Birds eat spiders. Snakes eat birds. Usually these two processes are unrelated, except in the deserts of Iran, rocky and barren, where a snake called Pseudocerastes urarachnoides lives. It looks much like any other desert-dwelling snake, with mottled scales that allow it to blend into its brown and gray background. By the early 2000s, scientists had seen a preserved specimen. They noted that it had an oddly lumpy tail with fingerlike extensions at the tip, but they didn’t know whether the protrusion was a malformation in the individual that had happened to be captured, or a natural part of the animal.1 Then they found a half-digested bird in one of the snakes. And they started to wonder.
It turned out that the knobby bit on the tail is a lure, used to attract prey. Several other snakes and lizards have body parts that resemble tasty bits of food—a worm, an insect, or other item that would appeal to something the reptile would eat. The Iranian snake is called the spider-tailed viper, because its tail not only looks but also behaves exactly, and I mean exactly, as if it were a spider, with appendages that alternately pause and scuttle like the real thing. The first time I saw a video of the snake using the lure, I was sure the filmmakers were illustrating how the tail lure worked by somehow placing a real spider on the body of a snake, an act that in retrospect seems improbable at best. But the mimicry is that good: the end of the tail has scales that are shaped to resemble the legs of a spider, each of which skitters over the body of the snake looking for all the world like they are connected to a spider body. It’s not just that it looks like a spider; the tail acts like a spider, and birds apparently perceive it as one too.
In his splendidly titled blog Life is Short but Snakes are Long2 (the name comes from a book review by David Quammen), Andrew Durso from Florida Gulf Coast University suggests that the tail “probably represents the most elaborate morphological caudal ornamentation known in any snake” (caudal means tail-end), and I agree, only without the qualifier “probably.” But the question is, how did such a precise and elaborate mimicry evolve? Not to put too fine a point on it, but snakes just aren’t that bright, and they don’t have very good eyesight, and they probably don’t spend any time looking at their own tails, and even if they did, they couldn’t simply will a spidery appendage into existence. Furthermore, the snake cannot possibly be aware of how a spider moves, and can’t know that a spider will attract prey, and can’t modify its tail to behave in an ever-more alluring manner based on the response it gets. And where would the snake have gotten the idea in the first place?
Before answering that question, consider another example, one that doesn’t combine quite so many phobias: bumblebees gathering pollen from flowers. Pollen is a rich source of protein for the adult bees and their young, but of course it is only available when plants are flowering. That means that if bees emerge from winter hibernation and are establishing their colonies too early in the season, before plants bloom, they risk starvation. A group of scientists from Switzerland noticed that some bumble bees were making tiny holes in the leaves of some of the plants, and that the damaged plants flowered much earlier.3 Experiments both in the laboratory and in the field showed that the bees were much more likely to make the holes when they were starving; well-fed bees left the plants alone to flower at their leisure. This meant that hungry bees got their pollen sooner. What is more, the plants did not respond to just any holes in their leaves—if the scientists attempted to mimic the bees’ damage, flowering was not accelerated nearly as much. Whether the bees have something in their saliva that induces changes in the plant’s reproduction remains unknown.
A commentary on the study gave a pithy summary: “Pollen-starved bumble bees may manipulate plants to fast-forward flowering,” and called the behavior “a low-cost, but highly efficient, trick.”4 A trick it may be, but no one, least of all the scientists who discovered the behavior, would suggest that the bees perform their bit of horticulture consciously: bees are hardly examining the landscape, fretting over the dearth of pollen, and selecting plants to chew on with the expectation that this will eventually yield food. So, as with the spider-tailed viper, how did such a complex behavior evolve? One could argue that the bees are exhibiting an even more sophisticated behavior than the snakes, because in the bees’ case, the reward is delayed, whereas with the spider-tailed viper, food in the form of a hungry bird arrives right after the lure is deployed.
Both examples certainly illustrate the futility of expecting a “gene for” any behavior given the number of nerve cells, muscles, and other tissues and organs that are involved. There are no genes that direct agricultural activities or manifest spiders from snake scales. Even with that caveat, the evolution of these behaviors is hard to imagine. But if we think about behavior the way we would think about a physical characteristic, it is easier to approach.
