Let us begin with Descartes, since arguably it is with him that the present story has its origin. In his Meditations on First Philosophy (1641), the great French philosopher and mathematician lays out a thought experiment meant to undermine the supposed foundations of his commonsense beliefs. Let me imagine, Descartes writes, that a powerful deceiver is the source of all my thoughts and perceptions. When it seems that I am opening my eyes and seeing the stove in front of me, my vision is the product of a cunning deception. When it seems that I am raising my arm to scratch an itch, my feeling of embodiment is another of the deceiver’s illusions. And so, too, for my sequences of logical and mathematical reasoning: these are the deceiver’s art. In the face of this hypothesis, Descartes asks, what could permit me to rely on my senses or my reason as ways of knowing how things are?

Along with this skeptical thought experiment, Descartes’s most famous contribution to the history of philosophy is the first argument he offers in response to it, which identifies one’s own existence as the first thing a person is able to know with perfect certainty. I think, therefore I am was Descartes’s formulation of the crucial argument in his Discourse on Method (1637)—and from this it follows that what I fundamentally am is not a human being, nor anything that takes up bodily space. In knowing himself, Descartes knows himself to be a thinking thing, and he knows this prior to his knowledge of anything that is bodily or “extended,” such as his eyes, his arms, and the furniture in the room around him. Indeed, Descartes reasons, I myself could exist as the thinking thing I am, with just the thoughts and perceptions that I know myself to have, even if everything “external” to me and my thinking were imaginary.

In contemporary philosophy, the most compelling reconstruction of Descartes’s skeptical scenario is in a 1981 article by Hilary Putnam, which explores the cognitive situation of a “brain in a vat.” In Putnam’s thought experiment, a human-like brain is sustained by fluid and connected to a computer that feeds it stimulation corresponding to what our brains ordinarily receive from our sensory systems, in a way that is responsive to outgoing signals from the motor system in the brain. This image avoids what is commonly taken to be the central embarrassment of Cartesianism—namely, that Descartes’s “thinking thing” is an immaterial substance that exists outside the influence of physical laws. At the same time, Putnam’s scenario renders in even more literal terms the idea that thinking, perceiving, and acting are internal processes that must be connected somehow to objects in a realm outside them. Descartes’s immaterial soul may have been a metaphysical extravagance, but Putnam’s brain in a vat provides a more scientific-seeming version of the same basic picture—the picture of our mental lives as confined to a special realm. For Descartes, that realm was the domain of the immaterial soul. For Putnam, it was that of the vat.

Descartes’s ideas marked a radical departure from the understanding of human nature that had been laid out by Aristotle and developed by his scholastic commentators. But the radicalness of these ideas is likely to be lost on us, as they have captured the popular and scientific imagination no less than the imagination of professional philosophers and their students. The Cartesian picture of human nature, where the mind is something inner with a problematic connection to things in the “external world,” including our very bodies, is for many people the default way of understanding who and what we are. Putnam’s thought experiment may seem implausible, but don’t most of us already think of ourselves as essentially brains in vats? Only the vats are our human heads, the cables leading in and out are our sensory and motor nerves, and what lies outside is neither a deceiver nor a computer but the “external world” itself, doing just the same work of stimulating our brains that a deceiver, or a computer, or a sufficiently realistic and immersive virtual-reality system could in principle do instead. So it is that many of us fail to see that this picture is only a picture, and that there may be alternative pictures that provide a superior understanding of what we are.

Reality+, the latest book from the philosopher David Chalmers, is an engaging and far-reaching exploration of this Cartesian picture, by way of the hypothesis that our mental lives are the product of a massively complex computer simulation. Its central thesis is that virtual worlds are real worlds, so that even if we are living in a computer-generated simulation, or are ourselves the products of such a simulation, this is no threat at all to our own reality, to the reality of the things we take to be around us, to the truth of most of our beliefs about those things, or to the value and meaningfulness of our lives. The origin of this argument is in Chalmers’s paper “The Matrix as Metaphysics” (2003), which used that famous science-fiction film to mount an updated version of Putnam’s “Brain in a Vat” thought experiment. Most of us watch The Matrix, with its rows of human bodies hooked up to a machine feeding them images of a world nothing like the one around them, and judge this simulated world to be an unreal illusion that its subjects should want to escape. In this reaction we are in agreement with Plato, whose Republic imagines a world of prisoners chained in place and observing a realm of shadows. Anyone who was freed from this bondage and escaped to see things in the light of the sun would be obligated to return to the cave and help bring others to enlightenment. If Chalmers is right, however, this reaction would be misplaced. The shadow world is no less real or meaningful than the world beyond, and there is no deception at all in being confined to it.

