The following excerpt is taken from Chapter 12 of Kevin Laland’s new book Darwin’s Unfinished Symphony. How Culture Made the Human Mind. (2017, Princeton University Press).
We have all experienced technological innovation during our lifetimes and, depending on our age, can remember the first appearance of iPods in 2001, the World Wide Web in the 1990s, mobile phones in the 1970s, or color TV in the 1960s. Each of these influential, recent innovations swept society as the cutting-edge advance of the day, only to be refined, elaborated, and improved upon by succeeding technology. The logic of cultural evolution is identical to that of biological evolution, even if the details differ. New ideas, behaviors, or products are devised through diverse creative processes; these differ in their attractiveness, appeal, or utility, and as a result are differentially adopted, with newfangled variants superseding the obsolete. Technology advances and diversifies by refining existing technology, which in turn has bolted the innovations of earlier times onto their predecessor’s standard. Through endless waves of innovation and copying, cultures change over time. The logic applies broadly, from the simplest of manufactured products like pins and paper clips, to the dazzling complexity of space stations and CRISPR gene editing, and even back through time to the stone knapping of our hominin ancestors and the creations of animal innovators from the distant past. Technological evolution is relentless for exactly the same reason that biological evolution is; where there is diversity, including diversity in functional utility and inheritance, then natural selection inevitably occurs.
Curiously, while the evolution of technology is apparent to many, the evolution of the arts is less widely accepted. Yet the production of artistic works, and the manner in which art changes over time, owes a substantive debt to imitation that goes far beyond the copying of styles, techniques, and materials. The film and theatre industries illustrate what architecture, painting, and sculpture affirm. That is, in the absence of a mind fine-tuned by natural selection for optimal social learning, art simply could not be produced.
Having examined the little appreciated dues that the art world owes to our biological heritage, we will go on to consider the evolution of dance. The history of dance is particularly well documented, and provides a wonderful case study with which to illustrate how human culture evolves. We will see that cultural evolution is neither linear (constantly progressing from simple to more complex over time), as envisaged by nineteenth-century anthropologists, nor treelike with independent lineages constantly branching, as Darwin portrayed biological evolution. Cultural evolution is more of a melting pot, with innovation often the product of borrowing from other domains, such that cultural lineages come together as well as diverge. This can be seen in the richly cross-fertilizing coevolution of dance, music, fashion, art, and technology, whose histories are intimately entwined.
We will begin with the movies. In the critically acclaimed film The Imitation Game, Benedict Cumberbatch received plaudits for his brilliant portrayal of Alan Turing, the eccentric genius who cracked the cyphers of the Enigma machine, used by the Nazis to send secure wireless messages during the Second World War; in so doing, Turing devised the world’s first computer. Turing’s machine endeavored to imitate the human mind and perhaps a particular mind; this was possibly the mind of his childhood friend and first love, Christopher Morcom, after whom he named his electromechanical code breaker. Turing was awarded the OBE by King George VI for his Bletchley Park services, which were estimated to have shortened the war by two to four years. But Turing’s life ended in tragedy. Prosecuted for homosexual acts, still a criminal offense in Britain in 1952, he endured two years of aggressive hormone “treatment” before committing suicide by eating an apple laced with cyanide shortly before his 42nd birthday. Only in 2013 did Queen Elizabeth II grant him a posthumous pardon and British Prime Minister Gordon Brown apologize for the appalling treatment that this brilliant scientist and war hero suffered at the hands of his nation.
Turing is widely regarded as the father of modern computing. According to artificial intelligence legend Marvin Minsky of MIT, Turing’s landmark paper of 1937 “contains, in essence, the invention of the modern computer and some of the programming techniques that accompanied it.” The metaphor of the mind has inspired artificial intelligence research for half a century, fueling countless advances in computing technology. As far back as 1996, human conceit took a humiliating hit when a supercomputer called Deep Blue defeated Garry Kasparov, perhaps the greatest-ever chess grandmaster, to exert the superiority of the mechanical over the organic mind. The world’s most powerful computer today, the Tianhe-2 supercomputer at China’s National University of Defense Technology, is the latest in a long line of refined imitations of pre-existing technology that can be traced all of the way back to Hut 8 at Bletchley Park. One day soon, quantum computers are expected to supplant today’s digital computers. Already the world’s most accurate clock is the Quantum Logic Clock, produced by the National Institute of Standards and Technology (NIST), which uses the vibrations of a single aluminum atom to record time so accurately that it would neither gain nor lose as much as a second in a billion years. Yet, in homage to their humble beginnings, such technologies are known as “quantum Turing machines.”
