Dr. Stephen Lea is my candidate for bravest man in the world. An emeritus professor of psychology at the University of Exeter in England, he published a paper in 2018 with Britta Osthaus1 titled, “In What Sense Are Dogs Special?” In this study, which I explain in detail later in the chapter, the researchers concluded that although dogs can perform some cognitive tasks perfectly well, they are not, as the title says, particularly “special.”
The reception to their work was not appreciative. “Your Dog Is Probably Dumber Than You Think, a New Study Says,” smirked a typical headline from Time magazine. “Shocking News for Dog Lovers: Canines Aren’t as Smart as We Think” bemoaned the Mercury News. One article with the headline “New Research Suggests Dogs Aren’t Exceptionally Smart” has a quotation from Lea as a subheading of sorts: “We’re certainly not saying that dogs are dumb,” which bespeaks a certain desperation. Lea also tried to pacify the dog people in another interview with “Dog cognition may not be exceptional, but dogs are certainly exceptional cognitive research subjects.”2 No one seemed placated by these qualifications. The paper by Lea and Osthaus could have been titled “Goats, Pigs, and Pigeons Are Special, Just Like Dogs,” but that lacks a certain zing.
Never mind that the study didn’t show that dogs were stupid. But it turns out that saying anything negative about our best friends is tantamount to insulting God, mother, and country all at the same time. Why is that? Why are we so obsessed with all matters canine, from the evolutionary origin of dogs to the way they can best be trained to stay off the sofa? What really does make for a good dog?
From my perspective in this book, the domestication of animals, and of dogs in particular, is one of the best ways to understand how behavior evolves, and how it is linked to physical characteristics like size and shape. Domestication has been a cornerstone for many scientific ideas over the centuries. Charles Darwin relied on domestication to formulate his ideas about evolution. He was particularly engaged by pigeons, which were being bred by British Victorians to have many elaborate feather colors and patterns. The birds helped him argue that even the most outlandish types of pigeons arose from a single species, the humble Rock Dove. Science journalist Richard Francis points out in his book Domesticated that “human civilization as we know it would not exist”3 without domesticated plants and animals, since they enabled us to leave lives of hunting and gathering to settle down. Similarly, other scientists have made sweeping claims about the impact of domestic species, dogs in particular. Certainly life as we know it would be vastly different without our pets, crops, and meat- or milk-yielding beasts. What is more, when it comes to domesticated animals, the hallmark of domestication is behavior, not appearance. Cows that gave more milk, or sheep that had thicker wool, would be of little use if they never stood still long enough for us to herd them into a barn.
But of course there is more to our interest in domestication than just the practicality of keeping animals. We obviously love our domestic animals, or at least highly value them, in the case of the less-cuddly ones like camels or reindeer. At the same time, they showcase our dilemma about the fixed versus malleable nature of behavior. If, on the one hand, instincts are immutable, how did early humans make a lap-sitter out of a fierce predator? But if behaviors are simply induced by the environment, well, a similar question comes to the fore: How could those humans have taken placid canine parents and counted on them to produce placid puppies?
Our understanding about the domestication of dogs has expanded enormously over the last few years, as scientists sample ancient DNA and use sophisticated statistical techniques to evaluate the similarity between modern breeds and dogs from thousands of years ago. You can now use ancestry websites to examine not only where your dog came from, but also its likely personality characteristics and whether you should do things like enroll it in agility training. But it is turning out that the simple story of a wolf that lurked at ancient campsites and then stayed because of mutual convenience is much more complicated than it first appeared. Virtually everything about what we thought we understood about dogs is now up for grabs—their origin, the way they were domesticated, their behavioral abilities and empathetic connection to humans, and even whether they are distinct from their wolf ancestors.
The Wolf That Came in from the Cold—or Didn’t
Where did dogs come from? The simplest answer to that question is wolves: domestic dogs share much of their genetic material with wolves, they look like wolves, and archaeological evidence suggests that dogs started out in places where wolves lived. After that, however, the story gets complicated quickly. Estimates of the time of the first dogs range from as many as 135,000 to as few as 10,000 years ago, an astonishingly large spread for an animal that has been so intensively studied.
Part of the ambiguity stems from what counts as evidence of a dog rather than of a wolf, which then leads to the equally thorny question of how the shift from wild creature to tame companion occurred. A 26,000-year-old mark in Chauvet Cave in southern France seems to show, next to the footprints of a young boy, the paw marks of—well, either a wolf or a large dog.4 But which? Other scientists looked at fossil bones associated with human encampments and suggested that something like a large dog may have been present in Europe up to 36,000 years ago.5
Fossil remains, however, are frustratingly limited in what they can tell us about the relationships between wolves and dogs, or about how various dog breeds themselves are connected. One has to infer a great deal simply from measuring the bones themselves. Because individuals differ in any species, generalizing from one bone from one animal to an entire species can be difficult. Just in the last few decades, scientists have been able to rely instead on DNA, either from modern dogs to determine how much they differ from each other and from modern wolves, or from ancient samples that come from bones excavated at archaeological sites around the world. Such studies have variously concluded that dogs originated in the Middle East, in southern China, and Europe.