How could this be? Isn’t a simulated or “virtual” world necessarily a realm of illusion, and wouldn’t we be deluded if we took its denizens to be real? Chalmers answers no, arguing that the labels “virtual” and “real” should not be opposed in the way this objection assumes. True, the products of a computer simulation are not part of physical reality, but there are other ways that something can be real. For example, consider the possibility that I throw a no-hitter in the World Series, and then contrast it with each of the following: my throwing a World Series no-hitter in the context of a simulated season in a video game, and my merely dreaming that I threw a no-hitter. There is a perfectly natural sense in which the middle possibility, where I play out a simulated season without ever leaving my living room, is one in which I really throw a no-hitter. (For example, if I brag about the exploit to my friends and they ask if I really did it, then I won’t be lying or misleading them if I say yes, as long as it is understood that we are talking about a video game.) Chalmers applies a sophisticated version of this line of argument to the idea that our entire world is a massive simulation. Having rejected the mistake of conflating the real with the physical, we can see that there need be nothing unreal about what happens within a virtual world.

But wait. In order for me to have thrown a no-hitter in a World Series game, many things would have to be true of me that are not—for example, that I had the ability to pitch effectively to professional hitters, and that at some point I stood on the mound in a major-league stadium. None of these things need to be true for me to have thrown a no-hitter only in a video game or only in a dream. Nor, of course, would these things need to be true for me to throw a no-hitter in a massive computer simulation—in which case I myself will have been not a human being on a pitcher’s mound (or in my living room or in bed), but a brain in a vat or a self-conscious virtual avatar made of nothing but digital bits. And doesn’t this show that there is an important difference in how things are with me, and in the nature of the activities I engage in, depending on whether or not I am a living organism who acts in the physical world by means of my bodily movements? Doesn’t it reveal an important difference between, on the one hand, what we are and do in a physical world and, on the other, what we are and do in a world that is merely virtual?

This is a point where Chalmers’s argument relies on his commitment to a fundamentally Cartesian anthropology, and betrays his failure to recognize that Descartes’s picture is at work in the background. Consider the following passage, in which Chalmers discusses what makes a given body “mine”:

I could lose pain or hunger and be unable to eat or drink, but this body would still be my body. My thinking could occur in a Cartesian mind, but this body would still be mine. And it’s not so obvious that the physical body is the locus of my existence. I could transplant my brain to a new body, or upload myself to the cloud, and exist without the old body. So it’s arguable that, like my avatar [in a computer simulation], my physical body is not quite the same as me.

This line of argument is supposed to show that if I throw a pitch in a virtual baseball game, I myself will be located on a pitcher’s mound no less and no more than if I do so in a physical game. According to Chalmers, what’s true in both cases is that “my body,” whether it’s a physical thing or a digital avatar, will be standing on a physical or virtual mound, and that my body will not be the same as me. This, of course, is exactly the conclusion of Descartes’s second Meditation, argued for in very much the same way.

Beyond the similarity to Descartes’s arguments, it is striking how each description Chalmers gives of what “could” happen is entirely unquestioned and unsupported. It is an observable fact that human beings can go into persistent vegetative states in which they lack consciousness and cannot eat or drink voluntarily. By contrast, the supposed possibility that human-like thinking could occur in immaterial minds or be transplanted via our brains into new bodies or uploaded into computer systems is so far only a philosopher’s fantasy. And outside the confines of this fantasy we have no way of saying what would happen if any of these “coulds” became real. Imagine that my brain is transplanted into a different body, which then begins to think just as I do. What reason is there to think that the resulting person would actually be me, rather than that a mental duplicate of me would have been brought into existence? The problem is even more severe in the case where “I” am supposed to have been uploaded to a digital cloud. Given that multiple versions of “me” may be created simultaneously in this way, why should any one of them have a claim to identity with the original? The answer, in each case, is that the only reason for saying what Chalmers does is that we are simply taking Descartes’s picture of the mind for granted.

Reality+ is in many respects a stunning success. It is well written, cleverly illustrated, and packed with useful distinctions and powerful arguments. It makes excellent use of both history and contemporary culture to help the general reader understand its key concepts. And it does all of this without sacrificing any of the rigor one would expect in an analytic philosopher’s treatment of these matters in an academic journal.

Still, I also find the book to be a revealing illustration of how one of the dominant contemporary approaches to philosophy can fall short of what the discipline demands. As long as our philosophical thinking is tethered only to fantasy or the farthest edges of scientific speculation, its conclusions will tend to reflect our casual and culturally contingent presuppositions. And as long as this work proceeds without awareness of the philosophical pictures that our thinking relies on, we will be unable to challenge the accuracy of those pictures, or search for potential alternatives to them.

Reality+
Virtual Worlds and the Problems of Philosophy
David J. Chalmers
W. W. Norton
$32.50 | 544 pp.