We readily recognize the role that imitation plays in technological evolution, just as we easily comprehend Turing’s attempt to imitate the mind’s computational power with a thinking machine. What is often overlooked is that, as every actor and actress earns a living by imitating the individuals portrayed, all movies are in the imitation game. The entire film industry relies on the ability of talented thespians to study their focal character’s behavior, speech, and mannerisms in meticulous detail and to duplicate these with sufficient precision to render their portrayal a compelling likeness, and leave the storyline credible. Cumberbatch convinces us that he really is Alan Turing, just as Marlon Brando persuaded us that he was Vito Corleone in The Godfather, or Meryl Streep is the quintessential Margaret Thatcher in The Iron Lady. The magic of the movies would dissipate instantly if this pretense ever broke down. Academy awards and Golden Globes are the ultimate recognition handed out to honor the world’s most gifted imitators. Tens of millions of years of selection for more and more accurate social learning has reached its pinnacle in the modern world’s Brandos and Streeps. Yet such extraordinary acting talent was clearly not directly favored by natural selection. No amateur dramatic productions were performed in the Pleistocene, and being a proficient actor did not bring reproductive benefits to early hominins. Acting is not an adaptation, but rather an “exaptation,”—that is, a trait originally fashioned by natural selection for an entirely different role. Acting proficiency is a byproduct of selection for imitation.
Among our distant ancestors, those individuals who were effective copiers did indeed enjoy fitness benefits, but their copying was expressed in learning challenging life skills, not the performing arts. We are all descended from a long line of inveterate imitators. By copying, our forebears learned how to make digging tools, spears, harpoons, and fish hooks; make drills, borers, throwing sticks, and needles; butcher carcasses and extract meat; build a fire and keep it going; pound, grind, and soak plant materials; hunt antelope, trap game, and catch fish; cook turtles, and make tools from their shells; mount a collective defense against ferocious carnivores; as well as what each sign, sound, and gesture observed in their society meant. These, and hundreds of others skills, were what shaped the polished, imitative capabilities of our lineage. Acquiring such proficiencies would have been a matter of life and death to the puny and defenseless members of our genus in their grim struggle to forge a living on the plains of Africa, the deserts of the Levant, or the Mediterranean coast.
Hundreds of thousands, perhaps millions, of years of selection for competent imitation has shaped the human brain, leaving it supremely adapted to translate visual information about the movements of others’ bodies into matching action from their own muscles, tendons, and joints. Now, eons later, we effortlessly direct this aptitude to fulfill goals utterly inconceivable to our forebears, with little reflection on what an extraordinary adaptation the ability to imitate represents. Imitation is no trivial matter. Few other animals are capable of motor imitation, and even those that do exhibit this form of learning cannot imitate with anything like the accuracy and precision of our species. For over a century psychologists have struggled to understand how imitation is possible. Most learning occurs when individuals receive “rewards” or “punishments” for their actions, like achieving a desired goal or else experiencing pain. This reinforcement encourages us to repeat actions that brought us pleasure and to avoid those activities that brought pain or stress, a process known as operant conditioning. The reward systems that elicit positive or negative sensations are ancient structures in the brain, fashioned by selection to train animals’ behavior to meet adaptive goals. However, when we learn to eat with chopsticks or to ride a bicycle by observing another individual, we have seemingly not received any direct reinforcement ourselves, so how do we do it? Even more challenging to understand, how do we connect the sight of someone else manipulating chopsticks, or peddling a bike, with the utterly different sensory experience that we encounter when we do these things? This correspondence problem has been the bugbear of imitation researchers for decades. Even today, there is little consensus as to how this is done. One conclusion is clear, however. Solving the correspondence problem requires links, in the form of a network of neurons, between the sensory and motor regions of the brain. Years ago, when a postdoctoral fellow at Cambridge University, working with eminent ethologist Patrick Bateson, I explored the evolution and development of imitation using artificial neural network models. We found that we could simulate imitation and other forms of social learning, provided we pretrained the artificial neural network with relevant prior experience that allowed it to create such links between perceptual inputs and motor outputs. Interestingly, our neural networks that simulated imitation possessed exactly the same properties as “mirror neurons.”
Mirror neurons are cells in the brain that fire both when an individual performs an action, and when the individual sees the same action performed by others. Mirror neurons are widely thought to facilitate imitation. As the brain expanded during human evolution, those regions now known to be involved in imitation, such as the temporal and parietal lobe, grew disproportionately larger. The parietal lobe is that precise region of the primate brain in which mirror neurons were first detected in monkeys, and brain-imaging studies show that the same regions of the human brain indeed possess these mirroring properties. Plausibly, the mirror neuron system was the direct product of selection for enhanced imitation among our ancestors. These cognitive abilities continue to allow us to learn new skills today—for instance, how to drive, wield a hammer, or cook a meal. However, those same cognitive abilities are also what permits Jimmy Stewart to impress us every Christmas as George Bailey in It’s a Wonderful Life.
