the mind. The problem of consciousness resists the traditional approach of science.

To be clear, science has made great strides in studying the brain, and no one doubts that brains enable consciousness. Scientists such as Francis Crick (who died in 2004) and Christof Koch made great strides in pinpointing the neural correlates of consciousness – roughly, the task of figuring out what sorts of brain activity are associated with what sorts of conscious experience. What this work leaves unanswered, however, is why conscious experience occurs at all.

There is no universally agreed-upon definition of consciousness. Awareness, including self-awareness, comes close; experience perhaps comes slightly closer. When we look at a red apple, certain neural circuits in our brains fire – but something more than that also seems to happen: we experience the redness of the apple. As philosophers often put the question: why is it like something to be a being-with-a-brain? Why is it like something to see a red apple, to hear music, to touch the bark of a tree, and so on? This is what David Chalmers called the ‘hard problem’ of consciousness – the puzzle of how non-conscious matter, responding only to the laws of physics, gives rise to conscious experience (in contrast to the ‘easy problems’ of figuring out which sorts of brain activity are associated with which specific mental states). The existence of minds is the most serious affront to physicalism.

This is where the zombie – that is, the thought experiment known as the ‘philosopher’s zombie’ – comes in. The experiment features an imagined creature exactly like you or me, but with a crucial ingredient – consciousness – missing. Though versions of the argument go back many decades, its current version was stated most explicitly by Chalmers. In his book The Conscious Mind (1996), he invites the reader to consider his zombie twin, a creature who is ‘molecule for molecule identical to me’ but who ‘lacks conscious experience entirely’. Chalmers imagines the case where he’s ‘gazing out the window, experiencing some nice green sensations from seeing the trees outside, having pleasant taste experiences through munching on a chocolate bar, and feeling a dull aching sensation in my right shoulder.’ Then he imagines his zombie twin in the exact same environment. The zombie will look and even act the same as the real David Chalmers; indeed:

he will be awake, able to report the contents of his internal states, able to focus attention in various places, and so on. It is just that none of this functioning will be accompanied by any real conscious experience. There will be no phenomenal feel. There is nothing it is like to be a zombie.

Imagining the zombie is step one in the thought experiment. In step two, Chalmers argues that if you can conceive of the zombie, then zombies are possible. And finally, step three: if zombies are possible, then physics, by itself, isn’t up to the job of explaining minds. This last step is worth examining more closely. Physicalists argue that bits of matter, moving about in accordance with the laws of physics, explain everything, including the workings of the brain and, with it, the mind. Proponents of the zombie argument counter that this isn’t enough: they argue that we can have all of those bits of matter in motion, and yet not have consciousness. In short, we could have a creature that looks like one of us, with a brain that’s doing exactly what our brains are doing – and still this creature would lack conscious experience. And therefore physics, by itself, isn’t enough to account for minds. And so physicalism must be false.

The zombie argument has recently been taken up by Philp Goff, who explores it in his book Galileo’s Error (2019). Once again, the issue isn’t whether zombies are actually walking among us, but rather, whether they could exist. Goff writes:

Nobody thinks that philosophical zombies exist, any more than they think flying pigs exist. But there is no contradiction in the idea of a zombie, and hence if our universe had been very different, perhaps if the laws of nature had been different, there could have been zombies roaming our planet.

In other words, it’s not just a question of what one can imagine; people can imagine all sorts of implausible things. As Goff put it to me during a recent Zoom call: ‘The question is, are they logically coherent, and ultimately, are they possible in this very broad sense of possibility.’

At first, this seems like a powerful argument. If you believe that zombies could exist, you’re forced to accept the possibility that matter-in-motion can’t explain everything. In particular, the thing we hold most dear – our actual experience of the world – is missing. And so physicalism falters.

