19. Panpsychism

Panpsychism is a very ancient theory. Traces of it can be found in the earliest Greek philosophers (who are often described as "hylozoists", that is, as holding that all things are animate). Aristotle reports of Thales (De anima 411a7; cp. Plato, Laws 899b) that he taught "Everything is full of gods". This may, Aristotle suggests, be a way of saying that "soul is mingled with everything in the whole universe", including what we usually regard as inanimate matter. This is the doctrine of panpsychism.

Among the Presocratic philosophers down to Democritus, panpsychism has a materialist or at least semi-materialist character in so far as the psyche, or mind, was regarded as a very special kind of matter. This attitude changes with the moral or ethical theory of the soul developed by Democritus, by Socrates, and by Plato. Yet even Plato ( Timaeus 30 b/c) calls the universe "a living body endowed with a soul".

Panpsychism, like pantheism, is widespread among Renaissance thinkers (for example Telesius, Campanella, Bruno). It is fully developed in Spinoza's treatment of the mind-body relationship, his doctrine of psycho-physical parallelism: ". . . all things are animate in various degrees". (Ethics II, XIII, Scholium.) According to Spinoza, matter and soul are the outside and inside aspects, or attributes, of one and the same thing in itself (or things in themselves); that is to say, of "Nature, which is the same as God".

A very similar yet atomist version of the theory is the monadology of Leibniz. The world consists of monads (= points), of unextended intensities. Being unextended, these intensities are souls. They are, as in Spinoza, animate in various degrees. The main difference from Spinoza's theory is this: while in Spinoza, the thing in itself is the (inscrutable) Nature, or God, of which body and soul are outside and inside aspects, Leibniz teaches that his monads — which are the things in themselves — are souls, or spirits, and that extended bodies (which are spatial integrals over the monads) are their outside appearances. Leibniz is therefore a metaphysical spiritualist: bodies are accumulations of spirits, seen from the outside.

Kant teaches, by contrast, that the things in themselves are unknowable. Yet there is a strong suggestion that, as moral characters, we ourselves are things in themselves; though there is also a suggestion that the other things in themselves (those which are not human) are not of a mental or spiritual character: Kant is not a panpsychist.

Schopenhauer takes up Kant's suggestion that as moral characters — as moral wills — we are things in themselves; and he generalizes this: the thing in itself (Spinoza's God) is will, and will manifests itself in all things. Will is the essence, the thing in itself, the reality of everything, and will is what, from outside, for the observer, appears as body or matter. One can say that Schopenhauer is a Kantian who has turned panpsychist. In order to carry out this idea, Schopenhauer emphasizes the unconscious: although his will is mental, or psychical, it is largely unconscious — completely so in inanimate matter, but largely so even in animals and in man. Schopenhauer is thus a spiritualist; but his spirit is mainly unconscious will, drive, and appetite, rather than conscious reason. This theory¹² 3exerted a great influence upon German, English and American panpsychists who, partly under the influence of Schopenhauer, interpret the chemical affinities, the binding forces of atoms, and other physical forces such as gravity, as the outward manifestations of the drive-like or will-like properties of the things in themselves which, seen from outside, appear to us as matter.

This may serve as a sketch of the idea of panpsychism.¹³ (An excellent historical and critical introduction by Paul Edwards may be found in his [1967 (a)].) Panpsychism has many varieties, and it offers what appears to its defenders a comfortable solution of the problem of the emergence of mind in the universe: mind was always there, as the inside aspect of matter. This seems to be the reason why panpsychism is accepted by several well-known contemporary biologists, such as C. H. Waddington [1961] in England or Bernhard Rensch [1968], [1971] in Germany.

It is obvious that panpsychism is, from a metaphysical (or ontological) point of view, nearer to spiritualism than to materialism. However, many panpsychists, from Spinoza and Leibniz to Waddington, Theodor Ziehen, and Rensch, accept what I called in section 16 above the physicalist principle of the closedness of the physical world. They believe,¹⁴ like Spinoza and Leibniz, that psychological or mental processes and physical or material processes run parallel, without interacting; that mental (World 2) processes can act only upon other mental processes, and that physical (World 1) processes can act only upon other physical processes, so that World 1 is closed, self-contained.

I will here present three arguments against panpsychism.

(1) My first criticism of panpsychism is that the assumption that there must be a pre-psychical precursor of psychical processes is either trivial and completely verbal, or grossly misleading. That there is something in evolutionary history which preceded, in some sense, the mental processes, is trivial as well as vague. But to insist that this something must be mind-like and that it can be attributed even to atoms is a misleading way of arguing. For we know that crystals and other solids have the property of solidity without solidity (or pre-solidity) being present in the liquid before crystallization (though the presence of a crystal or some other solid in the liquid may help in the crystallization process).

Thus we know of processes in nature which are "emergent" in the sense that they lead, not gradually but by something like a leap, to a property which was not there before. Although the mind of a baby develops, gradually, from a pre-mental state to full consciousness of self, we do not need to postulate that the food which the baby eats (and which in the end may form its brain) has qualities which can be, with informative success, described as pre-mental or as in any way even distantly similar to mind. Thus the pan-element in panpsychism seems to me gratuitous and also fantastic. (But I should not say that the idea shows much imagination.)

(2) Panpsychism accepts, of course, that what we usually call inanimate or inorganic matter has a very much poorer mental life than any organism. Thus to the great step from non-living to living matter will correspond a great step from pre-psychical processes to psychical processes. It is therefore not very clear how much panpsychism gains for a better understanding of the evolution of the mind by assuming pre-psychical states or processes: even on the panpsychistic account, something totally new enters the world with life, and with heredity, if only in several steps. Yet the main motive of post-Darwinian panpsychism was to avoid the need to admit the emergence of something totally novel.