First, let’s review the way that any nonbehavioral characteristic evolves. Most of us understand the basics of evolution via natural selection. Imagine a population of living things—let’s say a rain forest–dwelling group of birds—that is naturally variable in plumage color, with some that are greenish, some that are blue, and some that are glossy black. The variation is there initially because genes produce lots of differences, both because of how they are combined when two parents reproduce and because mutations continually arise by chance. In our case, let’s say the green birds that match the trees are the most likely to evade detection by predators and survive, which means that the genes associated with green feathers are more likely to be passed on by green-feathered parents.
Eventually the population contains more green birds than the other colors, since natural selection has winnowed out the more conspicuous prey. Even after most of the population is green, any modifications that make the birds better camouflaged—say a break in the pattern so that the feathers more closely resemble leaves, or a spot that looks like a bit of decay—will still mean that the parents with better protection are more likely to have offspring. Note that mere survival via the more cryptic plumage isn’t enough; the leaf-resembling birds also have to successfully reproduce, or else it won’t make any difference to the gene pool in subsequent generations. Selection doesn’t need to know the genes involved, just their product, as in this case the color of the plumage. And it seems self-evident as well that a bird doesn’t need to know what color it is, or actively try to change its feathers.
When we talk about behavior evolving, however, the process can seem a bit more indirect. Behavior comes and goes, unlike a tail feather, so it’s hard to see how individuals who did something an hour, a day, or a year ago are selected to have more babies later in life. Furthermore, behaviors seem to require agency, an internal urge to do something, and it is hard to see where that comes from. Along those lines, the Minneapolis Star Tribune has a regular birding column, and one week it featured nest building. In it, columnist Val Cunningham mused:5
Each species [of bird] has its own nest style, and here’s an amazing thing: No bird has ever observed its parents building their nest, yet in her very first season a female bird (it’s usually the female) builds exactly the nest characteristic to her species. How can that be?
How indeed? The article says that it’s “instinctual, hard-wired into their brains.” But that doesn’t answer the question about behavior, any more than saying that snakes wiggle their tails and bees punch holes in leaves because they have tail-wiggling or leaf-punching instincts. Where did the instinct come from in the first place?
A more concrete answer is that animals performing intermediate steps in complex behaviors had an advantage. In the snakes, say that the ancestors of the spider-tailed vipers had caudal lures, and already vibrated their tails when hunting, as many snakes do. The vibration seems to distract potential prey, which then attend to the tail and ignore the lethal end of the snake, to their doom. Then imagine that a few of the ancestral vipers happened to have small projections on those caudal lures, like warts. If the warty tails were more likely to attract birds than non-warty ones, the wart-bearers ate more, were more likely to survive, and had more warty babies (sorry for the mental image, particularly for the snake-averse). The snakes that happened to combine ever-wartier tails with movement got even more birds; the ones with movement that looked more spidery did even better, and so on. It’s not hard to figure that some snakes are just wigglier than others; after all, even humans vary in how much they fidget, a characteristic that some scientists think is linked to our metabolic rates.
In the story about the spider-tailed viper, scientists call the prey (birds) agents of selection. The birds find certain tails more alluring than others, so they choose them, without the snakes having to do a thing. It takes many, many generations for the snakes with the spider-tipped tails to predominate, but then snakes are of a very old lineage, having arisen over one hundred million years ago. That gives them a long time for extremely small changes to accumulate.
We can construct similar scenarios for the bees that tear leaves, and for the birds that build the best nests, with similarly small variations that confer an advantage to the bearer. Bees already bite at plants sometimes, and some of the dinosaur ancestors of birds appear to have gathered materials in their surroundings when they laid their eggs. Those rain forest birds similarly grow greener and greener as their feathers accumulate the right kind of pigment. The important point is that it doesn’t matter why the variant is produced—it’s all about the consequences. So if a twitchy tail means the snakes eat more and are around to have more babies, a twitchy tail will appear more frequently in the population, whether the snakes are aware it’s twitchy or not. The same process applies to behavior that has worked to produce extraordinarily complicated structures such as the eye (a favorite target for creationists, who often assert that such organs could not have arisen via evolution); again, a series of intermediate steps, each of which yields an advantage to the bearer, is all that’s required.
Behavioral Family Trees
That series-of-intermediate-steps answer, of course, pushes the rise of these incremental behaviors back in evolutionary time without really answering the original question. Where did caudal lures come from? Why did dinosaurs build nests? And did they all do so, or just some? It is a truism that behavior doesn’t fossilize and become memorialized in the geologic record, but it is still possible to understand the evolution of behavior over deep time.
