G.Edelmann/G.Tononi
THE UNIVERSE OF CONSCIOUSNESS.
How Matter becomes Imagination
Basic Books 2000
Justus de Swart
The Biologica! Conditions of Consciousness
A Review of Edelman and Tononi's "A Universe of Consciousness"I: Introduction
Although there is little empirical doubt of the cerebral base of consciousness, it still has an unapproachable quality about it.
Gerald Edelman and Giulio Tononi (hereafter ET)' offer a hypothesis that should give us the tool to start disentan-gling the 'world knot', an image Arthur Schopenhauer used to describe the prob-lem of the origin of consciousness.
Their primary focus is not the richness in everyday experience, but the conditions that allow us that experiential richness— a difficult enough task, as most would admit reading the book.
It shows how the cranial barrier can be overcome by new observation techniques, producing a plethora of experimental data in support of the concepts and theories described in the book. This material leads to the formulation of the dynamic-core hypothesis which describes the necessary biological conditions for consciousness.
II: Methodological Quandary
ET have taken the trouble to clear the gap between subjective and objective aspects of consciousness.
First, they make three assumptions which serve as a guide:
the existence of the physical world;
the validity of the evolution theory;
the existence of the private character of qualia.For consciousness this means that it has a material basis, is rooted in our biology and has a strictly private aspect.
There follows another trio of sub-assumptions, underlining the proper evolutionary order of conscious phenomena:(1) being and describing: being has primacy over describing, which is a derivative mental function of being,
(2) doing and understanding: we learn by doing and acting in a world, thus all understanding is a lived understanding (although ET admit that our linguistic faculty is able to invert this order); and
(3) selection and logic: logic is not the hidden web that underlies reality; natural selection has primacy over every human artifact and offers a deeper understanding of humans and their consciousness.A second point addresses the nature of consciousness and how to deal with it.
It is better to look at properties of consciousness that are very persistent and general among humans. ET instance two properties that stand out prominently:
unity of consciousness: consciousness seems to present a coherent picture of the world, and
informativeness: consciousness reduces uncertainty or—better—gives us supposed certainty about what is around us.Take for instance REMsleep behav-iour disorder. These patients are convinced that their dreams are real to the point that they act them out, with undesirable consequences as one may guess. The persistence of these properties reveals something relevant about the problem of con-sciousness.
The study of consciousness must therefore take its clues from these kinds of fundamental properties instead of introducing vague metaphors which could hamper every scientific effort. By focusing on key properties, ET try to avoid the subject-object problem and the likelihood of being mired in the rich-ness of phenomenological world.
III: From Immunology to Neural Darwinism
The new book is, as one might expect, heavily loaded with the legacy of Edelman's evolutionary theory of the mind and brain, a theory which is laid out in his trilogy Neural Darwinism (1987), Topobiology (1988) and The Remembered Present ( 1989).
This obviously influences their presentation and interpretation of new neurological data. In 1972 Edelman together with Rodney Porter received the Nobel Prize for Medicine for their work in immunology. Their findings led to the confirmation of a theory developed by MacFarlane Burnet and Niel Jerne in the 1950s, stating that the immune system contains a full population of lympho-cytes ready to be selected by bacterial and viral intruders. This confirmation was a blow for the then reigning theory of Linus Pauling which, instead of a diversity of lymphocytes, proposed one type of lymphocyte with the ability to pick up infor-mation from an antigen. So the idea of selection prevailed over instruction.
Stretching the idea of selection within initial variation to the way the brain works is called 'somatic selection' or 'neural Darwinism'. In the brain we have neuronal groups instead of lymphocytes and the selection mechanism has been replaced by perception.
The opposition selection-versus-instruction looks like the above mentioned difference between selection and logic. ET acknowledge the role logic plays in higher cognitive functions and planning of behaviour, but would that suggest a toning down of the sharp rejection of formalism in explaining human behaviour in Edelman's Bright Air, Brilliant Fire (1992)?
Nothing whatever seems to point that way. Logic is built on selectionism and as such is an artifact of life. Logic like all human behaviour, is ultimately selected in the course of time within an inherited variation of behaviours.IV: Consciousness in The Individual Brain
According to Edelman's theory of the brain and mind the global anatomy of the brain is determined by natural selection.
