11 The Origin of Sentience

To be sentient is to be “able to perceive or feel things” (OED). It is connected to the senses: when an organism acquires senses, it becomes sentient. The senses have been around for a long time, much longer than language or even thought. The sense of touch is probably the most ancient sense. Even the lowliest organisms, such as worms, have a sense of touch—they feel things. They have sensations when they make contact with things—at any rate, many simple organisms do. There was a time on Earth at which sentience had not yet evolved, when senses had not yet appeared—the time of plants and single-celled organisms. How did sentience arise from this state of things? What was the preadaptation that formed its launching pad? We were able to come up with suggestions for the origins of language and thought—can we do the same for sentience? For that we would need to cite a trait that got converted into sentience by intelligible incremental steps. Let me say up front that I have no idea how that could have happened: this is one of the most profound and difficult questions facing biology, physics, and philosophy. The trouble is that any adaptation that might lead to sentience already contains it, and any that does not contain it will not lead to it. A gradualist explanation looks ruled out by the very nature of the evolved trait.

But I do not intend to debate that issue here. In fact, I am going to assume, for the sake of concreteness, that a panpsychist explanation is on the right lines—that is, sentience evolved from a kind of proto-sentience found in all matter.¹ The preexisting form for whole-organism sentience is simply small-scale sentience—sentience at the molecular and atomic level. The preadaptation for the advent of massive three-dimensional animal bodies is simply the prior existence of matter—a lot of that stuff was lying around when evolution began, and evolution tapped into it. We therefore have no problem explaining where the matter of animal bodies came from; it sprang not from nothing but from the antecedently existing material universe. Well, for the panpsychist, a lot of sentient stuff was lying around in matter, and the evolutionary process simply tapped into that abundant resource. Sentience did not spring from nothing either; it already existed in elementary form in all matter. Panpsychism thus has an answer to the question of how sentience evolved: ancestral preservation and incremental adaptation are respected. Nothing I say here will depend on the truth of panpsychism, but for expository convenience I will adopt it pro tem. I am more interested in the question of why sentience evolved: why did organisms start to sense and feel things? Why did the primitive sentience latent in matter become organized into organism-level sensory experience? It might have just smoldered harmlessly in matter for all eternity, having no adaptive potential, as it did for millions (billions!) of years before animal life evolved—so why did it acquire adaptive significance? What function does it serve? This is the same as the question: Why did the senses evolve? Why didn’t life remain at the level of insentient existence? Evidently it did so for a very long time, so what changed? Or, to put it differently, why are some contemporary life forms sentient and some are not? What does sentience do for an organism? Why isn’t the planet populated by completely insentient organisms, mindlessly going about their reproductive business, as it was for several billion years?

The theory I shall propose I call the predator avoidance theory. Let me expound it in a roundabout way, as follows. The theory itself is quite intuitive, but the rationale for it takes some spelling out. Before life of any kind began on Earth, when the landscape was basically just water and rock, the Sun beat down on its surface, just as it does today. The sunlight contained huge reservoirs of energy, in the form of electromagnetic radiation. This is the Stage Zero phase of life: there is a source of energy available (Earth is not shrouded in impenetrable darkness), but no life form exploits it. After a while, bacteria and elementary plant life evolved (for reasons that remain obscure), which used the Sun’s radiant energy as a form of nutrient—photosynthesis began. This is the Stage One phase: naturally selected organisms living off the Sun’s energy but having no sentience or accompanying senses. Stage One organisms, let it be noted, lack mobility, staying fixed in one place, unless moved by an outside force. The reason they don’t move from place to place is simply that sunlight is available virtually everywhere—you don’t need to travel to a particular place in order to tap into it. We can imagine a planet, call it Dearth, on which things are different: on Dearth the atmosphere is largely opaque, but holes occasionally open up in varying locations, so that sunlight reaches the surface in different places from time to time. Then plant life on Dearth will perish if it stays in one place—it needs to move to where its source of nutrition is. These sun-dependent organisms will therefore experience selective pressure to acquire the ability to move to where the sunlight is to be found (suppose a shaft of it moves steadily across the landscape at a rate of four miles an hour, with the rest of the planet in complete darkness). On planet Dearth we might well find mobile plant forms: consumers of electromagnetic radiation that follow their “food source” around. But on planet Earth the sunlight falls indifferently and abundantly pretty much everywhere, so there is no real point in moving from place to place to find it. And developing the ability to move around is a metabolically costly adaptation, so best left undone ceteris paribus. Foraging plants have no raison d’être here on Earth.²

