Reprinted from NOESIS, vol.XII, 1986, pp.129-133

In the history of philosophy and of scientific thinking there are known some theories of mind. Among these may be mentioned:

• the monist theory, for which the fundamental reality is mind;
• the dualist theory, for which mind and body are of a different nature;
• the interconnectivity theory, for which the mental activity of the brain is a consequence of the huge number of interconnections among neurones;
• the identity theory, for which "mind is the brain" or " brain is the mind", a theory based on a biological philosophy [1];
• the data processing theory (informatic theory), for which the mind is the software of the brain and for which there is no fundamental difference between natural intelligence (NI) and artificial intelligence (AI);
• the theory of non-existence of the mind, for which the mind is a system of semantic structures ( or meaning structures [2]) through which a person is in contact with the physical and social worlds.

Some of these theories have many points in common and, of course other theories may be mentioned or may be advanced.

Perhaps the most important problem to be clarified, both for a theory of mind and for the artificial intelligence theory is the relation between natural and artificial intelligence.

Alan Newell, a well-known scientist and specialist in artificial intelligence, believing many years, together with Herbert Simon, that NI = AI, writes:

"One of the world's deepest mysteries - the nature of mind - is at the centre of AI [...]. Its discovery which will no more be a simple act thant he discovery of the nature of life, or the origins of the universe, will be a major chapter in the scientific advance of mankind. There will be a coherent account of the nature of intelligence, knowledge, intention, desire etc., and how it is possible for the phenomena that cluster under these names to occur in our physical universe" [3].

AI practice cannot solve the mind problem. But it can help to acquire a new understanding of both mind and AI theory.

Concerning, for instance, the modelling of space and time in data-processing programs and in AI programs, Bernard Meltzer, an expert in expert systems, observes that these programs

"seem to model a ghost, ethereal world in which not only there is no space or time, but not even physical objects: a solipsistic world[...] [4]".

Something is lacking in AI to have all what we know about human mind, i.e.: it has not the feeling of continuity of the objects the man sees, continuity of itself, feeling of continuity of time, intuition, genuine creation and so on. We think that such phenomena are not purely structural, but have components that we call purely phenomenological.

Mind processes are without any doubt, informational processes. We may question only the nature of mental information, in what degree it is identical or similar to the structural information of computers and AI programs.

We may think that in a mind, information has not only structural but also yet unknown phenomenological components. And these phenomenological phenomena (related, in a way, to the phenomenology of Husserl) cannot manifest outside matter, but in a special form of matter, that we called informatter. It is not possible to present here all the ontological bases [5][6][7] of such an idea, but for the purpose of this paper it is sufficient to say that any biological body is formed of the usual structural substance and informatter, the latter offering to a mind all what AI has not: continuity and phenomenological senses of which every mind is aware.

If we take this theses (I) as an initial point, we may derive a lot of consequences, shortly presented as follows:

II. Mental information has both structural and phenomenological components.

III. The brain is both structural-physical (the usual body) and phenomenological-physical, that is the brain is physical and alive, that is biological.

IV. The brain is a substratum for structural-phenomenological informational processes.

V. The mind is the entire informational activity of the brain, structural and phenomenological.

VI. Mind is not the brain, but the brain contains the mind.

VII. The brain is therefore a physical-informational device.

AI has nothing phenomenological (in the sense explained above). Being intelligent, an AI system has a psychology but its psychology is not mental in itself. AI offers only an illusion of a mental psychology, although, we must recognise, sometimes, a very good illusion.

How to test by experiment the difference between AI and NI ?

A criterion might be the test of creativity, but this is delicate even for NI; otherwise, to ask where a piece of knowledge is placed inside the intelligent system (artificial or natural). An AI system will point to some locations of memory. A man cannot indicate such location. It may be objected that man has an associative memory and has not sensors for his own brain, but we think that it is not only this, because the brain cannot be reduced only to the network of neurons, also cannot be reduced to its molecules, especially those inside the neurones, which may have a role in the computational processes of the brain. If we take into account its informaterial component, this one not having a spatial character, the mental processes will never have a totally spatial localisation. Some philosophical systems of antiquity spoke, perhaps not without reason, about infinity of mind.

Never will a non-mental intelligence say like Anna de Noaille:

"Je t'avais regardé. Le regard est un contact plus nete et dur que le mineral. C'est un chosc qui ebranle deux esprits..."[8].

Never will a non-alive artificial intelligence transmit or receive something from mental to mental, as in the above example, a reciprocal mirroring of the eyes of two beings.

Some considerations concerning the human mind may be drawn from a theory of a problem [9] suggested by the Romanian philosopher Lucian Blaga.

Some other considerations concerning the human mind may be deduced from Kant's theory of intellect and reason, and in summary we may say that AI has not reason in the kantian sense, only a human may have this reason, or in the future an alive AI.

In fact, we think that a structural-phenomenological theory of mind is possible, which might explain the similarities and the differences between AI and NI.

We may say that the semantics of AI is only structural, and the semantics of NI is structural-phenomenological.