This global brain anatomy can be recog-nized through the so-called topological brain structures. These are innate species--typical brain structures which are shared by all humans. The brain structures can be likened to a musical composition. Each human is a kind of conductor perform-ing more or less the same composition. Yet, each conductor makes its own arrangement. The topological brain structures are formed by neuronal groups, which are the basic elements of the brain.Learning then means continually rear-ranging these basic elements within the initial variation of neuronal groups through a process called somatic selection.
The somatic selection process is expressed in the individual thoughts and actions during life. So the conductor's effort does count. He is not a passive bystander, transcribing notes into music, but an active participant reliving notes into music. The somatic selection process con-sists of three processes largely based on the principle of selection within variation of neuronal groups:( 1 ) Developmental selection. This process starts with the actual unfolding of the genetically determined topological structures and the selection of the elemental neuronal groups with their global and fixed functionality. From the embryonic phase onwards person-environmental interactions cause individ-ual arrangements of neuronal groups. Perceptual categorization, i.e. percep-tion, plays an important part in this process.
(2) Experiential selection, which partly overlaps (1), strengthens synaptic con-nections within the brain to form functional neuronal groups. These func-tional neuronal groups are accomplished by cerebral acts, i.e., ordered sequences of brain activities in a domain of perception, movement, or speech that, in time, lead to a particular neural output.
Memory from this point of view is non-representational: all memories are a result of mental acts in the present and active recategorizations. Memories are fluid instead of precise replications of deep-seated things of the (evolutionary) past. Taking develop-mental and experiential selection together, one may say that our life is a per-petual striving between the natural push (our heritage) and somatic pull (our existence).
(3) Reentry. A dynamic process of interchanging brain areas allows the individ-ual nervous system to carve up the amorphous world into objects and events. Reentry seems to offer a solution for the binding problem: how does a mel-ody arise out of a countless clutter of neurons in the head of the conductor?
Somatic selection constitutes the basis for the integration and differentiation of neuronal groups giving rise to conscious experiences. Integration and differentiation are key concepts of the dynamic-core hypothesis.
The hypothesis goes roughly as follows:
(1) A group of neurons makes consciousness possible if it forms part of a distributed functional cluster which by reentrant interaction (reentry) in the thalamo-cortical complex achieves in a few hundred milliseconds a high degree of integration.
(2) The functional clusters must be differentially expressed in a high degree of complexity to sustain consciousness.(1) and (2) taken together form a dynamic core, the biological condition for consciousness. In other words, consciousness requires a certain synchronization of functional groups, which have both a certain unity and complexity and at the same time need to last a certain duration: (1) makes possible the unity of consciousness and (2) the informativeness of consciousness.
Reentry looks like a homunculus. How does it take care of the integration of the neuronal groups in different brain areas? Through functional structures and patterns of certain brain areas, leading to synchronization of brain activities which under the above mentioned conditions lead to consciousness.
There is neither a morphological nor material explanation intended because a process is involved. Neural dynamics, the way patterns of brain activities change in time, and nonlinearity are important factors within this process.Reentry itself does not restrict the selection processes. This is taken care of by another special brain structure, the value systems. Within these systems all kinds of impulses are initiated, from dispositions, like feeding and fight-or-flight behaviour, to intentional behaviours.
Value systems seem to provide a solution for the temporal binding problem: they provide the glue for the many bits of consciousness and cause a feeling of being present in a particular time and place. Let us take a closer look at these systems.V: Value Systems
The brain acquires through value systems a steady state. Selection within variation would go on endlessly without them. Value systems are divided into two components:
(1) a relatively rigid component, the body of a human, and
(2) a flexible component, consisting of two parts.The first is a set of brain structures arising out of the brain stem fanning out to the midbrain and cortex. This part controls the basic human needs.
And the second part comprises the modifiable value systems, which are closely related to the so-called experiential selection process which creates values during life.ET propose that neurotransmitters which play an important role in emotions (serotonin, noradrenaline), memory processes (acetylcholine) and learning (dopamine), are also playing an important role in these value systems.
Value systems are evolutionarily important because they demonstrate the nesting of selection principles. This nesting refers to the notion of contingency on different levels of life systems.
The principle of natural selection cannot control the contingencies of modern life, while the principles of somatic selection can. In this respect the difference between primary and higher-order consciousness must be mentioned.
Primary consciousness can also be found in animals. It plays a significant role in perceptual processes.
