10 Selective Cognition and the Mouth

In the previous chapter I accepted that possessing a hand is not a necessary condition of possessing thought. The reason is that many living animals have thoughts but no hands. It is not a biological law that thinking animals are handed animals. Thus thought must be an evolutionarily accessible trait for an animal independently of any preadaptation in the form of a hand. But this admission does not rule out a theory of the origin of thought in which prehension is crucial, because the hand is not the only prehensive organ. The mouth is also prehensive, and it is far more widespread in the animal kingdom than the hand. I want, then, to consider the idea that oral prehension might lie at the root of thought: that is, the mouth was the preadaptation that gave rise by intelligible modification to thought. The mouth is what made thought possible, by providing a platform from which thought could evolve.

Let me emphasize at the beginning that this is a highly speculative proposal, which I do not expect to meet with instant acceptance. I am myself by no means confident of it. But I think it is a useful theory to consider, because it illustrates the kind of evolutionary question we should be asking: namely, how can a certain trait have intelligibly arisen? The constraints on explanation are quite clear and sharp (we must eschew saltations): we need to find a preadaptation for every adaptation. So what preceded thought in the biological world that could serve as a preadaptation for thought? Thought had to come from something—but what exactly? At least the theory I shall consider makes an attempt to answer that question.

I must immediately make two terminological disclaimers. The first is that I do not intend to explain all features of what we call “thought” this way: I do not intend to explain the consciousness of thought via the mouth, nor the logical form of thought, nor its role in reasoning, nor its connection to other psychological phenomena. I intend to explain only one feature of thought, which I call selective cognition (it could also be called cognitive selection). By this I mean the ability of thought to concern a single object in abstraction from surrounding objects: to pick one object out and focus on that object to the exclusion of others. In sensation, many objects are presented simultaneously, as parts of a manifold; in thought we can select one of these objects, ignoring the rest, and predicate things of it. There is a singularity that thought possesses that is not possessed by sensation. Psychologists speak of “selective attention”; I am speaking of something similar—the ability of thought to select one object as cognitive focus. Instead of being assailed by a plurality of impinging objects in sense perception, the organism can choose one object to think about. I assume that this basic cognitive ability evolved well before the full capacity we now call “thought,” which brings in many other features; indeed, it may have evolved originally in creatures now extinct (like so many existing adaptations). So think of some ancient creature of the deep that first possessed the cognitive ability to select—I am trying to explain how that creature acquired the ability in question. The current human ability to select cognitively will have evolved from this very early adaptation, but it might be as different from the first form of the ability as the human eye is different from the first eye (from which our eye is also descended). Our selective cognition descends from the earliest form of the capacity, but the modifications and refinements will doubtless be enormous. We might refer to the early form as “proto-thought” or “proto-cognition,” if we are squeamish about applying the usual terms to these relatively primitive creatures.¹ In any case, I take it that a kind of mental selection evolved long ago in quite primitive animals (though ones capable of sensation)—and I want to explain this without postulating saltations or ungrounded leaps forward. I want to identify the evolutionary precursor. That is, I am asking what needs to be added to basic sensation (which may be regarded as a form of consciousness or sentience) in order to derive selective cognition.

The second disclaimer is that I do not mean any old mouth. I do not mean just an aperture through which food passes into the animal’s digestive system. The mouth must be significantly prehensive—some sort of hinged affair capable of seizing and gripping an object. It need not have lips and teeth, but it should open and shut and close around objects. It should be able to take in some given object to the exclusion of others, as with food selection. The idea, then, is that selective cognition evolved from oral prehension—loosely, “thought came from the mouth.” I will argue for this theory by presenting the formal features of selective cognition, as I understand it, and then comparing them with the formal features of oral prehension.² Then I shall make a suggestion as to how the former developed from the latter.

