Rodney Cotterill’s No Ghost in the Machine (London, 1989) is a scientist’s attempt to prove that there is no such thing as a soul. I often wonder exactly what people who write this way think they are arguing about. These Enlightenment arguments resemble like a photographic negative the claims of believers as to the existence of supernatural domains like Heaven and Hell. First, an “immaterial,” “supernatural,” “transcendental” form of being (or Being) is described; then it is said (not) to “exist.” But arguments about existence all wind up referring to something like the material conditions by which we recognize existence in our own world. Thus the soul is said to be immaterial, but to inhabit the body and leave it at death, which is to say, to behave just like a physical body (e.g., the breath-psyche). Heaven is said to be supernatural, yet we conceive it as a place where the blessed reside.

GA avoids this kind of language by explaining it and offering an alternative to it. The model for all transcendental entities is the linguistic sign, which does not “exist” in the spatio-temporal world. Words are not mortal, but they do not “inhabit” their speakers and need no Elysian fields on which to disport themselves. Nor is this linguistic reference merely a more sophisticated version of the Enlightenment denial of the spiritual. Language is not merely a cognitive operation of our brains; it is a source of sacralization that arises in order to defer mimetic violence. To represent by means of a sign is to cut off from worldly action. The supernatural quality of what is designated by the sign does not arise from the formal reality of the sign’s existence in a different world from that of things; it is rather this formal difference that arises from the the human community’s “absolute” need of putting the desired object beyond the reach of its potentially contentious members.

Thus the first word is the name-of-God that both designates the object of desire and at the same time does not signify merely this mortal object but a sacred Being that will subsist after the object has been destroyed (in the simplest scenario, divided up and eaten by the group). The animal that the word designated is no more, but the word and its peace-bringing power remains, and God, or the sacred, is what it means. The word’s “immortality” contrasts with the worldly mortality we depend on to support our appetites.


The major bugbear of Cotterill’s book is “dualism,” understood as the Cartesian doctrine that mind is separate from body–in particular, the brain. But this way of stating the problem merely perpetuates misunderstanding. Mind as “the ghost in the machine” has no possible material referent, yet is derided for lacking one. The image conjured up in the reader’s mind is of some kind of preternatural ectoplasm, that is, once more, a “supernatural” substance that is not of this world yet manifests itself in it. It is easy enough to dismiss such concepts as obscurantist holdovers from primitive times.

But if we think of dualism as the doctrine that signs are not things (not even “one thing substituted for another”), then these objections vanish. A word is not a thing, but neither is it some kind of ghostly emanation. It is an atemporal form, not some kind of immortal substance. Nor does it “exist” in my brain. This is a point that deserves some clarification.

Recent research into the workings of the brain, some of which has been remarked on in these Chronicles, has gone far toward understanding how language functions in the individual language-user. But that the word “tree” in my brain is instantiated by a set or “node” of neurons connected on the one hand to mental images of trees and on the other to the sounds [t], [r]… and the letters T, R… does not mean that the word is a “thing” in my brain. The very nature of the circuitry makes the word more a “disposition” than a thing; but the crucial point is that a sign cannot be understood as existing à la Herder in a single mind. As soon as its existence is thus defined, it becomes no more (pace Chomsky) than a Skinnerian association. What makes language more than a reflex is that it exists as a mode of communication among separate individuals. On this point, I am happy to cite a prominent theoretician of mimesis and language origin–one who, sad to say, seems unaware of GA:

The integration of morphophonological addresses [i.e., sound patterns] into a larger descriptive [i.e., lexical] system is an inherently social activity, and one is tempted to predict that this process could not be confined within the isolated brain; that is, one should not expect to find the ‘language acquisition device’ that Chomsky predicted entirely inside the individual brain. Rather, the emergence of language depends on a community of brains in interaction.
(Merlin Donald, “Preconditions for the Evolution of Protolanguages,” in The Descent of Mind, Michael     Corballis and Stephen Lea, eds., Oxford UP, 1999), p. 150)

To the extent that the word “tree” “exists,” it does so not in my brain but within the collectivity of English speakers. Any private associations I may develop with the word can become part of the word’s meaning only if they are communicated to and repeated by others. There is no simple limit to what a word communicates; even the most rigorously defined meanings are “fuzzy.” That is the result of the way signs exist: changes of meaning in one mind do not instantly propagate themselves to other minds. For this reason, Locke warned of the danger of assuming that others use words with the same meaning as ourselves. But the point of language is not to arrive at absolute transparency but to defer violence; the originary sign designates not a rigorous concept but the sacred source of this deferral.


Materialists are likely to tax this conception with mysticism. There is, they claim, no such thing as “the word ‘tree,’” only individual sounds and marks–and trees. That a group of people associate a set of these sounds or marks with mental images of trees is no different from Pavlov’s dog associating the bell’s ring with dinner.

What is my answer to these objections? That ontologies–monist, dualist, or what have you–are not determined objectively on the ground; they reflect whatever distinctions the ontologist considers essential to his model of the world. Just as human beings are “just” collections of quarks, so words are “just” sounds. But the materialist dismissal of dualism jettisons, along with the ghostly soul, the specificity of human language and culture. This is less easy to do when one has made the connection between, on the one hand, the soul and the family of ideas associated with the sacred, and, on the other, the phenomena of language and “secular” culture. Just as ideas are now thought to reside in the brain not in specific neurons but in interactions among neurons, so do the signs of language and culture exist not in individual minds but in the interaction among human beings.

Let me clear up a possible objection. The existence of categories as opposed to individual beings did not begin with language, or even with life. What then is the difference between saying the word “tree” is atemporal, since it does not depend for its existence on any given instantiation, and saying the same thing of the category “proton”? In Saussurian terms, “tree” is part of the langueeven if no one speaks or thinks it, but is not the langue itself just a set of nodes in the brain, or in a community of brains? My answer is: yes, but the point is not to show that words are “immortal,” but that they have a different relation to temporality than things, one that is the basis for our concept of immortality. Gods too may be considered mortal in the sense that when their believers are gone, they are no longer “real” gods, but this form of “death” is not directly comparable to the decay of the flesh. An electron can “know” the category “proton,” but not even a chimpanzee can reflect on the non-mortality of the sign, because no sign has ever been necessary to preserve his society from mimetic violence.

It is only after explaining why human communities have a stake in the existence of the signs of language and culture that one can explain why these communities create the idea of the soul, of God, the sacred, mana, karma, and so on. I wonder how long will we have to wait before the dialogue des sourds between science and faith gives way to an anthropological science fully cognizant of the specificity of its human subject-matter.