Saturday, July 11, 2015

Animal Intelligence

[Edit 02/11/2017: The final version of this essay can be found here.]

Chimpanzees communicating with sign language and lexigrams. Rhesus macaques accurately adding Arabic numerals. Dolphins employing sponges as protective gear during foraging. Crows bending wires into food hooks. Octopi transporting and manipulating coconut shells for camouflage and shelter. These and many similar examples are showcased regularly by today's scientists as the evidence that animals—meaning the animal species other than Homo sapiens—possess intelligence, intelligence comparable to and in some cases rivaling that of humans.

But when these examples are examined not for what they are supposed to demonstrate, but instead for what they actually demonstrate, they contradict the scientists' conclusion, and what they reveal instead is a widespread misunderstanding about what intelligence is. For all but one species on this planet the phrase animal intelligence is essentially meaningless, a null set. And for the exception—for the human species—the phrase animal intelligence is more akin to a contradiction in terms, the left-to-right juxtaposition of this species' sudden and surprising turn from the first word of the phrase to the second.

The many offered examples of animal intelligence fall invariably into one of just two categories.

The first category comprises those instances typically originated in the laboratory, such as primates displaying language skills and numeracy, birds employing an assortment of knick-knacks as tools, rodents navigating ever more complex mazes, etc. And more broadly, this category also encompasses the many domesticated animals and their activities that are frequently portrayed as intelligent, such as responding to a called name or consistently unlatching the hook from the pen. Such behaviors, when performed by humans, are taken as instances of human intelligence, and so it is an easy next step to assess such behaviors, when performed by animals, as instances of animal intelligence. Further bolstering this line of reasoning is the widespread agreement that human intelligence is produced via the human brain. Animals clearly have brains too, and so what could be more natural than to characterize their sophisticated laboratory and domestication behaviors as expressions of animal intelligence produced via the animal brain. What arises then from all these seemingly straightforward considerations is the widely held notion that not just humans but the other animals too must possess a neural proclivity for intelligence, and although this intelligence might differ in degree from species to species (for presumably evolutionary reasons), it certainly does not differ in kind.

Nonetheless, this line of reasoning masks a crucial omission: it too conveniently forgets the context of all these examples. By definition, every instance of a laboratory- or domestication-based animal intelligence behavior manifests exclusively within an artificial environment, which is to say that it manifests exclusively under human-generated conditions. Chimpanzees do not respond to chimpanzee-invented words and lexigrams, they respond instead to human-invented words and lexigrams. Crows do not manipulate crow-generated artifacts, they manipulate human-generated artifacts. Rats do not navigate rat-engineered mazes, they navigate human-engineered mazes. The inevitable presence and influence of human-originated features within the settings of all these examples calls immediately into question the applicability of the phrase animal intelligence, because it remains highly uncertain that animal has anything to do with these cases. It would be more legitimate to say that the animals in these examples are displaying not a form of animal intelligence so much as they are simply redisplaying well known forms of human intelligence, a kind of intelligence irrevocably tied to human-specific circumstances. This distinction is not merely semantic, because in fact its consequences are both consistent and real. Without exception, when the human-originated features are removed from the setting of any one of these examples, the corresponding intelligence behaviors disappear right along with them. No primate ever employs an abstract symbol or lexigram in the wild, and no cat answers to its given name in the great outdoors.

Perhaps even more tellingly, the presence of human-like animal intelligence behaviors within artificial settings, and the complete absence of such behaviors within natural settings, assails the very notion of an intelligence produced via the brain—be it animal brain or human brain. The fundamental distinction between wild animals and animals under human influence is certainly not genetic or neural, not unless one is prone to believe in biological magic (as some scientists would seem to be). No, and so the question must be asked: if laboratory- and domestication-based animal intelligence behaviors are the evidence of a capacity for intelligence within the animal brain, then why does this capacity not express itself within non-artificial settings? If for instance the primate brain possesses enough intelligence to produce abstract language and arithmetic within the laboratory, then it certainly possesses enough intelligence to produce abstract language and arithmetic within the jungle—and yet it never does so. Thus it is a logical mistake to attribute animal intelligence behaviors to the animal brain, because in fact the animal brain does not differ from setting to setting, and yet animal intelligence behaviors can differ vastly from setting to setting.

