Cooperative Intelligence: The Precursor for Language?

In my previous post, Why did language evolve? I considered the evolutionary motivations for the emergence of language: I argued that language evolved in order to coordinate the human conceptual system, for purposes of encoding and externalising thought. In this post, I consider the evolutionary precursors of language. I ask: what enabled language to get off the ground to begin with?

Intelligence, tools and other minds
Our understanding of the mental capabilities of our nearest primate cousins, chimpanzees and bonobos, has undergone a revolution in the last two decades or so. Our hominin lineage separated from the African great apes—tailless primates, that evolved for life amidst the arboreal canopy of sub-Saharan Africa—some six million years ago. But we now know that many of the essential ingredients, in terms of, at the very least, aspects of our interactional intelligence, is something that we share with other great apes.
An important symptom of intelligence, and hence, of a sophisticated conceptual system, is tool use. In the earliest stage of human prehistory, Homo habilis (‘handy man’), a very early species of ancestral humans, living in what is today Ethiopia, developed the ability to manufacture rudimentary stone tools. In particular, ancestral humans worked out that by striking or ‘knapping’ a hard stone against the fault line on other stone materials, like flint, quartz, or obsidian, a sharp, flake-like edge could be created. And these flakes could be used for cutting the skin, flesh and even bone of dead game. By around 1.8 million years ago, with the advent of the more advanced Homo erectus (‘upright man’), ancestral humans began creating other tools, out of wood, bone or antler, to work the flakes, creating more effective cutting edges. Archaeological finds, dated to this period, include stone hand axes and cleavers.
Tools are also widely deployed by a wide array of species for acquisition of food and water, grooming, construction and even defence. This ranges from sea otters who have been observed to transport small rocks to break open clams and shell-fish, to moulting brown bears that use rocks to exfoliate, to elephants who use sticks and leaves, held in their trunks to swat away flies. But it is amongst apes and monkeys that we witness the most sophisticated non-human tool use. In the last couple of decades, primatologists and comparative psychologists have made important breakthroughs in the study of the cognitive tool-manufacturing capabilities of great apes, especially chimpanzees.
Chimpanzees are particularly adept tool users: they regularly use stone tools to crack nuts. But not only do chimps make use of objects they find, they often fashion tools from naturally occurring entities. Chimps carefully strip leaves off twigs they find or break from trees and use the stems to fish for termites or other insects in the ground. Moreover, we now know that chimps can make use of several tools to help them prepare food for consumption. For instance, chimps from the Nimba moutains in Guinea, Africa make use of both stones and pieces of wood to act as cleavers to cut up treculia fruits—about the size of a melon—into more digestible portions. Intriguingly, they additionally make use of stone anvils as a chopping surface—combining two tools for food preparation, a feat once thought to be beyond the ken of all non-human species. Chimps even create tools for hunting. For instance, chimps from Senegal have been observed to manufacture rudimentary spears to hunt for bushbabies.
Up until the 1980s, ethologists thought that of the great apes, it was only chimps that were capable of manufacturing tools in this way. However, we now know that a range of species are capable of this. Both bonobos and chimpanzees are known to create sponges from moss and leaves. They use these sponges to absorb water before washing themselves. Like Chimps, orangutans also manufacture tools to fish for termites. They do this by fraying the end of a twig, and then inserting into termite nests in trees. Moreover, mandrills—Old World monkeys closely related to baboons—have been observed to manufacture tools for personal grooming. In the wild, mandrills regularly create tools to clean their ears. And in one instance, a mandrill was even filmed stripping down a twig before using it to clean away dirt from under its toenails.
Another feature, and arguably, the hallmark of interactional intelligence, is the ability to recognise that, like you and me, others have thoughts, feelings and wishes that are not dissimilar from our own. This amounts to an understanding that other members of one’s species are intentional agents. Until relatively recently, it wasn’t clear whether the great apes—gorillas, orang-utans, chimpanzees and bonobos (pygmy chimps)—understood other members of the same species to be intentional agents. In philosophical circles this idea is sometimes referred to as having a ‘theory of mind’: understanding others in one’s species as having minds, which, like our own, we can attempt to interact with and influence, via communicative strategies. Moreover, having a theory of mind, means that we can recognise the intentions of others, based on visual cues such as facial expressions, body posture and so on. Recognising others as intentional agents affords an intention-reading ability: we can modify our own behaviour, and attempt to influence the behaviour of others, as a consequence of reading their intentions. For instance, if I were to see a small child at the edge of busy road, seemingly about to step out, I might react by inviting the child to wait, or to find a safer place to cross. And this unremarkble act arises precisely because I recognise the child’s intention—its desire to cross the road—and the potential harm that might arise, due to the child’s relative immaturity, and hence its failure to fully anticipate the potential danger.