The formal features of thought (selective cognition) are as follows: object uptake, object retention, cessation of object retention, reuptake of object, object switching, volitional character, internality, and mode of presentation. Thus a person might start to think about an object at a given time, keep thinking about it over a period of time, stop thinking about it, start thinking about it again, switch to thinking about another object, and do all this as a matter of will, with the process proceeding “internally” (“in the mind”) and so that the object is presented in a specific way to the person. Bleached of the irrelevant current connotations of the word “thought,” we can say that primitive selective cognition involved cognitive uptake, retention, cessation, reuptake, switching, agency, internality, and mode of presentation (perceptual). An object is selected, dwelt upon, dropped, brought back, ignored in favor of another object, and so on. The process is a dynamic and purposeful direction of cognitive resources onto one object at a time, abstracting away from the sensory manifold. The mind “homes in” on a single object, from among many, and holds it in “thought” for a while. It is important that the process is purposive, because thinking about an object is subject to the will but perceiving an object is not.³ One of the things we are trying to explain is how thought differs from perception in its object-directedness: thought is active and willful but perception is passive—I can decide what to think about but not what to see (given that my eyes are open, etc.). So our creature of the deep is capable of these active feats of cognitive selection—it is not simply at the mercy of what floods in through its indiscriminate senses.

It is easy enough to see that the formal features of oral prehension mirror those of selective cognition. The organism is surrounded by an array of objects, of which it has sensory awareness, but it purposefully selects only one object to hold in its jaws (imagine this is one clam chosen from a bunch of other clams). It holds this object in its mouth for a while, perhaps tasting it; then it may release the object. Later it may grip the object in its mouth again, or it may grip another object. All this is done purposefully. The object is inside the animal’s mouth for the duration of the gripping (“internality”). Also, the object will be perceived in a certain way while in the mouth—a sensory mode of presentation will accompany the gripping. The object is temporarily inside the animal’s oral cavity, presenting itself in a certain way, as a result of the animal’s purposes. Dynamically and structurally, then, oral prehension mirrors cognitive apprehension. Abstractly considered, the processes are notably isomorphic. Thus, if you wanted to construct a new trait with a certain structure on the basis of an already existing trait, this would be a good way to go. The basic design features are already present; you just need to make certain modifications and upgrades. We have a kind of mind-mouth isomorphism, so the one could be the launching pad for the other. We can build the new structure around the design of the old structure, because the basic engineering specs are the same. Since evolution tends to be economical and conservative, this conversion of function will appeal to it. We have here a possible case of exaptation, the deployment of an old trait in a new setting.

But how exactly does the mouth, an external physical organ, get turned into a “mental organ”? That looks like a big transformation, a saltation no less. It is not as if the mouth became progressively smaller over evolutionary time, until it vanished and transformed itself into the mind! Nor did the animal’s brain develop a region with actual prehensive powers—a weird brain-mouth. Where are the missing increments in our story? Where is the continuity we seek? To answer that question, we must turn our attention from the mouth itself to the brain machinery behind it. The brain machinery must somehow code the actions of the mouth—it needs suitable motor schemata or programs. It also codes for the sensory aspects of the mouth, including proprioception. The brain needs an elaborate “mouth map,” as it needs a map for other parts of the animal’s body. This takes the form of an abstract representation, not a little homunculus mouth in the animal’s head. So in addition to the mouth adaptation itself, there must be a computational-machinery adaptation in the part of the brain that runs the mouth. Think of it as a set of instructions for doing things—such as seizing objects, holding onto them, letting them go, and so on. Then that is the adaptation that functions as a preadaptation with respect to selective cognition. The brain machinery for oral prehension is co-opted as brain machinery for selective cognition. Logically, the case is like the jawbones of the reptile being recruited as the ear bones of the mammal: a new function is assigned to an old structure, with suitable small modifications.⁴ The brain circuits that underlie the functionality of the mouth are redeployed to serve the functionality of cognition, with the same abstract structure preserved. Instead of controlling only oral actions, they now also control mental actions—and these actions are adaptive, so selected for. The same program is used to run different applications, in effect. The program that was useful for controlling the dynamics of the mouth turns out to be useful for controlling the dynamics of cognition. The old brain circuits are “kicked upstairs” to perform new tasks. Thus an evolutionary step forward is taken. Our creature of the deep, with its active and questing mouth, takes the first faltering steps toward what will finally become thought, by developing the capacity to “hold” an object in its mind. Given ancestral preservation, our human capacity for thought therefore owes its origin to such distant goings-on, involving a fish’s mouth and brain (assuming our story is on the right track). We inherited the relevant brain anatomy from this groundbreaking fish.⁵