First, although behavior itself doesn’t turn to stone, it is still possible to infer what animals were doing in the past from their bodies as well as the ways in which their bodies are preserved. For example, we think that some dinosaurs took care of their young because groups of a single adult with several juveniles were found fossilized together. While virtually all modern lizards lay their eggs and then abandon them, the descendants of dinosaurs—birds—are champions of parenting. And fossilized animal footprints or burrows can tell a great deal about how an animal moved and what it ate, as can the structure of body parts like teeth and limbs. One group of scientists claimed that the grooves on a tooth of one Tyrannosaurus rex were made by another individual of the same species, which they concluded meant that the famed dinosaurs were cannibals. In a similarly gruesome example, the fossilized remains of a ten-foot-long snake from seventy million years ago was discovered encircling a crushed dinosaur egg in a nest of otherwise unbroken eggs. Michael Benton of the University of Bristol agreed with the authors of the study that the “snake was waiting and snatching juveniles as they hatched,”6 which gives new meaning to the term cold-blooded.
It is also possible to reconstruct behavior by carefully analyzing fossil skeletons: an animal’s stance, its jaw formation, and the relative sizes of the bones in its legs can all reveal a great deal about what it ate and how it behaved. The development of 3D scanning and printing techniques has led to ever-more sophisticated models of ancient life. To return to T. rex, that ferocious epitome of dinosaurs, scientists, Jurassic Park aficionados, and six-year-olds have all long wondered how fast the predators could run. Estimates made over decades had ranged from over forty mph to eleven mph, the latter being about the speed of a human long-distance runner. Who is correct? A recent study7 pointed out that at very high speeds, an animal as large as a T. rex would have risked toppling over, and furthermore that the muscles needed to power its hind legs would have had to comprise 86 percent of its body mass, a virtual impossibility that would have left no room for any other body functions. Hence, its likely maximum pace was close to the lower estimate, which still means the dinosaur might have been able to catch a fleeing caveman—except that, luckily, none would be available for at least sixty-five million years after T. rex became extinct.
Another way to infer how behavior evolved doesn’t use fossils at all. One of the most intuitive ways to understand the evolution of any characteristic is to think about its similarities in other living things, and how those similarities came to be. Evolutionarily speaking, objects—or behaviors—can resemble each other for one of two reasons. To illustrate, imagine an array of limbs from different animals: a whale flipper, a bird wing, a human arm, an octopus tentacle, and a starfish arm. Which of these is not like the others? They are all used for movement, but we know that the first three limbs share more than the same function. They have bones that are similar to each other because they are inherited from a common ancestor. In an X-ray, you can see a humerus, that long arm bone that extends from the shoulder, in each, though the shape and position are modified (and whales don’t have shoulders, exactly). The bones started out the same, in an ancestral vertebrate millions of years ago, but became modified through natural selection by the particular circumstances in which each limb found itself, whether in water, or air, or on land. This type of similarity is called homology.
In contrast, both the starfish and octopus arms not only lack bones, but these animals have not shared a common ancestor for far longer than any of the vertebrates noted here. They evolved from more recent, and armless, ancestors. Selection favored extensions of their bodies in each case, but the resemblance between the limbs happened independently, through a process called convergence, or convergent evolution.
Similarity because of a common ancestor or because evolution produces similar structures through different pathways are both common. North American flying squirrels and small Australian marsupials called sugar gliders are adorable, large-eyed, nocturnal mammals with furry membranes between their front and hind legs that are used by the animals to glide between trees. But they resemble each other not because of a mutual gliding ancestor, but because of convergent evolution.
But back to behavior. People have noticed for a long time that species that look somewhat alike often act similarly as well. Take hummingbirds, for example. Many of these tiny New World birds use acrobatic aerial displays, sometimes called skydancing, to attract mates. A male hummer will ascend into the sky at high speed, then zip up and down or back and forth in a U-shaped or oval pattern. The wings—not the vocal apparatus—of the bird make snapping or whistling noises as part of the display. Each species has its own variety, with Anna’s Hummingbirds flying in a tall, narrow oval; and Costa’s Hummingbirds using a much shallower path. The two species look somewhat alike, and their behavior is similar as well.