Now suppose that Stage Two organisms evolve: these creatures derive their food source from Stage One organisms, that is, plants—they are vegetarian. These organisms have a fairly stationary lifestyle too, because their food source is abundant and can’t run away from them (wherever there is sunlight, there are plants lazily basking in it); but they may develop some rudimentary ability to move around to deal with the fact that the local vegetation might become depleted. Think of them as slow-moving grazers and berry pickers—slothlike creatures. At this point, there is some evolutionary pressure to be economical as to size—if you move at all, as a matter of survival too much bulk is a liability (as is a cumbersome body structure). Trees can be as bulky and unwieldy as they like, because they don’t have to move themselves anywhere. Stage Two organisms will have to be better designed for movement, more streamlined, less bogged down. But still rapid locomotion is not at a premium; indeed, survival can be achieved while sticking to the same general area for one’s entire life. And why go to the trouble of moving, given that this is a metabolically costly activity? Why not just lazily munch away in the same comfy spot?

Now we get to Stage Three organisms: their food source is Stage Two organisms—they are carnivores. Energy is transmitted from the sun into the plants, then into the plant eaters, and finally into the eaters of the plant eaters. Stage Three organisms throw the cat among the pigeons, quite literally: they are predators. The Stage Two organisms need to avoid them. How do they do that? They need to sense when the predators are near and then move promptly away. In particular, they need to sense contact from predators, so that they can move their bodies to another place in the nick of time. Thus sentience arises from predator avoidance, as a correlate of predator-detecting senses. Costly sentience exists because being eaten by predators is costlier. This seems plausible enough so far as it goes, but there is an awkward question: why didn’t the plants develop sentience too, so as to avoid their predators? The answer to this question is somewhat complex. In the case of plants they have a stationary lifestyle, so size isn’t much of a problem (hence very big trees). But this makes them vulnerable to predation—they can’t run away from the voracious plant eaters. What then is their strategy of survival? It is two-pronged: being made of (a) invulnerable parts and (b) vulnerable redundant parts. In the case of a tree the invulnerable parts are the trunk and the thick branches—no normal plant eater can destroy the plant by biting these in two, which would indeed kill it. But the leaves and buds are very vulnerable to being eaten, so how does the plant survive their destruction? It does so because the leaves are highly redundant—you can destroy a lot of them, maybe all, and the tree still does not die. The leaves are not vital parts of the tree (unlike the trunk), in the sense that damage to them leads to the death of the whole organism. Thus it doesn’t matter to the tree’s survival that its leaves can be easily eaten, so it is not necessary strenuously to protect against this. Accordingly, there is no need to develop senses that record when the leaf is being “attacked.” Of course, having lots of redundant parts means the organism will have trouble moving; but trees don’t move, given their food source, so this problem does not arise for them.

But things are different with the Stage Two organisms: they do move. They need to move to follow their food source in case of depletion, though not very quickly, but they also need to move to avoid their predators, a good deal more rapidly. One strategy for defeating their predators would be to be composed of redundant parts, so that if a piece of the body is bitten off this has no serious effect on the survival of the organism.³ There are plenty more legs and heads where that came from! But that strategy creates bulk and weight, and Stage Two organisms need to move, which is metabolically costly. In addition, it is inherently risky, because the predator might be powerful enough to eat the whole organism (this never happens with trees). It also involves difficult engineering problems. Once an organism has to move to survive and reproduce, it cannot afford to have redundant parts, because these will need to be carted around somehow. Thus we observe a marked difference between plants and animals: animals are far more vulnerable to damage to their parts than plants are. All the parts of an animal are pretty vital—the ears are not dispensable like leaves, say. It wouldn’t matter if the vital parts were themselves invulnerable—say, if they were made of unbreakable material or flesh so poisonous that it would instantly kill any predator. But for various reasons this happy setup is impossible. So the animal has vulnerable vital parts, while the plant has vulnerable nonvital parts (as well as some invulnerable vital parts, like the trunk of a tree); and this traces back to the role of movement in the organism’s life, which itself reflects the nature of the food source.

