The Pinocchio Theory

Here is the current draft of the Introduction to the book I am trying to write this summer, Discognition: Fictions and Fabulations of Sentience. Of course it is subject to revision.

What is consciousness? How does subjective experience occur? Which entities are conscious? Or, to put things as particularly as possible: what is it like to be a bat? — as Thomas Nagel famously asked. For that matter, what is it like to be a dog, a robot, or a tree — or even a human being? Is it like anything at all to be a rock, or a star, or a neutrino? How do we explain the very fact of being aware? What does it really mean to be conscious, to think, to feel, or to know? And what is the difference — if any — between thinking, feeling, being aware, and knowing? Such questions might seem to have obvious answers — until we actually try to answer them. Then we discover that we don’t have a clue, and that these questions have never come close to being plausibly answered. Still today, there no consensus whatsoever upon any of these topics: neither among scientists and philosophers, nor among the general public. We are clearly sentient, and yet we do not know what sentience is, how it can exist, or what it means.

Whenever I come across such intractable problems, my impulse is always to turn to science fiction. Perhaps we will be able to imagine what we are unable to know. Science fiction is a special kind of literature — or better, paraliterature, as Samuel R. Delany calls it — that operates through speculation and extrapolation, and that takes place (conceptually, if not grammatically) in the future tense. It is a kind of thought experiment, a way of entertaining odd ideas, and of asking off-the-wall what if? questions. But instead of approaching its issues abstractly, as philosophy does, or breaking them down into empirically testable propositions, as physical science does, science fiction embodies these issues in characters and narratives. By telling stories, it asks questions about all sorts of things: consciousness and cognition, the future, extreme possibilities, nonhuman otherness, and especially the deep consequences — the powers and limitations — of both our ideologies and our technologies.

The method of science fiction is emotional and situational, rather than rational and universalizing. Philosophical argumentation and scientific experimentation both endeavor to prove and to ground their assertions, however counter-intuitive these may seem to be at first glance. Science fiction also proposes counter-intuitive scenarios; but its effort is rather to work through the weirdest and most extreme ramifications of these scenarios, and to imagine what it would be like if they were true. Where philosophy is foundational, science fiction is pragmatic and exploratory. And where physical science seeks to settle upon predictable and repeatable results, science fiction seeks to unsettle and singularize these results, and to provide us with unrepeatable histories. Science fiction does not ever actually prove anything; but its scenarios may well suggest new lines of inquiry that analytic reasoning and inductive generalization would never stumble upon by themselves.

In Discognition, I look at a series of science fiction narratives in order to raise questions about consciousness and thought — or better, about sentience. I prefer this latter term, because it does not presuppose that mental processes and experiences are rational, nor even that they are necessarily conscious. When certain philosophers elevate human “sapience” over mere animal “sentience,” they are indulging in dubious feats of self-congratulation. For in fact, there is far more of an evolutionary continuity than a sharp distinction between the way that my dog thinks, and the way that I think. I have many unique qualities of mind that he can never hope to possess; but the inverse of this is also true. Understanding and intelligence (which Robert Brandom lists as the characteristics of sapience) are in fact deeply rooted in such features of sentience as sensory awareness, reality testing, irritability, and arousal. The difference is one of degree, rather than one of kind.

Brandom is therefore wrong to scornfully dismiss what he calls the “merely sentient” condition of animals. My dog may not be able to “offer and inquire after reasons,” as Sellars and Brandom would wish — just as he cannot figure out how to extricate himself when he gets tangled up in his leash. Nonetheless, he exhibits a wide range of moods and feelings. He is is quite good at posing and pursuing many sorts of complicated goals. And he is highly skilled at expressing his desires, in ways that I am able to understand; and at comprehending — and responding flexibly to — my own moods and desires. Thinking is a far more common and widely distributed process than we are sometimes willing to recognize.

The narratives that I discuss in this book offer us speculation — fictions and fabulations — about sentience. There is something oddly recursive about this, since sentience itself is arguably a matter of generating (or being able to generate) fictions and fabulations. We ought to resist the all-too-common equation of sentience with cognition. We often find this assumption taken for granted in contemporary philosophy of mind, as well as in neurobiological research. But mental functioning and subjective experience need not themselves be cognitive — even though cognition seems impossible without them. Sentience, whether in human beings, in animals, in other sorts of organisms, or in artificial entities, is less a matter of cognition than it is one of what I have ventured to call discogniton. I use this neologism to designate something that disrupts cognition, exceeds the limits of cognition, but also subtends cognition. My working assumption is that fictions and fabulations are basic modes of sentience; and that cognition per se is derived from them and cannot exist without them.