The earliest animal behaviorists, such as Konrad Lorenz, who helped develop the theory of imprinting, were fascinated by these similarities between behavior and appearance, and drew elaborate diagrams showing how such displays differ across species. But perhaps because people often think behavior is different from physical characteristics, given its fleeting nature, as I discussed in the last chapter, using such homology to understand the evolution of behavior has invited skepticism over the years. Some scientists thought behavior was just too variable to use in evolutionary studies, while others were concerned about the lack of fossilized behavior. Peter Klopfer, mentioned in chapter 1, found behavior “too malleable” to draw any conclusions about its evolution. And the late, famed paleontologist Stephen Jay Gould flatly said that “it might be interesting to know how cognition (whatever that is) arose and spread and changed, but we cannot know. Tough luck.”8
Many of these objections, however, could equally be raised about understanding many other characteristics, not just behavior. Physiology and its associated tissue don’t fossilize, or at least not very well, but we can draw many conclusions about similarities in digestion by comparing animals eating different things. Hearts are not preserved in stone, but we are pretty sure that the four-chambered variety in mammals arose from a common ancestor with crocodiles and alligators, who also have the same kind. And when you actually measure the variability in behavior, as scientists Alan de Queiroz and Peter Wimberger did9 by examining both physical and behavioral characteristics in animals ranging from wasps to newts to birds, it turns out that behavior isn’t any different from leg length or tooth size in how easy or hard it is to describe or how much it differs among individuals.
With that cleared out of the way, it becomes possible to use homology of behaviors in a different way. Instead of asking how behaviors that we see could have evolved, given a set of previously determined evolutionary relationships, we can see if behaviors shed light on those relationships themselves. It is kind of like the chicken and egg question, but with actual chickens. Or at least actual birds.
Let me explain. In this example, we’ll be looking at manakins, small songbirds that live in the tropical forests of Central and South America. About forty different species of manakins exist, and the males engage in elaborate courtship displays, doing fancy dances showing off their brightly colored feathers and making sounds that, a bit like the hummingbirds, are produced by the males’ wings, not their vocal system. Some species also have unusually thick bones and strong musculature associated with moving the wings. Kim Bostwick, an ornithologist at Cornell University, has studied manakins for many years. She was particularly intrigued by the Club-winged Manakin; it has an extraordinary display in which, among other acrobatics, the male turns away from the female, bends over, and shuffles backward while keeping his rear end in the air. If that sounds bizarre, well, it is, but it is also a display that has some similarities with another kind of manakin. Bostwick figured that the ancestor of the two types of manakins must have had the roots of the display, and both species then inherited versions of it.
However, when scientists looked at the anatomy of the two types of manakins, they were not that similar, and the species were judged to be evolutionarily rather far apart. So Bostwick painstakingly amassed information on both the mating displays and the bones, feathers, and other physical features of not just the two manakins in question, but of as many members of the group for which she could find good specimens.10 Then she constructed evolutionary trees that showed how one aspect of the display, or one thickening of a bone, could be ancestral to the others, with further modifications as the species evolved. The trees revealed that the behavior and the appearance of the manakins evolved together, so that as the mating display got more sophisticated, the manakin’s bones became heavier, which enabled the exaggeration of the wing noises that were used by ancestral manakins. The various parts of the display are homologous in the different manakin species, each related to the other, meaning that the behavior can be used along with the physical attributes of the bird to reconstruct the evolutionary history of the group. This combined approach is better than a history that uses only the anatomy of the birds.
Just like limbs, behaviors can exhibit either homology or convergence. Among snakes, defensive behaviors such as hoods or open-mouth displays seem to be the result of convergence. In crocodiles, parental behavior is common, and is thought to be related to the care of young seen in modern-day birds, so that is a homology. What all of this means is that we can trace the evolution of behaviors in much the same way that we trace the evolution of jaws or feathers. We can also think about how quickly or slowly behavior evolves by examining how persistent behaviors are over long stretches of evolutionary time. Some behaviors, like tool use, seem to have arisen relatively quickly. Others appear to have remained unchanged for much of evolutionary history. Virtually all four-limbed vertebrates, from birds to mice to lizards, scratch their heads by lifting a rear foot over the front leg, or over the wing (humans are an obvious exception, but then we have those handy fingers). Eminent animal behaviorist John Alcock speculated that grooming behavior isn’t subject to selection as animals compete and enter new habitats or stay ahead of predators—what works to scratch an itch on a mountain will do equally well in a forest and won’t have an impact on the animal’s survival or mating.11 On the other hand, a behavior that allows an animal to evade attack in the desert will fail miserably in the marshes, so one might imagine that such behaviors change as selection on them changes as well.
In summary, it isn’t a matter of which came first, the behavior or the physical appearance—like the chicken (or the manakin) and the egg—because you always need one to produce the other.
