The dog-human connection in evolution
Posted by gregdowney on August 23, 2010
Evolutionary theorists have long recognized that the domestication of animals represented a major change in human life, providing not just a close-at-hand food source, but also non-human muscle power and a host of other advantages. Penn State anthropologist Prof. Pat Shipman argues that animal domestication is one manifestation of a larger distinctive trait of our species, the ‘animal connection,’ which unites and underwrites a number of the most important evolutionary advances of our hominin ancestors.Shipman’s proposal is discussed in a recent forum paper in Current Anthropology and is the subject of her forthcoming book, The Animal Connection. The paper is interesting to us here at Neuroanthropology.net because Shipman indirectly poses fascinating questions about the evolutionary significance of human-animal relationships, including the cognitive abilities of both and how they interact.
As Shipman puts it in the Penn State press release about the research, if we only think about what domesticated animals do for us as a species, we miss the truly curious thing about our relationship to them:
No other mammal routinely adopts other species in the wild — no gazelles take in baby cheetahs, no mountain lions raise baby deer…. Every mouthful you feed to another species is one that your own children do not eat. On the face of it, caring for another species is maladaptive, so why do we humans do this?
Although researchers working on symbiotic inter-species relationships might highlight that the support of other species hardly requires adopting their young and feeding them canned kitten food (a critique Travis Pickering levels in his comments), Shipman’s statement highlights nicely that human-animal inter-species relationships seem to extend beyond merely treating them as tameable prey or means to a human end. But then again, this super-instrumentality could be ascribed to a large number of human traits.
The domestication of animals wasn’t merely about capturing a buffet-on-the-hoof, from Shipman’s perspective, but the continuation of a long-term evolutionary project by our species to study animals, first when we were prey for them, and later as predators ourselves.
One of the clinchers for her argument is that the first animals domesticated were not food sources, but a fellow predator and scavenger: the wolf (dogs being descendants of wolves, even a subspecies by some reckoning). Clearly, domestication wasn’t first about eating the animal:
Shipman suggests, instead, that the primary impetus for domestication was to transform animals we had been observing intently for millennia into living tools during their peak years, then only later using their meat as food. “As living tools, different domestic animals offer immense renewable resources for tasks such as tracking game, destroying rodents, protecting kin and goods, providing wool for warmth, moving humans and goods over long distances, and providing milk to human infants” she said. (again, from the press release.)
Overall, Shipman’s thesis is an interesting one, stronger in relation some eras than in others, in part because the implications, at an abstract level, can be quite vague. But I think her analytical framework is very much worth considering as it shifts some of the evolutionary theoretical discussions about hominin advances. Often, evolutionary theorists can treat the advent of tools or the domestication of plants as a rational advance brought on by purely cognitive means: an exclusively human community made a kind of technologically-available-means-ends calculation.
Shipman’s discussion, instead, focuses our attention on how humans might have been perceiving the world, and their place among a variety of species, not just in relations to other hominids. Considering how complex relations with animals might influence human foraging strategies, technology, and perhaps even symbolic functions is a useful new angle on a number of questions.
I’ve been thinking a lot about dogs lately (more on this), so considering the cognitive dimensions of our evolutionary relationship to animals is particularly relevant to me, but most specifically about dogs rather than other domesticates. I’m not sure I’m wholly persuaded by Shipman’s main point, which seems to be that the animal connection is driving the last several million years of human evolutionary development, but I do like very much how her perspective opens up a whole set of really intriguing theoretical questions.
Dog blogging and my canine connection
Dogs are on my mind, not merely because they’re my best friends (really), but also because I have an honours student doing a year-long research project on dog-human interaction in sheepherding trials.
One of the great things about being an academic advisor is working with intensely bright undergraduates. The collaboration can help remind the jaded researcher of just how great we have it, pursuing our curiosities and intellectual obsessions wherever they might lead. This year, I’m in the enviable position of working with three extremely sharp undergraduates doing their honours theses, which is like a senior year of the undergraduate career almost wholly devoted to a single long original project.
‘Dog man’ is Paul Keil, an impressive interdisciplinary thinker who, to our delight, is completing his thesis in the Department of Anthropology, but also works as a research assistant in the Macquarie Centre for Cognitive Science. Besides his project on dogs, he’s also helping on some research on collective memory and other issues in cognitive science.
Paul first brought the Shipman article to my attention when he must have gotten a pre-publication look at it from someone connected to Current Anthropology. This is the abstract:
A suite of unique physical and behavioral characteristics distinguishes Homo sapiens from other mammals. Three diagnostic human behaviors played key roles in human evolution: tool making, symbolic behavior and language, and the domestication of plants and animals. I focus here on a previously unrecognized fourth behavior, which I call the animal connection, that characterized the human lineage over the past 2.6 million years. I propose that the animal connection is the underlying link among the other key human behaviors and that it substantially influenced the evolution of humans.
In the introduction to the article, Shipman suggests that three key innovations are normally treated as hallmarks of our species: tool use, symbolic thought, and food domestication.
I hypothesize that a fourth trait, the animal connection, is an equally important and diagnostic behavior of humans and that the animal connection unites tool making, symbolic behavior and language, and domestication into an adaptive package.
In fact, I’d say that there’s a few other hallmarks, but some of these are less clearly linked to the ‘animal connection’ which is Shipman’s ultimate goal; one could cite bipedalism, adaptation to heat and long-distance movement, dietary versatility, complex social structures…
The exercise of isolating the human-ness from all other animals ‘essences’ is not a terribly interesting one to me. Just because other animals do something (like use tools or communicate) doesn’t mean that humans’ behaviours are any less odd or interesting.
Shipman sees this animal-human connection as beginning first with hominin observation of animals as prey (although also, presumably, our ancestors would have also worried about animals predator, to avoid or deter).