Even those who aren’t swayed by the zombie argument acknowledge its intellectual allure. ‘It’s elegant because it’s a very simple argument,’ says Keith Frankish, a philosopher with appointments at the University of Sheffield and the University of Crete. ‘It seems like you can get to a really big conclusion – a big radical conclusion – from a couple of fairly straightforward and attractive premises. That’s the dream of philosophers – to have these revolutionary arguments on the basis of premises that you can ascertain just in your armchair, just by thinking about it … If that isn’t seductive, I don’t know what is.’

As one begins to dissect the zombie argument, however, problems arise. To begin with, are zombies in fact logically possible? If the zombie is our exact physical duplicate, one might argue, then it will be conscious by necessity. To turn it around: it may be impossible for a being to have all the physical properties that a regular person has, and yet lack consciousness. Frankish draws a comparison with a television set. He asks if we can imagine a machine with all the electronic processes that occur in a (working) television set taking place, and yet with no picture appearing on the screen. Many of us would say no: if all of those things happen, the screen lights up as a matter of course; no extra ingredient is required.

Turning back to consciousness, Frankish adds: ‘I think if you really could understand everything the brain is doing – its 80 billion neurons, interconnected in goodness knows how many billions of ways, supporting an unimaginably wide range of sensitivities and reactions, including sensitivities to its own activity … If you could really imagine that in detail, then you wouldn’t feel that something was left out.’ (At the very least, this objection highlights how careful we have to be when we say that we ‘conceive’ of something. Can any of us really conceive of 80 billion of anything?)

Clearly, a great deal rests on the issue of ‘conceivability’. Sean Carroll, a physicist at the California Institute of Technology who weighed in on the zombie issue in a recent paper, gives an example from mathematics: ‘If you went back 10,000 years and explained to someone what a prime number is, and asked: “Is it conceivable to you that there’s a largest prime number?” Well, they might say “yes”; as far as they can conceive, there could be a largest prime number. And then you can explain to them, no, there’s a very simple mathematical proof that there can’t be a largest prime number. And they go: “Oh, I was wrong – it’s not conceivable.”’

It’s asking us to picture a bird that walks like a duck and quacks like a duck, and yet is not a duck

In a similar vein, geometers long imagined that it might be possible to ‘square the circle’, a task that was eventually shown (in 1882) to be impossible. The philosopher Massimo Pigliucci, reflecting on how once-conceivable things often get demoted to the realm of the inconceivable, has written that ‘conceivability establishes nothing’. At the end of the day, Carroll finds the idea of conceivability too fuzzy to do what philosophers want it to do. ‘I think that conceivability is just a misplaced concept to use in arguments like this,’ he says, ‘because it is leveraging fuzziness to reach sweeping conclusions far beyond what is warranted by one’s state of knowledge.’

A closely related issue is the problem of accepting the zombie thought experiment’s premises at face value. We’re told that the zombie is just like us, and yet lacks consciousness. Let’s put this into practice: we meet a creature that looks and behaves just like a human, but a philosopher assures us that it’s actually a zombie. What would we make of their claim? Rebecca Hanrahan, a philosopher at Whitman College in Washington State, argues that in such a situation we would not, in fact, accept the claim that the creature lacked consciousness. ‘If I go to another world and see a creature that looks like me and acts like me, then I’m going to have to conclude that it also has the same phenomenological sensations that I do,’ she says. In other words, the first premise of the zombie thought experiment never gets off the ground: Chalmers asks us to accept a human duplicate who lacks consciousness as though this is a straightforward request – but it is not. To put it somewhat crudely, it’s asking us to picture a bird that walks like a duck and quacks like a duck, and yet is not a duck.

The zombie argument seems to belong to a class of arguments that Daniel Dennett calls ‘intuition pumps’. These are arguments – typically thought experiments – that lead the reader toward a certain appealing but not necessarily warranted conclusion. (Problems involving the mind and the brain seem to spawn more than their fair share of these problematic thought experiments; a well-known example is John Searle’s ‘Chinese Room’ argument against the possibility of explaining the mind in terms of information-processing; Dennett has shown convincingly where the argument falters.) In the case of the zombie argument, it’s suggested that we can easily picture a creature that has all the outward attributes of a normal, thinking human being, yet is one that lacks consciousness. But it turns out that conceiving of such a creature is no mean feat.

