Developing Intelligence

There are many theories of how human behavior came to differ so profoundly from that of even our closest primate relatives – language, recursion, theory of mind, and enhanced working memory are just a few of the “critical components” that have been proposed as enabling human intelligence. A very different perspective, advocated by Tomasello and Carpenter, suggests that it is simply humans’ extreme propensity for social interaction that is at the core of the evolution of human intelligence.

In their article, Tomasello and Carpenter focus on four effects of this social tendency, which they also term “shared intentionality.” The first of these relates to how shared intentionality changes the dynamics of early life experiences.

Consider that chimpanzees “know what others see,” – in other words, they follow the gaze of others, will retrieve food only if the dominant chimp cannot see it, and will thereafter hide this food from the dominant chimp. However, they will not utilize this knowledge in ways that even 1-year-old babies do. That is, human babies but not chimps seem to infer that a companion has discovered something new about an object if they continue to look surprised while gazing at this object. Based on this and other evidence, Tomasello & Carpenter suggest that humans are driven to share attention and interest with others from a very early age, whereas primates are not.

The authors suggest that a second effect of “shared intentionality” is a fundamental difference in communication. The authors leave aside language in favor of analyzing pointing habits: 14-month-old babies, but not chimpanees, will infer that an adult is pointing towards the contents of a bucket rather than the bucket itself if they had previously been playing a “hiding game” together. Likewise, infants but not chimps will point to unfamiliar objects in their environment, as if simply to engage in shared attention.

This relates to Tomasello & Carpenter’s third effect of “shared intentionality: collaboration. Whereas chimpanzees are exquisitely sensitive to competition, they do not show the same sensitivity to collaboration. Even group hunting can be viewed as involving relatively little shared goal- and plan-making, instead consisting mostly of monitoring others’ hunting strategies and responding appropriately. Further support comes from studies in which infants but not chimps will attempt to re-engage an adult in a game which the adult had abruptly stopped playing.

Finally, the authors suggest that a fourth effect of “shared intentionality” is a change in the processes of teaching or instruction. For example, adult chimpanzees seldom explicitly demonstrate specific actions to younger chimpanzees, who seem to learn instead by modeling those actions without the elder’s knowledge. In contrast, human infants are acutely aware of the social aspect of learning, in the sense that they will precisely model an adult’s demonstrated actions if the adult is watching, but will just reproduce the result achieved by the adult in the absence of social attention.

While the behavior of human infants and chimpanzees can seem very similar, Tomasello & Carpenter conclude that they critically differ in their social dynamics. Specifically, human infants can use “shared common ground” as a pragmatic reasoning tool in social interaction, in terms of both learning and play with both various objects and others. Whereas, infants are particularly motivated to share their mental experiences, chimpanzees seem to engage in such activities as only a means to an end.

Although humans and non-human primates differ in a variety of seemingly more important ways (e.g., language and enhanced working memory being just two), Tomasello suggests that “shared intentionality” is the most critical: a propensity for social interaction allowed each of these other capacities to evolve to their current state. In contrast, other theorists have hypothesized other “core differences,” for example the use of recursive symbols. Unfortunately, it’s difficult to imagine how some of these theories could be falsified, since we have little ability to infer “recursion” or “social intention” from the archeological record of early humans, and even if such data did exist, these differences could be incidental as opposed to causal factors in the development of human intelligence.

Related Posts:
Symbols Language and Human Uniqueness
Smarter than the Average Primate
Simulating Emotion and Emotional Robotics


  1. #1 Stephen
    January 5, 2007

    Also these ideas, “shared intentionality”, “recursive symbols” are presented as binary. You have them or not. However, it’s much more likely that there is a continuum. Some people pick up math easier than others. Some people’s social prediction is better than others. In a human group, this variety can be of benefit to the group as a whole. The group doesn’t need everyone to be a seer or mathematician. So, it’s even possible that there is no critical skill. The diversity itself can be a critical group phenomenon. No one’s brain needs to be that huge and expensive, and yet the group can outperform other groups.

    So why is it that when you get 100+ very smart people together, you get a group that is as stupid as Congress? It’s because in a big group, it’s still the highly skilled individual that’s important. The other 99 people should recognize the one who has the right skills for the task and use just that one opinion. But what Congress has is 100+ A-type personalities that think their opinion always matters. And yet, at the moment, this is the best we can do.