However, chimps, we now know, appears to have, on the face of it, something approaching a theory of mind. The psychologist, Dale Peterson, recounts a telling example of chimp deception, which betokens an ability to gauge what other chimps may be thinking. One young chimp, nick-named Dandy, was courting a female. The form that courtship takes, amongst chimps, is to directly reveal just how arousing a male chimp finds a female to be. So, Dandy revealed, to his paramour, his erect penis. But just then, a higher-ranking male approached, making advances towards Dandy’s love interest. As Peterson reports: “For Dandy, that meant big trouble. Worse, the young male's forbidden interest in the female was being communicated honestly by something he seemed to have little control over, an erect penis—which he quickly covered with his hands, apparently hoping thus to deceive the higher-ranking male.”
What this strikingly reveals is that while Dandy had no voluntary control over his erection, nevertheless, in the presence of an older, stronger male, his sexual arousal, and the potential threat it suggests to the older male’s attempt at sexual activity, is unwelcome. Moreover, Dandy’s aroused state might even be detrimental to Dandy’s well-being, given the older male’s physical prowess, and prior territorial claim. The only explanation, available, for Dandy’s subsequent behaviour—attempting to hide his erection from the competing male—is precisely because Dandy understands that fellow chimps have wishes and desires of their own—that other chimps are intentional agents. Dandy wished to both avoid causing offence, and to being physically chastised as a consequence. Chimps seem to understand that other members of their species have minds of their own, with their own wishes, desires and beliefs. And as we saw earlier, recognising a member of one’s own species as a minded creature—an intentional agent—is foundational to the human meaning-making capacity: our interactional intelligence—and its distillation, in the human lineage, as a sophisticated cooperative intelligence—depends on it.

The cooperative species
Interactional intelligence is a necessary precursor to the emergence of a genuine cooperative intelligence, with the concomitant pro-social impulse evident in our species. For reasons that are, perhaps self-evident, something like language would not have been possible without the cooperative turn that emerged at some point in the minds of our forebears, sometime in the last two million years or so. After all, while we now know that chimps understand that other members of their species are intentional agents with whom they can interact, their life-style remains primarily individualistic, rather than cooperative. And they don’t have language.
In contrast, our species is inherently cooperative: we achieve far more, working together, than we ever could alone. Our societies are inherently cooperative—in modern society, we pay taxes to ensure the upkeep of welfare states of various stripes, to maintain a free state-education system, and in the United Kingdom, a publically-funded National Health Service, free at point of access. There is the rule of law, overseen by a police service, a judiciary and other state-run agencies. We share streets where we live, laid out by adherence to and sanction by public planning bodies—and pay local taxes for refuse removal, street lights, and so on. We drive on communal roads, adhering to a highway code and other traffic laws, and we agree upon a political infrastructure—sometimes through democratic elections, sometimes through other means, such as monarchic-feudal governance structures, or a combination of the two: the UK, for instance, has an unelected sovereign as Head of State. And these cultural institutions administer and drive many aspects of our national life, which ultimately, affect each of us in our daily lives. This cooperative mode of living is qualitatively distinct from any other extant species.
My claim is that language could only have arisen in such a cooperative milieu. It isn’t sufficient just to understand that other members of our species are intentional (and sometimes rational!) agents that we can engage with. Cooperative intelligence entails more than merely co-action, alongside others. It additionally involves shared and devolved responsibilities for the greater good, which we all, ultimately, hope to benefit from. It requires pooling resources, and agreeing who will do what in order to facilitate obtaining and then the sharing of resources—food, shelter, and the opportunity to mate being chief amongst these.
But cooperation entails more than interaction, and the division and sharing of labour in order to achieve common goals. It requires a means of communication, not least to ensure coordination of the goals we cooperatively attempt to achieve. In short, without this new, cooperative way of living, there would—and probably could—not have been language. And this is why our species—and lineage—developed language, whilst our nearest primate cousins have not. Our lineage, and species, developed a new form of cooperative intelligence, built upon the interactional intelligence shared, more or less, with our great ape cousins. And consequently, what is different about our species is not just the nature and range of concepts that we hold in our conceptual systems; we additionally have language that allows us to make use of these self-same concepts for meaning-making and communication.
The comparative psychologist, Michael Tomasello, proposes that the development of cognitively modern humans, with our contemporary smarts, has followed an evolutionary trajectory, in terms of the stages that have led to our cooperative intelligence. And crucially, Tomasello’s natural history of human thinking centres on the notion of intentionality—our awareness of the thoughts, wishes and feelings of others—and how far and to what extent we engage with the intentionality of others.
For the most part, chimps, and other great apes, operate at the level of ‘individual intentionality’. Chimps, for example, spend most of their waking hours “in small bands foraging individually for fruit and other vegetation, with various kinds of social interaction”. While chimps are social creatures, and recognise hierarchical and social relationships, all essential for group living, their interactions are primarily self-centred and self-serving, as in the case of Dandy’s attempt to hide his erection. Moreover, when chimps do engage in collaborative hunting, their actions remain largely individualistic, rather than truly collaborative. For instance, chimps sometimes engage in group hunting of monkeys. In this situation, when a monkey has become separated from its group, the chimps opportunistically surround and proceed to capture the monkey, a strategy that can only work through collaboration. While one chimp initiates the chase, others move to block possible routes of escape. But only one chimp actually captures the monkey, and that is the chimp who consumes most of the meat.