A study published in 2013 by Robert Wayne, an evolutionary biologist at UCLA specializing in dog genetics, and his colleagues compared DNA from fossilized bones, from modern-day wolves, from several dog breeds, and from coyotes—using the latter as what is referred to as an “outgroup,” a more distant relative that serves as a benchmark for the other comparisons.6 Their analysis yielded an estimate of dog origins in Europe from 18,000 to 30,000 years ago. The scientists imagined that as wolves interacted with humans, they migrated along with them, thus becoming separated from other wolves and subject to humans seeking and selecting the more sociable associates. This research also cast doubt on the suggestion that dogs arose in East Asia, because those studies, Wayne and his colleagues attest, used DNA not from dogs but from modern-day hybrids between wolves and dogs.
But other scientists, most notably Peter Savolainen at the Royal Institute of Technology in Stockholm, weren’t convinced. They pointed out7 that Wayne’s samples hadn’t included enough individuals from Asia in the first place; you can’t find something from a place where you haven’t looked for it. A rather contentious debate ensued, with arguments about who was sampling appropriately, who was using the right DNA, and what parts of the fossils to measure. Paleontologist Mietje Germonpré measured a 32,000-year-old skull from a museum in Brussels and identified it as a dog, which would have put the origin of the domestic animal rather early, but her work too was hotly contested. In an article in the journal Science, she is quoted8 as saying, “It’s a very combative field. More than any other subject in prehistory.” In case you have never met any archaeologists, let me point out that this is really saying something.
A new slant on the story appeared in 2020,9 when a group of scientists published the results of a study of sled dogs, those hardy workaholics of the Far North. Sled dogs seem like a niche breed, but they have ancient origins; carved bone and ivory tools resembling dog harness fasteners have been discovered at sites in Siberia dating back nearly 10,000 years. The researchers compared the DNA from ten sled dogs from Greenland, a bone from a 9,500-year-old Siberian dog that was associated with evidence of sleds, and a wolf bone also from Siberia that was dated to about 33,000 years ago. The ancient dog bone was similar to modern sled dogs and to the ancient wolf, but not to modern American wolves. This suggests that the sled dogs in the Iditarod can trace their ancestry back to Siberia, and that they are not a modern breed originating only a few centuries ago, as one might have thought given their current use.
Perhaps even more intriguing, the genetic similarities between old and new sled dogs suggested just how selection for domestication changes the genome. Most domestic dogs have multiple copies of a gene that is useful in digesting starches, widely viewed as a way that dogs became better able to live with humans and our carb-reliant diets. Even among humans, populations such as those from East Asia, where rice has been a staple for many thousands of years, have more copies of the gene for starch digestion than those populations from places where starches are not as major a food source.10 Indigenous populations in the Arctic fall into the latter category, and, no surprise, so do the sled dogs. The sled dogs do have genes that help them digest the fatty acids that are prevalent in a diet heavy in well-larded meat. One of those genes is similar to a gene found in polar bears, though of course the bears and dogs have not shared a common ancestor for a very long time, illustrating the way that evolution uses different solutions to solve the same problem.
How Did Dogs Become Dogs?
The ability of dogs (at least in certain areas) but not wolves to digest starches leads to another hotly contested place that is relevant to dog evolution: the garbage dump. The conventional wisdom is that tens of thousands of years ago, humans and wolves were bitter enemies. As Richard Francis puts it, “All wolf-human interactions were overtly hostile. We competed fiercely for the same prey and probably killed each other at every opportunity.”11
Then, the story continues, some wolves started hanging around human encampments and scavenging food from the trash heaps. The wolves that were less aggressive toward people, perhaps bringing their pups with them, gained by being fed. The humans—though the literature is less clear here—initially simply tolerated their presence and then later favored some individual wolves and even took in their puppies, raising them as pets. This scenario is suggested to have taken place sometime around the beginning of agriculture, perhaps 10,000 years ago. After many generations of people selecting the tamest wolves, the animals changed not only in their aggressiveness but also in their appearance.
Alternatively, another theory suggests that the wolves domesticated themselves. Again, the friendlier individuals were spending time near humans, although no one deliberately tried to raise puppies or even encourage interactions. Eventually, however, the self-domestication meant that ever-tamer wolves gained more advantages from humans, and dogs were the result.
