If I seriously believe that there is no "I" within me who exercises free will and consciously, deliberately chooses, how can I decide what to do? The answer is to trust the audience called the meme. The selections of genes and memes will determine action, and one need only accept that there is no need for an additional "I" to intervene. If I want to live honestly, I should just step aside and let the decisions make themselves.
I say the result is unsettling. At first, it is strange to observe whether some action is being performed or not. I had two possible routes home: a main road, and a prettier road that took longer to reach the cabin. Driving up to the junction, I would often be frustrated by indecision. How could I decide? Which would be most satisfying? Which would be best? One day it suddenly hit me that the "I" did not need to decide. I sat back and paid attention. When the light turned, the foot took the pedal and the hand changed gears. The choice was made. Certainly I never crashed into a stone wall or another car. And whichever road I took, it was fine. Over time I noticed that more and more decisions were like this. Letting that many decisions just happen brought tremendous liberation.
However, "NeuroPath" does not offer us such an easy escape hatch.
179
Because Neil claims to have actually achieved a state of being without the illusion of a responsible selfhood, Neil embodies the riddle of performative contradiction. "In the end, you still experience something," Neil says. "Just an extremely different experience, an experience much more sensitive to the fragmentary truth of our soul, an experience without will, purpose, selfhood, right or wrong." At first glance, Neil seems to have "gone past the obscured… he thinks he has seen his way through the illusion of consciousness." In that state, Neil no longer operates from "the perspective of motive, goal, reason." Rather, as Neil himself describes: "I shut down certain performance-inhibition circuits… that nonsense you psychologists call anxiety, fear, and so on. That nonsense is now nothing more than a memory to me. But I shut down some more deceptive circuits as well. For example, I know I do nothing by will. I'm no longer fooled by the idea that 'I' am doing something."
We can think of this state as the rational "impoverishment of selfhood" that Brassier sees as essential to the scientific method, which "makes it possible to operate in the space of reason governed by concepts." We can also regard it as a kind of Zen no-self of the sort that consciousness researcher Susan Blackmore describes.
178 Decognitive
How could it be that "the category of narrative... has been rendered cognitively unnecessary by modern science"? We cannot escape from reasons and explanations. Even more so when what we are trying to explain is the inevitable failure of reasons and explanations. And similarly, there is no way to go beyond the genre conventions of science fiction or detective fiction. Especially when we try to criticize or weaken those genres.
To put it another way, fiction strains to reach an ultimately unreachable point as it explores the performative contradiction. That is, fiction proceeds toward a point that cannot be expressed or dramatized within itself. That point is precisely Neil's own consciousness. What is it actually like to live and experience the world in a way that renders Nagel's question "What is it like?" meaningless? Even when Thomas was strapped into the puppet machine, "Thomas was, on the whole, awestruck — he even felt a kind of reverence. What would it be like to advance, with a plan indistinguishable from impulse, raising another incident in the shipwreck of a thoughtless world without a self or a conscience? What would it be like to act not as something pitiful and broken like a human being, but as a vehicle without a self, the bearer of all that has come before?" The novel forces us to confront this question, but at the same time forbids us from answering it.
Chapter 4 — Thinking like a human being 177
are always frustrated.
Repeating these clever maneuvers, NeuroPath becomes a kind of anti-detective novel. The detective genre operates by moving through a series of particular hypotheses or explanations, each of which is initially proposed and then later loses credibility. Each failed hypothesis leaves room to be replaced by a better one, and this continues until the case is resolved at the end of the novel. As Sherlock Holmes said: "After eliminating the impossible, whatever remains, however improbable, must be the truth." But NeuroPath drives this process into a malignant infinity. The novel never arrives at a definitive explanation, and so it weakens the very tendency to attribute explanatory meaning to what is happening. We do not know to the very end what motivates Neil's actions, nor even whether to say that Neil is persecuting Thomas, or rather that he is trying to enlighten him. The lesson we have learned through Argument A itself is that such matters should never be explained from the standpoint of psychological motivation.
