Commentary on Daniel C. Dennett and Marcel Kinsbourne (1992) Time and the observer:  The where and when of consciousness in the brain.  BBS 15:183-247.

Descartes' fundamental mistake:  Introspective singularity

Joseph E. Bogen

Department of Neurological Surgery, University of Southern California, Los Angeles, CA 90033

The thoughtful target article about consciousness by Dennett & Kinsbrourne (D&K) (1992) contains many different contentions.  This commentary is directed toward one of their specific arguments, that there is no midline structure in the encephalon where one could localize consciousness, hence consciousness must be diffusely distributed.

            My use of the term "encephalon" follows that of John Hughlings Jackson (1874) who wrote:

The nervous system, I repeat, is double…I use the word brain to include the cerebral hemisphere and the subjacent motor and sensory tract. [Whereas] I use the word encephalon to include all parts of the nervous system within the skull.

            I am in complete agreement with D&K’s assertion that there is no midline structure whose neural composition, connections, and activation could be responsible for various perceptions, ideas, feelings, and so on becoming momentarily conscious.  But it does not follow (even if it were to be true) that structures with such composition and connections are nonexistent anywhere, hence that the neural correlates of consciousness must be diffusely distributed, as they assert.  Their argument is erroneous because (1) it ignores some well-known facts, and (2) it puts excessive emphasis on introspection.  To show how they have gone wrong, some introductory background is necessary.  As we will see, although they claim to criticize Descartes, they have implicitly joined him in his most far-reaching error.

            One of the most obvious facts about the cerebrum is that, in almost all respects, its parts exist in duplicate.  For every structure within each hemisphere, its near twin exists in the other.  This fundamental fact was well known to Descartes.  Why then did he look for an unpaired midline structure?  And why the pineal?  To the first question he gave a single, psychological answer.  To the second he gave three reasons, anatomico-physiological.

            Descartes looked for a single midline structure because of his introspectively based conviction that consciousness is single.  He wrote, “cogito ergo sum,” both verbs in the singular.  In addition, since almost all parts of the brain are double, “inasmuch as we have only one solitary and simple thought of one single thing during the same moment, it must necessarily be that there is some place where [the two come together]”  (Clarke & O’Malley 1968, p. 471).

            In other words, Descartes gave his psychology greater weight than he gave the anatomical fact of duality.  His decision was complemented by three other observations, all concerning the pineal gland.  First, the pineal gland is (as he said) “in the middle” and “the most internal.”  Second, its destruction (he believed) resulted in death.  And third, it is ideally situated to control (he thought) the flow of ventricular fluids (Clark & O’Malley 1968; McHenry 1969).

            Descartes’ proposal has been discredited for a variety of reasons including his metaphysical dualism (see, for example, Ryle 1949), but his dualism is irrelevant.  A dualist could say that the soul is manifest throughout the encephalon, or one could say, as Fritsch and Hitzig (1870) said of their discovery of the motor cortex, “such facts show that the origin of at least some function of the soul is bound up with circumscribed parts of the brain” (Von Bonin 1960), p. 78).

            We should understand that the metaphysical argument (dualism versus materialism) is orthogonal to the localization argument (special locales versus wide distribution).  Descartes, taking a solidly dualistic position, was a bit less committed to a special locale than is often thought; he wrote, “although the soul is joined to the whole body, there is, however, a certain part in which it exercises its functions more particularly than in all the others” (Clarke & O’Malley 1968,. 471).  Similarly, one could take a solidly materialistic position and believe that all thalamocortical interaction is potentially conscious but that there is a special locale where this potential is particularly realized.

            For those who follow Descartes (and D&K) by insisting on the supremacy of the usual introspection (consciousness is single), the pineal is now thought to be a ludicrous candidate.  Contrary to the information then available to Descartes, pineal destruction need not result in death.  Above all, our current view holds that it is the flow of information between neurons, not ventricular fluid flow that subserves the processes we call, in the aggregate, mentation.

            It is easy for us to ridicule Descartes now, but we should remember Berry Campbell’s criterion:  the greatness of a man’s contribution can be measured by how long he has held up progress.  Half a millennium is surely world class!  It is the subsequent accumulation of empirical facts, not better philosophy, that has rendered the pineal a poor choice for those who are still looking for one or more special locales.

            D&K recognize the nondependence of Cartesian localization on metaphysical dualism. They mainly criticize what they call “Cartesian materialism” for the assumption that there is a place (in their words) “where it all comes together.”  And they insist that the assumption of such a single place continues to pervade, as a “bad habit,” most people’s thinking about consciousness.  Indeed it does!  And it does so for D&K too!  Their error is to suppose a single consciousness, and in that, they follow Descartes.  And for D&K, there is a single place:  “The brain itself is Headquarters” (p. 185).  They state, “As a matter of empirical fact, nothing in the functional neuroanatomy of the brain suggests such a [central] general meeting place” (p. 185).  They should have said “of the encephalon” rather than “brain” of which (per Jackson) there are two.  Their sentence seems relevant only because they refuse to allow Hughlings Jackson, and they do so because they insist that their introspective perception of the oneness of consciousness should have greater weight than not only the fact of anatomical duality but the hemispherectomy and callosotomy evidence for physiologic duality.

            The fundamental mistake of Descartes was to assume singularity, based on his own introspection.  And a fundamental error made by D&K is to follow him in this respect.  Whatever the neural substrate of consciousness, localized or distributed, we must infer that it is double from evidence both anatomic and physiologic.  First, the anatomy exists almost entirely in duplicate.  Second, if either hemisphere is removed, the remaining unpaired hemispheric structures can sustain consciousness.  This is a frequently observed fact, as contrasted with Kinsbourne’s fantasy that “there is no reason in principle why several or many aware neural systems might not result [from repeated subdivisions of the brain]” (Kinsbourne 1988, p. 252).  Any psycho-neuro-philosopher who, in these days, does not wholeheartedly recognize and utilize these two facts (one of them unknown to Descartes) is in the same epistemological ark as theorizers on evolution who do not wholeheartedly accept the multibillion-year age of the Earth.  In each hemisphere there are a number of candidates (structures with widespread connections) for consideration as a “general meeting place” for those who wish to localize.  For those who do not wish to localize, there is no evidence that a piece of a hemisphere can be conscious.

            The behaviors of callosotomized cats, monkeys, and humans are most reasonably interpretable on the view that consciousness can proceed simultaneously and differently in each hemisphere.  As for the intact human, we can reasonably expect that there will be occasions when the anatomic potential for duality becomes a physiologic actuality, for shorter or longer times (Bogen 1990).  To believe otherwise is to make what I call the “Assumption of Perpetually Obligate Synchrony.”  Suppose we agree to this assumption.  Suppose we accept, for purposes of argument (in the nonsplit case), that a combination of callosal and various subcallosal mechanisms keeps in essentially complete synchrony the neural activities of the two hemispheric systems for consciousness.  Doing so does not allow one to ridicule localizationist materialism on the basis that there is no midline meeting place.  Once the duality of consciousness is recognized, a neo-Cartesian view has no need for such a single encephalic meeting place, indeed it explicitly denies it.  That D&K use this argument shows how embedded within their own approach is the idea of what Kinsbourne has acclaimed as “the intact individual’s unified awareness” (Kinsbourne 1988).  This grand delusion (common to us all) permeates their own position no less than that of those they criticize, or of Descartes himself.  Until D&K (and others) give up their belief in conscious singularity they will only be arguing about the structure of their own preconceptions, rather than about the physio-chemical activities subserving consciousness.