I am interested in the origins of the capacity to make inferences about the goals and intentions of others, based on observable behavior. The comparative approach that I will take to answering this question, though innovative, is unorthodox but, I believe, has the potential to be extremely illuminating. I will investigate the origins of these cognitive inference skills in dogs and wolves, though future work may also include wolf-dog hybrids and domestic cats for greater comparative analysis.

Hare and colleagues (e.g., Hare et al., 2002; Miklosi et al., 2004) have made an extraordinary discovery: on tasks that require the inference of intention from observable behaviors, such as inferring what object is being referred to based on a pointing gesture, dogs perform substantially better than both chimpanzees and wolves, who fail at the task. This is remarkable because of what the phylogenetic distribution implies about the origin of the skill. Humans, even human children, perform well on this task, yet their closest evolutionary relatives, chimpanzees, do not. On the other hand, while dogs were domesticated from wolves only a few tens of thousands of years ago, dogs perform well on this task, but wolves do not. Hare et al. (2002) propose that this difference is a result of domestication, and that the selective forces involved in domestication in some way parallel those involved in the evolution of human social cognition. To support this, they recently found a similar difference between domesticated and wild silver foxes on the same task (Hare et al., 2005).

The social carnivores present a test system for understanding the evolution of social cognition that is superior in many ways to the human / chimpanzee system. The main reason is the multitude of changes that have occurred in the human lineage since divergence from the chimpanzee lineage, including not only changes in human cognition and behavior but also social organization and ecology. Significantly, we now have language, diverse forms of social scaffolding during childhood, mass media, and more, which render comparisons of cognitive skills with chimpanzees problematic, because it is difficult to pinpoint reasons for the difference.

Comparing dogs with wolves as I propose here, on the other hand, controls for many of these differences in species ecology, especially because I will be using human-reared wolves in this study (see below). In addition, the time in which dogs evolved from wolves is relatively short (~10,000 to ~30,000 years--see, Lindblad-Toh et al., 2005) compared to the time since the human and chimp lineages diverged (~7 million years), so the number of changes in brain architecture is probably smaller, allowing us to pinpoint more closely the cognitive changes that domestication has favored.