"There are no reasons," the novel repeats again and again. "There are only causes." But at one point, as Neil reminds Thomas, the novel also says: "Reasons can be a fraud… but they still function." In other words, NeuroPath does not eliminate narrative entirely; rather, it leaves the failure of narrative expectations.
176 Decognitive
As we cannot agree with Argument A through any "game of giving and demanding reasons." Argument A is acknowledged — if "acknowledged" is still the appropriate word — only when Neil has actually established it by putting Argument A into practice; only when Neil has violently forced Argument A onto Thomas, who cannot accept Argument A in any straightforward way. Brassier argues that when we are "bound" by a rational scientific claim, "this binding is not something passively submitted to by the object, but something voluntarily performed by the subject." But NeuroPath describes a process that has no room for that voluntariness. Put differently, we cannot eliminate narrative (as Brassier hopes). It is only through narrative that we can reach the point at which narrative's own illusion of establishing and stabilizing meaning is undone.
Rather than philosophically overcoming or reducing the performative contradiction, we need to embrace our performative contradiction. Properly contextualized, performative contradiction is essential, even unavoidable, and indeed necessary. Let us call this a kind of expansive or paradoxical pragmatism. The only way to realize the implications and consequences of an idea is to enact it (and that, as William James suggests, is the place where an idea finds its "truth"). Within a present situation that includes our cognitive faculties themselves, the idea demands that we conduct a kind of experiment on ourselves, and as a result we obtain it.
174 Decognitive
are completely blind to deeper processing… The neural correlates of consciousness do not have access to the real neurophysiological power-holders that lie at depth.
In this respect, we are in a worse situation than the mythical prisoners in the cave of Plato's Republic. For the prisoners in the cave at least had the slim hope of becoming aware of their plight and moving toward the light. But we are never freed from our illusions even when we come to recognize that our experience is illusory. We can never see "the shadows lying behind the obscured frame." That is, the very "frame" that limits my awareness is itself "obscured," so I cannot realize that my experience is in fact limited and partial. Because I cannot perceive the boundaries of my experience, I cannot grasp that my experience is, far from being comprehensive, limited. This is why, for example, the novel observes that my "field of vision" seems to "just slip away without going through any visible edge."
Furthermore, Argument A weakens our common intuition about our own inner senses, or what philosophers call qualia. I, like many people, tend to think there is a certain vividness and intensity to my inner life. And yet this qualitative dimension of experience is something I cannot grasp and put into words. For example,
Chapter 4 — Thinking like a human being 153
argues.
Even if not entirely so, then largely so: within a network of media technologies that operate predominantly outside the human modes of awareness (consciousness, attention, sensory perception, etc.), human complex entanglements are bringing about fundamental transformations in present human experience.
Argument A in NeuroPath suggests that this is not only about computerized microsensors and other forms of what Hansen calls "21st-century media," but applies broadly from the perspective of the brain's relationship with itself. "The brain is simply not set up to track itself… It lacks the processing capacity… At best, what the brain can do is draw a cartoon about itself." Argument A says that the everyday life we experience is no more than such a "cartoon." Inevitably, we live in "Disney World." A world that "covers vanity with comfortable vanity… and settles on psychological needs rather than physical facts." This is inevitable for the following reasons.
Most of the brain's processes lie outside the experiential. They do not exist for consciousness, nor even as something absent from consciousness… Our brain… cannot follow the brain that leads it
152 Decognitive
even down to chemical events.