An organism can in principle survive predators by a number of strategies: it can have no vulnerable parts at all, in which case it need not move to avoid predators; it can have vulnerable parts but all of them are redundant, in which case again it does not need to move to avoid predators; it can have a combination of invulnerable and redundant parts, so that again no movement is necessary; or it can have vulnerable vital parts but can move efficiently away from predators. The last is the strategy of Stage Two organisms in the presence of Stage Three organisms. But in that case the organism needs a way of detecting the presence of predators, so it needs senses, and hence sentience. Sentience arises when an organism must move to avoid predators; it is not necessary otherwise. Thus there are three necessary conditions for sentience to arise: vulnerable parts, vital parts, and a mobile lifestyle. Together these are sufficient (given the antecedent presence in matter of potential sentience).⁴ Note that the first two conditions are not sufficient on their own, since there is no point in sensing predators if there is nothing you can do about it (you live by a hope-for-the-best strategy). True, if you are organism that can poison your enemies with ease, or give them a severe electric shock, then there is no need to move away from them; but this is really an example of invulnerability, and very hard to achieve given the adaptability of predators. This is why the vast majority of prey animals choose the strategy of moving away from their predators. It is not a bit surprising that predator avoidance typically involves movement and perception.

It is true that having senses would have some prima facie utility for Stage Two animals even before the Stage Three predators came along, because senses could serve in the search for plant food sources. The animal could see or smell a particularly delectable leaf or berry from afar and hence head right for it. But two points can be made about this possibility. The first is that the selective pressure to develop such senses would not be very intense, because the food source would not itself be a moveable feast; so acute fine-grained sensing, especially of motion, would not be required—as it is when dodging a fierce and agile predator. Second, it is not at all clear that genuine sentience would be needed to secure the advantage in question: the animal might get by with a merely insentient means of detecting the presence of food in its environment, as the most primitive organisms presumably do. It could make do with suitable photoreceptors and chemoreceptors without the full panoply of perceptual sentience. But it is another matter, again, when avoiding fast wily predators: here the organism needs the benefit of full-blown high-tech sentience—seeing, hearing, and smelling with high acuity. So I suspect that sentience only evolved on Earth once carnivores did—that was the urgent and nonnegotiable impetus for becoming genuinely aware of the environment. Before that it was possible to live as a sleepy and slow-moving sloth, with barely a glimmer of consciousness (maybe just a feeling of emptiness in the belly when food was needed). Given that fine-tuned perceptual sentience, like any sophisticated biological adaptation, has its own operational costs, in terms of energy consumption and upkeep of the underlying neural architecture, it will only evolve if the need for it is pressing enough—and escaping a tiger’s jaws is the kind of stimulus that is especially pressing. Without the need to avoid carnivores, animals could have survived quite nicely without going to the trouble of developing the kinds of bodies and brains that make senses and sentience possible. The reason an antelope is sentient is not so as to locate the grass it eats but to avoid the predators that threaten to eat it. The senses developed in the battle against predators not in the location of a vegetarian diet. That was when the benefits of the equipment started to outweigh its costs.

It might be wondered how this theory explains the existence of sentience in the predators themselves. Predators don’t need mechanisms of predator avoidance, do they? Two answers to this question suggest themselves. The first is that predators are in turn prey. This is obvious for mid-range predators, like hyenas and humans, because they are vulnerable to more powerful predators. But it is also true for top-of-the-line predators, like lions and tigers, because in youth and old age they become potential prey for lesser species. Also, they are vulnerable to competitive attack (sometimes cannibalistic attack) by members of their own species. So they need their injection of sentience too. The second reason is more interesting: once their prey develop capacities for predator avoidance, an arms race develops, because the predator animal needs to be able to detect and capture the evasive prey animal. The predator must develop sentience because the prey has. It would be different if the prey were insentient and immobile, like a tree; then the predator could get by with very minimal detection methods. But if the prey can spot you at 100 meters and make off, then you need yourself to be able to identify and track the prey over time and space. So predator sentience evolves because prey sentience evolves—predator avoidance on the part of prey is the root of sentience in the predators themselves.