Also overlooked, but far less obvious, is how reliant film, theatre, opera, and even computer games are on the audiences’ abilities to imagine themselves part of the action, to experience the fear and the tension, and to share the main character’s emotions vicariously. These capabilities were also likely fashioned in the sweaty heat of the African jungle, where the ability to take the perspective and understand the goals and intentions of those occupied in important tasks helped the observer to acquire the relevant technology. Here too, the ancestral sharing of emotions in social settings, such as responding with anxiety to the fear of another or drawing joy from the laughter of a child, helped to shape the empathy and emotional contagion that makes the movie a heartfelt experience. These sensitivities are also reliant on forms of social learning with adaptive functions, such as helping individuals learn the identity of predators or circumvent other dangers. In the absence of these social learning abilities, we would all watch movies like sociopaths, utterly indifferent to the lead character’s trauma, equally unmoved by the Psycho shower scene or Rhett Butler and Scarlett O’Hara’s kiss. Global box office revenue was estimated at $40 billion in 2015. Without the human ability to imitate there would be no movie industry; and for that matter, no theatre or opera.
In fact, when we start to think about it, connections emerge between a surprising number of the arts and the imitative and innovative capabilities that drove the evolution of the human brain. Consider, for instance, sculpture. In order to complete his statue of David in 1504, Michelangelo had to solve a correspondence problem of his own. Rather than moving his own body to match David’s posture, Michelangelo had to move his hands and arms, skillfully wielding hammer and chisel to transform a block of marble into an exact replica. To do this, Michelangelo had to translate the visual inputs corresponding to the sight of the male model into motor outputs that generated a matching form in the stone. That he exceled in this challenge, and produced one of the greatest masterpieces of the Renaissance, is testament not just to his talent but also to many years of practice in stonework. Michelangelo began his artistic training at the age of 13, and spent time as a quarryman in Carrara, where he learned to brandish a hammer to good effect. Those years of experience functioned to train the neural circuitry of his brain (just as we trained our artificial neural networks) to be sensitive to the correspondences between the movements associated with his masonry and the physical results in the stone. That training, however, could only be effective because Michelangelo possessed a brain uniquely designed to generate rich cross-modal mappings between the sensory and motor cortex when given the right experience; this was a legacy of ancestral selection for imitative abilities.
Admiring marvelous sculpture like David or the Venus de Milo can be a startlingly sensual experience, especially when one considers that we are confronted with, in essence, a block of stone. One is often left secretly wanting to reach out and touch the beautiful forms. Some cultures, such as the Inuit, even make small sculptures that are solely meant to be handled, rather than seen. That we should experience such sensations again draws on those cross-modal neural networks. These connect physical representations of objects in our minds to the objects themselves, and from there to a pre-existing network of associations and, often intimate, memories.
Only a very large-brained species could ever have produced works of sculpture fashioned with such precision. Such works require meticulous and controlled hand movements, manual dexterity that evolved along with increased brain size. Mammalian brains changed in internal organization as they got larger, inevitably becoming more modular and asymmetrical with size, as described in chapter 6. With increasing overall size, larger brain regions typically become better connected to other regions and start to exert control over the rest of the brain. This occurs because neurons vie with each other to connect to target regions and this competition is generally won by those neurons that collectively fire the target cells, giving large brain regions an advantage. The net result is an increase in the ability of the larger brain regions to influence other regions. The dominant structure in the human brain is the neocortex, which accounts for approximately 80% of the brain by volume, more than in any other animal. In the primate lineage to humans, the neocortex (the thinking, learning, and planning part of the brain) has become larger over evolutionary time, and has exerted increasing control over the motor neurons of the spinal cord and brain stem; this has led to increased manual dexterity and more precise control of the limbs. The cerebellum, the second largest region of the human brain, also plays an important role in motor control and has enlarged during recent human evolution as well. This motor control is what makes humans exceptional at finely coordinated movements. If I am correct, and innovation and social learning have driven the neocortex and cerebellum to become larger over evolutionary time, then this natural selection may simultaneously have generated human greater dexterity, which could be expressed not just in painting and sculpture, but also acting, opera, and in particular, dance.
The motor control that allows humans to produce artistic works and performances spontaneously is a capability that no other animal shares. Granted, the internet is awash with reports and YouTube footage of artistic animals, but these have not stood up to close scrutiny from animal behavior experts. You may well be able to buy painting kits for your cat or dog, and your pet may well enjoy the experience, but little that is genuinely artistic is produced. Like most other animals that have been handed a paintbrush, dogs and cats lack both the inclination and motor control to produce representational art, and I strongly suspect that any abstract beauty observed in the colorful product is strictly in the eye of the pet owner.
