Herbert Feigl, The "Mental" and the "Physical": The Essay and a Postscript (1967).
E. "Mnemic", "Holistic", "Emergent" versus "Non-Mnemic", "Atomistic", "Compositional".
This bundle of contrasts has often been associated with the distinction of the mental and the physical. Fortunately, except for one facet of the emergence issue, discussion can be quite brief. The mnemic as a criterion of mind was stressed especially by Bertrand Russell. But long before him, the physiologist Ewald Hering (and his disciple Semon) considered the mnemic as a general property of all organic matter. Even in inorganic matter there are more or less permanent modifications of dispositional properties which cap be effected by various influences. Certain features of elasticity and of magnetic hysteresis are "mnemic" in this sense. And of course the storage of information in present-day computing machines clearly shows that mnemic features, just as the "purposive-intelligent" features, need not coincide with mentality in the sense of sentience or awareness.
The holistic aspects of the phenomenal fields were brought to the fore by the Gestalt psychologists. But almost from the beginning, this school of thought (especially ever since W. Köhler's book on Physical Gestalten, 1920) emphasized the idea of the isomorphism of phenomenal with neurophysiological configurations. Thus again, without the addition of the criterion of immediate experience we do not obtain a distinction between the mental and the physical configurations or "organic wholes" or "dynamic Gestalten."
Inseparably connected with holism and the Gestalt philosophy is the doctrine of emergence. The old slogan "the whole is greater than the sum of its parts" has of course no very clear meaning. Much of its obscurity is due to the lack of a definition of the phrase "the sum of the parts". Recent analyses1 of the still controversial significance of "organic wholeness" and of "emergent novelty" have contributed a great deal to the clarification of the issues. There is no imperative need for us to enter into details here. It will be sufficient for our concerns to realize that in modern natural science no sharp distinction can be made between resultants (as in the composition, i.e. vectorial addition of forces or velocities) and emergents. In the explanation of the properties and the behavior of complexes and wholes we always need laws of composition -- be they as simple as the straightforward arithmetical addition of volumes, masses, electric charges, etc., or slightly more complicated as is vector addition, (or just a trifle more involved as is the relativistic "addition" formula for velocities), or extremely complex as are the so far not fully formulated composition laws which would be required for the prediction of the behavior of organisms on the basis of a complete knowledge of their microstructure and the dynamic laws interrelating their component micro-constituents.
Modern quantum physics, on a very basic level, employs laws which have "organismic" character, as for instance the exclusion principle of W. Pauli2 which holds even for single atoms. It is conceivable that much of what is called "emergent novelty" on the chemical and biological levels of complexity may ultimately be explained in terms of the organismic or holistic features of the laws of atomic and molecular dynamics; and that, given those basic micro-laws, the only composition laws (which scientists often take for granted like "silent partners") are simply the postulates and theorems of geoirietry and kinematics. This is indeed my own, admittedly risky and speculative, guess; that is to say, I believe that once quantum dynamics is able to explain the facts and regularities of organic chemistry (i.e. of non-living, but complex compounds) it will in principle also be capable of explaining the facts and regularities of organic life. But no matter whether these conjectures prove correct or incorrect, emergent novelty from a logical or methodological point of view simply means the impossibility of the derivation of the laws of complexes ("wholes") from the laws that are sufficient to predict and explain the behavior of their constituents in relative isolation. Thus, the laws that are sufficient to account for the motion of free electrons (as in cathode rays, and traversing electric or magnetic fields) are clearly insufficient to account for the behavior of electrons when they are constituents of atoms.
It stands to reason, that in order to "glean" (i.e., to ascertain) the laws of nature, scientists can't afford to stop their investigations on a very low level of complexity. In some cases we are lucky in that from such a very low level of complexity upwards to higher complexities of any degree, no new physical laws (but only geometrical composition laws) are required. This holds, for example, for the law of the lever which remains applicable even for the most complex system of pulleys. It also holds for the law of gravitation and the laws of motion (both in their Newtonian form). The "many bodies problem" is unsolved only in the mathematical sense that no single set of simultaneous equations has as yet been found for the prediction of the motions in complex star systems. But successive approximations can be computed to any desired degree of accuracy. In other eases (as with the behavior of electrons) we could never glean all the relevant laws below a certain level of complexity. And I have admitted (in section II) that it is always logically conceivable that our scientific theories may have to be amended and enriched by the introduction of new basic concepts (variables), and this is of course tantamount to the introduction of new (lawlike) postulates and/or existential hypotheses.
We have seen that the mnemic, teleological, holistic, and emergent features are not adequate as criteria of mentality, because these features characterize even inorganic structures and processes. Emergence as conceived by most dualists, however, refers to the evolutionary novelty and the (physical2) underivability of sentience or raw feels. The whole issue therefore turns again upon the criterion of subjective experience. The issue can be brought out by questions such as the following: Suppose we could predict the detailed chemical structure of an entirely new perfume which will be manufactured in Paris in the year 1995. Suppose, furthermore, that we could equally exactly predict the neurophysiological effects of this perfume on the mucous membranes of a human nose, as well as the resulting cortical processes in the person thus smelling the perfume. Could we then also predict the quality of the experienced fragrance? The usual answer to this question is in the negative, because it is assumed that the fragrance in question will be an "emergent novelty." But behaviorists, and physicalists generally, need not take such a pessimistic view. For given the presuppositions of our questions it should also be possible to predict the answers to questionnaire items like "Is the fragrance more similar to Chanel 5 or to Nuit d'Amour?" That is to say, we should be able to predict the location of the quality in the topological space of odors, provided we have a sufficiency of psychophysiological correlation laws to make this particular case one of interpolation or (limited) extrapolation.
The issue can however be made more poignant if we are concerned with the prediction of qualities within an entirely new modality. In the case of the congenitally blind who by a cataract operation suddenly attain eyesight, the experience of colors and (visual) shapes is a complete novelty. Suppose that all of mankind had been completely blind up to a certain point in history, and then acquired vision. Presupposing physical determinism we should (according to my basic conjecture) in principle be able to predict the relevant neural and behavioral processes, and thus to foretell all the discriminatory and linguistic behavior which depends upon the new cortical processes (which correspond to the emergent, novel qualities of experience). What is it then that we would not or could not know at the time of the original prediction? I think the answer is obvious. We would not and could not know (then) the color experiences by acquaintance; i.e., (1) we would not have them; (2) we could not imagine them; (3) we could not recognize (or label) them as "red", "green", etc., even if by some miracle we suddenly had them, except by completely new stipulations of designation rules.3
I conclude that the central puzzle of the mind-body problem is the logical nature of the correlation laws connecting raw feel qualities with neurophysiological processes. But before we tackle this difficult question, a glance at one more issue is required.
Notes 1. Schlick (299); Nagel (232, 235); Henle (153); Bergmann (28, 34); Hempel and Oppenheim (152); Rescher and Oppenheim (277); Pap (244).
2. Cf. the clarifying discussion by Margenau (208); and the stimulating, but perhaps somewhat speculative, ideas of Kaila (169).
3. Cf. Pap's discussion of absolute emergence (244).