Can dogs think? Maybe yes, and maybe no. What dogs do quite well, though, is make people think that dogs can think

Natural History, Feb, 2005 by Bruce Blumberg, Raymond Coppinger

How Dogs Think: Understanding the Canine Mind by Stanley Coren Free Press, 2004; $26.00

If Dogs Could Talk: Exploring the Canine Mind by Vilmos Csanyi North Point Press, 2005; $25.00

Where is the dog owner who hasn't wondered just what might be going on in that canine head? Those who train dogs professionally are certainly moved at times, in moments of frustration, to doubt that anything is going on in there at all. If dogs are so smart, why can't more than a small percentage make it through training as agility dogs, customs dogs, guide dogs, or obedience dogs, to perform reliably at tasks that seem quite straightforward?

Yet when it comes to presuming that there are such things as canine minds or canine thinking--as the books under review do in their titles--ordinary skepticism seems to take the night off. If only one could just sit down with a dog over a glass of sherry and ask, What's on your mind? Maybe that's the key to the matter: dogs seem to possess some special talent that makes people think about them as drinking buddies. It is the rare dog lover who doesn't have a story about some astonishingly humanlike behavior in a dog. Dog owners all seem to want to prove their dogs are special--veritable geniuses in the world of animal cognition, or at least, like the children in Garrison Keillor's Lake Wobegon, all above average.

Even animal experts, who are supposed to know better, can find it hard to resist the temptation to explain canine behavior with fanciful "just so stories." Dog stories almost invariably imply that the dogs in question can think--and think like people.

But stories are one thing, science another. In science, the best one can do is observe the behavior of dogs in natural and controlled settings, and then, on that basis, make testable inferences about canine cognition. Yet despite the popularity of dogs--or maybe, in large part, because of it--the scientific investigation of canine behavior and cognition has lagged far behind that of many other animals. The reasons are not hard to fathom. Ethology, for instance, is the study of the behavior of animals in their natural habitats, but for domesticated canines, the "natural habitat" is the somewhat artificial one people have created. Another reason for the dearth of scientific studies is the deep emotional bond between dogs and people, which poses a constant challenge to the objectivity of the investigator. Particularly when people's tendency to anthropomorphize their dogs is coupled with such thorny concepts as "mind" and "thought," the need to be aware of unconscious bias and unfounded interpretation is paramount.

Two recent books tackle the problem head on. In How Dogs Think: Understanding the Canine Mind, Stanley Coren, a psychologist at the University of British Columbia and a noted author on dog-human interactions, concludes that the scientific evidence just doesn't support ascribing humanlike mental activity to dogs. In If Dogs Could Talk: Exploring the Canine Mind, Vilmos Csanyi, an ethologist at Eotvos Lorand University in Budapest and a noted dog expert, seems equally convinced of the opposite conclusion.

Most dog books fall roughly into two types: the ones that focus on training, and the ones that tell dog stories. And though Coren and Csanyi serve up healthy portions of both, they also strive mightily to go beyond those formulas and report what rigorous research has to say about whether dogs have minds and can think--and, if so, to what degree they think like people.

Coren begins his book by tracing the history of the concept of mind as applied to canine cognition--what he calls "the battle between 'dog as thinker' and 'dog as machine.'" In Coren's account, philosophers such as Plato and Diogenes exemplify the former, whereas strict behaviorists such as B.F. Skinner insist on the latter. Expert opinion then just swings back and forth between these two poles like a pendulum.

Is the field of canine cognition doomed forever to repeat this seemingly endless dispute? In his book Wild Minds: What Animals Really Think, Marc D. Hauser, a psychologist and neuroscientist at Harvard University, suggests a different approach. Hauser sees an animal's ability to solve problems--to act intelligently--as indicating that it possesses a set of "mental tools." Those tools arise out of a complex interplay of genes, development, and learning. Hauser further suggests that, to the extent a species faces specific problems, its tool kit may include unique or specialized tools, of greater or lesser quality, that do not occur in the tool kits of other species. Species specialization does not make one species "smarter" than another, according to Hauser, just "wonderfully different."

Applied to dogs, Hauser's approach cuts through the vague but passionate disputes about whether dogs think, or just act like robots. Instead, he suggests that research focus on relatively straightforward comparisons between the mental tools deployed by various species. For example, how well do dogs compare with other species in remembering objects that are out of their sensory range? This mental tool, known as "object permanence," presumably must be brought to bear if a dog is to' answer that age-old question, "Where is your ball?" Mental tools that deal with object permanence, as well as counting, mapmaking, knowledge about the nature of objects, and on and on, help an animal solve problems such as where to find hidden caches of food.