The more logical and direct alternative would be to attribute human-like animal intelligence behaviors to what is actually distinct between the circumstances of the laboratory and the circumstances of the wild, namely the presence of human-generated artifacts and influences within the former setting and the complete absence of such artifacts and influences within the latter setting. The human-originated environment—mostly non-biological, highly constructed, and abundantly suffused with underlying elements of pattern, symmetry, repetition, structure and form—this is the only distinguishing feature to which to attribute human-like animal intelligence behaviors, behaviors that manifest within its presence but disappear upon its absence. This human context—this human intelligence context—cannot be overlooked in the description of laboratory- and domestication-based animal intelligence behaviors, because it is this context alone that emerges as the precipitating factor for the induction of such behaviors.

Furthermore, if human-like animal intelligence behaviors are attributable solely to the presence of artificial and structural features within the surrounding environment of such behaviors—and not attributable to the animal brain—then there is little reason to expect the situation is any different for humans themselves. It is of course not possible to observe directly the comparable behaviors of modern humans versus wild humans (even the few remaining humans yet living under the most primitive of conditions are still thoroughly ensconced from birth in artificial features and circumstances—abstract language, weapons, constructed shelters, etc.). Nonetheless, the evidence from anthropological history reveals that for a very long period of time Homo sapiens individuals lived as little more than animals themselves, engaged almost entirely in a struggle for survival and procreation, and if one were to compare the language, arithmetic, tool usage, shelter creation and other intelligence behaviors between those ancient humans and modern humans, one would recognize that here too, just as was the case in the comparison of wild animals versus laboratory- and domestication-based animals, the gulf in intelligence behaviors is vast—a nearly complete absence in the former case and an overwhelming presence in the latter. But again unless one is prone to believe in biological or evolutionary magic (as some scientists would seem to be), one cannot attribute this vast gulf in intelligence behaviors to genetic or neural causes, because in fact the genetic and neural distinction between ancient humans and modern humans is minuscule, a comparison of a species against itself. Instead, the only non-magical, differentiating influence to which to attribute the vast gulf in intelligence behaviors from ancient humans to modern humans is the large and obvious distinction in their environmental circumstances. For ancient humans their immediate surroundings were as natural as natural can be, void of all artificial constructions and features, the equivalent of a wild animal's domain; while for modern humans the natural world has been practically eclipsed from view, crowded out by surroundings mostly non-biological, highly constructed, and abundantly suffused with the underlying characteristics of pattern, symmetry, repetition, structure and form. For modern humans, the source and inspiration of their many and varied intelligence behaviors can be found literally everywhere close at hand, while for ancient humans such sources and inspirations were literally nowhere to be seen, and this vast difference in setting, coupled with a near equivalence in biological underpinning, strongly challenges the notion of an intelligence produced via the brain.

Thus for this first category of examples offered in support of the existence of an animal intelligence—examples originated in the laboratory, the barnyard and the home—it can be seen that the widespread and unquestioned conclusion that such behaviors are the evidence of animal intelligence produced via the animal brain overlooks what these examples actually convey. The presence of such behaviors exclusively in human-originated circumstances means that this form of intelligence is more human-inspired than it is animal-inspired, and means that far from being produced via the animal brain—a brain that is just as functional and just as present in the wild—these intelligence behaviors are more directly attributable to the human-centric features and influences that constitute the behaviors' surrounding environment, the only type of environment in which these behaviors manifest.

The second category of examples offered as evidence for animal intelligence reverses the circumstances of the first category and consists solely of those cases that occur entirely within natural settings, completely removed from any artificial (that is, human) influence. This includes instances such as dolphins and mollusks employing found objects for shelter and tool, migratory birds navigating by landmarks and stars, squirrels employing deception to safeguard their cache of food, and so on. Indeed the enumeration of examples from this second category would appear to be potentially without end, because it could reasonably be argued that any behavior successfully advancing an animal's quest for survival and procreation might be a viable candidate for inclusion in this category. Scientists are of course apt to concentrate only on those behaviors that have a similarity to well known human behaviors—behaviors that are considered by the scientists to be humanly intelligent—but this clearly reflects some anthropocentric bias and effectively reduces the phrase animal intelligence as applied to this category to be little more than a synonym for animal similarity-to-modern-humanness. Yet even under this tilted approach the ambiguity would still remain. For although a human might cleverly apply deception to safeguard his valuables in certain circumstances, he might also smartly employ brute force when different circumstances arise; and so when the squirrel protects its stash of acorns via deception within the tree and the lion preserves its kill via brute force upon the savanna, are both behaviors to be described as intelligent? In this second category of examples, quite in contrast to the first, the phrase animal intelligence is no longer controversial with regard to its first word, but the controversy now rages full tilt around the aptness of the second word. Since all the examples within this second category reflect natural behaviors, behaviors forged through evolution and generally well ingrained into the species, what justifies the application of the word intelligence to particular instances of these behaviors, and how would these particular instances be distinguished (could they be distinguished) from all other natural behaviors that effectively serve the purpose of advancing an animal's quest for survival and procreation?