Scales, and a Head Full of Lizard
Evolution gives us convergence, so that structures that appear similar can have different evolutionary ancestry. It also gives us homology, so that structures that appear different can have a common origin. We also know that simpler forms gave rise to more complex ones. That’s true for appearance—an amoeba is less complicated than a kangaroo—and for behavior—those spidery-tailed snake displays are more elaborate than simply waving the end of a tail back and forth.
From those principles, it’s easy to develop a very common misconception, one that reveals itself in all those cartoons that show a fish sticking its head above water on the shore, followed by a reptile, then a four-footed mammal, then an ape, then a caveman (almost always with a spear), and finally a human doing something like eating a cheeseburger or typing on a keyboard. The idea of the drawing is that we are progressing ever onward, with each form more advanced than the last. Humans, then, are the pinnacle of evolution.
This belief in a hierarchical classification of living things, termed the scala naturae, is an old notion. Aristotle arranged all living beings along a scale with (predictably) humans at the top, and the other creatures beneath us in decreasing complexity. In one rendition, angels were included, and were seen as closest to God, followed by humans, and then other animals. Similar ideas have been perpetuated over the centuries, with one of the more recent versions being a “ladder” of evolution, so that again, humans are at the apex, preceded by other mammals, which are themselves preceded by reptiles, fish, and on down to the invertebrates, each with its own rung. It is as though living things are in a gigantic military, with microbes or worms as the privates and people as the generals.
Persistent though it is (one can still find references to an evolutionary ladder in some modern textbooks), the scala naturae is not just outdated, it is completely erroneous. It is true that some groups of organisms, including humans, arose more recently in evolutionary history than others. But recent evolution isn’t an award, it’s just an attribute. Domestic dogs arose more recently than wolves, but also more recently than humans. The novel coronavirus arose more recently than leprosy, and both of them are newer than people. So who is at the top? Evolution leads to a bush, not a ladder, with the branches of the bush representing change from one form to the next over time.12 The tips of the branches can be thought of as the living things that are present now, each one as highly evolved as the other.
The scala naturae rears its head when it comes to behavior, because it can make us think that if everything is always getting more complex, then behavior must exist on a continuum, with simple actions like the tumbling bacteria giving way to a tail-wagging dog, a nest-building bird, and eventually, a brain-surgery-practicing human. And since the brain ultimately produces our behavior, it can be tempting to see behavior evolving the same way, as if mammal brains are like reptile brains, but with added complexity. If so, then as newer, more sophisticated parts of the brain evolved, they were layered onto the older ones. It’s like the notion of human ice cream bars I alluded to in the last chapter.
This concept might sound familiar to anyone who has heard—or made—a reference to a “lizard brain” that makes decisions based on emotion or instinct rather than reason. Called the triune brain theory and developed by psychologist Paul MacLean in the 1940s and 1950s,13 the lizard brain was also popularized in Carl Sagan’s famous book The Dragons of Eden.14 MacLean postulated that the modern vertebrate brain had three units: an atavistic reptilian one that only takes itself into account, an early mammalian one that is emotional but potentially unselfish, and a more rational later mammalian one. The interaction of these components produces the often-contradictory behaviors and impulses we see enacted.
As neurobiologist Anton Reiner noted in a review of one of MacLean’s books,15 the triune brain idea has a lot of appeal. It allows us to classify less-desirable behaviors, like flying into a rage at the checkout line in the supermarket, as somehow nonhuman, and perhaps therefore something we are not responsible for. It helps fuel that sense of human exceptionalism I discussed in chapter 1, as though even our brains represent the latest, most effective model, the one that superseded the clunky old-fashioned brain our distant ancestors were forced to use. It encourages us to dismiss “instincts” as a holdover from our past, separate from a more reasoned way of thinking. And, as Reiner says, MacLean’s ideas “are also appealing because they are simple; after a ten-minute exposition of them one can feel equipped to explain much human behavior with the force of science behind one.”
The problem is that the triune brain, like the scala naturae, doesn’t exist, or, as the title of a 2020 paper puts it, “Your Brain Is Not an Onion With a Tiny Reptile Inside.”16 (Whether you find the onion or ice cream bar metaphor more appealing is, I imagine, a matter of personal taste.) Modern neurobiology shows quite clearly that brains do not possess “newer” parts affixed onto “older” ones, and furthermore, that brain and behavioral complexity do not map onto any kind of evolutionary sequence. And we cannot assign whole classes of behavior, such as territoriality or aggression, to a particular part of the brain, since the brain and nervous system are much more integrated than the triune brain model implies. Finally, MacLean considered parenting to be part of the more advanced behavioral repertoire, and hence associated with the mammalian part of the brain, even though crocodiles, birds, and at least some dinosaurs took care of their young.