Fictions and fabulations are often contrasted, or opposed, to scientific methods of understanding the world. But in fact, there are powerful resonances between them; they are both processes of speculative extrapolation. In other words, constructing and testing scientific hypotheses is not entirely different from constructing fictions and fabulations, and then testing to see whether they work or not, and what consequences follow from them. For science is far more than just a passive process of discovery, or a compiling of facts that are simply “out there.” Rather, science must actively approach things and processes in the world. This is the reason for making hypotheses. Science needs to solicit and elicit phenomena that would not disclose themselves to us otherwise. It must somehow compel these phenomena to respond to our questions, by giving us full and consistent answers. All this is necessary, precisely because things in the world are not cut to our measure. They have no reason to conform to our presuppositions, or to fit into any categories that we seek to impose.

The modern empirical scientific method is sometimes described as a process of “torturing nature to reveal her secrets” — a phrase often wrongly attributed to Francis Bacon. Philosophers of science also like to quote Isaac Newton’s Hypotheses non fingo (“I feign no hypotheses”). But a much better account of actual scientifc practice is the one proposed by Bruno Latour and Isabelle Stengers, who say that scientists work by negotiating with nonhuman entities, and by entering into alliances with them. Scientists do not get very far by treating the things they are interested in as mute and inert objects to be dissected. They do much better when they are somehow able to collaborate with the very entities that they seek to observe and explain.

Alfred North Whitehead, a major inspiration for both Latour and Stengers, notes that if the “rigid… Baconian method of induction” had been “consistently pursued,” it “would have left science where it found it.” Nothing new would ever have been discovered. The same can be said for Newton’s claim of making no hypotheses. Whitehead insists that science needs, not just empirical observation and induction, but also “the play of a free imagination, controlled by the requirements of coherence and logic.” That is to say, a certain degree of speculation is always necessary in scientific research. This speculation has to be “controlled” in some manner; it cannot be altogether arbitrary and unbounded. But without speculation, science is caught in a rut. It cannot stretch beyond the given, immediate facts, in order to provide a plausible explanation for these facts.

The speculative process described by Whitehead is roughly similar to what Charles Sanders Peirce calls abduction. For Peirce, abduction stands in contrast to — and supplements — both deduction and induction. Deduction starts with conditions that are already given, and traces out a chain of logical consequences for those conditions. Induction, for its part, generalizes on the basis of an already given set of particular observations. According to Peirce, neither deduction nor induction can actually suggest anything new. Abduction, in contrast, makes a sort of leap into novelty. It shifts register: suggesting a higher-order explanation for the circumstances with which it is concerned, or positing a possible cause for the effects in view. Science is often praised for having — as other human disciplines do not — an intrinsic self-correcting mechanism. But without first engaging in abduction or speculation, science would never come up with any material to confirm or deny, or to self-correct.

Because it requires flights of speculation, as well as because it requires collaboration among many separate entities, science can never be purely human, nor purely rational. This is why efforts to place science on a pedestal, radically separating it from other forms of thought and endeavor, are so deeply mistaken. Empricial science and rational discourse are largely continuous with other ways of feeling, understanding, and engaging with the world. These include art, myth, religion, and narrative, together with the nonhuman modes of inference exhibited by other sorts of organisms.

We should therefore always be alert to the deep bioligical roots of scientific experimentation and discovery. As Björn Brembs points out, there has recently been a major change of paradigm in neuroscience: a “dramatic shift in perspectives from input/output to output/input.” We can no longer be satisfied with the old stimulus/response model, according to which animals (and other organisms) passively respond to prior, incoming stimuli, and learn by means of conditioning (or associations among these stimuli). For this is only one part of the story. In addition, and probably more importantly, biological entities are active reality-testers. They are always busy “probing the environment with ongoing, variable actions first and evaluating sensory feedback later (i.e., the inverse of stimulus response).” Output tends to come first. Organisms engage their surroundings with spontaneous actions, rather than just waiting for and responding to sensory inputs.