Therefore our psychological self-explanations are best understood as fictional narratives, or confabulation. Here you may recall Benjamin Libet's famous experiments on "decisional precursors." These experiments are a frequent topic in both affect theory and the philosophy of mind. And they are explicitly mentioned at one point in NeuroPath. Libet found that the "readiness potentials" associated with a subject's intended action are established in the neurons of the brain 0.5 seconds before the subject consciously decides to perform that action. In effect, our decisions have already been made — for us, or at least by non-conscious processes within us — before we became aware of making such decisions. Here the third person is one step ahead of the first person. Mental events that cannot be accessed through first-person introspection can be measured and recorded by the scientific tools that objectify them. Under such conditions, my sense of "exercising free will" is just a self-defeating attempt to attribute to myself a cause for an event that has already taken place in my brain. To borrow the novel's words: "Will is… a kind of addendum, something that comes to us only after the fact." It is hard to fully grasp "how there is conscious experience after the fact."
This temporal lag is decisive. It means that our minds can never know themselves. The media theorist Mark Hansen
Chapter 4 — Thinking like a human being 151
here, is both neurobiological and sociohistorical. Brassier warns that human subjectivity no longer offers a refuge to slip out from science's relentless demystification and disenchantment of the world.
In fact, Argument A in NeuroPath says that things are even more extreme. Recent research in neurobiology and cognitive psychology shows that we have no conscious access to most of the neural processing that takes place in our brain. This is why our actual awareness is so misleading and incomplete, prone to illusion and error. This is even more thoroughgoing than what Freud imagined. Attempts to self-diagnose through introspection are at best incompetent, and at worst delusional. Most of what we think about ourselves is biased and inaccurate. Argument A in NeuroPath pushes the negative results of this recent research as far as the most provocative consequence possible: not merely in fact but in principle, the human mind cannot understand itself at all. Things like intentions, meanings, and purposes "only seem real because we are not looking at neural processing." And because we are inevitably committed to artifacts like meaning or purpose, the actual material causes of our thoughts and actions remain inevitably invisible to us. Structurally, we can only chase our own mental states without producing the brain's electrical
150 Decognitive
use weather satellites to track storms instead of relying on Thor.
The aspect of Argument A in "NeuroPath" aligns well with the naturalistic current in contemporary philosophy. Speculative realist philosopher Ray Brassier develops a similar point, drawing on both continental and analytic philosophy. According to Brassier, due to the success of physics over the past several centuries, the world's "intelligibility
has been peeled away from the question of meaning. With modern science, conceptual rationality has cut itself loose from the narrative structures still pervasive in theology and theologically-tainted metaphysics. There is no author of the book of the world, and no encrypted story hidden in the fabric of reality. Nature does not unfold any story."
Such observations are no longer even matters of debate. Max Weber's thesis on the disenchantment of the world has become our manifest condition. The "contemporary culture" described in NeuroPath has long "accepted" the meaninglessness of natural events, the fact that they are indifferent to all that is human.
But what happens if we apply the scientific method not just to the physical world around us, but to ourselves, especially to our minds? This is a question that we are still wavering on. As Argument A claims, if naturalism is true,
147
tells the story of Thomas Bible, a 40-year-old psychology professor. Thomas is the author of "Across the Dark Brain," a book that puts forward a revolutionary new theory about the human mind. The book was so extreme that it failed to make a great impression on the public. "The reviews were harsh. The book quickly went out of print." But everything in the novel focuses on Thomas's theory, referred to throughout as "Argument A." Although NeuroPath is science fiction, this is not just because the novel is set in the near future. More importantly, as the author's note for "Across the Dark Brain" explains, the Argument A presented in the book is "extrapolated from real trends and findings in neuroscience, psychology, and cognitive science." These trends and findings all point to the uncomfortable truth that "we are not what we think we are."
