(Based on book of that name by Gerald M. Edelman and Giulio Tononi, Basic Books, 2000, 274 pp.)

According to the book, consciousness is a process, not a thing. [A verb, not a noun.] It does not depend on a single brain area or on a static cluster of neurons. There is no “cartesian theatre” (Dennet) or “pearl of the mind” or a “Homunculus” who “reads” the sensual input and issues orders to the muscles. Yet Dennet (“Consciousness Explained”) finally postulates a central von Neuman linear computer (like a central processing unit, CPU) where everything comes together (and the “demons” compete for access), and Edelman and Tononi postulate a “dynamic core” in the functional cluster of neurons in the thalamocortical brain region. “These concepts seem to approximate a Cartesian theatre. I suppose that the difference is that it is dynamic, ever changing, not static and permanent.

The mind-body problem has been dealt with by many philosophers. Our subjective experience and the objective outside world seem to be realities apart. Descartes spoke about “Res Cogitans” and “Res Extensa”, connected only by the intervention of God, who synchronizes them. This was strict dualism. He also assumed that animals have no minds or souls; they are automatons. Some think that the “hard problem” of how the mind links to the brain will never be solved, even when we know all the details of the operation of brain systems. Schopenhauer called the “hard problem” “the World Knot”, a term adopted by Edelman and Tononi.

Subjective experience (phenomenology) and the external world are perceived very differently. We can examine brain neurological events and interactions from the outside, as is done in science, and we have progressed considerably along this road, as this book shows. In this way, an observer (scientist) examines someone else’s brain. But we also have an everyday experience of our own mental events, which in a fundamental way are the only things we really know directly, without any intervening processes of induction or deduction, even at a primitive level. The recognition of objects as permanently continuing structures comes early in an infant’s life, but it is only an inductive conclusion that can never be proved. Before this happens, the infant perceives the continued existenc~ of its own mind, the sense of “I”.

We could explain the difference as an internal and an external view of the same reality, the two sides of a coin. Perhaps, if we could be a “mind” inside an electron or a photon, we would see the quantum world very differently, maybe without paradoxes. Two sides of a strip of paper (not a coin) can be melded into one by constructing a Moebius strip. I don’t know what would be the equivalent for combining the interior and exterior view of the mind/brain system. Alexander the Great is reported to have “solved” the Gordian knot problem by cutting it apart with his sword – a solution inadmissible in topology (but typical of a warrior, to whom the sword is the solution of last resort). The World Knot cannot be legitimately cut apart. The sword, we are reminded, is not the solution to any problem.

According to Edelman and Tononi, the fundamental properties of consciousness are integration (non-decomposability) and differentiation (informativeness or complexity).

(Integration and differentation are my own favorite metaphors for world structures of subsidiarity from local to global levels of government, derived~~~he embryonic development of organisms. The terms also invoke a resonance from the two mathematical operations so named.) The same two fundamental properties, the authors argue, must be (and are) reflected in the neural substrate that supports consciousness.

The brain is a product of evolution in more than one sense. The organ as a whole developed by natural selection because it gave the individual better survivability and opportunity to reproduce. But secondly, during fetal development, neurons compete with each other to make the right connections to other neurons; those that get lost in the maze are programmed to die. (This is similar to what happens to immune cells as they get “educated” in the thymus or the bone marrow not to “recognize” (i.e. react to) self, the cells of its own body.) And thirdly, by means of experience, mainly after birth, in childhood and adulthood, there is a selection with regard to the strengthening of some synapses and not others (Hebbian junctions: “cells that fire together wire together” is the rule). This is called memory and is involved in learning.

This is why even identical twins are not identical. Genetically they are, but they experience different events in fetal and subsequent development. Thus the phases are: evolution (of whole species), development (of the fetus), and experience (lifetime learning). It is evident, therefore, that the genes do not entirely determine who we are and how we feel, think and behave.

So memory is non-representational; it is recorded as patterns of variously strengthened synaptic connections, not as repeated sensory experiences, pictures, music, or fragrances. (Though the latter can often trigger memories.) Human memories can be false, elaborated by the left-brain integrator which tries to fill in missing information. The higher the organism, the better it is able to tell lies. That is the faculty of imagination, the basis of creativity — like creative book-keeping.