For instance, fruit flies (the special focus of Brembs’ own research) only have tiny brains; but they actively compare the actual results of their reality-testing with what can only be called their prior expectations. They also engage in spontaneous (non-deterministic and unpredictable) actions, so that their behavior “is notoriously variable, even under identical sensory conditions.” The same applies, not just to animals with neurons and brains, but also to non-animal forms of life, like trees, bacteria, and slime molds. That is to say, living organisms are continually engaged, in their own particular ways, in processes of speculative extrapolation and experimentation. When scientists perform experiments and develop theories, actively soliciting responses from the world, they are fundamentally doing the same thing as fruit flies and slime molds — albeit in a far more sophisticated manner, and on a more reflexive meta-level.

Among human beings, speculative extrapolation is not only the method of science. It is also what art in general does — and what science fiction does in particular. As the philosopher Eric Schwitzgabel puts it,

Increasingly, I think the greatest science fiction writers are also philosophers. Exploring the limits of technological possibility inevitably involves confronting the central issues of metaphysics, epistemology, and human value.

In this book, I seek to explore the potentials and implications of sentience by turning to fictions and fabulations — and in particular to written science fiction narratives. Some of the texts that I look at are set in the very near future, and trace out the potential implications of already-existing technologies and research programs. Others are set in a more distant future, and involve more radical flights of extrapolation. Some of these stories can be described as reductionist and eliminativist, in the sense that they seek to demystify and discredit our common sense assumptions about how our minds work. Others might be described as expansive, in that they seek to show that phenomenal consciousness is irreducible, and more widely spread than we sometimes imangine. Some of the narratives deal with human intelligence and consciousness in particular; others propose radically alien sorts of mentality. In all cases, I seek to follow, and extrapolate from, the suggestions expressed by the narratives themselves — rather than viewing them with suspicion, or working to critique them.

More specifically, the hypothesis, or speculative wager, behind this book is that science fiction narratives can help us step beyond the overly limited cognitivist assumptions of most recent research both in the philosophy of mind and in the science of neurobiology. This is because narrative fictions nearly always extend beyond cognition. They are about connecting how and what we know to how we feel, and to how we might act— to what is it like? in short. Even the most reductionist SF stories still work, not just to explain, but to entangle us within their grim scenarios. In this sense, works of art are forms of — or occasions for — rehearsal, as Morse Peckham argued long ago. With their extrapolations, they allow us to respond vicariously to situations that might be extremely dangerous and painful, were they actually to exist. Art readies us for evaluation and action under conditions of uncertainty. In the aesthetic register, Peckham says, “responses are redundantly maintained in situations in which nothing is at stake.” This is precisely what allows narrative (and other forms of art) to explore exteme possibilities.

Psychoanalysis and cognitive science both tell us — albeit for vastly different reasons — that consciousness is only a very narrow and specialized part of mental activity. Most thinking takes place nonconsciously, outside of our attention or awareness. Even more of our thinking slips away — it cannot be retained in memory, or in the form of concepts. Fictions and fabulations can provide us with a sort of feed forward — to use a phrase of Mark Hansen’s — of those mental processes that are not available to introspection. Hansen emphasizes the (quite science-fictional) way that computational microsensors are now able “to stand in for consciousness, to take the place of sense perception in the operations of registering sensory data.” Things beneath or beyond the reach of phenomenal perception are thus made accessible to us, albeit belatedly and indirectly. I want to suggest that fictions and fabulations, whether articulated by human beings or by other entities, are also forms of indirect, nonphenomenological access to nonconscious forms of sentience.

Through fictions and fabulations, we learn that there is more to thought than consciousness. But there is also more to thought than the nonconscious computations of which cognitive science speaks. Before it is cognitive, let alone conscious, thought is primordially an affective and aesthetic phenomenon. This is best grasped as a process of what Alfred North Whitehead calls “feeling.” Whitehead uses this word, he says, as “a mere technical term” in order to designate “that functioning through which the concrescent actuality appropriates the datum so as to make it its own.” What this means, in more familiar language, is that every entity becomes what it is by “appropriating” what is left behind by other entities that precede it. Most crucially, an entity perpetuates itself by appropriating its own prior states of existence. But an entity also appropriates other entities in its surroundings. It picks up whatever it encounters: whatever affects it, or provides conditions or resources for its own continued existence.