If the standard use of the word intelligence within scientific discourse is to be given any weight at all, then there is no question it is a definitional mistake to apply the word intelligence to any biologically natural behavior, be it successful or otherwise and be it similar to human behavior or not. This is seen most clearly and directly by inspecting the contents of that preeminent tool for measuring intelligence, the IQ exam. Although this feature is often overlooked, forgotten or ignored, an IQ exam deliberately and categorically excludes many types of behavior from its jurisdiction. For instance, a test-taker's athletic ability never comes into play—one's ability to run, leap or throw neither helps nor hinders one's performance on an IQ exam. And more germane to the discussion at hand, an IQ exam never assesses a test-taker's ability to survive or procreate under primitive conditions—one's ability to vanquish predator or prey, and one's likelihood to foster a prodigious lineage neither helps nor hinders one's performance on an IQ exam. What an IQ exam does measure is a circumscribed and biologically foreign set of capabilities and behaviors, targeting a test-taker's capacity to respond productively to a series of challenges constructed almost entirely out of artificial components, components carrying the underlying characteristics of pattern, symmetry, repetition, structure and form. It is by these circumscribed and biologically foreign means that an IQ exam can capture the type of human-like intelligence behavior that was the focus of attention under the first category of examples offered for animal intelligence, but it is also by these same means that an IQ exam excludes the type of biologically natural behavior that is the focus of attention under this second category of examples offered for animal intelligence. By design and by intent, biologically natural behaviors are to have no influence on intelligence as measured by an IQ exam, and thus any resemblance of natural behaviors to behaviors that are measurable by an IQ exam must be taken as nothing more than an accidental coincidence.

The fluidity of an IQ exam's contents, along with the corresponding fluidity of what those contents measure, provides still more justification for uncoupling entirely all intelligence behaviors, which are quite malleable, from all biologically natural behaviors, which are not malleable at all. For instance, it has been well documented that due to the Flynn effect intelligence exams must be re-engineered on a regular basis, recasting questions to be more sophisticated and challenging as time goes on. Questions assessing a test-taker's general knowledge, which would have been quite localized in the past, today must encompass a global, indeed a universal, scale; and questions covering topics such as mathematics, logic, vocabulary and grammar would have looked quite different if set hundreds of years ago versus how they are set today, and will morph still further when presented in the future (think of the way in which electronic communication is altering the rules of vocabulary and grammar even today). Plus it is not just the shifting nature of the IQ exam that attests to this fluidity of intelligence; in everyday usage and in general application it can be seen that intelligence behaviors have an inherent tendency to change over time. A human of the past who could use his scythe to harvest grain would have been described as reasonably intelligent, but today's farmer who cannot advance beyond the scythe to the mechanisms of the combine will be assessed as far less so, and the engineer of the future who cannot transcend both the scythe and the combine to master the intricacies of the automated process will be seen as intellectually left behind. In short, intelligence behaviors do not stand still, they do not solidify into long-term habit or an enduring nature. Intelligence behaviors are generalizable, they can be quickly advanced. Intelligence behaviors are promulgated rapidly and then widely transformed.

By contrast, biologically natural behaviors are characterized precisely by the fact they have become so deeply ingrained, the predictable and enduring aspects of the species in particular and the animal kingdom in general. Forged through evolution and constrained by biological pressures, natural behaviors transform on only the rarest of occasions and under the most extreme of circumstances. Almost every offered example of animal intelligence that falls within the domain of this second category—dolphins employing sponges as tools, birds navigating by landmarks, intricate nest building, insect communication dances, coordinated pack hunting, etc.—every one of these behaviors would have been observable exactly as it is today a hundred thousand years ago, and will be observable exactly as it is today a hundred thousand years into the future, with no generalization, no advancement, no promulgation, no transformation. If there were exams for measuring the capacity for any of these natural behaviors the exams would never need to be revised but could serve their purpose faithfully millennium after millennium after millennium.

