During my time as a social science student, some six years later, I wrote a paper on pathological narcissism. In it I explained how instincts could become meaningful, Freudian-like, psychic entities and drives. I treated the brain as, what I called, an “association system,” by which in retrospect I meant a neural network. I explained how sensory “activation patterns” in such an association system can, in principle, become organized automatically around very simple instincts like sucking. In this way groups of interconnected activation patterns can be formed, which can be analyzed as being, for instance, sexual drives or a superego. The superego may, for instance, form itself around the instinctive recognition of a pair of darker spots, which in real live will most often be the eyes of our parents. Because our instincts are connected to the association system, this instinctual recognition of a pair of spots is automatically coupled to children’s recognition of their parent’s eyes, for the simple reason that they often occur together. Around this pattern of recognition children form an image of their parents. Because of this instinct, this pattern may later become something which we call the superego. In this way this instinct begins to influence our thinking and behavior. In my paper I explained that, in principle, in order for these psychic entities to come into existence, the instinct has only to stimulate the association system a little, at a specific spot or area. That’s all. The rest goes ‘automatically,’ from a technical, association system point of view – socially, of course, this is anything but automatic.

I was bewildered by the high esteem and words of praise that my teachers gave me for this paper. I was rather uncertain about it myself, because it seemed rather technical to me. It was the writing of a technician, not the writing of a social scientist, I feared. I put a lot of effort in making the connection between the technical and the social-psychological realm. This uncertainty has stayed with me. For me it is also hard work to put something, which I feel to understand very well, into words. I guess I’ll only understand why this is so in retrospect, when I am another 20 years older – I do not:-/

Unfortunately mathematical prove is beyond the scope of this book, even though I will come close to it now and then. In the end one should be able to write down everything which is mechanical in mathematical terms. Only this proves that one does not forget anything. This said, a mathematician knows more than anyone that his mathematical symbols and formulae do themselves not give any meaning at all to the world. Only the words in natural language which go with them, and our own imagination, do. , for example, does not mean much if one does not give meaning to its symbols and its operators. The concepts of ‘force’ and ‘mass’ in this formula have a history of centuries. Ones this meaning is crystallized out, the formula is hanging in the air like a ripe apple one might say – in reality this is a more dialectical, or even solipsistic process, of course, but I hope you see my point. To gain and create a meaningful base for my theory my starting point will be imagination, meaning, and natural language. From there I will crawl back to the borderline after which mathematicians can take the rule.

In this period I also read a chapter on consciousness in a book by George Mandler on emotion and cognition.[^]

[^]George Mandler. Mind and Body. Psychology of Emotion and Stress. W.W. Norton & Company, London/New York, 1984.

In it he proposes that, with regard to visual things, one is only conscious of five things at a time. Unfortunately this chapter on consciousness was not very well integrated within the rest of his book. It simply fell out of its cognitive scheme of thinking, I think.

A television program is a very ‘temporal’ thing. I soon discovered that when, while looking at a television program, you close your eyes and count the number of things you just saw on the screen, you can easily remember three things, and with difficulty five, but that’s it – except when you are familiar with the program, of course.

I found that the structure of a movie on television is actually built up completely out of these elements of three to five interconnected items. If one sees, for instance, the same car or house twice, then a connection is made between an earlier and a later event in the movie. By such ‘connections-in-time’ a ‘structure-in-time’ is built up.

I call this structure a ‘temporal image.’ Understanding temporarily this way may seem simple. The difficulty comes when you realize that people have gained a non-temporal image of this temporal image too. It is this non-temporal or static image which gives the temporal image a more permanent structure. That is one thing this book is about.

If you understand me well, you’ll see a heartfelt consequence of this. In structuring and explaining this phenomenon I will explain the essence of being and consciousness, and I will prove that machines can, in principle, have consciousness too. For me this has very important ethical implications, but I can hardly speak about them while I have not yet convinced you of the fact of the machine being conscious.

The things foregrounded cannot stay hanging in the air, of course. They must be connected somehow, further on in the movie. This is where the art of story telling comes into play, and this is also where, in our time, ‘realism’ is created. Lots of home movies, shot of real life events, are not realistic for the simple reason that they do not have a ‘mental story’ to them. To create a so-called ‘realistic’ movie, the camera must really ‘look’ like a mind’s eye and foreground those things within a structure that is humanly interesting to watch. This is especially true with television since there is little room on its screen. It is the invisible but real person behind the eye of the camera, watching the scene which we watch too, who’s ‘watching’ we experience when we watch a realistic movie. We can experience the same beauty or ugliness when someone who is close to us, or who is a good storyteller, tells us a story. At such moments we watch the world through the eyes of someone else and we experience their inner beauty or ugliness.