Tomasello argues that “chimpanzees are engaged in a kind of co-action in which each individual is pursuing his own individual goal of capturing the monkey.” The pursuit can only be achieved by collaborating in a cooperative way. But the fact that the meat is not shared—those that don’t actually do the capturing are left with scraps and remnants—suggests that chimps are not quite operating beyond a single intentionality mode. The hunt is not fully cooperative. Chimps recognise the need for interactional co-action, but not fully-fledged cooperation, which involves performing specific individual roles in the hunt, in exchange for the promise of a fair share of the spoils.
At stage two, individuals achieve what Tomasello calls ‘joint intentionality’. This involves an agreement—or at least a recognition—that the greatest spoils can be achieved by, for instance, hunting in a truly collaborative manner. But of course, if individuals are to work as a team, there can still only be one captor. But to work as a team, and to be able to bring down large game, requires those individuals that are not actually involved in the kill being rewarded appropriately: their role is, after all, crucial to the successful hunt, such as driving the game into a trap where it can be subdued and killed. And it is this understanding—greater spoils result from pooling resources, and subsequently sharing food—that betokens this more cooperative cognitive strategy.
Tomasello proposes that following the separation of hominins from the great apes, and for the first several million years of the hominin lineage, through the genus Australopithecus—essentially an upright ape—the cognitive strategy was essentially individualistic. It most likely operated on a similar plane to that of other great apes. Whilst australopithecines most likely had some aspects of human-like interactional intelligence, they lacked the essential ingredient of being able to set aside their individualistic motives for a more cooperative approach to procuring food. They were still operating at the level of single, rather than joint intentionality.
During the emergence of the genus Homo, a little over two million years ago, ancestral humans began the move to joint intentionality, and a fully-fledged cooperative intelligence. According to Tomasello, this most likely culminated with the species Homo heidelbergensis, the presumed common ancestor of our species, Homo sapiens, and Homo neanderthalensis. Fossils for Homo heidelbergensis (or a closely related species, Homo antecessor), date from 800,000 years ago, at a site in Spain; this suggests the species may have existed perhaps over a million years ago in Africa. Archaeological finds of weaponry, such as sophisticated, wooden throwing spears, also reveal that Homo heidelbergensis may have been the first human species to be able to bring down large game, through collaborative hunting. And this could only have been possible if they hunted cooperatively. Moreover, there is archaeological evidence that this species of ancestral humans brought the food back to a base to be shared out, including by those, especially women and children, that hadn’t taken part in the hunt. Behaviour of this kind would, indeed, seem to suggest the kind of cooperative intelligence that Tomasello associates with joint intentionality.
Our capacity for joint intentionality appears to be a defining feature of human cognitive organisation. It is for this reason that we share in the collective upbringing and education of our children. Chimps and other great apes don’t. We point things out to others, in an unsolicited fashion: those things that we believe might be helpful for them. Again, chimps don’t do this. And as I observed earlier, we take collective decisions about our shared, and individual lives, a feature absent from the lifestyle of chimps. These types of behaviour, as well as language, are only possible with the advent of cooperative intelligence—joint intentionality.
While language is not something that we are born with—it is not innate—human cooperative intelligence is. Numerous experiments in the psychology lab demonstrate that pre-school children—children with little or no language—are naturally cooperative, in a way that chimps, in the same lab settings, just aren’t. For instance, in one study, children aged just three were invited to play a game with one adult. Once they were engaged in the game, a second adult invited them to join in another even more compelling game. Before leaving for the new game, the infants either paused, or handed the instrument they were using to the first adult, seemingly sensing that a joint activity entails a joint commitment to see the game out. In contrast, there is no evidence that chimps of the same, or indeed any age, perceive joint activities in this way.
In the third and final evolutionary stage of cooperative intelligence, our lineage has developed what Tomasello terms ‘collective intentionality’. This type of cooperation involves the extrapolation of joint intentionality in order to develop rituals and other social structures that support, maintain, and even constrain our collaborative way of life. Social constructs such as money, marriage, and so on are all cultural artefacts that result from intelligence of this kind. And with this final, extended type of cooperative intelligence, evident in Homo sapiens, and arguably in the social behaviour of the extinct Homo neanderthalensis, full-blown language of the kind we recognise today would have had to be in place, in order to support the rich culture that this sort of collaborative living entailed.
In the final analysis, our species is uniquely cooperative in the way that no other species is. Once ancestral humans began the move away from single intentionality—the type of interactional intelligence which early hominins shared with great apes—rudimentary forms of communication, perhaps via gesture and pantomiming, would have already been on the march. And the way of life that this gave rise to, would have further selected for enhanced collaboration, leading, ultimately, to the development of full-blown language.