The animal connection comprises an increasingly intimate and reciprocal set of interactions between animals and humans (i.e., members of the genus Homo) starting ∼2.6 million years ago (mya). The animal connection began with the exploitation and observation of animals by humans. Over time, regular social interactions were incorporated into the animal connection. This trait is expressed today in the widespread adoption, or cross-species alloparenting, of animals—including dingoes, possums, bandicoots, raccoons, deer, moose, bison, fruit bats, lizards, bears, tapir, monkeys, sloths, coatimundis, antelopes, zebra, tree kangaroos, rabbits, weasels, ferrets, rodents, and birds, cervids, felids, and canids of all types…—as members of the family.
The list of animals with which we sometimes have alloparenting relationships is impressive, suggesting not merely that humans have been capable of domesticating all sorts of animals, but also that we’re pretty generous about providing a space in our communities and flexible enough to support a wide variation in animal behaviour. To put it bluntly, we’re pretty easy for an animal to counter-domesticate (or whatever you want to call what they’re doing to us as we’re domesticating them).We can be trained by animals of many sorts to provide them with appropriate housing, food, interaction, and other support. As Shipman describes, we will go to all sorts of extremes, even breast-feeding animals, pre-chewing their food, or purchasing outrageously expensive gourmet dog biscuits and fuzzy little sweaters with matching booties in order to keep our animals from running off to seek for more congenial living arrangements.
This two-way argument has also been made by theorists like Peter Bleed (2006), pointing out that focusing on the animal-human relationship as wholly driven by human domination underestimates the degree to which the little buggers have us where they want us (I mean, look at Spud – you want to say ‘no’ to that?). As early as the work of Zeuner (1963), theorists pointed out that the human-dog relationship was not like other dynamics of domestication, and that dogs themselves may have initiated a process that led to their eventual domestication by living commensally, or following along with humans and slowly adapting to life with our type, rather than by simply being the passive victims of human projects to dominate and shape animals.
Shipman’s approach is helpful in thinking about dogs in that she doesn’t draw such a stark divide between ‘wild’ and ‘domesticated’ animals, stressing instead the continuity in human ability to understand, observe and use animals. Similarly, Peter Bleed (2008) points out that a reconsideration by paleoanthropologists and evolutionary theorists of the transition between foraging and farming has also found continuity rather than an absolute and categorical divide.
Other hallmarks of humanity
Shipman goes on to review three other, more widely accepted hallmarks of human distinctiveness: tool use, symbolic language and domestication. She repeats the classic anthropological discussion by Leslie White of tool use as ‘extrasomatic adaptation,’ or outside-the-skin forms of adaptation:
Tool using in the broad sense is an extrasomatic adaptation (White 1959) of humans: a means by which humans evolved behaviorally without adapting physically.
As some long-time readers will realize, I’m not entirely comfortable with White’s terminology because I feel the hard-and-fast line between behaviour and biology, technological and physiological adaptation, obscures the ways that tools shape the human nervous system and body. To say that a tool is ‘extrasomatic’ or outside the body denies the ways that tools influence bodily development, whether it’s skeletal asymmetry or increased sensitivity in the thumbs of videogame players.
Stone tools are not easy to use or to manufacture, I would argue; chimpanzees, for example, have a very hard time manufacturing stone tools even when they are shown how to do it, because the manual skills involved – bringing high energy percussion precisely onto the tool’s surface – require great dexterity, eye-hand coordination, and trainability. But this is an argument for a different post – back to animals.
Shipman discusses a wide range of symbolic behaviour – ‘ritual, language, art, objects of personal adornment, and the use of ochre and pigments’ – as indicative of a distinctively human way of life. The difference between human symbolic behaviour, especially language, and animal communication has been discussed extensively in the anthropological literature, and a wide range of traits has been suggested as distinguishing the two. The arbitrariness of linguistic signs, the ability to generate greater complexity with grammar, the disambiguation possible with spoken words, the greater abstractness and ability to discuss non-concrete subjects or absent objects, rather than just index elements in the immediate environment, all stand out as critical.
Finally, beginning as early as around 15,000 bp, humans domesticated plants and animals, giving our ancestors control over food sources, making sedentary life possible, and eventually producing specialization and surplus labour, the foundation for ‘civilization’ and more rapid technological change.
As Shipman points out, following Clutton-Brock (1999), domesticating animals is significantly different from domesticating plants as it requires shaping animal behaviour as well as morphology, learning to interact with the animal’s instincts and capacities: ‘Domesticating an animal is fundamentally developing a means of communication with that animal’ (Shipman 2010). In contrast, humans do not have ‘reciprocal, individual relationships’ with the plants that we grow.
The connection between humans and animals must have been quite intimate, Shipman argues, as zoogenic diseases crossed over into human populations, and we know from ethnographic research that humans and domesticated animals often live under the same roof or in the same compounds.
The modern human-animal connection
Shipman argues that this long evolutionary interaction with other animals has reinforced a distinctive openness to mutual relationships in humans, which we can see in a host of contemporary traits. Shipman draws out evidence of the positive health benefits of pet ownership, for example, to highlight the role animal-human relationships play in our lives, and points to the more intensive examples of helping dogs and therapeutic pets.
From the contrast between the rarity of alloparenting between species among other wild mammals and of feral children being raised by animals, Shipman concludes that the animal-human connection is essential to our species:
From this evidence, I conclude that adopting and nurturing individuals of another species is an extremely rare behavior among nonhumans, whereas the animal connection is a universal human behavior. In summary, the animal connection clearly is a universal human trait with a fundamental and enormous effect on human well-being.
Shipman cites statistics on household ownership of pets and expenditures on pets as demonstrating the ‘universality’ of the human-animal connection. In fact, the picture on this, I think, is a bit more ambivalent and even trending in the wrong direction for her argument.
In Japan, for example, highly urbanized populations have lower pet ownership rates: the rate is as low as 20% of households. In parts of the US and Australia, rates are much higher (around 62% of households in the US and 63% in Australia have pets), but these rates have fallen from as high as 85% in some rural regions. (I’m pulling some of my information off the Australian Companion Animal Council report, Contribution of the Pet Care industry to the Australian economy, 7th edition, 2010, and the 2009/2010 AAPA National Pet Owners Survey (US), not a free download.)
