Cognitive Science and Psychiatry

Henrique Schützer Del Nero

Abstract:

Cognitive Science has been dealing with a very large class of phenomena, but problems tied to pathology have been omitted. It is clear that the problem of reduction x emergence of the mental regarding the brain is at the very core of the problem. Psychiatry as it is conceived nowadays represents a dilemma because one interferes with both extremes of the chain: with synapses and with stories. Is it possible to accommodate these two extremes in terms of representation proposals? Is it time to begin a discussion of the nature of the mind considering dysfunction as a privileged arena where both the brain, not only as a token-implementer, but a type designer, and the individual's program play important and complementary roles? Kantian notions as the 'forms of intuition' may enlighten these problems when we consider functions and not contents as the real core of cognition. Consciousness as a function, fed by mathematical and computational concepts as structural stability, may represent something new, surpassing old sterile debates and launching important ideas that may orient research in pathology.

Functional Psychiatry

If Cognitive Science seeks to enlighten the very nature of cognitive phenomena it has do deal with pathology too. Arguments that it is early to think on higher functions do not proceed because it is the essence of cognition, be it normal, be it deviant, that is on play. A de-affectize and de-contextualize Cognitive Science could excuse for not having dealt with affective disorders (mainly manic-depressive disorder) and with neuroses, but it would not be excusable for not dealing with psychoses (schizophrenia and others). Even if one defends that Cognitive Science deasl only with cognitive phenomena, then disorders of thought -- psychoses -- must be considered.

Schizophrenia is a very complicated pathology and all those that defend a cautious position could say that there is damage in the hardware. Then, even if token-identity is assumed to exist and to play the major role, it is not important to Cognitive Science because there is an implementational trouble.

Let us forget the major psychotic syndrome, schizophrenia, and think on many other disturbances that may affect thinking processes.

There may occur thought-disorders in :

a) Affective disorders: a state called 'catathymical' delusion with a mood disorder that underlies it. (in the tradition of the famous german psychiatrists, Karl Jaspers, 1920, and Kurt Schneider, 1948,, both authors of general treatises on psychpathology). Treatment can be done with anti-psychotic drugs, but permanent relief will be obtained through anti-depressants. There are also delusions in some cases of manic psychoses. This can be considered if one wants to dissociate thinking and 'emotion'.

b) Personality disorders: where a genetic troubles and environmental events play an important role.

c) Psychogenetic psychoses: particularly in hysterical psychoses (Ganzer syndrome) where it is supposed to exist some abnormalities at the level of processing things but a major contextual event triggers a psychotic event. (in Roy 1982)

We are not here considering many taxonomic problems, nor being very strict with the diagnosis table we are referring to. Of course a critic based upon DSM (Diagnosis Statistics Manual), or in RDC(Research Diagnosis Criteria) could hold, but at the expense of privileging labels and misconsidering arguments, clinics and history.

Briefly, there exist many affective, context-driven, biographical, problems that disturb cognition, scaffolding delusions, hallucinations, psychopathic conducts regarding social values, etc. Moreover, drugs usually are very promptly recommended in these cases as psychoteraphy, re-socialization, re-programming techniques, etc.

If drugs can affect thinking disorders, rearranging thoughts in a certain stable way, even if the cause of the pathology is a 'psychic' one, then:

a) how can one defend representations to exist, playing such an enormous and type-dissociated role regarding to the brain?

b) how can token-identity be considered if recover is so convergent regarding normal structures of evaluation, judgment and critics?

What I propose here is that Psychiatry represents a very extreme case for Cognitive Science because it acts at the same time in the two supposed to be separated levels: at the level of implementation and at the level of program. I.e., at the brains level and at the subject's level.

If functionalism was held to exist, Psychiatry would become Psychoanalysis, having almost nothing to do with the brain except for these pathologies where a clear lesion exists. As lesion is an exception within today's pathology, at least in terms of sufficient and necessary conditions, Psychiatry is almost impossible within the conceptual scope of functional emergence of the mental.

Another way to look to these problems would be to recruit connectionist models to take account of pathology. This is half true, because the semantical blocks that one chooses to interpret nodes and basins of attractions, etc, are the very core of the problem. Connectionist ways of processing don't enlighten our problems because the problem is tied to the objects and to the meaning of them and not primarily to the rules of connection. Even if one offers a connectionist model of psychoses, categories like judgment, ideas, coherence, justification and so forth are pre-required, what makes the model impossible to be projected (radically translated in logical terms) in the Central Nervous System.