Another problem centres on what consciousness actually does. As a philosopher would put it, what causal role does it play? Does it cause matter to move about? Or to put it another way: does consciousness impact behaviour? By Chalmers’s account, the zombie is supposed to behave exactly like us – even though we have conscious experiences and the zombie doesn’t. The implication seems to be that conscious experiences play no causal role in the world. But in that case, why even postulate its existence? The usual response is that consciousness is something we immediately experience; we can’t be wrong when we claim to be conscious. But when we reach for a glass of water, aren’t we doing so because of the conscious experience of being thirsty? If we are, then consciousness does, in fact, seem to impact behaviour; and if we aren’t, then consciousness seems to be nothing more than what philosophers call an epiphenomenon, a kind of secondary phenomenon. As Hanrahan puts it, consciousness would be like the humming sound that your computer makes – it’s always there when the computer is on, but it has no bearing on what the machine is actually computing.

Carroll’s objections to the zombie argument focus on precisely this point. ‘The zombie concept is only coherent if you think that none of our conscious experiences have any influence whatsoever on our behaviour,’ he says. Goff disputes this point; in Galileo’s Error, he argues that there is ‘no contradiction in the idea that something with the same physical nature [as a human being] could lack an inner subjective life’ [Goff’s italics] and that ‘there is no inconsistency or incoherence in the idea of a zombie.’

Zombies are either inveterate liars or, at a minimum, they’re extremely confused about their condition

The difficulty comes to a head when we look at the things we say about our conscious experiences. If I’m sad, I’ll say that I’m sad – but the zombie, in the same situation, would also say it’s sad (if it didn’t, we’d spot it due to this difference in behaviour). For Carroll, this stretches the argument to its breaking point. ‘If someone says “I’m sad”, and you say “Describe to me your sadness” – well, if you believe in the possibility of zombies, and the conceivability of zombies, then that experience of sadness can’t actually be influencing or informing what you say about your sadness,’ says Carroll. ‘And whatever you think about consciousness, that’s not consciousness as I understand it. When I’m sad or when I’m seeing red or when I’m feeling hot, that influences how I talk and move and behave in the world.’

Again, Goff sees the situation differently. After a prolonged back-and-forth with Carroll on a recent episode of the Mind Chat podcast, hosted by Goff and Frankish, Goff Tweeted: ‘The same software can be run on different hardware, it obviously doesn’t follow that the hardware doesn’t do anything … Likewise, the thesis that human behavioural functions could be realised in non-conscious zombie stuff doesn’t entail that human consciousness doesn’t do anything.’ Carroll replied in a blog post, arguing that, sure, the same computer program can be run on different machines (this is what philosophers refer to as ‘substrate independence’) – but he notes that the substrate doesn’t affect the outcome of the calculations. Analogously, he writes, those who want ‘to differentiate between the software of reality running on physical vs mental hardware cannot claim that consciousness gets any credit at all for our behaviour in the world.’

However one frames the relationship between minds, brains and bodies, there seems to be no getting around the problematic nature of the descriptions zombies give of themselves: they’re either inveterate liars – they insist they’re enjoying the taste of a delicious apple even though, by the terms of the thought experiment, they’re experiencing nothing at all – or, at a minimum, they’re extremely confused about their condition. And if the zombie is confused about what it is or is not experiencing, perhaps we are too. In fact, with just a little effort, one can enlist the zombies in support of physicalism: however sincerely we might say ‘But I know I’m conscious; I feel it; I cannot be wrong about this,’ we must bear in mind that the zombie would utter the exact same words in the same situation.