Like I said, I’m sympathetic to Shipman’s discussion, but I still run up against my own reservations about declaring things ‘universal’ when they are, in fact, quite rare in some societies. I’m just not sure why we need to make such an easily falsifiable over-reaching claim when the observation of patterns of human-animal nurturing is, regardless of universality, incredibly interesting.
The animal connection as evolutionary driver
In her account of human evolutionary developments, especially the advent of tool use, symbolic activity and animal domestication, there’s little serious to dispute; although some dates are still subject to debate, which commentators like Richard Klein point out, and other theorists might share my reservations about specific terms (like ‘extrasomatic adaptation’), the general outline is relatively well agreed upon and many familiar landmarks appear in Shipman’s outline.
Where Shipman’s evolutionary account starts to depart significantly from a consensus in the field is to point out the presence of animals and animal-human relations in the other three evolutionary milestone processes. She makes the case, not just to include animal-human connection among other defining traits, but as a foundation for the other major changes in hominin lifeways.
I hypothesize that as an ancient, diagnostic trait of the human lineage, the animal connection had a major influence on human evolution, genetics, and behavior. This hypothesis predicts that the fossil and archaeological record will include abundant evidence that (1) humans were intimately and persistently connected with animals, (2) human adaptive changes were causally linked to the animal connection, and (3) a meaningful adaptive advantage of the animal connection can be identified in each stage of human evolution.
Shipman first points to the advent of stone tools themselves, the fact that their development would have made our ancestors as much predator as prey, as the first step in the emergence of the animal-human connection. As hominins became more predatory, they would have to study both prey animals and fellow, competing predators, both to better acquire food and to fight off rivals.
The change in diet from primarily herbivore or opportunistic, carrion-eating omnivore to confident hunting predator was an extraordinary shift in relation to other animals, as Shipman points out in her response to other commentators:
The human-animal connection propelled human evolution later, according to Shipman when between 200 kya and 40 kya, evidence of technological and cultural change suggest the development of more sophisticated modes of communication, including language and symbols. Shipman again argues for the influence of animals on human evolution, as more sophisticated, broad-based foraging required the memory and transmission of more extensive information. The process is most evident in cave paintings, Shipman suggests, where the evidence is not simply that many, if not most of the paintings are of animals:
Ecologically, making a transition from a largely herbivorous or even omnivorous niche to a predatory one is of tremendous importance. There are few examples in the mammalian world of species that have made such a transition, perhaps because it has such momentous consequences.
What is not depicted in any recognizable fashion is also key: there are no landscapes, no depictions of geographic features (mountains, water sources), no dwellings or shelters, and nothing about climate or weather. Rarely depicted subjects include humans, insects, small animals, birds, plants, reptiles, nuts, fruit, berries, or tubers… Thus, the overwhelming frequency of animal depictions, coupled with the expanded exploitation of animal resources, indicates the increasing value of the animal connection.
Shipman points out that images of animals in cave paintings took more than just a little effort – pigments were costly in terms of resources as they often were carried long distances, mined, mixed with binding agents that might also be rare. Their use suggests, in the words of other theorists, that our ancestors were ‘informavores,’ my new word for the day.
Although most current discussions of the evolution of human intelligence focus on its usefulness in intra-species social interaction (for example, Robin Dunbar), Shipman highlights that these early representations do not feature human social interaction – though they could – but rather symbolically salient animals. From cave paintings, we see a community intelligence and visual code focused intensely on large-bodied animals, or inter-species communication, with a startling degree of detail and observational skill.
Shipman sums up:
The first and second stages of human evolution reveal a trajectory in behavior that is marked by an intensifying focus on the behavior and ecology of animals, accompanying a progressive broadening of the human predatory niche. When joined by increased sophistication in tool making, the animal connection enabled some human populations to procure more animal resources from a wider range of species and habitats—an obvious evolutionary advantage. Knowledge of the animals without the tools, or possession of the tools without the knowledge of the animals, was unlikely to have been advantageous.
I agree with Shipman’s basic point that tools without knowledge are unlikely to have been terribly helpful, but I don’t think that this necessarily proves humans have a special ‘connection’ with animals. After all, virtually every predator gets really good at either going hungry or understanding on some level, even if not at all conscious, where to find prey and how to bring it down.
But her restatement of some well-known facts, such as the predominance of animals, not all of which were primary food sources, in cave paintings in light of her argument about human-animal relations does demand some consideration. Was the animal-human connection driving human evolution? I’m not so sure. Is the predominance of animals in painting interesting? Hell, yes, and I’m glad she’s drawing this out for further thought.
Rather than seeing the ‘connection’ as a driver of human evolution, I think we can see it as a very intriguing product of having such a large brained, interpretively creative animal becoming a predator. The European cave paintings, often of animals that did not predominate in the painters’ diets, for example, may not have been caused by a ‘connection’ with animals, but by the challenges of passing on information about animals in human communities or of the creative elaboration of the relationship between predator and prey into new, ritualized, even magical or proto-religious forms.
The absences are as interesting as the fact that animals are present. For example, the fact that dogs and wolves do not appear more often in the paintings is intriguing, especially if dogs were already socially interacting in rich ways with humans by the time of the paintings. One suggestion is that dogs, like humans, were somehow tabooed from depiction, as neither species appears with anything like the frequency we might expect (if the depictions were simply drawn from daily life without ideological or cultural biases).
In historical hunters and gatherers, we often find very elaborate theories and even spiritualities around the predator-prey relationship: animals are treated with certain forms of respect and complex rituals structure elaborate ‘relationships’ that, in empirical terms, are pretty one-sided. The hunters may be staging all sorts of rituals in honour of their prey, but it’s not clear that this is because of any actual connection to the animals other than in the perceptions of the hunters. More likely, it’s an extraordinary projection that creates a rich relationship, lived experientially only by the humans involved, but one that may contain pragmatic as well as mystical information.
Starting at least 15 kya, humans began to have long-term interactions with wolves that transformed them into dogs. Shipman goes with the earliest dates provided by archaeological evidence of the ‘Goyet dog’ from Belgium, dated to 32 kya, but these remains are still fairly controversial, in part because the intervening evidence is fairly sparse. Moreover, genetic evidence tells a radically different story, pushing back the possible divergence of domesticated dogs from wolves to as long ago as 135kya (Vilà et al. 1997; see also Zeder et al. 2006; for a discussion, check Anthropology.net, but see Shipman’s piece in American Scientist).
