Herbert Feigl, The "Mental" and the "Physical": The Essay and a Postscript (1967).
D. "Purposive" versus "Mechanical.
Along with'direct experience, it is perhaps intelligence which makes up the most important characteristic of the commonsense concept of mentality. And intelligence is usually and most basically characterized as the capacity of utilizing means toward the attainment of ends. One trouble with this characteristic is that common language is apt to describe as "intelligent" even the instinctive behavior of many animals. In the case of, e.g., social insects (termites, ants, bees, etc) the behavior is stunningly purposive, highly organized, and intricate; arid yet we hesitate to ascribe sentience or subjective experience (raw feels) even only remotely resembling our own to these entirely different organisms. Moreover, the current scientific use of the word "intelligence" tends to be restricted to those evolutionary levels and species in which learning combined with ingenious (inventive) and symbolic behavior plays a dominant role. Pigeons, rats, cats, dogs -- those favorite laboratory animals of the behavioristic psychologists -- show (in each species) marked individual differences in the speed and the scope of their learning. Anthropoid apes, like the chimpanzees, are famous (ever since W. Köhler's original experiments) for their inventiveness -- in addition to their commonly known capacities for imitation. Genuinely linguistic behavior, involving syntactical, semantical, and pragmatic features, seems to be restricted to homo sapiens; the so-called language of the bees (which is apparently instinctive and lacking in syntactical and semantical flexibility) does not seem to be an exception.
If intelligence or just purposiveness were chosen as the sole criterion of mentality, then it would be hard to draw a sharp line anywhere within the realm of organic life. Even in the kingdom of plants we find processes whose teleological characteristics are not fundamentally different from the features of purposive behavior in the lower animals. Of course, if one deliberately makes the (often suggested and no doubt helpful) distinction between two types of teleology, one of them involving conscious aims, and the other excluding them, and designates only the former as "purposive," then the empirical evidence suggests (but does not force upon us) the decision to call "intelligent" only the behavior of the higher animals, or perhaps to restrict the label "intelligence" to human beings (i.e., if and when they behave in a genuinely sapient manner).
It becomes clear then that the scope of the two criteria (sentient and sapient) is not necessarily the same. The two concepts are not coextensive. The situation has been further complicated in our age by the construction of "intelligent" machines. Logical reasoning, mathematical proofs and computations, forecasting, game playing, etc. are all being performed by various and usually highly complex electronic devices. Here the temptation to ascribe "raw feels" becomes even weaker than in the case of the lower animals.1 Inductively it is plausible that sentience requires complex organic processes.
Descartes was perhaps not completely wrong in restricting mentality to human beings. If "mind" is understood as the capacity for reflective thought, then indeed we may have reason to deny minds (in this sense!) to animals (and perhaps even to electronic computers!). The issue is difficult to decide, because the connotations of "reflective thought" are numerous and indefinite. But if it connotes a conjunction of sentience, learning capacity, spontaneity (free choice), purposiveness (in the sense of goal directedness), original inventiveness, intentionality (in the sense of symbolic reference), and the ability to formulate rules of behavior (practical, moral, linguistic, etc.), then mind (in this sense) is clearly the prerogative of man.
All the foregoing considerations need not disturb us. They merely lead to the scarcely surprising conclusion that the term "mental" in ordinary and even scientific usage represents a whole family of concepts; and that special distinctions like "mental1", "mental2", "mental3", etc. are needed in order to prevent confusions. (We shall return to a brief discussion of "intentionality" in subsection F.)
As far as the original distinction of purposive versus mechanical is concerned, it scarcely helps in the definition of the mental versus physical distinction. If "purposive", despite our warnings, is taken as synonymous with "teleological", then we have a distinction, which, though it becomes rather irrelevant to the mental-physical issue, is not useless in the natural sciences and in technology. But then it can no longer be considered as either sharply exclusive, or as particularly enlightening. The flow of a river toward the sea is a mechanical and non-teleological phenomenon, but the functioning of servomechanisms is mechanical as well as teleological, and the functioning of the heart is teleological and presumably "mechanical" in the same (wider) sense in which complex servomechanisms operating by negative feedback are regulative physical devices. In short, the phrase "teleological mechanisms" in our age of cybernetics is no longer a contradiction in terms.
Notes 1. Cf. however, the remarkable and stimulating discussion of the robot problem by Scriven (304). We shall return to this issue in connection with the scrutiny of the analogy argment for "other minds" in section V.