The reason that animals develop advanced abilities to sense their environment, such as we observe all around us, thus traces ultimately back to the existence of predation. If there were only plants and vegetarians on planet Earth, there would be no sentience to speak of; at most there would be a very minimal kind of mechanical detection of plant food sources on the part of the plant eaters. Things only heat up sentience-wise when predators come on the scene. Then it becomes imperative to be able to detect, quickly and efficiently, what is going on around you, and take appropriate action. Advanced sentience, of the kind we see in most animals today, results from fear of other animals. The tranquil vegetarian, prey to no animal, could be to all intents and purposes unconscious—certainly not the sensory expert that populates Earth today. It is the fear of other animals that wakes him into fizzing, full-bore sentience—now he must be on his sensory toes.⁵

It is really the lack of sentience in plants that calls for special explanation. This is quite puzzling. Why don’t they move away from their predators too? If they did, sentience would be useful to them. There must be selection pressure that favors a mobile plant (and some do move to some extent), because a species of plant that can escape its predators has a greater chance of survival than a species that cannot escape. Whole species of plants can be eaten into extinction, and individual plants certainly often are. Why don’t plants evolve little feet to help them escape being eaten? Aren’t there possible planets on which some plant species have developed locomotion? If they did, they would have a use for sentience too. The answer must presumably lie in a fundamental trade-off between the advantages of immobility and the advantages of mobility. Given the fixed solar food source, and the extra resources needed for locomotion, the plant (or its genes) makes the calculation that moving to avoid predators is just not worth the trouble. That is, the genes for its construction survive better, given its lifestyle, than in an alternative lifestyle in which locomotion plays a role. True, locomotion and sentience would be useful in the avoidance of plant eaters, but it is just too biologically costly to install the necessary machinery. Instead the plant settles for a mixture of invulnerability and redundancy. But animals that live off plants and have predators have an overwhelming need to develop senses that enable them to survive in a dangerous world.

Let me now turn to a different question: the character of specifically human sentience. Human sentience is ultimately inherited from fish sentience (the lobe fish family in particular). Genes for sentience are passed down the generations, going back to its first glimmerings. There is an ancient preexisting form of piscine sentience of which the human kind is a modification. This is as true for sentience as it is for human hands or human intestines. It is natural then to ask whether any remnants of its earliest forms survive in us to today—relics of our ancestors that no longer serve any function in us. Given that sentience first evolved in the oceans, it was adapted for that milieu; but we live on the land. Of course, many modifications of the original form of sentience have occurred—trichromatic color vision is one of them. The eye originally evolved in the water, so is there anything about the eye today that reflects this lineage? Well, one point is that sunlight is less bright in the ocean than on land, so the existing mammalian eye was originally adapted for less intense light. To come out of the murky depths into glaring sunlight would have exceeded the acceptable thresholds of ambient light—the animal would be effectively blinded. Of course, there was a process of adaptation to stronger sunlight over many millions of years, but isn’t it still true that the mammalian eye is not well adapted for powerful sunlight? We can’t look into the Sun; nor can other mammals—yet the Sun is a perpetual perceptual presence. Our eyes just don’t respond well to glare. Not being blinded by the Sun would be an adaptive advantage. Might this be some kind of holdover from the days of ocean-going vision? Something like this is at least conceivable.