Both of these pathways take for granted the mutual enmity expressed by Francis, and the active role played by humans, and humans alone, in molding the friendly attributes of dogs out of the vicious clay of wolves. But what if those assumptions are incorrect? Ray Pierotti, a biologist at the University of Kansas, and his colleague Brandy Fogg believe that the image of the Big Bad Wolf is, as the name suggests, a fairy tale, and a Eurocentric one at that. In their book The First Domestication,12 they propose a more cooperative origin of dogs. Drawing from indigenous accounts, mainly from North America, Pierotti and Fogg argue that many tribal groups cooperated with wolves as equals, hunting together and viewing the predators as colleagues, not enemies. They contrast the language used to refer to wolves by Western scholars—with descriptors such as “ferocious”—with the stories by indigenous people, which mention “companion” or “grandfather.” This appreciation of wolves suggests to the scientists that wolves and humans evolved together. They cast doubt on the “dumpster” origin story, pointing out that wolves’ outstanding hunting ability would make it unnecessary for them to seek out human refuse to eat.
Instead, Pierotti and Fogg speculate that perhaps a solitary wolf that was driven from its pack joined the only other available hunters in the area, namely people; if those hunters were willing to partner with the wolf, both parties could benefit. The wolf-assisted humans could then expand into parts of the world that did not previously have wolves, and the appearance and behavior of wolves could have diverged into a more doglike form. This set of events could be repeated in various parts of the world, and we could see the emergence of dogs without having to posit a relationship in which humans did all the manipulating and wolves were simply receptacles of selection.
Other scientists are also rethinking the story, with an eye toward mutual hunting benefits rather than exploitation. It has even been suggested that wolves helped early humans extend their range by enabling them to bring down large prey. And anthropologist Pat Shipman argues13 that people worked with the earliest forms of dogs to hunt large numbers of mammoths, beasts that would have been hard for humans to kill with the tools they had at the time. The genetic and fossil evidence about dog domestication is still being gathered, with the boundary between wild and domestic still blurred.
My, What a Big Brain You Have
Pierotti and Fogg also maintain that wolves and dogs are not nearly as different as people like to think. In central Asia, for example, you would be hard-pressed to tell dogs from wolves, at least from their outward appearance. And archaeological remains can be tricky to interpret; what looks like a wolf to one can be a dog to another. To Pierotti and Fogg, all dogs are wolves, but not all wolves are dogs, meaning that the boundaries between the two are not fixed; a poodle is clearly a modern invention (most of the newer dog breeds only date back a couple of hundred years at most), but they are still sufficiently enfolded within the Venn diagram circle of wolves to make a hard-and-fast distinction questionable.
The authors suggest that as humans and wolves cooperated in hunting, wolves first changed their behavior to become more doglike, in a way that left the appearance of the two animals indistinguishable. Indeed, the evolution went both ways, with human behavior changing even as there was little effect on human physical characteristics beyond, intriguingly, a reduction in the skeletal shape and size differences between men and women. According to this view, wolves made early humans more cooperative in a way that distinguishes us from our closest evolutionary relatives, the chimpanzees. Their position is not widely shared; humans seem to have evolved to be more pacific than chimps for a variety of, again hotly contested, reasons, including the notion that humans themselves are self-domesticated.
Complicating this issue are popular, rather than scientific, conceptions about just how similar wolves and dogs really are. On the one hand, we breed dogs like bichon frises or Pomeranians, in which anybody would be hard-pressed to see a resemblance to a wolf. On the other hand, we (and the pet food companies, or at least their advertising writers) seem to be convinced that dogs are still, at heart, wolves—witness the response to an article in the New York Times suggesting vegan diets for dogs. The article noted that in part because of the aforementioned changes in genes allowing for the easier digestion of starch, it is possible to give dogs a plant-based diet, although experts disagree about whether a dog that eats a vegan diet will get all of its nutritional requirements for long-term health. Be that as it may, readers had strong opinions.14 One suggested taking a vegan dog to a sheep farm and then “watch how fast the ‘vegan dog’ becomes a wolf.” Concern about “forcing dogs to eat a diet they were not designed for” was met with the rejoinder that if the mere ability to consume flesh were the only consideration, humans should be eating each other, since cannibalism has been seen in many societies. Another flatly stated, “Dogs are not meant to be vegetarians. If you can’t deal with that, get yourself a goat instead.” Running through these reactions about inner wolves is the idea that animals—and perhaps people—have an essential nature—in this case, an internal wolfness, so to speak, that is dictated by their genes. That essentialism, as I noted in the previous chapter, will always lead us astray.