This redeployment process can be compared to the evolution of mental imagery. First animals simply perceived the world, with no mental images. The brain contained circuits that controlled and served the outer senses. When mental imagery evolved, it did not proceed from nothing—it exploited preexisting traits, namely the perceptual systems. That is, mental imagery evolved from sensory perception: visualizing evolved from vision. The former capacity is built on the latter, with suitable modifications. But it did not do so by successively modifying the sense organs themselves, as if the eye became smaller and smaller and pushed back into the brain. There is no little homuncular eye in the brain. It did so by exploiting the preexisting brain machinery, not the outer eye. The same circuits that were involved in human vision become used in the production of human visual images (with some modifications). Thus something “outer” became something “inner.” Traits of the brain are as much adaptations as traits of the body, and equally susceptible to enhancement and redeployment. Similarly, the brain circuits involved in selective cognition are borrowed from their original role in controlling the mouth. The obvious resemblance between perceptions and images suggests an identity (or at least overlap) of underlying machinery, and the formal resemblances between selective cognition and oral prehension suggest a comparable identity (or overlap) of underlying machinery. In neither case is there any objection of principle to evolution proceeding in the manner suggested. Things certainly could have gone this way, as a matter of logical and nomological possibility. This is an intelligible story in light of all that we know about the evolutionary process. Whether it is empirically true is another matter. We are speculating freely about very remote matters, about which firm data are unlikely to emerge. We certainly have not observed the oral brain circuits of a fish being co-opted into functioning as brain circuits for piscine selective cognition. This is a theory—a conjecture, a hypothesis—not an established fact. What mainly matters for our purposes is that the theory has the right methodological and explanatory properties—specifically, ancestral preservation and incremental adaptation. And what other theory meeting the requisite conditions of adequacy can be suggested? We wanted to show how thought could arise from nonthought, in a quasi-mechanistic manner, with no miracles or unexplained leaps—and this theory has the look of something that fits the bill, even if it is difficult to verify as a statement of prehistorical fact.⁶ As the prehensive hands formed the core preadaptation for language, so the prehensive mouth formed the core preadaptation for thought (or one feature of it)—so we are conjecturing.

What we think of as belonging to the “mental” side of our nature evolved from something belonging to the “bodily” side (though this dualism should probably be rejected and anyway looks very artificial when considering ancient and primitive biological forms). It would be better to say: one system of brain machinery evolved from another system. The machinery for cognition evolved from the machinery for oral prehension: bits of the second machinery form components of the first machinery. Selective cognition did not suddenly get installed one day from no antecedent organic structure, as a result of a fluky mutation that miraculously had just the right properties. It emerged from apt and suitable preexisting traits. And once this adaptation occurred there was room for further subsequent adaptations, no doubt taking many millions of years, until we reach the biological trait called “human thought.” What I have sketched here is the initial decisive step—how cognition was born (or might have been born) in its most primitive form.

I must now take up the question, postponed from earlier, of how literally we can take the prehensive theory of thought. Is it literally true that the mind “grasps” things or is this just a flaky metaphor? Does the mind “grasp” things in the same sense that the mouth and hand grasp things? The question is not easy to resolve. There are three possible theories of the concept of grasping: (i) the concept applies originally and literally to an action of the hand (or mouth) and only by extension and nonliterally to the mind; (ii) the concept applies originally and literally to an action of the mind and only by extension and nonliterally to the hand (or mouth); (iii) the concept applies equally and literally to both, being intrinsically neutral between the two—they are simply both species of a (unitary) genus. I suspect many people will plump for theory (i) without much hesitation. But two points should be borne on mind. One point is that if this is so then talk of the mind grasping something is pure metaphor, like calling Juliet “the Sun”: but it doesn’t seem like pure metaphor, dead or alive. How else would we literally describe what the mind does? Isn’t it simply true that I grasp the meaning of “The cat sat on the mat”? This is not some creative figure of speech, some easily paraphrased bit of poetry. Second, it is not strictly accurate to say that the hand grasps; the person grasps with the hand. In the same way, it is the person who throws or holds or squeezes or signs his name—the hand doesn’t do these things. To perform these actions, an agent must have certain intentions or purposes; it is not enough that his hand moves in certain ways. So the notion of grasping, even when applied to the body, is implicitly mentalistic, not purely “physical.” It involves the notion of a person with purposes.⁷ I would say that a robot with no inner life—no intentions and so on—does not literally grasp anything, even if its “hand” performs a clenching motion. It is only as if it grasps—just as it is only as if the Sun grasps the Earth in its gravitational field. Real, literal grasping, like real, literal name signing, requires suitable intentions (or at least some viable notion of purpose). Thus I grasp your hand (intentionally) when I take it in mine; it is not just that my hand closes vicelike around yours. It is a category mistake to suppose that the hand, considered in itself, grasps anything—as it is a category mistake to suppose that legs walk or the mouth sings. The concept of grasping is not a purely behavioristic or mechanical concept.