Intriguingly, the Humane Society of the United States recently organized a Chimpanzee Art Contest, to which six chimpanzees submitted “masterpieces.” The winner, Brent, a 37-year-old male from ChimpHaven in Louisiana, received a $10,000 prize from the stately hands of Jane Goodall. Brent, apparently, produced the work with his tongue, rather than bothering to use a paintbrush. The original works were then auctioned off on eBay with the many thousands of dollars raised going to support primate sanctuaries. Yet, however much one admires this charming, clever, and well-motivated funding initiative, the claim that the chimpanzees concerned are artists, in any meaningful sense, is greeted with skepticism by animal behaviorists and art scholars alike. A generous reading of the artistic pretensions of these animals would at best acknowledge some pleasure in generating colorful compositions.
Elephants are considerably more interesting because to the astonishment of thousands of gullible tourists, they regularly produce realistic paintings of trees, flowers, or even other elephants in sensational public performances at several sanctuaries in Thailand (figure 11). The artwork, which the elephants sometimes even sign with their name, sells in droves. However, all is not as it seems. Each paintbrush is placed in the elephant’s trunk by its trainer, who then surreptitiously guides the trunk movements by gently tugging at its ears. The elephant has been trained to hold the brush to the paper and move it in the direction to which its ear is being pulled. At the very least, one has to acknowledge an impressive piece of animal training, and one cannot help but admire the precision and control that the painting elephants exhibit with their trunks. Yet, a trick has taken place, and the trainer gets away with it by cleverly positioning himself behind the elephant. The tourists nonetheless typically go home happy, even those who spot the ruse, since no one can say that their “priceless” artwork was not painted by an elephant!
Figure 11. Painting elephants are becoming a major tourist attraction in Thailand. The elephants regularly produce realistic paintings of trees, flowers and other elephants in impressive public performances. However, all is not as it seems, and the tourists are being hoodwinked. Copyright Philippe Huguen/AFP/Getty Images.
Representational art is a uniquely human domain. That elephants can, with guidance, produce these pictures is nevertheless fascinating, precisely because it demonstrates that with training, they too are capable of building up cross-modal neural networks in their brains that translate tactile sensory inputs into matching motor outputs. The painting elephants have solved a correspondence problem of their own. It may be no coincidence that an Asian elephant from South Korea called Koshik was recently shown to be capable of vocal imitation, including mimicking human speech, while Happy, another Asian elephant at the Bronx Zoo in New York, was shown to be able to recognize herself in a mirror. Almost certainly, these capabilities are related. Like sculpture, producing paintings (and mirror self-recognition) makes demands on the circuitry of the brain involved in imitation.
Our big brains not only afford precise control of our hands, arms, legs, and feet, but also of our mouth, tongue, and vocal chords, which is what endowed our species with the vocal dexterity to speak and sing. Without that cortical expansion, members of our species could neither have fashioned a work of art, nor vocally expressed their admiration for it. The evolution of language is surely central to the origins of art, since art is rife with symbolism. As described in chapter 8, symbolic and abstract thinking are widely regarded as foremost features of human cognition. The use of arbitrary symbols allows humans to represent and communicate a wide range of ideas and concepts through diverse mediums. We possess minds fashioned by natural selection to manipulate symbols and think abstractly through spoken language, but we also express this penchant for symbolism in numerous artistic endeavors.
Architecture is one such domain. Victor Hugo’s 1831 masterpiece, Notre Dame de Paris, contains an extraordinary chapter entitled, “This Will Destroy That”; it echoes the enigmatic words of the evil Archdeacon Frollo, who rants against the invention of the printing press. Frollo expresses the terror of the church in the face of a rising new power—printing—that threatens to supplant it. The concern was not just that people might start to rely on books rather than priests to acquire their knowledge and advice, but also that the cathedral’s magnificent gothic architecture, already in disrepair, would lose its power and symbolism:
It was a premonition that human thought … was about to change its outward mode of expression; that the dominant idea of each generation would, in future, be embodied in a new material, a new fashion; that the book of stone, so solid and so enduring, was to give way to the book of paper.
To the modern reader such fears appear irrational. Yet, in the preliterate world, powerful institutions literally wrote their authority in stone. From the Pyramids to St. Peter’s Basilica in Rome or the Palace of Versailles, the magnificence, scale, wealth, and beauty percolated with the symbolism of God-given command and assuredness.