At the same time, we eagerly look to studies that show us the ways that dogs and wolves differ in their behavior. For example, a widely cited study from 201715 examined the ways that dogs versus wolves cooperated in a task at the Wolf Science Center in Vienna, Austria. Pairs of both species—raised in as similar a way as the investigators could manage—were presented with a tray of food on the other side of their cage. Two ends of a cord were threaded through rings on the tray such that if both were pulled, one by each member of a pair, they get the snack. But if only one did so, there were no treats for either. Though neither canid did spectacularly at this test, the wolves surpassed man’s best friend by a wide margin, with five of seven pairs managing to acquire the tray on at least some of their attempts, while only one of the eight dog pairs ever figured it out.
Much was made of this difference, with the experimenters concluding that it showed how dogs are not natural cooperators, unlike the pack-dwelling wolves. Perhaps, the researchers speculated, dogs lost their ability to work in a team as they forged their bonds with humans; you don’t need to collaborate to find a discarded carcass at a refuse site the way you do when you attempt to take down a moose. On the other hand, Laurie Santos, a cognitive scientist from Yale University, suggested that the dogs are perfectly good at cooperating, but they perceive the physics of the problem differently, and so can’t solve the problem.16
If the whole let’s-pull-on-a-string-together test sounds too much like some kind of team-building exercise from a corporate retreat, how about a nice game of fetch? A 2020 paper17 by Christina Hansen Wheat and Hans Temrin from Stockholm University determined how eight-week-old hand-raised wolf puppies responded to investigators—charmingly called “puppy assessors”—throwing a tennis ball. First, the puppy had a chance to play with the ball on its own, and then the assessor “encouraged” the puppy to retrieve the ball.
The results were, again, not stunning; just three of the thirteen wolves brought back the ball even some of the time, and only one, named Sting, managed to do so on all three of the trials. And an eminent researcher in dog genetics, Elaine Ostrander, pointed out that it’s not clear what the test revealed, since many dogs won’t retrieve balls either. Nevertheless, this accomplishment was lauded in the popular media as indicating an ancestral willingness to play with humans residing deep in the wolves’ DNA. In turn, that suggested to some that you don’t have to work that hard to domesticate a wolf—they have had it in them to be house pets all along.
Except, of course, when they don’t. Although Pierotti and Fogg point out that many of the instances of vicious behavior by wolves kept as pets are actually cases of dogs that simply look like wolves, it’s clear that, as yet another article about wolves and dogs in the New York Times put it, “No matter how you raise a wolf, you can’t turn it into a dog.” Or as Patricia McConnell, a noted animal behaviorist who consults on dog behavior, wrote: “Dogs aren’t wolves, pure and simple. Except, uh, they are. Sort of. Sometimes.” She goes on to explain that while we all see the similarities in behavior, and we know that dogs evolved from wolves, wolves can rarely be housebroken, and they use different cues to assess friendliness or aggression than do dogs.18
She also debunks the notion that wolves live in strict social hierarchies, with a single alpha pair, and that the dominance relationships in a pack are carried over into dogs, with humans the “alpha” individuals. First of all, social status in animals is complicated, particularly for big-brained species like wolves. Dominant individuals don’t always get their way, and different wolves can approach being alpha differently. And even in feral packs, dogs do not replicate the social system of wolves, which means that dog trainers who urge owners to constantly assert their dominance, as though life with a terrier in the suburbs replicated the Call of the Wild, are, well, barking up the wrong tree.
What this means is that, while wolves can certainly come to associate with humans without eating them (a basic requirement for domestication in anyone’s book), they are not simply dogs that haven’t been given enough attention while young. A great deal of research on wild wolves has shown that they have markedly different behavior from dogs at the get-go, including care of the young by both parents, not just the mother, as is the case for dogs. Though both use their noses in ways that make humans seem effectively scent-blind, wolves and dogs emphasize smell differently, with wolves using it as their exclusive way of interacting with the world as pups, and dogs relying on a combination of sight, smell, and sound. As I detail later in this chapter, many researchers are using the latest tools in genetics to better understand the ways that genes contribute to the changes we see in domestic dogs compared with their wolf ancestors.
Another much-touted difference between wolves and dogs is a dog’s ability to recognize when a human points to something. If you present the animal with two bowls of food whose contents aren’t visible to it, and point to the bowl with the snack, will the animal go where it’s directed? Again, sometimes. Dogs generally do better than either wolves or chimpanzees at this task, particularly when they are young (and are on a par with two-year-old children), although if the wolves have spent a lot of time with people they catch on more quickly. A key to success seems to be the ability of the animal to look the demonstrator in the eye, something wolves are simply not as prone to do. The results have caused researchers to wonder about whether dogs and wolves have similar abilities, but the ages at which they take hold are different in the two animals.