Higher-order consciousness is linked to an individual (a self), language skills and conceptual tasks.Planning of behaviour and goal direction require higher-order consciousness.
Value systems play a role during emotions, learning and memory and may influence their conscious concomitants. They express what the authors call the necessary prejudice of the organism, i.e., a necessary behavioural bias towards a world largely unknown.
Organisms, besides taking part in evolution, are also part of the physical world through their bodies. This physical tie to the world may be logically known, but to be able to find one's way one must have a kind of blind certainty, a necessary prejudice towards reality whatever that might be. The value systems seem to provide this certainty, a bridge between the subjective and objective world.
Neural Darwinism leads to the notion of the primacy of values. To strive for something is mostly born out of necessity. The necessity to survive seems to be the highest value and opportunism the best ethic. Modifiable value systems and higher-order consciousness however could lend our behaviour a certain objectivity and reasonable efficiency. So neural Darwinism does not discard the notion that there are true and valuable ideas which exceed the individual's self-interest.
VI: Soaring Meanings
A Universe of Consciousness is a fascinating book telling a gripping story. Some readers may have some trouble following ET in their arguments, the more so as the legacy of Edelman—which could scare some readers away—has been condensed into the book. The authors seem to realize this and interweave their argument with lucid analogies.
The Dynamic Core Hypothesis is aimed at explaining general qualia, like the impression of unity and the informativeness of conscious phenomena.
Properties of consciousness locked into the personal sphere with a singular character are not relevant enough for scientific explanation. The argument behind this view goes as follows. Properties such as informativeness and unity of consciousness are shared in more or less the same way by every one of us, but not everyone has the same experience when hearing an idea of some sort.
The first are relevant enough to be a scientific subject, science wants to know how they emerge or being caused, the latter are too personal or irrelevant to be a scientific subject.
From this perspective the hypothesis is indeed general. Still, choosing this point of view and not allowing any semantic element to be their problem means that their hypothesis lacks predictive value. ET do not see this as a problem because semantics is not a scientific subject. But what then are we to think about a specific problem such as perception blindness, in which a visual image is not recognized, but can be painted after hearing a word referring to the same image?
In the transition from perception to conception, from syntax to semantics, from form to substance, there could lie the real problem of consciousness.
How exactly perceptual categorization works does not become clear, and ET feel their shortcoming. Their premise is grounded on evolution theory and its principles, but robotics could eventually give us more insight, once it is possible to see how an artifact 'similar' to us perceives the world.
The question remains however, how it is possible that a theory based on the principle of selection could accept the validity of a machine not based on the same principle. More seriously, logic would be primary, for how is one to build such a machine in the first place?At the end of the book Willard Quine is cited saying that the thought 'thinking about Vienna' is a process grounded in the material world, but itself is not a material thing. ET agree with this.
Although the book describes how meaning is also defined by our value systems, they seem to soar above the material world in a web of relations, but this should not be confused with a form of dualism.
Yet, meanings can influence our behaviour. The question is how meanings can influence our thinking through consciousness if they are immaterial.
If one grasps a meaning it seems one is in an immaterial world. However, how can this be reconciled with ET's materialistic view?
They distinguish two forms of mental life. There is Mental Life 1, corresponding to primary consciousness, perceptual categorization and the biological underpinnings, in short the principle of somatic selection. And a Mental Life 2, corresponding to higher-order consciousness, the use of metaphors, conceptions and creation of meaning, in short the principle of experiential selection which is nested within the somatic selection principle.
Thus both are rooted in the material world. How can we grasp immaterial meaning in a material body? Modifiable value systems which constrain experiential selection could refine this process in such a way that mental processes such as thinking and meaning are experienced as a kind of voluntary self-selection.
Is it reasonable then to call for an extra selection principle, besides natural and somatic selection the principle of cerebral self-selection? ET do not hint at such an idea.
Accepting the idea that modifiable value systems influence our thinking and meaning brings back the notion of necessary prejudice.
How can we ever know that value systems can do this baffling trick when we ourselves are part and parcel of this process? One may be doubtful that a combination of relatively old neurotransmitters like serotonin, dopamine, noradrenaline and acetylcholine can do this trick.
Still, even if we discover that more refined and complex peptides are involved, it still does not mean we have put our finger on it. It may feel that our modifiable value systems, our mental kernel, can bridge the gap between the subjective and the objective, but logically we must always be sceptical.
HOME