One of the more prominent examples of this tendency to mistake an ingrained natural behavior for an intelligence behavior comes from the history of humans themselves, in their usage of stone tools. Although widely accepted as at least a precursor to intelligence, the ancient employment of edged choppers nonetheless went ungeneralized and unchanged for hundreds of thousand of years and thus was more akin to something like nest building in the birds than to anything artificial or modern. It was not until the human tool set suddenly transformed, transformed in material and categorization (and quite recently in human history and accompanied by dozens of other behavioral changes) that the species found itself rapidly rearranging its environmental circumstances and marching hurriedly towards an intelligence age. If there were a man today who could master no more than the flaking of some flint while at the same time being utterly dumbfounded by hammers, pliers and awls, he would not be regarded as intelligent, and neither would his long line of descendants if they were to somehow become stuck on these same stone choppers generation after generation after generation.

In comparison to biologically natural behaviors, intelligence behaviors are strange, fleeting and foreign, traveling almost exclusively in the company of artificial constructs; intelligence behaviors share essentially none of the enduring evolutionary characteristics that define biologically natural behaviors. Intelligence behaviors are not driven solely by a need for survival and procreation, intelligence behaviors do not become deeply entrenched, and intelligence behaviors were nowhere to be seen on this planet until humans quite recently and quite suddenly and quite prodigiously took them on. The many offered examples of animal intelligence that fall within this second category are in fact contradictions to the word intelligence, their biological and evolutionary underpinning meaning ipso facto they are to be excluded. That these instances are so frequently offered as examples of animal intelligence can be attributed primarily to a cause that is both obvious and quite benign, namely that these examples bear an accidental resemblance to modern human behaviors and scientists are unable to suppress their anthropocentric bias.

Thus for the animal species other than Homo sapiens, the phrase animal intelligence finds no meaningful application, its legitimate instances forming a null set. All the examples that would fall under the first category of animal intelligence—examples originated in artificial settings—fail on the word animal, and all the examples that would fall under the second category of animal intelligence—examples originated in natural settings—fail on the word intelligence. On Earth, the phrase animal intelligence currently attaches to humans and to humans alone.

Nonetheless, even with humans, the phrase animal intelligence has the most uneasy and paradoxical application, its two words representing two fundamental and opposing aspects of human nature. Indeed modern humans might be more articulately described by the phrase animal versus intelligence. This conflict arose of course historically, for man was once nothing more than animal himself and like the other species was a complete stranger to intelligence, a complete stranger to any artificial construct composed out of pattern, symmetry, repetition, structure or form. In those ancient yet not-so-long-ago days (not so long ago, that is, on any biological or geographical timescale), evolution was still the primary master, and the quest for survival and procreation was still the sole motivator. But today, for modern man, the tables have been nearly turned: nature has been practically eclipsed from view, survival and procreation have been mostly tamed by artificial means, and the evolutionary process has been turned completely on its head. The one great unanswered scientific question of the present day (a question only humans are capable of asking) is how can this sudden and prodigious transformation be described and explained—what are its characteristics, what brought it about, and what are its ultimate consequences?

The popular and conventional solution to the problem of explaining modern man and his burgeoning intelligence is to claim everything must have arisen as an evolutionary event. Some would describe this event as gradual (to meet the requirements of biological and evolutionary principles) and others—Richard Klein, for instance—would describe this event as sudden (to meet the requirements of man's surprisingly rapid turn), but in any case the essential requirement is that human intelligence be depicted primarily as a product of biological evolution, something akin to a genetic mutation producing a cascading neurological effect, because of course all animal transformations are products of evolution, are they not, can there really be such a thing as an exception?

But in fact intelligence is the exception. Not an exception, but the exception. Intelligence contravenes evolution, intelligence is thoroughly anti-evolutionary in its process, cause and effect. On Earth, man with his newfound intelligence has become an anti-evolutionary creature, an anti-evolutionary force, producing astounding environmental impact, observable literally everywhere close at hand.