I am intrigued by the idea that cultural biases have shaped the way we view domestication, and I think that vilification of wolves has probably clouded our opinions about how dogs evolved. But I am not convinced that it is only our Eurocentrism that makes us hold wolves apart. Instead, the dog-wolf divide illustrates yet again just how intertwined genes and the environment are in producing behavior. Yes, you can get a wolf puppy to play fetch, and be petted, and walk on a leash, so wolves obviously possess whatever genetic material is required for such interactions. But you have to live with those puppies 24/7 for several weeks before that becomes feasible, whereas a dog will oblige you with affection with a fraction of that type of contact. At the same time, you cannot simply grab a feral stray dog off the street and expect it to sit, stay, or even refrain from biting you. And the ability to play with an object, or rescue people in distress, varies widely across breeds, and individuals, and when and where it happens. These behaviors are more likely to occur in dogs than wolves, but they are not universal. This variability also casts some doubt on the generalization about one species or the other from a round of tests using just a dozen or so individuals.
The thing is, regardless of whether you are talking about a dog or a wolf, there’s no such thing as an innate tendency to fetch a ball, or a stick, or to pull on a string to get a tray of food. The differences between dogs and wolves illustrates the point I have been making all along: behavior isn’t something that is unilaterally drawn from a gene or group of genes, and it isn’t something that emerges from an experience during puppyhood. Instead, it’s both.
Domesticated in Life, Fur, and Paws
This combination of genetic change and early experience is part of the complex story of how domestication shapes behavior and appearance at the same time, but still can’t alter everything about an animal. Not only can’t you make a silk purse out of a sow’s ear, you can’t really change a sow’s ear much at all, at least without some unforeseen consequences.
To understand this, we need to delve deeper into what domestication really means. Many people use the word as though it is synonymous with tameness, or lack of aggression, hence the aforementioned tendency of dogs to be friendly and wolves not so much. Anthropologist Richard Wrangham points out19 that aggression comes in two basic flavors, reactive and proactive, a distinction first made by psychologist Leonard Berkowitz in 1993. Reactive aggression is the kind that occurs when a wolf snaps at a rival during a fight, or a child hits a classmate who snatched a desired toy. It is immediate, emotional. Proactive aggression is planned, the kind manifested by premeditated murderers. Wrangham argues that the two have different functions and are subject to different selective forces, as I will explain in the next chapter. Tameness, then, is related to lower reactive aggression, and it seems certain that people would have bred, or at least wanted to associate with, the wolves with the lowest levels of such aggression.
But domestication is more than tameness. An individual animal can be tame, because it has become accustomed to being handled by humans. Species differ in the degree to which they will respond to such treatment. People can approach the hand-reared wolves I mentioned earlier, and these wolves will even show them affection. But they are still unpredictable in their behavior. It took many generations of selective breeding, in which the friendliest wolves had puppies, and then the friendliest of their puppies were chosen to breed, and so on, until dogs came to be reliably tame.
So are dogs exactly like wolves, except with more docile demeanors? We all know the answer is no; dogs differ from wolves in many ways, both physically and in their behavior, and the links between the two provide some of the strongest evidence for my contention that both evolve in concert. Charles Darwin, in his work on domestication,20 was one of the first scientists to note that in addition to being less aggressive, domesticated mammals such as dogs also have floppy ears, white spots in their fur, curly tails, and white feet. At least a subset of these characteristics also appears in domesticated cows, horses, and cats when compared with their wild forebearers. Domesticated animals also tend to be smaller, with shortened faces and relatively smaller brains, than their wild cousins.
This constellation of characteristics is called the domestication syndrome, and its significance has been puzzled over since Darwin. Why should a white tail have anything to do with aggression, or with a smaller jaw? One explanation is that all the traits evolved as the various species lived with humans and they did not have to survive on their own in the wild. Hence white patches were seen as attractive by humans, or were useful in telling animals apart. Floppy ears were not so much selected for as not selected against—if floppy ears make an animal’s hearing less acute, that wouldn’t matter much if that animal is living a sheltered life and doesn’t need to be ever vigilant for the sound of a predator. Modern breeds, of course, do not all share these signatures of domestication, but those more individual departures from the syndrome might have come later.
The problem is that this seems like special pleading for what is really a rather motley collection of attributes. If pressed, one can come up with an evolutionary explanation for almost anything, and these just-so stories, as these explanations are called, after the Rudyard Kipling tales, often have little behind them other than a surface plausibility.
