Riddle me this: the craft and concept of animal mind.(Humans, Animals, and Machines)

Intentionality is defined by John Searle (1984, 14-15) as "that feature of certain mental states or events that consists in their ... being directed at, being about, being of, or representing certain other entities and states of affairs." In Searle's own animal example, he asks us to

consider the case of a lion, moving in an erratic path through tall grass.

The behaviour of the lion is explicable by saying that it is stalking

a wildebeest, its prey. The stalking behavior is caused by a set

of intentional states: it is hungry, it wants to eat the wildebeest,

it intends to follow the wildebeest with the aim of catching, killing

and eating it.

Daniel Dennett, meanwhile, uses the term "free-floating rationale" (1996, 121-33) to characterize such rudimentary higher-order intentionality among birds and animals. He also draws attention to a particular characteristic of dogs, namely, that among domesticated species, only they "respond strongly to the enormous volume of what we might call `humanizing' behavior aimed at them by their owners" (1996, 165). Because of our experience of working with dogs, we have chosen to make them the focus of our concerns here.

The working Border Collie is, perhaps, the most extreme example of canine exceptionalism. According to Stanley Coren (1994), an American psychologist and dog trainer, they are the smartest of dogs in terms of working intelligence and obedience. They are also comparatively familiar, not least through televised sheepdog trials in series such as One Man and His Dog, now in its twenty-fourth year.

While they have become a highly developed form of rural entertainment, sheepdog trials, which are particularly prevalent in upland areas of Britain, are also a serious dog business. Dog and handler compete as a team in front of the public and competition judges. Their purpose is to move groups of sheep through a series of obstacles and complete a number of maneuvers which are characteristic of the demands made by practical work. From the scores available for each element of the work, points are deducted for mistakes and, perhaps, infelicities of working style. The handler stands near the audience and judges, often leaning on a shepherd's crook. The dog appears to move the sheep according to the instruction of the handler, who sends him (most successful trials dogs are male) either clockwise or counterclockwise to "collect" the sheep, tells him to lie down, drive them on, move them left or right, and so on. Instructions are normally by means of whistles and sometimes through voiced commands. After standing at a distance, the handler walks into the center of the field for the finale, holding open the gate of a small pen into which the sheep are maneuvered by dog and human.

The notable features of the successful trial performance are often emphasized by an accompanying commentary. The successful dog shows both precise obedience and an intense interest in the sheep, while the handler typically displays a particular calmness and authority. The sheep frequently appear as somewhat unwilling though calm participants in the whole drama, hanging together in a close-knit group under the authority of the dog. The level of this accomplishment is, of course, immediately apparent as soon as one sees it being done badly with sheep looking harassed and forever scattering as a consequence. For the judges and experts, all attention focuses on the subtleties of the dog's performance, and his ability to control the sheep in a finely tuned manner. For while the trials performance is inevitably stylized and formalized, its point is still to produce a simulation of the practical skills which the working dog needs on a daily basis on the farm. Sheepdog trials may have become a form of popular rural entertainment, but their main aims are still to select particular dogs as breeding stock and to demonstrate their abilities. Indeed, such elements of simulation in the sheepdog trial can often disguise more subtle aspects of dog work which become more apparent when the dog is working in the less supervised environment of the farm. The trial must be seen as merely the tip of a large and many faceted iceberg.

Our own "epistemic authority" (Fuller 1995, 161) in the canine domestic sphere derives from our practical experience of working dogs (Cox 1980, 1987, 1991, 1994, 1995a, 1995b; Cox and Deeley 1995). It is, therefore, a form of experiential expertise based upon sheep farming and working in the shooting field with gundogs. Often, needless to say, such expertise remains substantially--and in many respects, perhaps, ineradicably so--unarticulated. Science, in contrast, has more concrete roles of epistemic authority. Such rules are not always, however, conventionally positivist. Claims for replicability and experimental control, those mainstays of traditional behavioral psychology, are in some instances substituted. For cognitive ethologists, in particular, epistemic authority entails the appeal to notions such as "observed in nature," or "equivalent to nature." Such rules may be more vulnerable to challenge, but they can be defended in a number of ways. Cognitive ethologists, who are often ready to recognize the speculative status of their work, typically introduce a broad mass of supportive evidence and combine or cross-reference it with elements of experiential expertise or "commonsense" understanding. Cognitive ethology therefore constructs its epistemic authority through combining elements of a scientific legitimacy founded on positivist concepts with elements based upon wider and more speculative observation, sometimes heretically including those narratives known as "anecdotes."

How might a sheepdog trial be characterized by the traditionally positivist methods of behavioral psychology? Such approaches frequently advocate a law of …