Most of the grounds for this determination have already been stated. When describing the circumstances in which intelligence behaviors (be they human or animal) exclusively manifest, it was noted that intelligence behaviors always appear within artificial settings and are always attached to constructed circumstances. Intelligence performance is measured primarily by an artificial instrument, its content composed entirely out of artificial components and its domain enjoined from measuring any athletic or instinctive ability. When comparing intelligence behaviors to biologically natural behaviors, it was noted that intelligence behaviors are fluid and accumulative, not conservative and ingrained, and intelligence changes are predictable via generalization and rapid promulgation, not random (as is the case with geological and gene mutative events). These already stated observations can be summarized into one readily apparent fact: the characteristics of intelligence are incompatible with the characteristics of biology and nature, the characteristics of intelligence are diametrically opposed to the characteristics of evolution.

This fact becomes even more apparent when comparing the dynamics underlying intelligence and evolution. Evolutionary dynamics are well known and straightforward to describe. Given a stable environment, organisms (as the genetic representatives of their species) undergo selection for that environment through a striving for survival and procreation. The organisms which are best suited to the environment will more likely emerge as dominant and established, and the organisms which are less suited to the environment will more likely diminish by being dominated and crowded out. Environmental change fosters some transition and churn, as does random genetic mutation, but since significant environmental change tends to be rare in typical circumstances and since random genetic mutation tends to be a long shot for increasing environmental fit, evolution tends to be a conservative and slow moving process, with significant alteration often taking place on the scale of hundreds of thousands or even millions of years.

The dynamics underlying intelligence are composed of these exact same components—environment, selection, mutation, survival and procreation—but these components are arranged in an alternative pattern, producing a fast-moving process that runs in evolution's counter-direction. Intelligence begins with just one species, a species that will not undergo any significant genetic mutation. The organisms of this species, once subject to natural selection as with all the other species, begin to circumvent selection by substituting instead its artificial counterpart, selectively mutating the organisms' environment, mutating it in such a way as to make the surrounding conditions more supportive of the organisms' survival and procreation (and generally less supportive of neighboring species' survival and procreation). The key to this deliberate, non-natural environmental mutation is an awareness of the environment's underlying characteristics, characteristics that are mostly non-biological and abundantly suffused with elements of pattern, symmetry, repetition, structure and form. The organisms of this species make use of these characteristics to mutate their environment in a self-preserving manner—think clothing, shelter, weapons, abstract words—and these mutations also serve as the prominent, long-lasting embodiment of the characteristics themselves, spreading knowledge and awareness to the other members of the species. And since these non-biological environmental mutations are not subject to any genetic or geological constraint, they can be attempted with accumulating impact and with increasing speed, a fact currently being demonstrated by humankind on a nearly daily basis.

Thus to make use of intelligence is to defy evolution; a species acquiring intelligence is turning evolution on its head. Or to make the comparison more direct, with evolution, the environment selects among mutating organisms for the best environmental fit, while with intelligence (anti-evolution), the organism selects among mutable environments for the best organism fit. Evolution and intelligence are opposing forces, they run in counter directions.

The recent history of Homo sapiens might tempt one to think that in the conflict between evolutionary animal and anti-evolutionary intelligence, intelligence must be emerging as the victor. Humans have been moving farther and farther away from their former animal circumstances and now live in settings where the artificial features outnumber the natural features on a scale of perhaps a hundred to one, maybe even a thousand to one. Intelligence has been measurably increasing population-wide year after year (the Flynn effect), and intelligence has grown so copious that a good portion of its augmentation is no longer directed to survival and procreation (understanding of the Big Bang for instance, intriguing though it might be, is not likely to impact human continuance anytime soon). The human transformation has been charting what seems to be a direct course, straight from all animal and no intelligence to all intelligence and no animal.