In another of her articles, Shipman offers an account of human domestication of orphaned wolves that is consistent with the idea that animals are ‘tools,’ and the implication that humans were the active agents in the relationship, that disagrees with the more mutual relationship described by Paxton and the Coppingers. Shipman explained in her 2009 article in American Scientist (‘The Wolf at the Door’):
How did this important change come about? Probably in the distant past, humans took in a wolf cub, or even a whole litter of cubs, and provided shelter, food and protection. As the adopted cubs matured, some were aggressive, ferocious and difficult to handle; those probably ended up in the pot or were cast out. The ones that were more accepting of and more agreeable to humans were kept around longer and fed more. In time, humans might have co-opted the natural abilities of canids, using the dogs’ keen noses and swift running skills, for example, to assist in hunting game. If only the most desirable dogs were permitted to breed, the genes encoding for “better” dogs would continue to be concentrated until the new domesticated species (or subspecies) was formed.
The wide disparity in likely dates of dog domestication and the difficulty of determining the earliest example of dog/wolf-human relations points to the complexity of these relationships, that they are not as simple as the scenario Shipman provides. The story she adopts is consistent with her focus on alloparenting and animals as ‘living tools,’ but not consistent with current thinking about how the reciprocal relationship between wolves and humans produced dogs.
A number of scholars have pointed out that the first wolves living in close proximity to humans might have been ‘self-domesticating,’ either following camps of nomadic hunter-foragers (see Paxton 2000) or living commensally near permanent human settlements to forage on human refuse (see Coppinger and Coppinger 2001). This pre-domestication stage is typically referred to as ‘commensal’ living, when two species benefit from living in close proximity and avoid overt conflict. The Copingers, for example, focus on wolf scavenging in human dump sites as a key point of contact, but the existence of ‘dumps’ requires sedentary lifestyles.
If, however, we start to think that the date of first ‘domestication’ could be earlier than around 15 kya, we run up against a problem, described by Schleidt and Shalter (2003: 65):
A word of caution, however; what do we mean by “domesticated”? In a most general sense: “no longer in its wild or natural state.” But, were our own ancestors back then, long before they built permanent houses for themselves, less “wild” than the wolves they associated with?… Is it not absurd to talk about the “domestication” of dogs by humans who had not yet any permanent domiciles (“domus”)?
The point is not merely semantic. If the date of wolves becoming linked to human communities is significantly prior to 15 kya, then Schleidt and Shalter are right: wolves were adapting to humans on the move, not eating refuse from human ‘dumps.’ The humans themselves were nomadic, likely following herds of prey if skilled in game hunting (and not clustering at coastlines, for example). In other words, the humans were likely living a bit like wolves, albeit with a host of cultural traits and technological capacities well beyond their canine colleagues.
Peter Bleed (2008: 8 ) points out that fully-fledged domestication required sufficient stability of the human communities involved and dogs and humans may have developed their ‘connection’ long before humans founded permanent settlements. From the point of view of the animal species, a domesticated form can only survive if the ecological niche that it will occupy, that is, the human-centred environment, is reliable. Bleed writes:
Most importantly, this new way of life gave people a new ecological standing in that, by the end of the Pleistocene, human groups were big and regular enough to provide a separate and reliable niche that could be “occupied” by other species. This niche was not a geographic space. Rather, it rested on the cognitive and cultural capabilities of anatomically modern folk. By 12,000 years ago, that niche was substantial and dependable enough to become important to the evolution of other organisms.
By saying that the niche was ‘cognitive and cultural,’ Bleed also can be read to suggest that it was not the sedentary nature of the population so much as the stability and continuity of community practices. For dogs, especially, even a traveling group of humans might provide a sufficiently stable niche as long as their practices reliably generated food, edible refuse and other necessary resources. A camp wolf needs a relatively predictable band of humans to follow, not a dump to inhabit.
In either commensal scenario, the later, permanent settlement version or the earlier camp following discussion, the human-dog/wolf connection would be especially difficult to detect archaeologically because there might have been little or no initial physiological change to the wolves involved and much less intimate engagement than in the orphaned wolf pup scenario that Shipman repeats. Moreover, the earliest relationships could have involved minimal behavioural modification, as well (as I’ll discuss), as wolves are so social that the behavioural jump to living alongside humans might not be that great.
In addition, commensality likely occurred when there were still large numbers of non-domesticated and non-commensal dogs, so the slow modification of a fragment of the population might easily be masked by periodic gene flow between population resevoirs. The semi-domesticated dogs could have interbred with their wild relatives for long periods of time (it’s hard enough to keep our dogs on the property, and they’re neutered and don’t get a whiff of a wild relative in heat to encourage wandering off at night).
Shipman doesn’t dwell on the specific theoretical and archaeological problems posed by the distinctive case of dogs, nor does she integrate alternative accounts of commensality. This skews her account of the human-animal connection, especially because dogs were likely the first animals domesticated and may have been engaged in tens of thousands of years of commensal living prior to fully-fledged domestication. To really explore the animal connection, we would have to spend a fair bit of time and thought on this earliest and longest relationship with its distinctive character and the species that, arguably, taught humans how to interact intimately with animals. This chapter in human animal relations, in which dogs and only dogs were in intimate relations with humans, was at least a few thousand years, but more likely tens of thousands of years (and perhaps even as long as 100k years).
Ironically, the dog is an ideal animal to use as a paradigm for an ‘animal connection’ in Shipman’s argument, for all kinds of behavioural and social reasons (which we’ll get to). Mietje Germonpre critiques Shipman a bit for the absence of dogs in his commentary on the target article, but Germonpre’s focus on animals as ‘vessels of symbolic meanings’ and potential sacrifices for ritual activity, in my opinion, is a distraction from the central issue of cooperative relations.