Ignoring phenomenological processes in mental thinking has far reaching consequences. The phenomenological processes may have a non-formal character, the structural processes, by definition, are formal, in the last instance are computational. If we conceive only a structural reality, this reality is then a computational process. The Universe, man and society would be computational processes.

If we recognise the phenomenological processes, the physical universe may still be a computational process, but not the man and not the society, although man and society have a great part of formal in them.

We are faced with two philosophical possibilities:

a) Either we recognise (like the science of today, especially physics and biology) only structural processes and then the entire reality is a computational process.

b) Or we recognise both structural and phenomenological processes as distinct and correlated, and then the reality is not entirely a computational process.

Michael Conrad, an emerging scientist in molecular computing, talking about the Turing-Church thesis on computability gives it the following interpretation:

"any system or process physically realisable must be effectively computable" [10].

By "effectively computable" one understands computable by a formal process with a succession of simple operations on strings of symbols, as it is the case, for instance, of a data-processing program. Conrad's extended thesis says that any realisable physical process is equivalent to a computation, even if is not effectively computable. Perhaps, at the molecular level, the brain succeeds to make computations that even if they are not of the effective type, they are still formal computations. Let us call integrated computations these noneffective computations. But taking place at the molecular level they are also structural and they cannot explain mind phenomena, although may be very powerful in computing and heuristics.

The first philosophical position (a) cannot explain mind phenomena. Today, neither physics nor information science can neglect mental phenomena, which are present in various forms perhaps in every biological body, under the form of structural-phenomenological processes. A mental process is essential in every living organism in order to have the phenomenological unity which is recognised, by biologists, above the structural unity of the body.

The mental is a fundamental phenomenon in the living matter and the recognition of this assertion is comnpatible, in our opinion, with the second (b) proposed philosophical position (which recognises a structural- phenomenological reality). Only the structural part of reality is computational.

Although the physical universe may be born by specific phenomenological-structural processes, the physical universe, if alone, is structural and remains and remains a computational process obeying the laws of physics.

The biological universe is not only structural, and in this second approximation even the physical universe is influenced by the mental behaviour of living organisms.

Living organisms, especially man, because of the great structural-phenomenological possibilities, have a huge informational disponibility. The human mind, being also phenomenological, is not determined completely by the laws of physics and biology.

We do not think that we might recognise information as a fundamental process in matter if we do not recognise phenomenological information. The structural information alone may be seen as a structural physical process. In such a case man would be determined by the laws of physics, society also.

We think that information may be taken into consideration, for the Universe, along with the four fundamental forces of physics. Concerning the profound matter from which the Universe emerged, the phenomenological information may remain the only fundamental phenomenon.

For man, information is added to the laws of physics to become man, and information is added to the laws of physics to make possible the society. And if information exists under structural form due to its initial phenomenological basis, then society could not be deeply explained without the fundamental existence of phenomenological information.

Therefore, contrary to the opinion that the entire world is working like a computer, an opinion accepted even by some philosopher like A.Sloman [11], we think that the world is working like a computer but not quite. We believe that philosophy and science too, may become structural-phenomenological one day.

Bibliography

1. D.O.Hebb, Essay on mind, Hillsdale, New Jersey, Lawrence Erlbaum Associates, 1980.

2. Lars-Gunnar Lundh, Mind and meaning. Towards a theory of the human mind considered as a system of meaning structures. Acta Universitatis Upsaliensis, 10, Upsala, 1983.

3. Alan Newell in eds. D.G.Bobrow and Patrick J.Hayes, Artificial Intelligence: where are we?, Artificial Intelligence, vol.25, 1985, p.377.

4. Bernard Meltzer, in eds D.G.Bobrow and P.J.Hayes, op.cit. p.406.

5. Mihai Drăgănescu, Profunzimile lumii materiale (The Depths of the Material World, Bucharest, Ed.Politică, 1979.*

6. Mihai Drăgănescu, Information, Heuristis, Creation, in the volume I.Plander (Ed.) ,Artificial Intelligence and Information,Control Systems of Robots, Elsevier (North-Holland), 1984, pp.25-29.**

7. Mihai Drăgănescu, Ortofizica (Orthophysics), Bucharest, Ed.Stiintifica si Enciclopedică, 1985.

8. Anna de Noaille, Exactitudes, Paris, Grasset, 2e édition, 1930, p,60.

9. Lucian Blaga, Opere, (Works), vol.8, 1983, p.99-1o7; 337-441; 365-371; works reprinted after the originals from 1933 and 1947).

10. Michael Conrad, On design principles for a molecular computer, Communications of the ACM, vol.28, 1985, May, p.465.

11. Aaron Sloman, The computer revolution in philosophy: philosophy science and models of mind, Hassocks, Sussex, Harvester Press, 1978.

Notes ( February 17, 1997).

* With an English edition on the web, The Depths of Existence, Bucharest, 1997, http://www.racai.ro/books/doe .

** See also on the author's Web-site: http://www.racai.ro/~ncristin/MD-Web/mdraganescu.html

dragam@valhalla.racai.ro