Primary consciousness is non-verbal and non-symbolic. All (at least higher) animals have it. It is pure phenomenology. Yet animals can make choices (primary will), have feelings, and can reason at a simple level. A critical new brain connectivity appeared in the transition from reptiles to birds and mammals. (This is why we talk roughly about the reptilian and the mammalian brain, layers still existing in the human brain below the cortex.) The innovation concerned re-entrant connections between the thalamus and the cortex. In higher animals, we talk about the thalamo-cortical reen- I trant system, which is the basis of primary consciousness. Also, specific thalamic nuclei were formed. The sense of self and signals from the world merge by the reentry mechanism to create primary consciousness. Then in humans the speech and symbolizing centres were added to the loop to create higher-order consciousness.

The dynamic core is a group of neurons within the thalamo-cortical system, constantly being added to and subtracted from at the edges. I have recently seen a lighted advertising display which reminded me of this: the central lights were always on, but the lights at the margins flickered on and off, seemingly chaotically. The dynamic core is the seat of the integration of all the inputs, yet maintaining their complex differentiation.

Consciousness is not the property of some neurons (as Penrose tried to imagine), nor of specific brain areas, nor even of neuron clusters scattered through different brain areas, but arises from the semi-organized semi-chaotic dynamic core. Typically, at the edge of chaos.

Edelman and Tononi then proceed to define “integration” and “differentiation” quantitatively by using concepts from information theory, such as entropy. units (such as neurons) that are interconnected achieve lesser entropy than independent units would. Thus the measure of integration is the entropy decrease due to connectivity. (Incidentally, there is an optimum level of connectivity, between none or very low (like “white music”) and complete or almost complete (like “brown music”) to a “Goldilocks” optimum (like “l/f music”). Cf. essay on “Very Large Graphs” in this volume.)

The concept of “functional clustering” is then introduced. This shows that functional integration can be brought about in less than a second. This provides a solution to the so-called “binding problem”. To quote from the book: “The activity of a group of neurons can contribute directly to conscious experience if it is part of a functional cluster, characterized by strong mutual interactions among a set of neuronal groups over a period of hundreds of milliseconds.”

There are three definitions of “information”, increasingly restricted:

(a) the measure of order in a system far from equilibrium;
(b) arising only with the origin of life;
(c) arising only in conscious observers.

These are, respectively, physical, biological, and psychological definitions. I would call them (a) “mere” information; (b) inherent meaning; and (c) explicit meaning. The latter two cover the whole spectrum of “minds” according to Jantsch: metabolic (enzyme), genetic (gene), epigenetic (homeobox), hormonal (messenger signalling), immune (“recognition”), neural (neuron), and human-symbolic (meme, phoneme).

Some attention is then given to colour discrimination, central in a discussion of qualia in general. Along 3 axes (red-green, blue-yellow, and light-dark), 3 neuronal systems can fire at rates between o (completely inhibited) and 100 Hertz (maximum, tending toward epilepsy or deep sleep spikes, both extinguishing consciousness), preferably at about 10 Hertz (another instance of the Goldilocks effect). The mix produces a particular colour perception.

These colour perceptions are joined with perceptions of shape and motion to form overall visual experience. There are similar clusters for sound, touch, proprioception, thought, etc., the total converging in primary consciousness. But some bodily properties, such as blood pressure and heart rate, are not normally perceived, although biofeedback methods can make them accessible, and even modifiable at will.

The dynamics of the core and related neural systems can be represented metaphorically as a tangle of connected springs under tension. (A diagram is provided.) Learned unconscious routines (such as riding a bicycle) are represented by other springs not in the central cluster, but connected to it at a few points. If unconscious routines, memories, intentions, and expectations join together to form an alternative functional centre (a splinter core or a new nucleus of a self), we may have an explanation of the multiple personality disorder. Obsessions and compulsions can be fixed, rigid unconscious routines that can only with difficulty be consciously resisted, as if certain ports in and out of consciousness were pathologically permanently open.

Finally, language and the higher consciousness liberate the imagination and open thought to the vast domains of metaphoric and symbolic existence.

The authors conclude that metaphysics (what really is) and epistemology (what we can know) should be grounded in biology, specifically in neuroscience. But I wonder if we are not locked into a circular argument, since what we know of neuroscience comes from our, possibly non-representational, perceptions. We study the brain as an “object out there”, and who are we as a subject or observer?