Human artwork has a long history, dating back some 100,000 years. It exhibits all the hallmarks of cultural evolution. While painting manifests multiple divergent styles, one ancient conceptual lineage sets out to represent the visual experience with accuracy. Consider, for instance, René Magritte’s famous painting The Treachery of Images, which shows a pipe that looks as though it is a model for a tobacco advertisement. Much to the puzzlement of millions of admirers, Magritte painted below the pipe, “Ceci n’est pas une pipe,” which is translated as “This is not a pipe.” At first sight, this appears completely untrue. What we momentarily forget, of course, is that the painting is not a pipe, but an image of a pipe. When Magritte was once asked to explain this picture, he apparently replied that of course it was not a pipe; just try to fill it with tobacco! Magritte’s point might appear trite to some, privileged as we are to live in an age where we can overdose on magnificent artworks that perfectly capture perspective and exhibit astonishing accuracy of portrayal. In the contemporary artistic movement of hyperrealism, the pictures of artists like Diego Fazio, Jason de Graf, or Morgan Davidson use acrylics, pencil, or crayons with such astonishing accuracy that they are almost always mistaken for photographs. Their work can be placed in a long-standing tradition that sets out to produce precise, detailed, and accurate representation of the actual visual appearance of scenes and objects. This movement flourished at various periods, and has been known as “realism,” “naturalism,” or (with appropriate reference to imitation) “mimesis.” Such hyperreal works allow the viewer to escape the correspondence problem by producing an image that exactly mimics what it represents. However, there can be no such escape for the artist, who must overcome this challenge in order to succeed.
Nowhere in the arts is the correspondence problem more clearly manifest than in dance, which again harnesses those same cognitive faculties that are necessary to integrate distinct sensory inputs and outputs. Following an excited conversation in a Cambridge pub in 2014, I recently began a collaboration with Nicky Clayton and Clive Wilkins to study the evolution of dance. Nicky is a professor of psychology at Cambridge University and expert of animal cognition; she is also a passionate dancer, and she merges this with her research as scientific director to the Rambert, a leading contemporary dance company. Clive is equally impressive as a successful painter, writer, magician, and also a dance enthusiast. We rapidly converged on the hypothesis that dancing may only be possible because its performance exploits the neural circuitry employed in imitation.
Dancing requires the performer to match their actions to music, or to time their movements to fit the rhythm, which can sometimes even be an internal rhythm, such as the heartbeat. This demands a correspondence between the auditory inputs the dancer hears and the motor outputs they produce. Likewise, competent couple or group dancing requires individuals to coordinate their actions, and in the process match, reverse, or complement each other. This too calls for a correspondence between visual inputs and motor outputs. That humanity is able to solve these challenges, albeit with varying degrees of ease and grace, is a testament to the neural apparatus that we uniquely possess as a legacy of selection for imitative proficiency. The same reasoning applies when individuals dance alone.
Contemporary theories suggest that while the potential for imitation is inborn in humans, competence is only realized with appropriate lifetime experience. Early experiences, such as being rocked and sung to as a baby, help infants to form neural connections that link sound, movement, and rhythm, while numerous experiences later in life, such as playing a musical instrument, strengthen these networks. The suggestion that taking up the piano will make you a better dancer might seem curious, but that is a logical conclusion to draw from the neuropsychological data.
The relentless motivation to copy the actions of parents and older siblings that is apparent in young children may initially serve a social function, such as to strengthen social bonds. However, childhood imitation also trains the “mirroring” neural circuitry of the mind, leaving the child better placed later in life to integrate across sensory modalities. Theoretical work suggests that the experience of synchronous action forges links between the perception of self and others performing the same movements. Whether because past natural selection has tuned human brains specifically for imitation, or because humans construct developmental environments that promote imitative proficiency—or both—there can be no doubt that, compared to other animals, humans are exceptional imitators. A recent brain-scanning analysis of the neural basis of dance found that foot movement timed to music excited regions of the brain previously associated with imitation, and this may be no coincidence. Dancing inherently seems to require a brain capable of solving the correspondence problem.
Comparative evidence is remarkably consistent with this hypothesis. A number of animals have also been characterized as dancers, including snakes, bees, birds, bears, elephants, and chimpanzees; the last of these perform a “rain dance” during thunderstorms, which has a rhythmic, swaying motion. However, whether animals can truly be said to dance remains a contentious issue, which depends at least in part on how dance is defined. In contrast, the more specific question of whether animals can move their bodies in time to music or rhythm has been extensively investigated, with clear and positive conclusions. Strikingly, virtually all animals that pass this test are known to be highly proficient imitators, frequently in both vocal and motor domains.
This ability to move in rhythmic synchrony with a musical beat by nodding our head or tapping our feet, for instance, is a universal characteristic of humans, but is rarely observed in other species. The most prominent explanation for why this should be, known as the “vocal learning and rhythmic synchronization” hypothesis, is broadly in accord with the arguments presented here. This hypothesis suggests that moving in time to the rhythm (known as “entrainment”) relies on the neural circuitry for complex vocal learning; it is an ability that requires a tight link between auditory and motor circuits in the brain. The hypothesis predicts that only species of animal capable of vocal imitation, such as humans, parrots and songbirds, cetaceans, and pinnipeds, but not nonhuman primates and not those birds that do not learn their songs, will be capable of synchronizing movements to music.