Yet the model has been hard to debunk, even among psychologists. The authors of the brain-is-not-an-onion paper, led by Joseph Cesario at Michigan State University, surveyed recent introductory psychology texts, and found some version of the triune brain misconception in nearly 90 percent of them.17 Cesario and his coauthors point out that contrasting an ancestral, impulsive, “animalistic” nature with a more long-term rational one lies behind research on willpower that touts the ability to forego reward with a more mature outlook on life. Instead, they argue that sometimes choosing the immediate reward is more beneficial than waiting; it all depends on the circumstances. Thus, they say, “The question of willpower is not ‘Why do people act sometimes like hedonic animals and sometimes like rational humans?’ but instead, ‘What are the general principles by which animals make decisions about opportunity costs?’ ”
Getting to the Genes, but Not the Way You Think
All of this talk about fossils and brains and similarities among different kinds of animals leaves out the link between genes and behavior itself, a link that obviously exists for behavior just as it does for physical characteristics. I will explain how we know which and how many genes are associated with specific behaviors in more detail in the next chapter, but before I do, it is worth noting just how indirect the connection can be between any one, or even many, genes and the behavior that we observe.
Let’s start by thinking about dogs. People quite happily acknowledge that while training and obedience classes can shape a puppy’s behavior, some of that behavior also comes from its breed. Retrievers are called that not because of how they look, but because of what they do. Dog owners are happy to wax rhapsodic about the behavioral quirks of particular breeds. The American Kennel Club18 acknowledges “personality” as one of the hallmarks of recognized breeds, and includes “may be stubborn” and “eager to please” as one of the search terms one can use for finding a pet, alongside “infrequent shedding” and “small size.” In fact, the behavioral selection criteria—for activity level, propensity toward barking, and trainability—outnumber the ones for physical characteristics.
Of course, people in the past who selected for specific traits in dogs did not know anything about genes. They just kept choosing the puppies who shed less or barked more, and kept doing so until those traits became pronounced in the breed.
At the same time, it is illuminating to understand just how convoluted the path from gene to behavior can be. As an illustration, consider a breed of dog called the Australian kelpie, a “working dog” that herds sheep and cattle. Kelpies are capable of doing their job with relatively little guidance, which is useful in the vast Australian outback. Like other breeds of domestic dogs, humans selected the individuals that behaved in a certain way—kelpies need to unhesitatingly muster the group of animals they are looking after, and they need to keep at it for hours at a time, often without food or water.
Claire Wade, a professor in animal genetics at the University of Sydney in Australia, has been studying genes in kelpies for many years. Using new technology, she can compare the DNA of kelpies that work as herders with those that are kept as pets, and she can also compare both to other breeds of dogs. It turns out that a crucial difference in the genes of the working kelpies is a section of DNA that is associated with, of all things, pain tolerance. The working dogs have higher pain tolerance than the other dogs. How could that lead to better herding?
Wade points out that “the ability to feel pain is a stop|go requirement for working Kelpies. In outback Australia, the ground coverings are extremely prickly. I always tell people of a story where I was visiting a friend and my dog ran out into the field but then froze and would no longer move. I needed to go and carry her back to the soft grass—we call that being a ‘prickle princess.’ Kelpies cannot be a ‘prickle princess’ or they never have a chance to demonstrate their ability for moving sheep.”19
So the dogs that felt less pain—because of a variant in their nervous system and the way that messages are transmitted among nerve cells—were the ones that were chosen for the breed. There is no gene for herding per se, but herding evolves nonetheless. Dogs that were not “prickle princesses” had more puppies than the ones that were carried out of the field, and so they became kelpie ancestors. Presumably, this kind of indirect manner of selection works for many creatures, not just domestic animals.
Does Behavior Lead and Evolution Follow?
Behavior may be continuous with morphology, and it may evolve like morphology, but we still recognize its fleeting nature. A dog wags its tail, a cheetah sprints after a gazelle, and then it’s over, the friendliness conveyed, the prey obtained or missed. Unlike a body part, behavior vanishes. How does the wagging or the sprinting get incorporated into the genes, a prerequisite for its evolution? Furthermore, many behaviors are at least partly learned, from the local dialect of a White-crowned Sparrow to the way a chimpanzee holds its termite-collection stick. So how does that learning fit into the evolutionary process?