But the temptation to think in this way is merely an illusion. The abundant increase in human intelligence, admittedly quite real, masks a reality that is just as important and just as essential, namely that the animal in man has gone nowhere at all, man's animal nature has diminished not one bit. The primary justification for this conclusion is of course the fact that the requirements of survival and procreation remain in full effect. Although intelligence and its many constructions have certainly eased the immediate challenges of survival and procreation for most people, and have sheltered the human species from a broad array of biological contingencies, these protections are nonetheless fragile and increasingly complex, and therefore not guaranteed to last. Nuclear arsenals, the profligate destruction of climate, eradication of supporting species, plus hundreds of other unseen vicissitudes—ruinous catastrophe seems to lurk around nearly every bend and along with it a return (at best) to an ancient and bestial existence. But even discounting this potential for civilization collapse, even assuming circumstances will continue as is, man's animal heritage must still insist on having its say. Greed, lust, rivalries, nepotism, revenge—these easily traceable holdovers from man's primate beginnings serve not only as the most captivating plot devices in popular forms of modern entertainment, they serve also as the most compelling motivators of day-to-day action in a modern human society. So resilient has been the primitive impulse within human temperament that the efforts of intelligence have often been the most successful not when confronting the lingering animal within man but instead when assimilating it, even sublimating it, and thereby channeling and dissipating much of its pent-up energy. Anyone who has ever witnessed up close the inner workings of a modern corporation, and has experienced first hand the often devious and sometimes brutal scramble towards executive office and boardroom key, will have a perfect acquaintance with the vestigial features of a hierarchical clan. Anyone who has ever attended the clamorous and furious taking up of sides in the giant arenas of battle—the home warriors clashing against the invading marauders—will have an intimate familiarity with the fears accompanying territorial battle and will have rediscovered a hunger for the spoils of conquest. Even when intelligence has been at its most powerful and progressive, even when building sophisticated constructions capable of advancing the entirety of the human species, intelligence has been nonetheless helpless against the visceral traits of that species permeating the final results. Witness one of the more recent triumphs of intelligence, witness the development of widespread electronic communication, capable of spreading advancement to literally all, and then witness the most popular and frequent use of that instrument, as the preferred and efficient conduit of gossip, fraud and pornography.

Some humans will no doubt feel an urge to pull back towards man's more familiar animal past, not comprehending that such a retreat means a return to the exigencies and stasis of a primordial existence. Other humans will desire that the forces of intelligence eventually conquer the animal within man, not recognizing that such an outcome puts an end to all vitality. What remains elusive is what might be the purpose behind the introduction of intelligence into an animal species—what could be the ultimate goal—but whatever that purpose or goal may be, the fundamental conflict so engendered appears to be at its most productive only while it is being waged, or perhaps while it is being transcended, and not when it is being won. The fate of modern humans is tied to this ongoing struggle between animal and intelligence, between evolution and anti-evolution, and it is in this way (and in this way only) that the phrase animal intelligence acquires legitimate and substantive meaning.

A remaining challenge regarding the human transformation from animal to intelligent being is to pinpoint exactly what was it that set this transformation in motion, and what continues to sustain its energy through the present day. One of the more logical answers to this challenge turns out to be both unexpected and provocative—far too unexpected and provocative to be taken up here—but one of the characteristics of this answer can be anticipated from the present discussion, anticipated from what has been said about the enormous opposing gulf that lies between animal and intelligence. Any species transitioning from animal to intelligence is performing a gigantic about-face, a turn from evolution towards anti-evolution, and is performing this turn against what must be an overwhelming inertia. There have been thousands and thousands of species come and go on this planet over millions and millions of years, and yet the human acquisition of intelligence has been an unprecedented occurrence, the rarest of biological happenings, the most atypical event since life began. So if one were to go in search of a cause for this atypical event, the places that one need not bother to look, the places almost guaranteed to supply essentially nothing in the way of meaningful information, would be the usual locations (the places where today's scientists are most likely to gather): normal biology, normal neurology, normal genetics, these are almost certain to be normal dead ends. And if one were to go in search of a cause for this atypical event within humans themselves, within the actual members of the population, then the one type of person one need not bother to consider, the one type of person almost guaranteed to supply essentially nothing in the way of meaningful information, would be the typical person (the type of person today's scientists are most likely to examine): the average primate neighbor, the average representative of the species, these are almost certain to be average dead ends. The atypical human transformation, the unexpected and provocative turn from animal to intelligence, the abnormal about-face from evolution to anti-evolution—if all this is to be explained, it will be explained not by what is usual but by what is unusual. If all this is to be explained by reference to humans themselves, it will be explained not by those who are normal but by those who are abnormal, explained by reference to those who are the most atypical members of the population.