Perspectives on memory research : essays in honor of Uppsala University's 500th anniversary

mathematical principles (see Runeson, 1977a. for more details). With regard to reputedly more basic variables, the planimeter is not especially efficient; it does not permit the straightforward measurement of, say, lines and angles, and such measurements may prove to be relatively inaccurate. But to be dismayed at the fact that the instrument is unreliahl~ --------with regard to these supposedly basic variables would be missing the point. The polar planimeter exemplifies smart devices in this sense: It is a specialpurpose instrument tailor-made for a particular kind of task and a particular kind of situation. Such being the case, its construction can take full advantage of the special features of the task it is to perform and the circumstances in which it is to operate (cf. Fowler & Turvey, 1978; Gel'fand & Tsetlin, 1962). Apart from bringing into question the hypostatizing of the basic variables of physics, the preceding examples and arguments are of importance in the following final respect. The suggestion is that the evolution of perceptual systems mdy be understood as the compiling of special-purpose, smart devices to directly register higher order variables or complex particulars. This would contrast with the understanding that evolution manufactured a few types of basic components, each computing a basic physical variable or simple particular, and achieved the computation of more complex variables through programs defined over the basic components. We might go further, and further than we have room for here, to delimit and evaluate the conceptual bequest of the preceding five centuries as it bears on the nature of perception. But what has been deliberated to this point must suffice; and in the contrast drawn in the last few paragraphs we have caught a 'We believe it prudent to avoid where possible the concrete-abstract distinction, where concrete refers to particular things and abstract to nonparticular things. In theview that we wish to promote, animals always perceive particulars. Some of these particulars may, as a matter of convenience, be deserving of the label complex in comparison with other particulars, but they are, nevertheless, particulars and thus ought not to be conceived of as reducible in the way that abstract amorphous entities might arguably be reduced to concrete particular entities. We thank Ed Reed for introducing us to this conception ofparticular, although our usage is somewhat different from his. glimpse of a different conceptual basis and one that we now try to develop. As postulated at the outset, the stock concepts of traditional and conventional perceptual theory follow from the assumed dualism of animal and environment. The promissory note to be cashed in is that an assumed synergy of animal and environment yields an orthogonal collection of concepts and in consequence a radically different understanding of perception. AN ECOLOGICAL REFORMULATION The Principle of Mutual Compatibility There is a contemporary understanding with which animal-environment dualism and the doctrine of intractable nonspecificity seem to be graphically at odds. It is that the universe is finely balanced with reference to life (Trimble, 1977; Wheeler, 1974), Characterizable in terms of a relatively small number of properties, it is becoming increasingly evident that the universe's hospitality to life depends on these properties havingjust the particular values they have, and no others. A fairly minor change in either a fine-grain property or a coarse-grain property of the universe, and life would be impossible. Thus, at the fine grain, a lowering of the electromagnetic force that regulates the structure and interactions of atoms and molecules would release electrons from their bondage to atoms, thereby making chemical reactions impossible; at the coarse grain, a rise in the rate of expansion of the universe would rule out the forming of galaxies and induce a uniform spread of matter (see Trimble, 1977). We recognize, therefore, the fitness of the universe for life. But more than this we will recognize that the universe and life are coimplicative: They have coevolved and they are codesigned. For the 17thcentury philosopher, Leibniz, it could not have been otherwise. Leibniz believed that substances could not interact. That which passes for causal interaction among substances was interpreted by Leibniz as a parallel coordination of state changes; the coordination arising from the inexorable unfolding of natural laws according to symmetry among substances. In his "principle of sufficient reason," Leibniz advocated that only those things may coexist-only those things are possible-that satisfy certain fundamental compatibility relations. By this principle-which we will refer to as the 'principle of mutual compatibilityw-compatibility relations among logically possible structures is the defining criterion of existence. We see, in short, that for Leibniz there must be a good or a sufficient reason for anything to exist. Darwin's theory of evolution and other evolutionary theories (e.g., cosmological theories) all begin by assuming the existence of a collection of things that evolve, but they do not address the issue of why that collection-or any other, for that matter-should exist at all. Leibniz sought a deeper understanding: Whatever exists does so because it is mutually