Argument A starts from the observation that science always understands things "in terms of quantity and function instead of quality and intent." In fact, "whenever science encounters intent or purpose in the world, it eliminates them. The world that science describes is arbitrary and random. Everything has countless innumerable causes, but nothing has reasons." Physics was a war machine, a weapon of mass destruction. As a result of physics' relentless progress over the past few centuries, "science has nearly scrubbed psychology from the natural world." We no longer believe that natural events carry omens or convey messages. We blame the thunder god Thor for an oncoming storm —
146 Decognitive
understand themselves through the terms of intention, desire, purpose, hope? But what NeuroPath reminds us of is, in fact, the following: "every thought, every experience, every component of consciousness is just the product of various neural processing." You may think you actively chose to do something. But "when we look at the facts — unfortunately, when we look at 'fact,' not 'speculation' — the brain has merely processed a series of sensory inputs… and then generated a particular behavioral output." The brain is "a machine that generates behaviors that are repeated or not repeated depending on how the resulting environmental feedback stimulates the pleasure or pain systems." We like to imagine ourselves as free, rational agents. But actually (to quote one of the most caustically cynical passages of the novel), "we are merely puppets of flesh fooled into believing that we live in a moral and meaningful world."
Brassier makes a similar point. As Nietzsche suggested in the 19th century, and as the existentialists held in the mid-20th century, the claim that today "it is human consciousness that gives nature itself the meaning that nature lacks" is no longer tenable. This is because we can no longer grant our own subjectivity a special status. "The very meaning produced by consciousness can be understood and explained as the product of an aimless, perfectly intelligible process."
Chapter 4 — Thinking like a human being 149
What is true of every other entity in the universe must also be true of human beings. What we have come to understand about the world must also apply to our own processes of understanding. And in fact, recent advances in neuroscience and cognitive psychology have taught us much about our brain. This has largely been due to new technologies — "functional magnetic resonance imaging" (fMRI), first developed in 1992, which allows us to track brain activity in real time, and "transcranial magnetic stimulation" (TMS), first successfully used in 1985, which allows us to influence specific parts of the brain to change a person's feelings, attitudes, and judgments. Through these technologies and advances in computational power, we have arguably learned more about the physical functioning of the brain in the last 30 years than in all of prior human history.
Of course, the relationship between brain and mind, or between the electrochemical processes of neurons and full subjective experience, remains highly contested. But Argument A in NeuroPath takes it as a given that science is now in the process of expelling psychology from the human world, just as it previously expelled psychology from the natural world. That is, somewhat paradoxically, the psyche itself is rapidly being depsychologized. Even psychoanalysis and deconstruction, which sought to decenter subjectivity, did not really prepare us for this contingency. We unquestioningly believe there are reasons for what we do. "Human beings
148 Decognitive
does not simply extrapolate from real advances in virtual world design and artificial intelligence. Chiang also extrapolates from the sociology of the internet, and from the economic conditions that allow software startups to actually exist today. The "life cycle" that takes part of the novel's title is not just about digients themselves. The phrase "life cycle" also applies to the company that designs, develops, and tries to sell digients. Quite practically, here "lifecycle" means a commercial "product cycle." Even though digients have sentience, they too, like other software, are threatened with obsolescence. Before the novel reaches its halfway mark, Blue Gamma goes out of business. The digients' "customer base" had "stabilized into a small community of enthusiast digient owners, and that alone was not enough to generate sufficient revenue to keep Blue Gamma running." So the company,
distributed a free version of the food-production software for customers who wanted to keep their digients running, but after that, the customers had to do everything at their own discretion.
At this point, most people simply "suspend" their digients, painlessly ending their existence. Because digients are not really alive in the first place, that is what "euthanasia could
are continuously moving forward. In fact, digients may be entities that never "reach maturity." The notion based on the biological model of a developmental plateau need not apply. As long as the digient continues to be in a running state, the digient's personality, too, may continue to evolve at the same rate.
So, "The Lifecycle of Software Objects" leaves open the question of whether digients can come to be "capable of making decisions taking responsibility for (their) future" entirely independently of the humans who care for them. Even by the end of the novel, although people "have spent years raising these digients," the digients are still closer to 'teenage' humans than fully grown adults.