Biologist James Mark Baldwin suggested that if an animal learned to do a new thing, and that new thing helps it survive and reproduce better, its genes are more likely to be passed along, even though there is no direct connection between the task that is learned and any particular gene or set of genes. This doesn’t mean that the animal wills evolution into happening, but, as the eminent animal behaviorist Sir Patrick Bateson put it, “Whole organisms survive and reproduce differentially and the winners drag their genotypes with them.”20
You could also think of this as behavior leading the way for evolution. When the environment changes, behavior, being inherently flexible, is how an animal first responds. That response then makes it possible for the animal to become adapted to its environment. This idea has been somewhat controversial, perhaps because as has become clear by now, behavior, physiology, and physical attributes are all very tightly linked. It is therefore difficult to pinpoint one of them as the obvious starting point.
These exchanges between behavior and the environment can be seen most vividly in the case of something rather stuffily termed niche construction. Ecologists refer to a niche as the sum of all the requirements of an animal (or plant) necessary for it to survive. A monarch butterfly, for example, needs to have milkweed to eat as a caterpillar, and the temperature must be above 55 degrees Fahrenheit for it to be able to fly. A mole has to have soil of the right consistency to burrow into and a nice selection of insects to eat, as well as a range of temperatures that are neither too hot nor too cold. Add up all those necessities, and they define the niche.
But what if the niche itself is influenced by the animal’s activities? One of the most well-known examples of such influence is that of beavers making a dam. A new dam starts when beavers bring tree branches to a stream and set them in such a way that the flow of the water is decreased. After the base is constructed, plant material and mud or rocks are used to build up the structure, with the beavers moving entire logs through the water. Eventually, the dam causes the section of river to flood, making a pond, and the beavers then make the lodge there to live in. Beaver dams are impressive constructions: beavers can move trees up to three feet in diameter to build them, and their dams can occupy several hundred feet of the habitat. The resulting pond provides protection from predators such as coyotes, and the dam also serves as a source of food in the winter, since beavers eat bark and woody plant materials.
Once constructed, the dam alters many things about the stream or river in which it is built, including its flow patterns, the way that leaves are decomposed in the water, and the kind of plants that can thrive on its edge. This means that the food sources of the beaver change, along with the habitat for birds, fish, and insects. Eventually, the whole area looks different because of the beaver dam. The beavers’ offspring then inherit the dam and its surroundings, which in turn means that natural selection acts differently on the generations that grow up with the altered environment than it did on the originators of the dam. That changed environment in turn has the potential to influence the evolution of the beavers’ behavior, and so forth.
Darwin and Emotional Evolution
Finally, what about the evolution of perhaps the most mysterious of behaviors, emotions; or as animal behaviorist Gordon Burghardt puts it, the “private experiences” that we, and perhaps some other animals, have?21 I will explore the evolution of intelligence and cognition in later chapters, but here it is worth thinking about just how something as hard to define but as important (at least to us) as an emotion could evolve.
Charles Darwin was extremely interested in how behavior, and emotions, could evolve. He was particularly fascinated by the idea that we could trace similarities in behavior across different kinds of animals much the same way we could see resemblances in their bones or teeth. His The Expression of the Emotions in Man and Animals22 is about the ways that animals show emotions, such as fear or anger, reflecting our common heritage, and has the now-famous illustrations of similar facial expressions in humans from across the world as well as in apes. He performed an experiment, advanced for its time, in which he showed guests at his home a series of photographs of faces representing different emotional states and asked them to identify the emotion depicted. (You have to wonder how this affected the likelihood of people accepting his dinner invitations, not to mention what his wife Emma thought of the plan.) Although the guests agreed on some of the emotions depicted in the photos, including happiness, sadness, fear, and surprise, they strongly disagreed about the more ambiguous emotions such as jealousy that the images showed. Darwin took this to mean that only certain basic emotional states are universal, as he had theorized. He further interpreted animal communication as a means for conveying emotions, so that signals like a dog raising the hair on its neck could have come about as way to show fear because it was connected to what he called “the direct action of the nervous system.”