Does this mean that consciousness is merely an illusion? Frankish believes it is; he describes conscious experience as ‘a fiction written by our brains in order to help us track the impact that the world makes on us’. Carroll, in his book The Big Picture (2016), takes a slightly different tack; he writes that consciousness is real ‘in exactly the same way as fluids and chairs and universities and legal codes are real – in the sense that they play an essential role in a successful description of a certain part of the natural world, within a certain domain of applicability’. Goff, in contrast, defends a view known as panpsychism – roughly, the idea that everything in the world has mental qualities, or, as he put it (along with two co-authors) in the Stanford Encyclopedia of Philosophy, ‘mentality is fundamental and ubiquitous in the natural world’.

Physicalists who aren’t swayed by the zombie argument are left pondering the question we began with: how, in a purely physical world, do minds arise? In Until the End of Time, Greene – as ardent a physicalist as they come – writes that the existence of minds represents ‘a critical gap in the scientific narrative … We lack a conclusive account of how consciousness manifests a private world of sights and sounds and sensations.’

Centuries from now (decades if we’re lucky), people will no longer speak of the hard problem as a great mystery

A step in the right direction, at least for thinkers such as Frankish, is to view consciousness not as a thing but as a process. Consciousness is something ‘that a very complex kind of organism does’, he says. He cites Dennett, who has pointed out that the cells in your brain are not fundamentally different from the cells in a big blob of yeast. ‘There’s no real difference between them,’ says Frankish; brains don’t contain some extra, special ingredient. ‘It’s just that the cells of a human brain are connected up in a very, very special way, compared to the cells in the bowl of yeast. And it’s what those cells are doing that makes the brain conscious.’

Carroll holds a broadly similar view. As he put it recently in an episode of his Mindscape podcast: ‘I think the world is made of stuff, obeying the laws of physics, and that’s basically it. Except when that stuff comes together to form complicated things, like human beings, there can be new, emergent phenomena that arise, and consciousness is one of those.’ Like Dennett and many others in the physicalist camp, Carroll believes the hard problem will eventually fade away – that is, centuries from now (decades if we’re lucky), people will no longer speak of it as a great mystery. Eventually, we’ll have learned enough about the workings of brains and their billions of neurons, says Carroll, that we’ll just say ‘Well, this is what happens when people have conscious experiences’ – adding: ‘And then the whole problem will just kind of go away.’

While the zombie argument, and the philosophical problems raised by it, may seem like mere pie-in-the-sky exercises that keep philosophers (and a few scientists) up at night, they tie into questions that have real-world consequences. Thinking about zombies forces us to think about how we deal with beings whose status as conscious entities is unclear – such as animals, for example, and foetuses, or some future versions of robots or artificial intelligences.

We all seem to agree that human beings are conscious, but how widespread is consciousness in the animal kingdom? ‘Is my dog conscious? Absolutely,’ says William Seager, a philosopher at the University of Toronto. ‘What about my parakeet? I think so. A rat? Probably. What about a snake, or a spider? Spiders act – they seem to want things. They form plans, they hunt, they seem to like to eat things, and they avoid situations that are dangerous. Are they conscious?’

The question is even thornier when we get to octopuses, which have a far more distributed neural structure than mammals. Since we don’t know exactly what generates consciousness, we struggle to determine who or what has it. Insects, for example, ‘are way simpler than us’, says Seager. ‘But that’s not fair; just because they’re simpler doesn’t mean they’re unconscious. So we have a kind of real-world zombie issue when we think about where consciousness cuts out, or where it turns on.’ Parallel questions inevitably come up when considering human development. At conception, a human embryo ‘is definitely not conscious, and at birth it’s definitely conscious’, says Seager. ‘Somewhere in the middle, consciousness turns on. We don’t really understand how that works. Again, we don’t know what it is about the brain that generates consciousness. So we have these conundrums.’

The zombie argument provokes for the same reason that the larger puzzle of consciousness provokes: it forces us to confront problems that stymied everyone from the ancient Greeks to Descartes and Galileo. Even the most hardened of the hardcore physicalists admit that the puzzle of consciousness is, well, puzzling. The zombie argument, flawed as it is, deserves credit for helping to bring difficult questions into sharp relief, even if it’s not the knock-down argument against physicalism that its proponents imagine it to be.