Briefly, what embarrasses Psychiatry in a Cognitive approach is not strictly ways of processing data but ways of building data. One must pursue a reductionist plan, that might shadow representation as it has been conceived by cognitive researchers, if one has to understand why drugs affect and treat psychiatric dysfunctions. Hypothesis tied to deficits of implementation are bizarre, because the personal own story is a powerful element of the diagnosis, treatment and recovery. Psychoanalysis, however, has almost nothing to say about major syndromes, what poses the question: how can Cognitive Science face the problem of considering the individual's context, story and motives, and at the same time consider affecting low-levels as synaptic clefts, with a very reasonable nomologycal predictability of response?

Psychiatry in my view can be considered the extreme challenge to Cognitive Science. It deserves a careful analysis on both extremes: on mental contents and reports and on brain implementatios and invariances. Were the brain only a token implementer, we would not have a stable psychiatric taxonomy. Were the brain these mere underlying processor, with no commitment to 'contents', drugs would not be so specific, in a certain sense, and so effective.

Consciousness could be considered a solution to these problems. Nevertheless, consciousness is so complicated that has been avoided in some discussions, assumed to be non-reducible to the physical or only a theoretical term. (well known Fodor's and Churchland's arguments)

I think that there is a way to prevent the intricacies of these arguments and the traps they contain. If one looks to consciousness as a content, one is immediately driven to introspection, qualia, and so forth. Intentionality, beliefs, self-reflection, etc. might be concepts that preclude any kind of reduction, but the function of consciousness might be not.

Consciousness considered as very specific evolutionary gain is tied to appearance of a very specific branch of the brain: the neocortex, mainly the frontal lobes.

Ethical values, social development, language, 'volitions', creativity, control, repression, responsibility, normative rules, etc., are apparently human traits. Consciousness, except for some forms of phenomenological and anticipating one, is also. Maybe these traits are firmly tied to consciousness and to pathology, due to the evolutionary gain they allowed.

What kind of taxonomy could be built if it considers functions and not contents, thinking on modes of computation that might underlie the voluntary-creative-'free' mode to exist?.

Kant thought on 'forms of intuition' and elected some a priori forms that could constitute knowledge. Newtonian mechanics and Euclidean geometry were these a priori forms. Is there any candidate to the condition of a priori form of a taxonomy that considers consciousness as a form, eo ipso its deviances, and explains how can a cognitive-behavioral act be sometimes conscious and sometimes automatic?

A functional account of categories should pay attention to this difference, because it represents the essence of the problem: is the object that is conscious or not or is it some computational peculiarity of its structure?

Structural Stability

Classical dynamics offers us a very intriguing phenomenon: deterministic equations may show for certain values of its parameters unpredictability of the next state. This is very important because the notion of determinism that it allows is one based upon structure. The equation obeys all the time a certain nomologycal description of sequence and causality, but for certain values in the parameters space, unpredictability holds.

Deterministic chaos and structural stability are concepts that derive from these peculiarities of classical physics. Suppose for a while a pendulum. It may be submitted to three types of friction: positive, negative and absent. If the pendulum is oscillating: a) if there is positive friction the amplitude of the movement decreases until it stops moving; b) if there is negative friction the movement increases infinitevely (suppose there is no string) c) if there is no friction it oscillates regularly for ever.

Structural stability is a concept closely related to parameters, in the case above described, to the friction parameter. Consider a small perturbation in a parameters value: if the solution or the physical behavior doesn't modify dramatically from the qualitative point of view, then we have structural stability regarding that parameter. If a dramatic qualitative behavior occurs, then we the system is structurally unstable (regarding that parameters value or interval of values). In the case of the pendulum for parameters within the positive and the negative range, small perturbations don't modify the qualitative character of the oscillation: either it tends to stop or to accelerate, however, for the parameter equal to zero, small perturbations change dramatically the behavior of the system. This is structurally unstable. Briefly: structural stability means that slight perturbations don't change the behavior of system from the qualitative (or topological) point of view. Structural non-stability is the opposite.

Parameter values that enhance structural stability are called ordinary. Those that enhance structural non-stability are called bifurcations.