Dogs as tools: my disagreement with Shipman
Shipman assumes that humans saw dogs as tools, treating the first domesticated animals much like stone choppers or wood spears, when she writes about the earliest examples of domestication. For Shipman, animal domestication was an extension of the logic of tools:
In essence, domestic animals are another kind of extra-somatic adaptation or tool that expands the resources humans can exploit. Transferring the concept of tool making and tool using from inanimate stone or wood to live animals was a fundamental advance in human evolution predicated on knowledge of biology, ecology, physiology, temperament, and intelligence of target species; of the selective breeding; and of communication techniques based on the animal connection.
Although I agree that animals are valuable renewable resources, the instrumental thrust of this account biases her article in two ways that I find difficult to reconcile.
First, because Shipman uses the idea of ‘living tool’ so strongly, the metaphor undermines the argument that the animal connection drives human evolutionary and technological development. Rather than privileging inter-species social relations and communication, the concept of ‘living tool’ subordinates the logic of animal domestication to the first great evolutionary development Shipman discusses: the creation of increasingly sophisticated inanimate technology. I’d like to suggest that animals are not tools, they have wills of their own, their own inclinations, instincts and patterns of reaction, so they demand more subtle cognitive abilities than inanimate tools.
Second, the account Shipman offers of human ancestors using animals to pursue human ends may exaggerate human foresight and refigure a collaborative dynamic as a human decision. Peter Bleed (2008: 8) warns that overestimating human awareness of what they were doing during domestication, over-emphasizing the instrumental account, and focusing too much on human agency can misrepresent the likely course of domestication. Bleed writes:
In spite of general understanding of the reciprocal nature of human ecological systems and Rindos’ warning that early human actors could not have been conscious of the implications of their actions, human choices and intentions tend to be given a central role in discussions of domestication. Even analyses offered from an explicitly evolutionary perspective present human cognitive capabilities as the factor that brought about changes in plants and animals.
A way of moving discussion of agricultural origins beyond narrow co-evolutionary approaches or explanations that rest on human intentionality is to ask how diverse nonhuman species might have been drawn into a new kind of relationship with people during the relatively brief period known as the Neolithic Revolution.
Shipman’s instrumental assumption — that animals are best thought of as ‘living tools’ —is not questioned by the commentators in Current Anthropology, but here I want to suggest that treating animals as ‘living tools’ creates several problems with understanding the course of domestication, especially of dogs. The irony, though, is that I think being more sensitive to the non-instrumental relations between humans and animals, especially dogs, actually helps Shipman’s argument and enriches the account of the animal connection by clearly distinguishing domestication from tool use.
Part of my resistance to thinking of animals as ‘tools’ arises from conversations with my honours student, Paul. I’ll let Paul report his findings (he’s writing up now), but one of the things that is clear from his ethnography is that dog handlers have to work with their animals, how the animal can’t simply be turned into a ‘tool’ that expresses the will of the human leader. Dogs, like other animals, have their own instincts, cognitive abilities, reactions and the like – handlers and dogs wind up much more like partners than in a one-sided or dictatorial relationship. Although Paul’s research is with modern examples, I think we have to reflect on what animals can do before we understand the tenor of early human-animal relations. (For example, some domestic animals may be so hard to handle that they are not ‘tools’ simply because of their obdurate-ness, whereas for dogs the issue is more than limitations or inclindations.)
Animals (huh!): What are they good for?
Shipman outlines a wide range of things that domesticated animals are good for:
At least 10 such resources can be identified as follows: (1) muscular power beyond human strength; (2) rapid transport of goods or people; (3) raw material (wool or hair) for making fabric, rope, and so on; (4) useful fertilizer, fuel, and building material (manure); (5) free disposal of refuse and ordure; (6) mobile wealth and storage for excess grain crops (which can be retrieved via slaughter); (7) high-fat and high-protein food (milk and milk products) for adults and weanlings, enabling a decrease in interbirth spacing; (8) protection for people, possessions, and dwellings; (9) tracking and killing of game or pests; and (10) combined traits that enable humans to live in new habitats. Examples of the last resource include the advantages camels offered in deserts, pigs and dogs in Oceania, reindeer and dogs in the Far North, yaks in the mountain regions of Asia, and alpacas and llamas in high-altitude South America.
When we focus on the likely long period of wolf-human commensality, an absolutely essential chapter in the biography of the animal connection, not all of these resources were in play. It’s unlikely, for example, that camp wolves or commensal dogs would have been any good for # 1, 2, 3, 4, or 7 in much of a way. #6 might have been a factor if humans kept commensal dogs to eventually eat them, but they would have made a terribly inefficient calorie storage device. #10 might be in play, but it’s not immediately obvious. The only things that commensal dogs would be good for in an indisputable fashion are functions #5, and to some degree, #8 and 9.
I do agree with Shipman that learning to handle and live successfully with animals would have been a selective advantage, but the point is that, for a very long period of time when humans were learning how to deal with dogs, the animals wouldn’t have been terribly useful measured by this standard of ten possible resources. Our ancestors would hardly have been able to get a camp wolf, an animal not entirely committed to hanging out with humans, to drag a sled full of gear. So what would dogs do for humans?
Sensing through animals: wolves, for example
Before dogs were useful for things like fetching balls and chasing fire trucks, I suspect one of the earliest capacities that they brought to neighbouring human communities was their perceptions. Wolves and other animals could perceive some sensations well before our ancestors, even if hominin sensory acuity in natural settings likely would have made their descendants seem positively disconnected from their environments (as I discussed in the post, Your Brain on Nature: Outdoors and Out of Reach 2). Wolves, for example, have especially acute hearing and night-time vision, both of which would be useful to humans.
The timescales of both models for a commensal transition to dog domestication suggest that their wolf ancestors may have first dwelled close by human encampments, living on waste and scavenging from humans. The wild dogs most likely to endure in this camp follower or garbage picker niche had to overcome both fear of humans excessive flight response, on the one hand, and too much aggression, on the other, or risk being driven off or injuring the source of scavengeable resources.