The intelligence of a digient is not different in any fundamental sense from that of an organic being. Chiang maintains this premise and carefully separates the issue of sentience from the issue of life. Digients have the former but not the latter. Just as we can do, digients can feel and sense, and they can reflect on what they have felt and sensed. But digients are not alive; they do not replicate or reproduce themselves. Because they have no hormones, digients are sexless. Whether digients have something like a "survival instinct," Spinoza's conatus, or any sort of drive toward self-preservation,
Chapter 3 — Thinking like an avatar 123
cognitive ability is necessarily limited. Cognition does not simply overpower the world around it. Rather, intelligence is constituted by finding ways to operate immanently within the world and in cooperation with the world's other entities.
Following Bruno Latour's idiom, intelligence can be said to act in coordination with and in alliance with other intelligence. Therefore intelligence is not something absolute but necessarily a matter of degree. The human characters in "The Lifecycle of Software Objects" possess greater flexibility and spontaneity than the Neuroblast digients. Even so, digients display flexibility and spontaneity that far surpass any digital agents that actually exist today. And today's digital agents, in their own way, are more flexible and spontaneous than older non-computational machines.
Chiang's account of AI is not far from the bottom-up, embodied, experience-based behavioral approach to intelligence favored by roboticists like Rodney Brooks. In early computer science, intelligence was usually defined as the ability to manipulate representations and symbols and to derive proper inferences from them. AI systems were therefore organized in a top-down manner, emphasizing propositional logic and the high-speed processing of large data. But by the 1980s this approach hit a snag. Researchers turned instead to connectionism and learning-based strategies somewhat closer to how biological brains actually develop.
120 Decognitive
Heuristics are also highly flexible. This means that a heuristic can simply be transferred from one experiential domain to another. And this transferability itself is the best sign that something like general intelligence really exists. Of course, the wide range of applicability of heuristics also means that when applied too generally or in inappropriate contexts, they tend to mislead us. Scott Bakker's extremely reductionist Blind Brain Theory (which I will discuss in detail in Chapter 4) argues that our intuitions about our own mental affairs cannot be trusted, precisely because "cognition is always heuristic." We no longer have a reliable source on which to ground insight. In other words, human beings are not even remotely close to the rational beings that so many philosophers and theorists have called them. And there is no reason to think that artificially intelligent beings would be more rational than we are.
So in "The Lifecycle of Software Objects," intelligence is heuristic. That at the same time means the intelligence is always finite, situated, and embodied. This is what makes intelligence a matter of overall sensibility rather than a particular cognitive skill. In the novel, mind operates within some particular physical-material context, and is specified as essential to that context. This is so regardless of whether the mind is virtual or biological. Cognitive
Chapter 3 — Thinking like an avatar 119
Hey! What I call overall sensibility is still being debated today. Just as Lewis reduces experience to a particular disposition, many philosophers and cognitive scientists also deny that something like a general, multipurpose intelligence exists. For example, Steven Pinker argues that the human mind consists of multiple "computational modules," each devoted to a particular cognitive task. This picture is not entirely wrong. But I find it doubtful for many reasons. First, even Jerry Fodor, one of the founders of the modular theory of mind, points out that this theory cannot explain how the mind decides which module to call up in any given situation. Second, the notion of a "module" supposed to run a particular algorithm is too formal and too linear to account for the messy way thought actually works, and for the gradual development of mental abilities. Considering the principle that "experience cannot be algorithmically compressed," we need something better than the modular theory to explain the flexibility, spontaneity, and creativity of intelligent behavior. Third, the notion of mental modules embedded in DNA seems to require concrete physical implementation, and thus seems to imply discordant correlations between specific regions of the physical brain and specific functions. That is, it would be a kind of neo-phrenology. The modular theory is too narrow to account for widely distributed processes.
116
Even if copies could be sold cheaply or distributed for free, each digient born of that process would live its own life. Each digient looks at a world it has seen before with new eyes, has its hopes realized or had them dashed, and learns what it feels like to lie and what it feels like to be lied to.