To say this is not, of course, to deny the fact that there exist not only unconscious mental states, but also many different degrees of consciousness. There can be little doubt that dreaming is conscious, but on a low level of consciousness: it is a far cry from a dream to a critical evaluation and revision of a difficult argument. Similarly, a newborn child has clearly a low level of consciousness. It probably takes years, and the acquiring of language, and perhaps even of critical thinking, before the full consciousness of self is achieved.

(3) Although there exists, no doubt, something that may be described as unconscious memory — that is, memory of which we are not aware — there cannot be, I propose, consciousness or awareness without memory.

This may be explained with the help of a thought experiment.

It is well known that, as a consequence of an injury, or an electric shock, or a drug, a person may lose consciousness (and that a period of time prior to the event may be extinguished from his memory).

Now let us assume that, by taking a drug, or by some other treatment, we may extinguish the recording of memory for several minutes or seconds.

Let us further assume that we are treated in this way repeatedly — say, after every p seconds — every time extinguishing our memory for a short blacked out span of q seconds. (We assume p > q.)

(a) We see at once that if the periods p were made equal to the extinguished periods q, no recorded memory would be left of the whole period of the experiment.

(b) Since the periods p are somewhat longer than the periods q, there will be a sequence of recordings left, each of the length p—q.

(c) Now assume (b); and further, that p—q becomes very short. I suggest that in this case we should lose consciousness for the whole period of the experiment. For after every memory loss (even upon waking up from deep sleep) it takes some little time before we can, as it were, re-assemble ourselves and become fully conscious. If this time needed to become fully conscious (say 0.5 seconds) exceeds p—q then, I suggest, there will not be any brief moments of consciousness or awareness whose memory is extinguished; rather, there will be no moment of consciousness or awareness at all.

To put my thesis in a different way, a certain minimum span of continuity of memory is needed for consciousness or awareness to arise. Thus the atomization of memory must destroy conscious experience and, indeed, any form of conscious awareness.

Consciousness, and every kind of awareness, relates certain of its constituents to earlier constituents. Thus it cannot be conceived of as consisting of arbitrarily short events. There is no consciousness without a memory that links its constituting "acts of awareness"; and these, in their turn, cannot exist unless they are linked to many other such acts.

These results of a purely speculative thought experiment are corroborated, as far as this is possible, by some of the results of brain physiology. I am told that some drugs used as total anaesthetics — that is, for producing unconsciousness — act in the manner described, that is, as more or less radical atomizers of memory connections, and thus of awareness. Some forms of epilepsy also seem to work in a similar way. In all these cases parts of the long term memory are kept intact, in the sense that upon the recovery of consciousness the patient can remember events of his earlier life or events up to losing consciousness; and it is this past memory (or so it seems) which makes it possible for the patient to preserve his self-identity.¹⁵

Now this thought experiment speaks strongly against the theory of panpsychism according to which atoms, or elementary particles, have something like an inside view; an inside view that constitutes the unit, as it were, out of which the consciousness of animals and men is formed. For according to modern physics, atoms or elementary particles have emphatically no memory: two atoms of the same isotope are physically completely identical, whatever their past history. For example, if they are radioactive, their probability or propensity to decay is exactly the same, whatever the difference in their past radioactive history may be. But this means that they have, physically, no memory. If psycho-physical parallelism is assumed, their "inside state" must also be one without memory. But then, this cannot be anything like an inside state: it cannot be a state of consciousness, or even of consciousnesslike pre-consciousness.

Memory-like states do occur in inanimate matter; for example, in crystals. Steel "remembers" that it has been magnetized. A growing crystal "remembers" a fault in its structure. But this is something new, something emergent: atoms and elementary particles do not "remember", if present physical theory is correct.

Thus we should not assign inside states, or mental states, or conscious states to atoms: the emergence of consciousness is a problem that cannot be avoided, or mitigated, by a panpsychist theory. Panpsychism is baseless, and Leibniz's monadology must be rejected.

It might be added that it now looks as if the beginning of human or animal memory is to be found in the genetic mechanism; that memory in the conscious sense is a late product of genetic memory. The physical basis of genetic memory seems to be within the reach of science, and the explanations that we have of it seem to make it totally unrelated to any panpsychistic effect. That is to say, instead of a straight progression from memory-lacking atoms to the memory of magnetized iron, and further to the memory of plants and to conscious memory, there appears to be a huge detour via genetic memory. Thus the results of modern genetics speak strongly against the view that there is any value in panpsychism — that is, against its explanatory power or explanatory prospects, although panpsychism as such is so metaphysical (in a bad sense) and has so little content that we can hardly talk about its explanatory value.

^{12 1} It seems to have been influenced by Goethe's novel Elective Affinities (Die Wahlverwandtschaften = The Chosen [Chemical] Affinities) in which sympathy, and attraction, were interpreted as akin to chemical affinity. Schopenhauer, who knew Goethe personally, was greatly influenced by him.

¹³  For a fuller discussion of the history of the mind-body problem, see chapter P5, below.

¹⁴  (Added in proofs.) Professor Rensch has kindly informed me that he disagrees with the view stated in the first part of this sentence, since he is not a parallelist, but an identity theorist. (But in my view the identity theory is a special case - a degenerate case - of parallelism; see also sections 22 to 24, below.)

¹⁵  See especially the remarks on the patient H. M. in Brenda Milner [1966].