Whether dog ancestors were de facto members of human groups or just lurking nearby, humans would have come to realize that the animals sensed things that the humans could not. Sounds and smells that didn’t raise an alarm in humans might cause dogs to start growling or to perk up to pay attention, especially at night. The process would hardly have been all that unusual to accomplished hunters and savvy observers of animals (like those who produced the cave paintings in Southern Europe). The realization would have likely built upon earlier awareness that animals reacted to each other, for example, when hunting, if prey startled.
Anyone familiar with animals in the wild learns to read bird movements, changes in the background noise of a forest and the like in order to anticipate threats or even simply perceive things that are beyond the range of direct sensation. For example, finding sources of water can be made easier by observing animals’ movements, the sounds of frogs, tracks, and a host of other cues.
Since dogs, like other social animals, signal amongst themselves when they feel threatened or sense something intriguing, humans surely learned to cue on this behaviour, likely before dogs were domesticated to any degree. Growling, hair standing on end, and stirring more generally among dogs, especially if the dogs were not pointed at the humans, would have been a good sign that something was amiss.
In this sense, dogs could become a kind of sensory prosthesis for humans, providing cues from sensations that were below the threshold of direct human perception. I’m reminded of this many nights when our dogs detect movement that I cannot perceive or when my wife asks one of them, ‘What’s wrong?’ when they behave oddly.
But the crucial dimension of this sensory commensality for the purposes of Shipman’s discussion of the animal connection is that this early realization that other animals sense, and the possibility of long-term dependence upon dogs or other animals as sensory prostheses, would involve a recognition on some level that animals ‘know’ things we do not. The humans would have to develop, in addition to an intra-species Theory of (Human) Mind, an inter-species Theory of (Animal) Mind, offering another important way that increasing intelligence might prove to be of adaptive advantage (see Flinn et al. 2005).
One of the key differences between dogs and wolves is that dogs can learn to pay attention to both their human interlocutors and to the things to which we pay attention. They cue off of our attention. For example, Miklósi et al. (2003) found that dogs are better social communicators with humans, looking to their handlers when running into an insoluble problem for cues like pointing, and hypothesize that dogs might even be better able to ‘read’ human faces than either wolves or other intelligent animals, like chimpanzees.
But the same hyper-social skill of looking to inter-species communication to understand the world is also a characteristic of humans, seen most clearly in the exaggerated case of guide (or ‘seeing eye’), tracking, and sniffer dogs. The fact that humans, not just dogs, can cue off the perceptions and non-verbal communication of other animals suggests that the animal connection is a manifestation of social intelligence, and thus part of a suite of advantages arising from increased encephalization (the reason I refer to Flinn et al. 2005). Sensory cuing, loosely under #8 of Shipman’s list of domesticated animal functions, might have been the first use dogs could have served to human communities, even before inter-species contact was intimate to any degree.
The reason I think sensory commensality would have been the first use of dogs is that it requires virtually no modification of wolf behaviour, except for proximity to humans. In addition, the human cognitive abilities required, the Theory of (Animal) Mind, is not a great leap for our ancestors either. Both human and dog adaptations are very much in line with the species’ hyper-sociality, and this sort of Theory of Mind activity and gregarious sociality with other species are not something found in a lot of other species. Even wild chimpanzees have trouble with things like pointing, which humans and dogs both figure out.
If we really want to think about how wolves might be useful to humans, one question to ask is what are wolves really good at doing: turns out, that’s hunting. Wolves are terribly young evolutionarily, not much older than humans; the available evidence suggests that a recognizably modern wolf may have only arisen less than 3 mya.
Prior to their partnering with humans, wolves were already enormously successful, behaviorally-flexible social canids, ranging across virtually the entire terrestrial Northern Hemisphere. They were the most wide-spread, large-bodied mammal, the top of the food chain, unless they came up against tigers or our ancestors. Ironically, once domesticated, Lupus became even more successful, tagging along with human colonization to every habitable continent.
Arguably, human ancestors displaced Lupus from its position atop the food pyramid, occupying a very similar niche of fast-moving, pack-hunting predator, well equipped to claiming ungulate game that moved in herds. Schleidt and Shalter (2003: 63) point out that human-dog collaboration brought together the two most social and successful hunters, remarkably similar in many ways once humans developed the necessary technology and skills.
Many later domesticates were complementary species, eating foods that humans could not digest and transforming them into animal protein (which we could). Dogs were different; dogs could be partners in doing something that our ancestors were increasingly good at doing themselves.
I take this analysis, in part, from Schleidt and Shalter (2003), who point out that both Homo sapiens and Canus lupus were extremely socially adept, pack hunting ominvores, as adept as other large carnivores at taking prey but less specialized or finicky about what they ate. Schleidt and Shalter (ibid.: 62) highlight how the social qualities of wolves, especially their cooperative inclinations and relatively nonviolent forms of intra-group organization, allow them to form “mixed, multi-species packs: humans, dogs, cats, goats, sheep, horses living in harmony.” One coming out of Africa, the other out of the Americas, these two mammals met up in the same niche with corresponding strengths in behaviour, and instead of fighting for competitive exclusion, ended up working together.
I’m not wholly convinced by the arguments that Schleidt and Shalter advance, but the idea that commensal wolves could hunt with humans seems to me to be more plausible than other initial uses from the list of ten potential resources. Wolves were already hunting in groups, using ambush and endurance-based techniques that might have been quite similar to contemporary humans, although humans likely brought some distinctive technological advances to the equation (such as projectile weapons).
Domestication from the dog’s perspective
Dogs are so unusual among domesticated animals, such an outlier, that the dog may be an instructive special case of domestication, also instructive to our ancestors in that the dog may have taught humans how to deal with animals. As Richard Klein suggests in his commentary on the Shipman article:
Only the special relationship between people and dogs may be significantly older [than 12k years], and almost everything about it, including its near ubiquity and its level of intimacy, suggests that it might be understood not so much from the human side but more from the dog’s as the human connection.
That is, if dogs are so unusual, maybe their domestication shows us more about the dog’s potential than about humans’ distinctive connections to animals. If dogs were self-domesticating, maybe they showed us how to do it, driving the rise of greater human intelligence about animals that would eventually lead our ancestors on to more challenging inter-species arrangements.