To the extent that digients in the novel are generated and grounded in this way, they must be characterized not by some particular bundle of skills, dispositions, or pieces of knowledge, but by what I call overall sensibility — that is, a general way of being in the world. The novel never shows us things from the perspective of the digient. But from the perspective of Anna, Derek, and the other human beings who encounter the digients, it seems clear that the digients have intentions, goals, preferences, and motives. The digients exhibit a considerable degree of self-awareness. Furthermore, the humans who care for them can converse with them in the way they converse with other humans, and at that level — at least, as one converses with a young child. Everything the digient says and does implies that it has a rich inner life. The novel implicitly demands that we adopt the generous principle that if some being seems sentient, we should accept that being as actually sentient.
115
(in Whitehead's usage) is conflating general possibility with real possibility. Logical possibility, or general possibility, encompasses everything not excluded by logical contradiction. Potentiality and real possibility mean more than this. They mean that there is a clear way of getting from here to there, that between the two there can be a trajectory or a "historical path" (Whitehead). Unlike mere logical possibility, potentiality or real possibility actually exist within the present. As Deleuze says, the possible is "real without being actual," and as Whitehead says, "the future is simply real without being actual."
Speculative extrapolation — or the inquiry into real possibility rather than general possibility — is a key foundation of both real scientific research and science fiction. Meillassoux says, "All science fiction implicitly upholds the following axiom: in the predicted future, it remains possible to subordinate the world to scientific knowledge." The flow of causality remains dominant. And science-fictional extrapolation also operates in this way. As Meillassoux shows by analyzing Isaac Asimov's short fiction, this is still the case in Chiang's fiction.
Overcoming the constraints of extrapolative speculation is not about engaging with science fiction itself, but about participating in what Meillassoux calls "extra-science fiction" — a nearly nonexistent genre. Meillassoux says of the latter genre, "its irregularity is sufficient to abolish science, but
Chapter 3 — Thinking like an avatar 111
considers only how that experience can be inferred. Experience itself is not the issue; only how it is cognized or explained is the issue. If the modernist poet T.S. Eliot once complained that "we had the experience but missed the meaning," all of these philosophers suffer from the opposite problem. These philosophers have figured out all the meaning, but somehow missed the actual experience.
How can that be? When philosophers wrangle over questions of the value and significance of phenomenal experience, the "properties" it has, the "dispositions" it reveals, even the question of whether it "exists" or not, they fail to consider this experience in any terms other than cognitive ones. There is a dimension of experience that philosophical accounts have missed. The reason they cannot help missing it is that this dimension cannot be conceptualized through philosophy. We might say that this missed dimension of experience is the aesthetic. In this sense, aesthetics is the other of cognitive philosophy. As Kant himself says in the Third Critique, aesthetic experience "contributes nothing to cognition... it is not based on concepts, nor does it aim at concepts."
Kant's account of aesthetic experience may seem to contradict his famous claim from the First Critique: "Thoughts without content are empty; intuitions without concepts are blind." But this contradiction may itself be seen as a difference between the function of the understanding and the function of the imagination. Or it may be seen as a difference between philosophical concepts and what Kant calls aesthetic ideas.
Chapter 1 — Thinking like a philosopher 51
It's a question of asking more directly. This question includes the issue of in which situations we can legitimately ask "What is it like?" — that is, the question of style. Given the conditions that we actually experience qualia and that physicalism is true beyond doubt, how can we account for qualia? More broadly speaking, where is experience's place in the philosophy of mind?
Most philosophical commentators on Mary's story dodge this question. Indeed, they tend to belittle or empty out the very notion of phenomenal experience. For example, on Lewis's account, Mary has not really experienced anything. What Mary gains by seeing a red patch or a red object is only the know-how, or ability, to recognize red when she encounters it again. Sensation itself becomes oddly empty. It points to future cases that go beyond itself but never "happens" in the present moment. Sensation has no "itself." Other thinkers go even further in this direction. Dennett deploys the following general argument: even though "phenomenology seems to be there" is an undeniable and universally evidenced fact, it does not really follow that there is phenomenology. Similarly, Scott Bakker, on the basis of his "Blind Brain Theory," suggests in his commentary on Mary's story: the inevitable blindness of the brain, which cannot control its own processes, entails as its essential consequence "the non-existence of things like affects, colors, and so on."