The many videos of birds, mostly parrots, moving to music on the internet are consistent with the hypothesis, but compelling footage of other animals doing the same is comparatively rare. Some of these “dancing” birds have acquired celebrity status; the best known is Snowball, a sulphur-crested cockatoo, whose performances on YouTube have “gone viral.” Snowball can be seen to move with astonishing rhythmicity, head banging and kicking his feet in perfect time to Queen’s “Another One Bites The Dust” or the Backstreet Boys (figure 12). Home videos can be faked, and parrots also have the ability to mimic human movements, so the footage alone cannot show that Snowball is keeping time to music directly. For this reason, a team of researchers led by Aniruddh Patel at The Neurosciences Institute in San Diego brought Snowball into the laboratory to carry out careful experiments. Manipulating the tempo of a musical excerpt across a wide range, the researchers conclusively demonstrated that Snowball spontaneously adjusts the tempo of his movements to stay synchronized with the beat.
Figure 12. Snowball, a sulphur-crested cockatoo, performs dances on YouTube that have thrilled millions. Careful experiments have demonstrated that Snowball adjusts his movements to keep time to the music. By permission of Irena Schulz.
Thus far, evidence for spontaneous motor entrainment to music has been reported in at least nine species of birds including several types of parrot, and the Asian elephant, all of whom are vocal imitators, and several of which show motor imitation. Entrainment has also been shown in a chimpanzee, a renowned motor imitator. The sole exception to this association is the California sea lion, which is not known to exhibit vocal learning. However, the fact that related species show vocal learning, including several seals and the walrus, raises the possibility that this capability or a relevant precursor may yet be demonstrated. Lyrebirds have not been subject to entrainment experiments, but males are famous for their ability to imitate virtually any sounds, including dog barks, chainsaws, and car alarms. They can match subsets of songs from their extensive vocal repertoire with tail, wing, and leg movements to devise their own “dance” choreography. Clearly, there is more to dance, at least social or collective dance, than entrainment to music. There must also be coordination with others’ movements, which would seemingly draw on the neural circuitry that underlies motor, rather than vocal, imitation. However, a recent analysis of the avian brain suggests that vocal learning evolved through exploitation of pre-existing motor pathways, implying that vocal and motor imitation are reliant on similar circuitry. The animal data provide compelling support for a causal link between the capabilities for imitation and dance. Whether this is because imitation is necessary for entrainment, or merely facilitates it through reinforcing relevant neural circuitry, remains to be established.
Dance often tells a story, and this representational quality provides another link with imitation. For instance, in the “astronomical dances” of ancient Egypt, priests and priestesses accompanied by harps and pipes mimed significant events in the story of a god, or imitated cosmic patterns such as the rhythm of night and day. Through dance, Australian Aborigines depict the spirits and ideas associated with every aspect of the natural and unseen world. There are animal dances for women, which are thought to function like love potions or fertility treatments to make a lover return, or to induce pregnancy, while male dances are more often about fishing, hunting, and fighting. Africa, Asia, Australasia, and Europe all possess long-standing traditions for mask-culture dances, in which performers assume the role of the character associated with the mask and, often garbed in extravagant costumes, enact religious stories. Native Americans are famed for their war dances, which were thought so powerful and evocative they were banned by the United States government—the law was not repealed until 1934. A variety of animal dances are also performed by Native Americans, and include the buffalo dance, which was thought to lure buffalo herds close to the village, and the eagle dance, which is a tribute to these venerated birds. This tradition continues right through to the present. For instance, in 2009, the Rambert Dance Company marked the bicentenary of Darwin’s birth by collaborating with Nicky Clayton to produce The Comedy of Change, which evoked animal behavior on stage with spellbinding accuracy (figure 13). In all such instances, the creation and performance of the dance requires an ability on the part of the dancer to imitate the movements and sounds of particular people, animals, or events. This reproduction contributes importantly to the meaning of the dance in the community, and imparts a bonding or shared experience. Such dances reintroduce the correspondence problem, since the dancer, choreographer, and audience must be able to connect the dancers’ movements to the represented target phenomenon.
Figure 12. Dancers from the Rambert Dance Company in The Comedy of Change. Several lines of evidence connect the ability to dance with imitation. By permission of Hugo Glendinning.
The most transparent connection between dance and imitation, however, will be readily apparent to just about anyone who has ever taken or observed a dance lesson; that is, dance sequences are typically learned through imitation. From beginner ballet classes for infants to professional dance companies, the learning of a dance routine invariably begins with a demonstration of the steps from an instructor or choreographer, which the dancers then set out to imitate. It is no coincidence that dance rehearsal studios around the world almost always have large mirrors along one wall. These allow the learner to flit rapidly between observing the movements of the instructor or choreographer and observing their own performance. This not only allows them to see the correspondence, or lack of correspondence, between the target behavior and what they are doing, but also allows the dancers to connect feedback from their muscles and joints to visual feedback on their performance, allowing error correction and accelerating the learning process.