For example, Schleidt and Shalter (2003: 57), argue that the ability to be domesticated is not merely a matter of intelligence, nor is it of animals being like us or having our capacities. Schleidt and Shalter ask:
Isn’t it strange that, our being such an intelligent primate, we didn’t domesticate chimpanzees as companions instead? Why did we choose wolves even though they are strong enough to maim or kill us?
Chimpanzees turn out to be really hard to socialize; dogs, not so difficult. For Schleidt and Shalter, the case of dogs suggests that we need to understand how dogs are shaping us as well as how we are shaping their selection, a ‘tool’ that shapes its user (a process that I argued is under-explored when treating culture as ‘extra-somatic adaptation’; I would suggest tools also become part of our species’ developmental niche and selective pressures).
Rather than seeing the dog as a ‘tool,’ Peter Bleed says that we also need to understand the animals as active partners in the relationship and humans as active, but often unconscious selectors. Bleed (2008: 9) nicely compares the idea that humans directed domestication comparable to saying that ‘petri dishes developed penicillin.’ Instead, he advocates recognizing the active role of the human partners and domesticates: ‘There is theoretical and academic utility in looking at “domesticates” not as passive resources or even as co-evolving species, but as influential occupants of a dynamic set of opportunities afforded by people.’ For the camp wolves, humans were a potential niche to occupy, one that they adapted to so well that they wound up piggy-backing on human global dispersal.
Our non-human social life
To wrap up, I want to reflect on the implications of considering dogs as first animal domesticates and not as ‘tools.’ Shipman’s ideas are intriguing, and I think that, if we recognize the distinctive way that our ancestors likely interacted with dogs, we see that animal domestication, in part, was not an extension of the logic of tools, but a growing sophistication on the part of humans’ social capacities. Shipman writes:
Clearly, humans who handled and lived with animals more successfully accrued a selective advantage in performing tasks that humans without animals could not achieve. Domestication was reciprocal, as the animals in turn selected for behavioral or physical traits in humans, such as better communication with animals and the continued functioning of lactase into adulthood.
I also agree strongly that domestication was ‘reciprocal,’ the presence of animals shaping humans just as humans shaped domesticated animals, but again, I think her own description undermines a too-simple reading of the animals as ‘tools.’
One of the most well developed argument for the function of increasing intelligence in humans is that our ancestors needed a lot of cognitive wattage to deal with complex social interactions among their fellow hominids (see, for example, Flinn et al. 2005). In a world of alliances, in-group sexual competition, cooperative hunting, demanding childcare situations, extended kin-based solidarity and long-standing inter-group rivalry, a human needed a fair bit of gray (and white) matter to sort out all the players.
The social intelligence hypothesis, sometimes referred to as the ‘Machiavellian intelligence hypothesis’ after a concept introduced by primatologist Frans de Waal, suggests that a wide range of human cognitive abilities might serve well in a swirling social world, helping us to accomplish such tasks as perceiving each others’ motivations, ascertaining the limits of each others’ knowledge, remembering allies and enemies, and recognizing deception.
But increased intelligence also would have allowed humans to sort out animals behaviours and cognitive capacities, realizing which animals were really dangerous, which could be frightened off, and how different prey behaved when confronted. All sorts of activities, from ambush hunting to choosing a good safe place to sleep for the night would have been made more effective with increased intelligence as our ancestors started to perceive more, not just of other humans’ motivations and perceptions, but also the capacities of non-humans.
In this sense, I strongly agree with Pat Shipman:
Domestication, she explained, is a process that takes generations and puts selective pressure on abilities to observe, empathize, and communicate across species barriers. Once accomplished, the domestication of animals offers numerous advantages to those with these attributes. “The animal connection is an ancient and fundamentally human characteristic that has brought our lineage huge benefits over time,” Shipman said. “Our connection with animals has been intimately involved with the evolution of two key human attributes — tool making and language — and with constructing the powerful ecological niche now held by modern humans.” (from the Penn State press release)
Again, she may be over-reaching, but she’s spot on when saying that, once animals become part of our intimate niche, they start to be part of the selective forces acting upon us, whether that’s as disease resistance to zoonoses (illnesses originating in animals, such as measles in dogs, mumps in poultry, tuberculosis in cattle, and the common cold in horses, as Sandra Olsen lists in her commentary) or social attributes that might make a person more likely to succeed in a human niche that includes other animals.
For example, a genetic abnormality that makes a person smell tasty like a leg of lamb might be increasingly maladaptive if you’re camping out with only-recently-domesticated wolves. Or, more seriously, being able to communicate with dogs, using one’s own emotions or activities in ways that dogs find predictable and directive, could help a hunter accompanied by camp wolves or a human band member looking to canines for protection.
Conclusion: From Machiavellian to Doolittlean intelligence
In this context, the dog represents a really interesting social problem: dogs can be pretty damn unpredictable, and the same species that can become your ‘best friend’ (in a proverbial sense) is also a pretty fierce adversary. Unlike other animals that can be pretty reliably classed into ‘predator,’ ‘prey,’ or ‘irrelevant,’ dogs, especially in that communal period, would likely be capable of nearly anything from a hominid point of view.
You wouldn’t just need Machiavellian intelligence, you’d also need Doolittlean intelligence (after Dr. Doolittle, that is): sharpened abilities to perceive the emotions of a non-human, to recognize animal behaviour patterns, and to develop ways to affect those behaviour patterns. These animal connection-related skills would be on top of other animal liked cognitive skills, such as those needed to stalk prey and avoid predators. Domestication would have substantially increased the cognitive challenges of dealing with animals by welcoming a whole new, more complicated class of actors into long-term intimate contact with human groups.
What I’m suggesting is not quite as strong as what Shipman posits — that the animal-human connection drove human evolution — but rather that the possibility of an animal-human connection demonstrates that the humans involved aren’t just capable of Theory of (Human) Mind perception. They’re obviously figuring out a Theory of (Dog) Mind as well, and perhaps a whole range of other feats of subjectivity shifting that allows them to ‘walk in the paws,’ if you will, of other species.