Biologists viewed Darwin’s work on animal emotions with unease for quite some time. Even well after his ideas about evolution by natural selection were accepted, many found his speculations on emotion to be anthropomorphic at best and downright squishy and embarrassing at worst. As Burghardt noted, “The reaction from human chauvinists in biology, the social sciences, and the humanities was swift and often brutal.”23 Klopfer went so far as to call Darwin’s enthusiasm for research into emotions a “scientific dead-end” and titled his review of a book attempting to reexamine research into animal emotions “Still Largely Where Darwin Left Us.”24
Part of the difficulty was and is that although people are fine—perhaps too fine, as I will argue in a later chapter—with seeing their dogs as being capable of jealousy, rage, or grief that is identical to that of humans, other animals are a harder sell. Elephants may mourn their dead, but what about, say, snakes? Tropical biologist Alexander Skutch said that snakes are creatures “in which we detect no joy and no emotion.”25 This is a sentiment with which many people would agree. But why? Do you have to have facial expressions to have emotions? Or at least eyelids, or lips?
The eminent primatologist Frans de Waal is a passionate proponent of the idea that nonhuman animals have emotions that are virtually identical to those of humans, although most of his work has been with chimpanzees and bonobos, creatures in which it is easy to see human behavior reflected. His book Mama’s Last Hug26 is filled with examples of animals exhibiting what certainly looks like jealousy, shame, and compassion. He also distinguishes between feelings and emotions. The former, he says, are the internal states that only the individual experiencing them can truly know, while the latter are “bodily and mental states . . . that drive behavior.” The emotions, De Waal suggests, are easily noted in other species. He finds the lack of willingness to acknowledge the similarities between humans and other animals a kind of arrogance, a form of the human exceptionalism that I noted in the last chapter.
I am sympathetic to the objection to human exceptionalism, and I certainly agree with De Waal about the continuity and the relationship between humans and other animals. At the same time, the solution shouldn’t be to lump us all together. It seems to me equally anthropocentric, perhaps even narcissistic, to assume that all—or any—other species experience life or emotions exactly the way that humans do. I study insects, and nothing is as humbling as the realization that I have very little idea about their emotions, or whether they have any at all that I could fathom. Their brains and nervous systems are completely different from our own—no homology there. And if I treat them like little people in exoskeletons, all I learn is how well I project my own emotions onto other beings.
It is tempting to start making categories here: primates are like people, and maybe along with all mammals, they should feel some form of emotions, so they go in the basket with humans. And maybe we should add crows and their kin, whose abilities I extoll in a later chapter. I realize that most people are fine with leaving me to neglect the emotional lives of insects. But eventually this sorting of creatures becomes unwieldy, and it also brings us back to that scala naturae, with a ranking depending on closeness to humans that then means some animals get to have emotions and others do not. But the choices are not all or none, human replica versus robot that feels nothing.
And with regard to those robotic insects, a 2020 article in the New York Times27 reported on a study of praying mantises, highlighting the insects’ ability to adjust their lethal strike depending on the speed of their prey. In a slow-motion video of a hunt, according to the author of the article, “We see the mantis pause and calibrate, almost like an experienced baseball catcher who has realized she’s dealing with a knuckleball.”
The subtitle of the piece is “New Research Shows These Ferocious Insects Don’t Just Hunt Like Robots.” Leaving aside exactly how we know the way robots hunt, the reaction of the readers to the article reveals a great deal about how we do, or don’t, see emotions in other species:
Why we should think that they, or any other living entity operates as if a robot is way beyond me. Do we really need to bring them into a lab to realise that they too are conscious, feeling beings?
Sorry to break this to you, fellow humans, but every pig you eat, and every bug you crush, had a life, a will to live, a sense of self, the capacity to feel fear, relief, hunger, lust, etc. Other animals may not have the same emotional range we have (although, who knows), but it’s likely that they experience some things, e.g., terror, even more intensely than do we . . . Praying mantises have awareness and can think, and are not, in fact, robots—any more than are ants, ant-eaters, hippos, worms, bees, snakes, or donkeys. So treat them accordingly, whenever possible.
If it is self-centered to assume we are the only animals to think and feel, why is it not equally self-centered to assume they think and feel as we do? The funny thing about seeing all animals—or all mammals, or all vertebrates—as sharing the same emotions as humans is that we have no such expectations of their many other characteristics. Hummingbirds’ hearts beat twenty-seven times faster than humans. Whales can dive to depths of two thousand meters on a single breath. Cats do not ovulate until after they mate. It is easy to come up with examples of animal functions that are vastly different from those of our own. And if it is reasonable for other animals to have vastly different reproductive or respiratory systems than people, why should we think they have the same emotions that we do? We should grant behavior the same evolutionary courtesy we do other characteristics, and try to understand how it sometimes came to be different in other species and sometimes the same.