To consider an alternative, we need to turn away from dogs for the moment. We also have to travel, mentally at least, to Siberia, and to a story that reads as much like a Cold War spy novel as it does a science experiment. In 1959, a Russian scientist named Dmitri Belyaev decided to try and untangle the puzzle of how dogs became domesticated by taking the then unheard-of tack of replicating the domestication process in real time.21 He and his colleagues took silver foxes, widely bred in vast Siberian farms for their luxurious pelts, and worked to turn them into friendly house pets. It was a deceptively simple process: take the puppies from only the friendliest foxes, breed them, and repeat. It only took ten generations to get foxes that licked the faces of their handlers, something that the original foxes would never do. What is more, selecting for tameness and tameness alone also produced genetic changes in all kinds of aspects of the foxes’ appearance: fur color, ear shape, and multiple breeding seasons per year, instead of the usual one. Belyaev and his colleagues argued that these changes were by-products of domestication, and part of the syndrome seen in other animals.
The experiment he started is still continuing, over sixty generations later, with biologist Lyudmila Trut in charge, and it is a landmark study in understanding domestication.22 The spy novel part comes in because Belyaev began his project as the Soviet Union was shaking off the disastrous influence of Trofim Lysenko, the director of biology under Stalin. In keeping with the idea that the government could engineer a perfect society simply by manipulating the environment—whether of crops or people—Lysenko rejected the burgeoning field of genetics and promoted agricultural improvement through the inheritance of acquired characteristics, a la Jean-Baptiste Lamarck, an eighteenth-century naturalist. Lamarck is best known for proposing the idea that animals (and people) could permanently change their traits through sheer effort, with the classic example being giraffes straining to reach high branches and passing on their stretched necks to their young. Lysenko applied the same idea to growing crops, and his efforts resulted in the failure of Soviet agriculture. But for decades he was highly influential, and Belyaev and Trut had to pursue their research in the face of lingering skepticism at best and the threat of imprisonment at worst. They persisted, and the work has yielded much data about how behavior changes over time.
But having identified a constellation of domestication-associated characteristics in body as well as behavior, Belyaev didn’t know how they became connected; genetics simply wasn’t far enough along to allow investigation of the mechanism. In 2014, biologists Adam Wilkins and Tecumseh Fitch, along with Richard Wrangham, published a paper23 suggesting that the domestication syndrome was due to changes in the migration of neural crest cells during embryonic development. Neural crest cells are part of the developing vertebrate embryo early on but disappear as the animal grows. The cells do not become any particular organ, but they influence the way that many different parts of the body develop, from tooth size to ear structure to the formation of the adrenal glands that produce hormones important in aggression.
This idea has been popular, and it does seem to provide a tidy theory uniting what had seemed to be a disparate set of characteristics. And Wrangham has extended it in his book The Goodness Paradox to suggest that the low reactive aggression is associated with the physiological traits of the domestication syndrome.
Recently, however, a snag in the story has emerged, with questions about the foxes used in the Russian study and hence the generality of the domestication syndrome itself. It started with a chance visit by the late biologist Raymond Coppinger to Prince Edward Island in Canada, which houses the International Fox Museum and Hall of Fame.24 Coppinger, who had written extensively on dogs, saw photos at the museum that looked like the foxes in Belyaev’s experiments, showing animals that had been captive-bred since the late 1800s. The foxes were already somewhat tame, with people walking them on leashes and holding them in their arms. These animals were used to establish the Russian fur farms in the 1920s. Belyaev himself had never claimed that he started with completely wild foxes, but he did refer to the starting population as “wild controls,” which may have given the impression that his study demonstrated how one could go from wild to domestic after a relatively small number of generations of selection.
Spurred by Coppinger, a group of scientists led by Kathryn Lord from the University of Massachusetts and the Broad Institute in Boston traced the history of the Russian foxes and reexamined the domestication syndrome in a wide range of animals.25 They analyzed the DNA of the experimental foxes and concluded that they derived from the Prince Edward Island population. Lord and her colleagues concur that Belyaev’s experiment offers rich fodder for the study of how behavior is inherited, but suggest that the rapid response to selection occurred because the foxes were already on a path to tameness, and hence the evidence doesn’t support Belyaev’s rapid process of domestication. It is worth noting that foxes seem particularly prone to domestication, making it hard to draw conclusions about other animals based on them; for example, Native Americans brought descendants of grey foxes from the Northern Channel Islands off the coast of California to the Southern Channel Islands and kept them as pets.26
As for the suite of characteristics said to accompany domestication, Lord and her colleagues examined nine kinds of other domesticated animals using three criteria to see how well the syndrome held up.27 First, the characteristic had to appear more or less at the same time as tameness during the course of domestication. Second, the characteristic needed to be more common in the selected population than in the population it came from. Third, the characteristic needed to be associated with tameness within an individual, so that the tamest dogs, or rabbits, or goats had to also be the ones most likely to show the white fur patches.
