Hume illustrates his conception of animal rationality with commonplace examples. For the most part his examples depict the function of learning on the basis of past experience: A dog is readily accustomed to come to attention when humans call it by its given name, or to respond with apprehension of pain when humans “lift up the whip”. An old greyhound much experienced in the hunt has learned where best to position itself as a member of a team of dogs, and horses, and humans cooperating to catch a hare.
Hume is rather myopically concerned with portraying such learning as endowing animals with new specific tendencies to “infer that the same events will always follow from the same causes”. His discussion is thus bogged down in confused Gothic conceptions of causality, necessity, universality, and inference. Hume is extraordinarily sensitive to certain problems posed by those traditional conceptions. Indeed his Enquiry is largely devoted to articulating and resolving those problems, though he only gets so far in that ambition.
In the present exercise, I attempt to extract and extend the most useful insights captured by Hume’s reflections on animal rationality by replacing his troubled talk in terms of causal inference with less burdened talk of behavioral dispositions, attitudes, and expectations. The shift in idiom may be traced through Ryle’s Concept of Mind. In my view, this transposition of Hume’s insight offers a richer conception of animal rationality while avoiding some of the philosophical problems associated with Hume’s treatment. For the purpose of the present exercise, I’ll largely ignore the letter of Hume’s account. I plan to consider his particular claims, along with problems embedded in his account, in a subsequent exercise.
In keeping with Hume’s account in section IX of the Enquiry Concerning Human Understanding, we may characterize human and at least many nonhuman animals along the following lines:
(1) The behavioral dispositions of intelligent perceptive agents tend to be informed by rational expectations acquired on the basis of experience.
(2) An animal who has acquired such expectations occasionally adopts, or undergoes occurrences of, what we might call attitudes of expectation, analogous to “propositional attitudes” (like hope, fear, wonder, doubt, belief) in human beings.
(3) Such attitudes of expectation tend to cooperate with other attitudes in the same animal we might characterize as attitudes of intention, impulse, or purposive action.
(4) Some of the dispositions acquired and exercised by these animals involve recognition and anticipation by one animal of relevant attitudes and behaviors of other animals.
(5) The exercise of capacities in one animal thus to recognize and anticipate the attitudes and behaviors of other animals is often a path or condition of coordination of attitudes and behaviors among individuals and throughout whole communities.
(6) The tendency of these animals to have their dispositions modified in this way seems to depend on their being constitutionally disposed:
(a) to distinguish various individual things in their environment on the basis of perception;
(b) to recognize a thing thus distinguished as the same thing encountered from various points of view and on various occasions; and
(c) to respond to various similar things thus distinguished as things of the same sort.
(7) At least some of these animals seem constitutionally disposed to develop particular social relationships with some individual animals thus distinguished.
It comes naturally to us to learn from experience along such lines. It’s not something we choose to do. It’s not something we could choose to stop doing. As Hume points out, our tendency to learn by trial and error is something like a “species of instinct” in us. It belongs to the natural spontaneity of rational animals.
The expectations acquired and refined in us on the basis of experience are “rational” insofar as they tend to track historical correlations in relevant objective states of affairs. The behaviors and intentions acquired and refined in coordination with these expectations are “rational” insofar as they tend to be informed by those expectations and to result in outcomes aligned with the motives and ends of the animal’s purposive behavior — for instance, catching hares and avoiding whips.
Such considerations inform a rich conception of what it means to speak of animal rationality, without getting bogged down in philosophical problems clustered around talk of sentience, and without getting involved in the “rope and pulley” problems that concern empirical scientists and artificial intelligence engineers.
I presume it’s not controversial nowadays to claim that the preceding characterization of rational animals applies to dolphins, dogs, and chimpanzees, as well as to humans. I take it for granted this is an empirical hypothesis which in our times is confirmed every day. I’m not concerned here to sort out the whole animal kingdom, nor to address the distinction between the “intelligence” of animals and the “intelligence” of plants, slime molds, or computer programs.
For all I know, most or all animals satisfy most or all of the criteria I’ve listed. The differences may be primarily matters of degree of complexity, variability, and scope.
[This essay is the second part of a series. The series begins here.]
WORKS CITED:
DAVID HUME, An Enquiry Concerning Human Understanding
GILBERT RYLE, The Concept of Mind
The Cabbage Farm 