Prospective new members of professional dance companies are given challenging auditions in which they are evaluated partly on their ability to pick up new dance routines with alacrity, an essential skill for a dancer. Dancing is not just about body control, grace, and power, but it also demands its own kind of intelligence. A key element in whether or not a dancer makes the grade essentially comes down to how good they are at imitating. A professional dancer at the Rambert once told Nicky and me that she had recently taken up sailing, and her instructor was flabbergasted at how quickly she had picked up the techniques involved. What the instructor had failed to appreciate was that dancers earn their living by imitation.
Imitation is not the only cognitive faculty that is necessary for learning dance. Also important is sequence learning, particularly in choreographed dances, which require the learning of a long, and often complex, sequence of actions. Even improvised dances such as the Argentine tango require the leader to plan a sequence of movements that provide the basis for the exquisite conversation between leader and follower. As we have learned, long strings of actions are very difficult to learn asocially, but social learning substantially increases the chances that individuals will acquire the appropriate sequence. Our ancestors were predisposed to be highly competent sequence learners because many of their tool-manufacturing and tool-using skills, as well as food-processing techniques, required them to carry out precise sequences of actions, with each step in the right order. The fact that these sequence-learning capabilities are clearly exploited in dance provides further evidence of the extent of the surprising connection between imitation and dance.
Culture evolves in two senses: the observation that cultural phenomena change over time, and the evolution of the capacity for culture. Evolutionary biology can shed light on these issues by helping to explain how the psychological, neurological, and physiological attributes necessary for culture came into existence. In the case of dance, evolutionary insights explain how humans are capable of moving in time to music; how we are able to synchronize our actions with others or move in a complementary way; how we can learn long, complex sequences of movements; why it is that we have such precise control of our limbs; why we want to dance what others are dancing; and why both participating in dancing and watching dance is fun. Armed with this knowledge, we can make better sense of why dance possesses some of the properties that it does, and why dances changed in the manner they did. As it is for dance, so it is for sculpture, acting, music, computer games, or just about any aspect of culture. Biology provides no substitute for a comprehensive historical analysis. However, our understanding of the underlying biology feeds back to make the historical analysis so much richer and intelligible.
 Mesoudi et al. 2004, 2006; Mesoudi 2011.
 This is perhaps because the arts place a particular premium on creativity and originality, and also possibly because evolution has a bad name in some areas of the humanities (see Laland and Brown 2011 for historical details).
 Morgan 1877; Spencer 1857, (1855) 1870; Tylor 1871.
 Darwin 1859.
 A story pervades that the apple logo found on iPhones and Macintosh computers is a tribute to Alan Turning, the father of modern computing, who died by biting into an apple laced with cyanide. While differing opinions abound, this story sadly would appear more likely to be an urban legend than the truth.
 Turing 1937; Minsky 1967, p. 104.
 Gould and Vrba 1982.
 Hoppitt and Laland 2013.
 Galef Jr. 1988.
 Strictly, the “most learning” referred to here should read most “instrumental” (or “operant”) learning.
 Pullium and Dunford 1980.
 For a recent review, see Brass and Heyes 2005.
 Laland and Bateson 2001.
 Rizzolatti and Craighero 2004.
 Striedter 2005.
 Iacoboni et al. 1999.
 This form of social learning is typically referred to as “observational conditioning.”
 Bronowski 1973.
 Striedter 2005.
 Deacon 1997.
 Striedter 2005; Heffner and Masterton 1975, 1983.
 Barton 2012.
 Prints of these are still available for $20 each (https://secure.donationpay.org/chimphaven/chimpart.php).
 For an accessible animal behaviorist’s assessment of elephant painting, see http://www.dailymail.co.uk/sciencetech/article-1151283/Can-jumbo-elephants-really-paint—Intrigued-stories-naturalist-Desmond-Morris-set-truth.html.
 Tourist camps in Thailand that boast “painting elephants” have attracted criticism from animal rights activists who express concerns that the training regimes may be cruel to the animals. The tourist camps, in response, claim that the elephants are mentally and socially healthy.
 Footage of painting elephants and chimpanzees can easily be found on YouTube.
 Stoeger et al. 2012.
 Plotnik et al. 2006.
 Consistent with this argument is the finding that rhesus monkeys can be trained to produce mirror-induced, self-directed behavior resembling mirror self-recognition, with appropriate visual-somatosensory training that links visual and somatosensory information (Chang et al. 2015).
 Striedter 2005.
 Hugo (1831) 1978, p. 189.