In addition, if humans are able to develop a Theory of (Animal) Mind that is not simply anthropomorphizing, this would suggest that, at least in these circumstances, Theory of Mind was not only a projection of one’s own awareness, but might also be something more malleable or imaginative. One can see this very clearly in Rane Willerslev’s (2004) remarkable ethnographic research with Siberian hunters.
I’m not entirely convinced of this last point. Most evidence suggests that humans close to dogs tend to anthropomorphize; surveys in the US, for example, reliably assert that their animals have human-like personalities at rates of around 90% or higher. Although anthropomorphizing animals, assuming that they have human emotions, can be a profound error in understanding our non-human neighbours, the tendency to perceive animals as having human-like emotions or motivations or personalities suggests a supple flexibility in that social mind that extends beyond just human life.
This runs against some ideas about how humans accomplish such feats as perceiving others’ intentions, for example, by projecting ourselves into the position of the other actor or by simulating their emotional reactions in our own emotional parts of the nervous system (for example, through mirror neurons). Instead, our ancestors would have to learn to perceive a range of non-human actors as having their own social signaling and non-verbal communication. These ancestors couldn’t get to that space simply by anthropomorphizing because the animals don’t have identical forms of expression (although there are interesting relations).
Shipman’s account of human-animal relations as part of human evolution, although I disagree on some of the finer points, is a really persuasive and intriguing way to rethink human cognitive development. Her article (and the forthcoming book) reminds us to include animals, not simply as predators and prey, in our thinking about the niche in which our ancestors developed. But treating this relationship as an extension of tool use likely overstates our ancestors’ foresight about the usefulness of partially-domesticated wolves and underestimates the degree to which humans and animals were in reciprocal relationships.
But I’m grateful to Shipman, not merely for the thought provoking article, but also for a great opportunity to post pictures of my dogs! So here’s the crew at our farm:
For more on dogs and evolution:
The press release about Shipman’s work on Penn State’s website: New Hypothesis for Human Evolution and Human Nature.
An earlier article by Pat Shipman on ‘The Wolf at the Door’ (American Scientist) which discusses both the evidence of early dog domestication and the issue of how difficult it is to pin down specifically domestic variation in Lupus.
Commentary by John Hawks on The Aurignacian dogs.
For a dissenting view, you could start at Bruce Bower’s piece, ‘Oldest dog debated,’ at Science News.
Evolution: The Curious Case of Dogs at Observations of a Nerd (great post, especially the discussion of Moscow feral dogs adapting to become great beggars on the city’s metros).
L. Case’s 2008 article from the Journal of Animal Sciences, ‘ASAS CENTENNIAL PAPER: Perspectives on domestication: The history of our relationship with man’s best friend.’ (86:3245-3251. doi:10.2527/jas.2008-1147)
See also Rodney L Honeycutt’s recent (2010) article, ‘Unraveling the mysteries of dog evolution,’ from BMC Biology (8/20, doi: 10.1186/1741-7007-8-20), available as an open source read at PubMed Central.
Science writer Carl Zimmer offers an account of dog intelligence, the opening of the Duke Canine Cognition Centre, and Henry the schnoodle figuring out what pointing means at Time Magazine: ‘The Secrets Inside Your Dog’s Mind.’
Retrieverman compiles a number of sources of accounts of ‘camp wolves’ living with Native American, Aboriginal, and other at least partially foraging peoples in his post, Camp Wolves.
Ridiculously cute photos of Louis, Roxy and Spud (a friend’s Groodle puppy), by the author.
Map of wolf distribution from Wikimedia Commons.
Created by Tommyknocker.
Cartoon. Not sure anymore. (If anyone knows, I’m happy to put a credit on this.)
Bleed, Peter. (2006). Living in the human niche Evolutionary Anthropology: Issues, News, and Reviews, 15 (1), 8-10 DOI: 10.1002/evan.20084
Clutton-Brock, J. 1999. A natural history of domesticated mammals. 2nd edition. Cambridge: Cambridge University Press.
Coppinger, Raymond, and Lorna Coppinger. 2001. Dogs: A new understanding of canine origin, behavior, and evolution. Chicago: University of Chicago Press.
FLINN, M., GEARY, D., & WARD, C. (2005). Ecological dominance, social competition, and coalitionary arms racesWhy humans evolved extraordinary intelligence Evolution and Human Behavior, 26 (1), 10-46 DOI: 10.1016/j.evolhumbehav.2004.08.005
Miklósi A, Kubinyi E, Topál J, Gácsi M, Virányi Z, & Csányi V (2003). A simple reason for a big difference: wolves do not look back at humans, but dogs do. Current biology : CB, 13 (9), 763-6 PMID: 12725735
Paxton, D. (2000). A Case for a Naturalistic Perspective Anthrozoos: A Multidisciplinary Journal of The Interactions of People & Animals, 13 (1), 5-8 DOI: 10.2752/089279300786999996
Schleidt, Wolfgang M., & Shalter, Michael D. (2003). Co-evolution of Humans and Canids: An Alternative View of Dog Domestication: Homo Homini Lupus? Evolution and Cognition, 9 (1), 57-72
Shipman, Pat. (2010). The Animal Connection and Human Evolution Current Anthropology, 51 (4), 519-538 DOI: 10.1086/653816
Vila, C. et al. (1997). Multiple and Ancient Origins of the Domestic Dog Science, 276 (5319), 1687-1689 DOI: 10.1126/science.276.5319.1687
Willerslev, R. (2004). NOT ANIMAL, NOT NOT-ANIMAL: HUNTING, IMITATION AND EMPATHETIC KNOWLEDGE AMONG THE SIBERIAN YUKAGHIRS Journal of the Royal Anthropological Institute, 10 (3), 629-652 DOI: 10.1111/j.1467-9655.2004.00205.x
Zeder, Melinda A., Eve Emshwiller, Bruce D. Smith and Daniel G. Bradley. 2006. Documenting domestication: the inter-section of genetics and archaeology. Trends in Genetics 22(3): 139-155. doi:10.1016/j.tig.2006.01.007
Zeuner, Friedrich Eberhard. 1963. A History of Domesticated Animals. Harper & Row: New York, Evanston.