But that is more a point about optics and receptors of the eye as a bodily organ; it is not a point about the very form of our phenomenology. It seems to me difficult to identify any feature of our sensory phenomenology that reflects our piscine ancestry; we might just as well have descended from some other ancestral stock, so far as our inner sentience is concerned. On the other hand, how much does contemporary fish sentience differ from mammalian sentience when you get right down to basics? Much the same senses are possessed, so the form of the sentience will not differ dramatically. It would be different if we possessed some remnants of a sense we no longer possess, simply because our ancestors possessed that sense. Suppose we no longer had noses and never perceived smells, yet in dreaming we experienced olfactory sensations in imagistic form: this might be the result of a remnant from the old days, still with a basis in the brain but with no functionality any more. That is the kind of thing I mean by asking whether any trace of our ancestral piscine sentience persists today: the ancestral fish had a sense of smell and we experience smells in dreaming just because of that ancestry, not because we have any use for a sense of smell in our current habitat (compare the appendix). But so far as I can see, nothing like this is the case. That is not because our sentience owes nothing to the sentience of our piscine ancestors—on the contrary, ours is inherited from theirs (with modifications)—but because we haven’t departed all that much from the original forms of sentience. The basic sensations of animals haven’t changed dramatically from the earliest life forms—visual, auditory, olfactory, gustatory, tactile. Maybe things are different on other planets where there is conscious life. In any case, our current sentience does not seem to contain any pointless evolutionary relics from our sea-dwelling past.⁶ It all seems quite adaptive. This is somewhat anomalous, given that the human body contains many obsolete relics of our distant evolutionary history (as our DNA does). Perhaps the consciousness of the fetus goes through a kind of pointless fishy phase, which is then replaced by something mammalian, as the body of the early fetus is said to display marks of our evolutionary past in the form of fishy attributes, such as gills.⁷ Or perhaps it is just all too long ago and no telltale remnants survive in us from that distant period, despite the fact that we originally evolved from fish, brains and all.

Summing up our discussions so far, we may now identify three great epochs in the evolution of what we call “mind.” First, after a long presentience period, in which bacteria rule the seas, basic sentience evolves, possibly because of locomotion and predator avoidance. This is then elaborated and refined into the senses with which we are now familiar. Later, cognition of the selective kind evolves, building on the prior existence of advanced oral prehension. This is then further elaborated into thought, as we know it today. Then, after a very long time, human language finally evolves, exploiting the hand and toolmaking and other developments. Thus we have: the Age of Sentience, the Age of Cognition, and the Age of Language. Each of these ages subdivides into further evolutionary innovations and modifications. The latter two ages were triggered and enabled principally by advances in animal prehension, involving the mouth and hand. Prehension precedes and shapes thinking and speaking. Did prehension play any role in the evolution of sentience? It might appear not, because that age was triggered by a predator-driven need for bodily senses; and indeed no prehensive preadaptation played any explicit role in the story I told about the origin of sentience. However, prehension might be seen to enter less directly into the story, in two ways. First, perception itself might be seen as prehensive in character—a mental reaching out and grasping of external things. Second, animal locomotion involves a further kind of prehensive act: the earth must be gripped somehow. The feet must grip the earth in order to propel the animal forward. This is evident when the plane of locomotion is an incline, as with climbing a tree or a steep hill. But even on flat land the feet must perform a certain amount of gripping—think of walking on a slippery surface. Claws and hooves grip the ground as the animal walks and runs. Fins can be said to grip the water in an extended sense. Wings likewise grip the air. Even in the case of snakes and worms locomotion requires a gripping action of some sort. So prehension—or proto-prehension—accompanies sentience, even if it is not a direct preadaptation for sentience. It may even be true to say that any organism that grips has sentience and vice versa. So prehension is part of the overall background of sentience, even if it is not central. Certainly the evolution of prehension marches in step with the evolution of “mind.” Body and mind do not evolve as separate entities but in a kind of continuous interplay.⁸

In humans, today, sentience, thought, and language are intermingled and inextricable. But viewed diachronically we can separate these three faculties and see how they might have evolved at different times and for different reasons. The broad category of “the mind” does not do justice to this variety. An evolutionary perspective thus enables us to make distinctions that are blurred from a synchronic standpoint. Then we can formulate our explanatory questions more clearly and fruitfully. Even if my speculations are misguided, I think they at least proceed from the right assumptions.