 The first use of perforated shells as beads is dated to over 100,000 years ago (D’Errico and Stringer 2011, McBrearty and Brooks 2000). The shells frequently have geometrical patterns cut into them, and have been colored with pigments. The use of red ochre as a painting material dates back further. Engraved ostrich shells dated to 60,000 years ago have been found in South Africa (Texier et al. 2010). By around 45,000 to 35,000 years ago, art was widespread (at least in western Europe) and highly consistent, and comprised pierced beads of ivory and shells, etched and carved stones, engraved decorations on bone and antler tools and weapons, and sculpted statues of animals and female figures, which were thought to be fertility symbols. However, the most evocative and striking images of Paleolithic artwork are unquestionably the magnificent cave art paintings discovered in several European countries (Sieveking 1979). Many caves are renowned for their artwork; the oldest include the spectacular paintings found at the Le Chauvet Cave in France, dated to 30,000 years ago. Perhaps the most remarkable collection of cave paintings is at Lascaux in Dordogne, France, where an incredible 2,000 painted images of horses, deer, cattle, bison, humans, and a 5-meter high bull, have been dated to 18,000–12,000 years ago. Also renowned is the beautiful painted ceiling of the cave at Altamira, in northern Spain. This was the first cave art to be discovered, in 1879. The art at Altamira, which has been dated to around 19,000–11,000 years ago, comprises stunning representations of bison, horses, and other large animals, with extraordinary use of colors and shading to indicate depth. The quaint story of its discovery details that the paintings, which are on a low ceiling, were initially missed by the team of archaeologists, but were spotted by one of the team’s 8-year-old daughter; she was the only individual small enough to stand erect and still look up at the ceiling (Tattersall 1995).
 There is the expected regional variation, with particular techniques, styles and materials used in specific locations, indicating that the art expressed particular meanings that were socially learned and shared by the members of the community (Zaidel 2013). The paintings record for posterity what dominated the minds of those peoples, the animals that they lived by and stalked, and the power and potencies that those creatures symbolized. The correspondence between those species that were painted and those that have been independently verified as present is sufficiently tight that ecologists now use paleolithic art to infer species distributions (Yeakel et al. 2014). There is also continuity over time, as the same methods and skills are reproduced throughout the millennia. For instance, the European cave art tradition lasts tens of thousands of years, while the use of pigments, such as red ochre, in rock paintings is still used today (McBrearty and Brooks 2000). These traditions were passed on from one generation to the next, picking up innovations from numerous creative, avant-garde, or radical individuals along the way, in a continuum that stretches back to the origins of our species, and forward to those exhibits found in today’s contemporary art museums. Finally, the observed patterns of change are historically contingent. Like technology, novel art does not spring forth fully formed from the mind of the maker, but rather is a creative reworking of existing artistic forms.
 The company was Ballet Rambert until 1966, and then Rambert Dance Company until 2013.
 Laland et al. 2016.
 Byrne 1999, Laland and Bateson 2001, Heyes 2002, Brass and Heyes 2005.
 Carpenter 2006.
 Heyes and Ray 2000, Laland and Bateson 2001.
 Brown et al. 2006.
 Some animals’ movements, such as the coordinated jumping and wing-flapping courtship of pairs of Japanese cranes, or the communication system of honeybees, possess some dance-like properties, but these are species-specific behavior patterns that have evolved to fulfil quite separate functions.
 Nettl 2000.
 Fitch 2011.
 Patel 2006.
 In contrast to this hypothesis, I also place emphasis on motor imitation.
 Doupe 2005, Jarvis 2004.
 Cacatua galerita eleonora
 You can see Snowball on YouTube at https://www.youtube.com/watch?v=cJOZp2ZftCw.
 Moore 1992.
 Patel et al. 2009.
 Schachner et al. 2009, Patel et al. 2009, Dalziell et al. 2013.
 Hoppitt and Laland 2013.
 Fitch 2013.
 Hoppitt and Laland 2013.
 Cook et al. 2013, Fitch 2013.
 Dalziell et al. 2013.
 Indeed, in a number of both classical and modern dance forms, motor imitation is key. Dancers are required to copy the process but not the product of the movement, and operate under socially constrained rules that depend critically on the technique and style of their particular school (e.g., Martha Graham vs. Merce Cunningham styles).
 Feenders et al. 2008.
 Clarke and Crisp 1983.
 Clarke and Crisp 1983, Dudley 1977.
 Clarke and Crisp 1983.
 Laubin and Laubin 1977.
 Clarke and Crisp 1983.
 Correction may also occur through manual shaping of the dancer’s body by the teacher or, to a lesser degree, through verbal instruction. In some dances, specific steps are given verbal labels, as in ballet in particular, which has its own elaborate glossary of terms, such as fondu, arabesque, chassé, and grand jeté, each with its own characteristic movements. Except in those cases, however, describing bodily movements with words is typically difficult. Hence, when dance instruction is given verbally, it is often through the use of imagery, where again an ability to relate one’s own bodily movements to another object, emotion, or entity is required.
 I am indebted to Nicky Clayton for drawing my attention to many of these points.
 Whalen et al. 2015.