48
that they all took for granted. Jackson took for granted both that qualia must have a physical cause and that qualia themselves lack any physical or causal efficacy. What Jackson seems to have set down himself and forgotten is the important claim that qualitative experience is embodied. Jackson states this point only as a passing observation. In a passage I have already cited, Jackson refers to qualia identifiable as "particular features of bodily sensations." The examples Jackson gives are "the painfulness of pain" and "the itchiness of an itch." But this can be generalized further. Qualia are not simply runaway mental events. Qualia arise from the body's physical activity and from the process by which it interacts with other things in the world. Qualia inevitably accompany our bodily operation. As Whitehead likes to remind us, "we see with our eyes, we taste with our tongues, we touch with our hands." Strictly speaking, then, we do not simply see. It is our eyes that see what they see. What William James says about emotions can also be a quale. It is the result or correlate of a bodily state.
I would also add that qualia experience neither does nor can arise from the absence of a body. Today almost no one would argue otherwise. In the case of phantom-limb pain or
46
there is no special turning point overall. There is no dramatic moment when artificial intelligence first becomes self-aware after crossing some threshold. Intelligence, like the process of growth, is rather a matter of degree. Digients' mentality continues to exist with the minds of less complex machines, as well as with the minds of animals and humans. Perhaps a digient can pass the Turing test. But there is no reason to subject digients to such a test, because digients function well in human environments without one. Digients' intelligence is broad rather than deep. And it is grounded in sociality rather than in solitude. Digients differ from human beings in many ways. But they can operate, even thrive, within the vast and complex context of Data Earth, and more generally within the context of the internet.
It is here that an important distinction casts new light on the problem of experience discussed in Chapter 1. David Lewis, responding to Mary's story, openly mocks the notion that "experience is the best teacher." Lewis argues that, insofar as the brain state of someone who has had a particular experience differs from that of someone who has not, anything that can bring about that change in the brain state — for example, "super neurosurgery" or "magic" — would produce the same effect as actually having that particular experience. It is essentially a matter of rewiring the neurons or rewriting the brain's software appropriately. Therefore Lewis argues that "finding out what it is like to have an experience"
108
need not be confined to the case of actually "having the experience." In some ways, Lewis's argument applies more strongly to digients than to humans. Changing a few lines in a software program is clearer and simpler than reconfiguring synaptic connections and neurons. There is much we still do not know about how neural connections work and how the physical brain is connected to the experiencing mind. But because digients are entirely determined by their source code, we don't need to be distracted by levels of functioning we don't understand.
Yet Chiang's implicit formulation of the problem also shows where Lewis's approach goes wrong. If "experience cannot be algorithmically compressed," then there is no way to shorten the time and effort, and no way to run the procedure more efficiently. To have an experience is precisely the manner in which the brain's electrochemical circuits are reconfigured or in which a digient's software code is rewritten. In Chiang's fiction, the digient's code can easily be copied or replicated. Similarly (as in Lewis's speculation), in many works of science fiction, brain states or mental states can be copied or transferred from one embodied entity to another, or even from an organic body to a distributed computer network. But none of this evades the necessity of producing in the first place — actually — code or brain states that have experience.
Chapter 3 — Thinking like an avatar 109
From its unreliability — or even from its inevitably delusional nature — they slide into the claim that it simply does not exist at all. But if subjective experience does not "exist" at all, how can it possibly be delusional?