The result? None of the species fulfilled all the criteria, though each showed some of the predicted associations. Another study by a different group of researchers focused on dogs, and examined seventy-eight breeds for correlations among the different aspects of the domestication syndrome—were white feet found with curly tails? Such associations would be expected if the syndrome occurs because the same mechanism governs all the various components. Again, the answer was no, and the authors concluded that dogs, at least modern breeds, do not illustrate the syndrome very well.28
What does that mean about the domestication syndrome? I am sure we have not seen the last of this debate, but it is clear that while behavior and appearance can both respond together to selection, whether imposed by humans or by nature, the relationship between them is complicated. And that complicated interaction between genes and the environment is what we expect for all characteristics, not just in dogs or in the syndrome of traits that accompanies domestication.
Who’s an Unexceptional Doggie? You’re an Unexceptional Doggie!
Dogs can do some amazing things. So can wolves. And the two differ in ways that are interesting to consider in light of dog domestication. But what about the question posed by Lea and Osthaus that I used to start the chapter—are dogs special?
No question, people love their dogs, and have for, well, as long as people and dogs have been together. And we think the feeling is mutual. Charles Darwin, who used domestication to understand how selection by humans could be a model for selection by nature, mused: “It is scarcely possible to doubt that the love of man has become instinctive in the dog.”29 One could see this mutual love in the flood of COVID-19 pandemic puppies, as people in the United States emptied out shelters and breeders to cozy up with affectionate canine companions.
And for reasons that aren’t clear to me, at least, people seem to have a stake in showing how their dogs, in particular, are geniuses. Dogs have been used in all kinds of cognitive studies, from the work of Pavlov on classical conditioning to tests of their ability to recognize a human in distress (spoiler: not very good). While some of that use of dogs is sheer convenience, I am willing to bet that there is more to it than just using dogs as a rat substitute.
One of the newest salvos in the how-gifted-are-dogs discussion has to do with speech. No one (yet) has actually taught a dog to talk, weird YouTube videos of dogs barking “The Star-Spangled Banner” notwithstanding. Still, people have long claimed that their dogs understand exactly what they say. To examine this claim, a group of Hungarian researchers set out to determine how dogs’ brains respond to human speech.30 They trained twelve pet dogs (border collies, retrievers, and a German shepherd) to lie down in an fMRI machine, and then had humans speak while the scientists recorded the responses in the dogs’ brains. The humans repeated either positive words, like “clever” or neutral ones, like “if,” with the same intonation.
The dogs responded differently to the two types of words, but the researchers hastened to say that this did not mean the animals understood the separate meanings of the words. Instead, they were interested in how the brains processed the sounds. The dogs seemed to use one part of their brain to respond to the emotional component of the words, and another part to respond to the word itself. Humans do a similar kind of partitioning when listening to meaningful versus meaningless sounds, which suggests that the ability to process vocal communication may be shared across a range of mammals.
Which brings us to the Lea and Osthaus study.31 What the scientists did was straightforward. Dogs may be smart, but are they any smarter than you would expect? To answer that question, Lea and Osthaus picked three groups for comparison. First, they looked at studies of other species that are related to dogs evolutionarily—members of the group Carnivora, meaning meat eaters, including African wild dogs (which are not the same as feral or street dogs) and cats. Then, they considered dogs as social hunters. Other groups of social hunters include dolphins and chimpanzees. Finally, they examined horses and domestic pigeons, both of which are of course domesticated animals like the dog, and which are also amenable to a lot of training by humans. Each of these lenses provided a different but complementary way to think about dogs’ abilities.
Any such comparison is bound to have an arbitrary element to it—one could argue for or against including any number of other species, from ferrets to elephants. But the idea was to see if the dogs performed in a way that was unexpected, given their place in any of the three grand schemes of animal life.
The result, as I’ve already intimated, was disappointing, at least if you are one of those my-dog-should-be-in-Mensa types. Yes, dogs do well at discriminating complex visual patterns, like telling human faces apart, but so do chimps and pigeons. Dogs are good at smells, but they are bested by pigs, which can even distinguish between the odors of familiar and unfamiliar people. Dogs are not particularly skilled at what Lea and Osthaus term “physical cognition”—recognizing the consequences of manipulating objects like strings attached to food, or understanding where a moving object will come to rest. Faint consolation may lie in knowing that cats do not do any better than dogs, though raccoons, unsurprisingly, excel at such tasks. Despite the heartwarming nature of movies like Homeward Bound, dogs aren’t even particularly good at navigating over long distances.