Pay attention to the fact that:

i) bifurcation is a concept tied to qualitative analysis.

ii) the more regular phenomenon like a pendulum oscillating free of accelerating or de-accelerating forces is structurally non-stable.

iii) if equilibrium means conservation of energy, and if a pendulum free of friction is a conservative system, then isolated, ideal and conservative systems are non-stable from the structural point of view.

iv) chaos may be a succession of bifurcations, in spite of being itself structurally stable. Small perturbations in a chaotic parameter don't change the multitude of states that characterize it. However, the system to be chaotic has to be non-linear and to have sensitivity to the initial conditions.

From these considerations I suggest that:

1) It is not necessary to recruit quantum formalisms to see the intricacies that may happen in the Central Nervous System.

2) Determinism is compatible with indeterminacy.

3) The same deterministic equation, i.e., one that describes a nomologycal event, can be stable or unstable, structurally speaking.

4) This depends on parameters.

5) Parameters are not exactly the major actors in the structure, they don't vary constantly as state variables do.

6) If structures are the same, if determinism is maintained all the time, something in the system may vary slightly and modify dramatically the system's behavior.

7) It is not the whole system that varies, it is the parameter.

8) Very harmonic behaviors, conservative ones, may be out of stability from the structural point of view.

9) Chaos, the name of necessity that dressed random, is structurally stable.

What could be extracted from the above said, if we are seeking a kind of formal invariance that could enlighten qualitatively a functional taxonomy to psychiatry and as side-effect a reductive-function approach to the relation between neurons and functional-dysfunctional representations and behaviors?

Things are conscious whenever they are new, subject to learning and creation. Judgment is conscious reflection and not a sub-liminal computation over a range of possible solutions. Whatever is sensed in the mental screen as conscious phenomenal is still unstable from the structural point of view, in spite of being harmonic and coherent (remember the pendulum). Whenever things become stable, setting parameters over ordinary values (due to synaptic plasticity), structural stability gates the mode to the automatic, non-conscious one. Psychoses may perfectly be a kind of half automatic, half conscious phenomena when coherence is loosen and stability is there. This may be the state of dreams and of psychoses. If chaos happens during sleep it is only an artifact that produces bizareness in order to grasp order from anomalous stable points. If it happens during wakefulness it is psychoses.

Here one have a minimum criterion to a Cognitive flavor that allows representations qua objects to exist, but that proposes a kind of underlying computational-formal mechanisms, based upon structural stability, as the source of consciousness, non-consciousness, dreams and psychoses.

Responsibility and Free-Will

The most important result of these ideas if they come be right and enable a series of experiments, is that whenever one adopts a deterministic causal approach to higher functions, one is forced to answer questions tied to freedom.

My opinion is that consciousness represents an evolutionary gain that allows language to represent, sentences to occur even if they are not true, values to be parameters of one's acts. Moral values, social contracts, deviance and suicide are characteristics of our mind. Cognitive Science must take that into account, and has to see that if causality still exists, even in a probabilistic and sometimes unpredictable way, determinism must face the problem of freedom.

We sense as if we were free agents, but if conscious objects seemed to be free, conscious forms are only structurally non-stable. Coherence, as the pendulum, is a very faint ideal that vanishes as things are learned and loose their novelty. After coherent and conservative ideal moments, there is only friction, energy lost or gain, automatic behavior indeed. Consciousness enables us to learn and to move and select. Were we free the contents would have the structure of freedom, but they have it only as a very faint parameter value. Every freedom in the realm of forms dissipates, being only unpredictable. Tragedy or ergodic are different sides of the same coin: necessity and law.

Then we suffer because we thought we were free, but there is something that precludes us to genuinely free. Pathology is the highest field of consciousness in a kind of double-bind syndrome: content is free, form is a slave of non-linear non-stable phenomena.

REFERENCES

Abraham,R. and Shaw,C. (1992) Dynamics: The Geometry of Behavior. Addison Wesley

Glass,L.& Mackey,M. (1988) From Clocks to Chaos. Princenton University Press

Jaspers,K. (1920) Allgemeine Psychopathologie. Berlin.Verlag von Julius Springer

Roy,A.(1982) Hysteria. John Wiley and Sons.

Schneider,K. (1948) Klinische Psychopathologie. Georg Thieme Verlag.Stuttgart.

Stein,D. & Young,J. (ed)(1992) Cognitive Science and Clinical Disorders. Academic Press

Searle,J. (1992) The Rediscovery of Mind. MIT Press

Talbott,J., Hales,R. Yudofsky,S. (ed)(1988) Textbook of Psychiatry. The americam Psychiatric Press