I think the problem here is related to the philosophical assumptions underlying the entire discussion. Daniel Stoljar and Yujin Nagasawa, in the introduction to a single-volume anthology of writings on Mary's story, say that "everyone agrees that something happens when Mary leaves the room" (emphasis Shaviro's). But they go further and suggest that the simple fact that "Mary has new experiences when she leaves the room" is no more than a banal "self-evident statement." Mary's actual experience has no essential significance. What truly matters to all such philosophers is rather something different. Jackson wonders what "information" Mary obtains as a result of her new experience. Tye seeks to subsume Mary's experience under the name of representation. Dennett and Lewis, each in different ways, regard Mary's experience as completely insubstantial. They consider the experience to be no more than the demonstration of a "disposition," or merely the product of an auxiliary capacity.
What unites all these philosophers is that they do not find Mary's experience itself in the slightest interesting or important. They focus on its grounds or consequences, or on what it gives us
50
as a consequence brings with it the "non-existence of things like affects, colors, and so on."
One obvious response to this kind of claim is sheer outrage. As Galen Strawson says of Dennett, it is nonsense to say that perceptual objects and affects merely seem to exist. "It seems that there is rich phenomenology or experience there because that phenomenology or experience is there." Phenomenal experience "exists," regardless of whether the appearing content is "real" or not, and regardless of whether what we can say about that content is true or not. Phenomenal experience itself "exists in appearance."
This claim is the minimal foundational form of Descartes' cogito. According to this claim, even if everything I experience is delusional, I can still legitimately say that I am still experiencing it. If we want to be even stricter than Descartes, it might be best to replace Descartes' overly intellectualized cogito ("I think") with a more primal sentio ("I feel"). Thus Deleuze and Guattari suggest that every "I think" already "presupposes a deeper level of I feel." And further, we may legitimately doubt whether to assign feelings to a stable, substantial "I" existing at a moment in time isolatable as "the present." But even with that kind of reduction, a minimal what-it-is-likeness remains. Lewis, Dennett, and Bakker, regarding subjective experience —
despite this, mystery remains.
Instead of simply appealing to experiences that are different from ours, Jackson makes qualitative experience itself unfamiliar from the familiar. Jackson's story suggests that there is a very fundamental difficulty even in describing "what it is like" for me to have my own inner senses. Qualia clearly cannot be captured in objectifying terms, nor can they be known in advance. This difficulty drives Dennett and Lewis toward the conclusion that asking "what is it like?" is meaningless.
At least when I attempt to speculatively reconstruct Jackson's argument, that is so. But unfortunately, Jackson himself does not push this approach. Jackson refuses to speculate in the way I would have liked. This is, as we have already seen, because Jackson formulates his question in terms of "physical information." Jackson asserts that this kind of information — what physicalists believe to be complete — is not "all the information there is." For Jackson, qualitative experience becomes a different kind of information from the physical. "There is something about (such) experience, about its character, that we are ignorant of."
But in saying this, Jackson never questions the murky conceptuality of information itself. Jackson does not raise the question of what it means to "have" a certain type of information, or whether experience can really be described as an entity with certain "properties."
44
does not. As a result, Jackson's argument increasingly takes us away from the initial enticing promise of making us think about "what it is like" to have sensory experience. Instead, Jackson leads us to think about something completely different. About the metaphysical claim of physicalism, about whether the supposed "properties" of experience are always "physical." Instead of marveling at "what it is like" to perceive red, we come to consider, as Lewis says, the criterion for "knowing what it is like" (emphasis Shaviro's) to experience red. The question moves from the affective domain into the cognitive domain.
So the story itself is fundamentally heading in the wrong direction. All the questions about physicalism — central interests for Jackson and all his respondents — are in fact irrelevant, completely missing the point. Even Jackson, while asserting the specialness of qualia and saying that they cannot be "reduced to the status of physical information," takes for granted, quite legitimately, that qualia in fact have a physical foundation. Jackson already accepts that "qualia are the result of what happens in the brain." "Qualia do not bring about anything physical, but are brought about by something physical." Even when trying to deny physicalism, Jackson has already committed himself to the assumption that qualia are no more than epiphenomena of physical processes (as reflected even in the title of his early paper).
Chapter 1 — Thinking like a philosopher 45