None of this means, as both authors point out repeatedly, that dogs aren’t wonderful companions. They can point (perhaps not literally, but that’s okay) to the ways that different species have different abilities. Osthaus said in an email to me that by emphasizing other animals’ skills, she was “doing my bit for animal welfare, really, as I pointed out pigs’ amazing abilities, as well as pigeons’.”32
A word about the bias of using human senses as a starting point is in order here. Many people have noted that so-called intelligence tests are biased toward the experimenter’s worldview. We think animals are smarter the more they are like people, and the more they rely on human senses, primarily vision, rather than, say, scent, which plays such an important role in the lives of dogs. Science writer Ed Yong pondered the famous mirror test,33 in which a spot is placed on an animal’s body in a place where it is invisible unless the animal sees it when it looks in a mirror. Most animals do not recognize their images in a mirror, with chimps and elephants as noteworthy exceptions that reach for the place on the body that has been altered. The results are sometimes claimed to show that a species has a “sense of self” and can recognize the existence of its own body. But dogs are not particularly visual creatures, and rely instead on smell. If we could give them the equivalent test in odors, would they excel?
Perhaps. But a larger point—and one that was recognized in the work by Lea and Osthaus—is that it doesn’t make sense to simply pick on an animal, no matter how beloved, and try to rank it according to a scale that only works in a single dimension or on human-centric traits. Yong quotes psychologist Alexandra Horowitz saying, “I think that all of these abilities, which we mostly decline to see in nonhuman animals until they’ve passed our tests, are defined in a far too binary way.”34 In other words, dogs are good at things that make sense for dogs to be good at, which is a rather unsatisfying answer but one that makes sense from the standpoint of evolution.
Dogs Rule with Their Drool
New techniques in genetics let us understand more and more about the link between how an animal looks and acts and its genes. What about combining the two? Elinor Karlsson is the head of a lab studying, among other things, dog genetics at the Broad Institute.35 Kathryn Lord, who led the work casting doubt on the domestication syndrome, is part of her research group. Karlsson has an ambition that verges on the grandiose, and I say that with all the respect in the world: she wants to sample the DNA of every single dog in the United States, and perhaps elsewhere, and link those genetic signatures to details of behavior.
In a project called Darwin’s Dogs,36 she is, as the project’s website says, “following the pawprints of evolution.” More prosaically, the lab is enlisting dog owners who send in a sample of their pet’s saliva (something dogs are more than willing to provide) and answer questions about the dog’s behavior. Does Fido cross his paws before lying down? Does Princess like to fetch balls? Using sophisticated techniques in genetics, some of which I outlined in the previous chapter, Karlsson and her colleagues are primarily trying to understand behavioral and psychological disorders such as canine compulsive disorder, a syndrome I will discuss in more detail in chapter 6. But the project also provides a powerful tool for understanding the evolution of behavior more broadly.
The project has already pinpointed a gene that seems to be important in narcolepsy, the sleep disorder, and is examining genes that might predict a dog’s success at becoming a working dog that can sniff out bombs or help people with disabilities. Despite many years of intensive training and breeding programs, the success of animals in the latter can be well below 50 percent, not the best return on investment. If Karlsson and her colleagues can understand the genetics behind the dogs that pass the test, that rate could be increased.
Other large-scale programs to understand dog behavior include Dognition,37 a program run by Brian Hare at Duke University. “Find the Genius in Your Dog,” its website proclaims. Owners perform a series of tasks like yawning in front of their dog to see if it yawns back (a test for empathy, or at least the strength of the bond between a dog and its owner), and send the results in to the site. In return, they receive a characterization like “charmer,” “maverick,” or “Einstein.” The idea is not to rank your dog against others, but to appreciate its unique qualities.
Along with his colleague Vanessa Woods, Hare points out that the last decade has seen a proliferation of books about the inner lives of dogs,38 as our interest in them has risen. Woods and Hare argue that cooperation between humans and dogs, as well as among humans, has been key to our mutual survival, and even our success at outcompeting the Neanderthals. Along with Wrangham, they see us humans having domesticated ourselves, becoming friendlier to one another in the process.
As someone who studies animal behavior, I applaud the increased recognition of the cognitive abilities of dogs. Right in our own backyards, or kitchens, or even beds, we have animals demonstrating the extraordinary power of evolution on behavior as well as appearance. We can even change each other’s hormones, with a recent study suggesting that gazing into your dog’s eyes can alter your—and their—oxytocin levels, a hormone associated with bonding and care.39 Dogs have become really, really good at being with people.
At the same time, increased scientific scrutiny of dogs has debunked some of the myths about their behavior. For example, a large-scale study of hundreds of breeds had a difficult time finding consistent differences in personality among them, despite frequent popular characterizations of which breeds are best for what. As Lea and Osthaus make clear, we actually know relatively little about the abilities of many animals, not just dogs. Once we know more about both those abilities and about the genes and experiences that produce them, we will be closer to understanding the animals that we first domesticated.