These reflections may tip us in the direction of theory (ii)—the concept of grasping applies originally and literally to an action of the mind, and only by extension and nonliterally to the hand or mouth—and I don’t think that possibility should be dismissed out of hand. Maybe the concept of grasping applies in the first instance to the mental kind of grasping, and is then projected outward onto the hand (as the concept of force is projected onto inanimate bodies from its original use in application to human agency). We therefore read grasping into the hand (or mouth); we don’t read it off the hand (or mouth). Don’t we read grasping and gripping into the roots of a tree, because of a similarity to other prehensive structures, even though we acknowledge that roots and trees don’t literally grasp or grip the earth? Maybe we impose the concept of grasping on the hand, viewing it as somewhat like the genuine mental kind of grasping. So it may be thought, but I also find this theory rather forced: it surely isn’t just a metaphor that hands grasp or that we grasp with them—real honest-to-goodness grasping is going on when I close my hand around a ball, say. I do literally grip your hand when I take hold of it. So theory (ii) does not ring true, conceptually speaking.

We might then find ourselves sympathizing with theory (iii). The issue resembles the concept of the visual as it applies to visual perceptions and visual images. It is initially tempting to suppose that “visual” applies literally only to visual perceptions, so that “visual imagery” is really just an extended metaphorical usage. But this flies in the face of an evident phenomenological resemblance. By contrast, some may feel that visual imagination is primary, because so-called visual perception is infused with it, so that “visual” applies primarily to visual imagery. This seems hard to accept too, because visual experience without visual imagination does not seem inconceivable. Thus we reach a type-(iii) theory of the concept visual: there is a genus of “the visual” that is neutral between perception and imagination, and this genus has two species, both equally and literally visual. I myself favor this type of theory for the concept of the visual, and it seems attractive also for the concept of prehension.⁸ There is a broad generic concept of prehension and the mind and the hand are both (literally) species of it. It is hard to articulate (in a noncircular way) the content of that generic concept without biasing it toward one or the other species of prehension, but it has to do with some abstract notion of selecting and holding (for this word too the question arises).⁹ We have the idea that the mind attaches itself to an object, maybe encompasses it—and the hand does something similar. An abstract schematic notion is thus specialized into two more concrete notions. The concept of prehension, then, is equally correctly used when we speak of grasping a meaning or grasping a ball, literally in both cases.

But we do not need to insist on the literalness thesis to maintain that selective cognition originated from oral prehension, though this would certainly lend support to that theory. What matters is that we have described a possible and realistic route to the formation of selective cognition based on a preexisting trait. This is not, of course, intended as a reductive thesis—to the effect that selective cognition is nothing but oral prehension. I take it such a view is absurd. We have only claimed that oral prehension (its cerebral basis) might form a natural evolutionary springboard for cognition in this primitive sense—the seed, the germ.¹⁰ In addition to oral prehension the organism must be an agent in some sense, and must also be sentient—without these background conditions I do not think talk of cognition is appropriate. The organism must be a perceiving agent already; only then can the mouth serve as a springboard to selective cognition. My picture of the ancestral fish is that it is already a sensing, purposive being, which has a prehensive mouth with the formal features I described; and then it makes the step to cognitive selection, with all these traits already in place. Incremental emergence over long periods by mutation and natural selection is the underlying picture, not the reduction of one trait to another. To repeat, evolution works by gradual intelligible steps, with every adaptation building on what was there before. We don’t reduce the later adaptations to the earlier ones; rather, we demonstrate a step-wise procedure that respects the limits of the process. There are no “inspired” adaptations or bolt-from-the-blue mutations. There are just painfully slow transitions constrained by the existing form of the organism. For any trait of explanatory interest, we must always seek the preexisting form from which it derived as a modification. This methodological requirement imposes strict conditions on any putative evolutionary explanation. Respecting those conditions has been the main purpose of this chapter.