This primordial act of feeling, or appropriation, happens before I know it, and often without my ever becoming aware of it. I can breathe without having to know anything about oxygen. Feeling, as Whitehead describes it, comes about prior to anything like understanding (in the Kantian sense), or cognition (in the current psychological and analytic-philosophical sense) or intentionality (in the phenomenological sense). Rather, Whiteheadian feeling is closer to Spinoza’s notion of affection (affectio), and to William James’ theory of emotion. Embodied response precedes, and does not require, intellectual apprehension.

In other words, feeling is something that happens without, or before, concepts. Here we can consider Kant’s dictum that “thoughts without content are empty, intuitions without concepts are blind”; Merleau-Ponty’s insistence that “unreflective experience” must itself be reflected upon, and that such reflection “cannot be unaware of itself as an event”; and Sellars’ attack on the “myth of the given.” All of these philosophers insist that there is no such thing as raw, unmediated experience. Our perceptions and emotions are always already conceptualized. Of course these arguments are in their own terms impregnable; if I want to insist upon a “feeling” that is prior to these modes of conceptualization and self-reflection, then I cannot go on to conceptualize it. I cannot assume its solidity as an idea, or as a point of presence. I must regard feelings, and characterize them, as fugitive and ungraspable; and perhaps also as non-functional, or even dysfunctional.

This means, in Kantian terms, that “feeling” is a matter for aesthetics, rather than for empirical understanding. Despite his strictures against “intuitions without concepts” in the First Critique, Kant nonetheless writes in the Third Critique of “aesthetic ideas,” which he defines as “inner intuitions” which are so powerful that “no concept can be fully adequate to them.” In phenomenological terms, we may say that feeling comes before, and falls short of, any sort of intentionality, or even of Merleau-Ponty’s reversibility. In cognitivist terms, finally, feeling has something to do with what Thomas Metzinger calls Raffman qualia: any such sensation is “available for attention and online motor control, but it is not available for cognition . . . it evades cognitive access in principle. It is nonconceptual content.”

In his recent book Plant-Thinking, Michael Marder credits plants with “non-conscious intentionality.” He means “intentionality” in the phenomenological sense: the idea that thought is of or about something. In this book, I argue pretty much the reverse: that living organisms, beyond and beneath their cognitive accomplishments, exhibit something like nonintentional sentience. Beneath intentionality, or before thought is about anything, there is a thinking process — an it thinks — that is nontransitive, without an object. When it thinks, it feels something; but it does not have any conception or representation of what it is that it feels. As Marder rightly points out, plants do not have anything like a unified or centered self. There is no “I” to a plant, no subject. But for this very reason, there is nothing — as far as a plant is concerned — like an intentional object either. My formulation is not an absolute reversal of Marder’s, because I do not equate sentience with consciousness. I think that Whitehead is right in speaking of the relative rarity of consciousness, and suggesting that most occasions of feeling are nonconscious. Plants are indeed sentient, as recent research has convincingly shown. But this does not necessarily mean that they are conscious. Plants feel, in Whitehead’s sense; they encounter the world. But they do not do so in any manner with which we are consciously acquainted.

In Discognition, I look at science fiction narratives — fictions and fabulations — that consider unusual forms of sentience, both in human beings and in other entities. The first chapter, “Thinking Like A Philosopher”, is not about a science fictional text per se, but rather about a counterfactual narrative — the story of Mary — that has become the focus of much speculation and argumentation among philosophers of mind. The second chapter, “Thinking Like A Computer,” discusses Maureen McHugh’s short story “The Kingdom of the Blind,” which contemplates the possibility of spontaneously arising machine sentience, or artificial intelligence. The third chapter, “Thinking Like An Avatar”, looks at Ted Chiang’s dramatization of the issues surrounding artificial intelligence in his novella “The Lifecycle of Software Objects.” The third chapter, “Thinking Like A Human Being”, considers Scott Bakker’s chillingly eliminativist view of human cognition, as expressed in his novel Neuropath. The fifth chapter, “Thinking Like A Murderer”, looks at Michael Swanwick’s short story “Wild Minds”, which was written before, but almost seems like a deliberate rejoinder to, Bakker’s novel. The sixth chapter,”Thinking Like An Alien”, examines Peter Watts’ First Contact novel Blindsight, which raises questions about the very nature of consciousness by imagining radical, posthuman mind alterations alongside a truly alien sort of intelligence. FInally, the seventh chapter, “Thinking Like A Slime Mold”, considers the strange mental powers of an actually-existing organism, the plasmodial slime mold Physarum polyycephalum.