The basic principles of neural networks were already proven to be right by other scientists – although no one around me knew – so I needed not to worry about proving this myself.

What is the essence of a neural network? One can make all kinds of neural networks but they all associate one or more patterns with each other or with themselves. If a pattern is associated with another pattern one speaks of “pattern association.” If patterns are associated with themselves it is called “auto-association.” Anything which a pattern associator can do, can be done by an auto-associator, so in my first, intuitive account of it I’ll skip the working of a pattern associator until a later chapter in this book. For the moment a neural network is an auto-associator to me. To me it is further more, also for the moment, a regularity detector.

What does a neural network do? It attempts to remember, or re-construct, within limits, all which is fed into it. This is called learning and remembering.

For instance, take a neural network, and attach a camera and a monitor to it. The auto-association is constructed such that the network tries to show that which is in front of the camera onto the monitor. It does this by correcting itself when what is on the monitor does not equal that which is before the camera.

Take some apples. Show them to the machinery a couple of hundred times. Then byte a piece out of an apple, hold it in front of the camera, and look at the monitor. What do you see? You will see an apple without anything bitten out of it.

What has happened? The neural network has completed the pattern of the apple. It has filled in the gap. This phenomenon is called ‘pattern completion.’ It is a form of remembering in the sense that the neural network re-constructs, from ‘memory,’ what it sees only partly.

Let us do the above again but now with a hundred apples and pears that are different from each other. If you look at the monitor during the learning process you will see strange things happening. In the beginning you will see every apple and pear on the monitor more or less as they actually look. At a certain moment however, pattern completion sets into work. No matter what apple you show, on the monitor you will see a ‘generalized’ or ‘typical’ apple/pear. If you keep showing apples and pears, however, the neural network will start to distinguish different kinds of apples and pears. It might, for instance, distinguish between apples and pears. No matter what apple or pear you show, the network will show a ‘typical’ apple or a ‘typical’ pear. Each item will fall in a certain category. Later the network might distinguish different types of apples, and different types of pears. Depending on the context these processes are called “generalization,” “abstraction,” and “feature detection.” Feature detection is a kind of specific generalization, or specific pattern completion.

We, humans, initially learn to know the world roughly in the same way, however unaware we may be of this, except that we have to deal with more things than apples and pears. Take for instance the learning by children of the past tense of English verbs. Children first learn all verbs as if they are irregular. As they master this, they discover, or are told, that there is a pattern in it. Next they tend to make all verbs regular for a while, including the irregular ones, which is known as “over-generalizing.” Lastly they discover the exceptions to the rule, and learn it right again.

Since I was – and am – striving for a monistic theory which is in conformation with all the facts that I know, I needed much time to think – you’ll find that I am certainly not eclectic in any negative sense. The consequences of my work with regard to our own beings, for instance with regard to explaining consciousness, made this need to reflect even bigger.

This said, it is clear that I had to draw borders somewhere. Without borders, any science becomes just a muddle of words. After one has distinguished these borders, however, one is not only free to pass them, one is obliged to pass them to and from, until one arrives at a truly monistic theory, integrating and incorporating everything around it. In this thinking and unification process the gray areas close to the borders should become as less ‘gray’ as possible.

How does one draw these borders? I do it, just like a neural network does, by abstraction. The object is to simply place those things which have more causal, psychic, or semiotic connections with each other – depending on the topic studied – into the same area. One can draw borders within an area and thereby create ‘sub-areas.’ With regard to neural networks I will, later on, call this process ‘gathering’ (or ‘collecting’). However silly it may sound, one can only gather by gathering. What I mean is, that it’s hard work.

Seen this way, the abstract areas created, become sites of meaning which will, through being giving names, form the symbols on which formal and syntactic-mathematical rules and definitions can be imposed. This is not to say that in our minds there are no interconnections if there is no syntax, it’s just that such knowledge cannot be communicated without a syntax – or, better, because those interconnections exist and can be uncovered in the domain of the real, it must be possible to create a syntax for it in the domain of language. The art of science is to create and to communicate meaningful entities by imposing a syntax, such as mathematical formula, upon them – it must be said though that this is impossible without the help of metaphors. If this communication is possible, and if its content is in correspondence with reality, then the meaning attached to the symbols that were created, will have proven to be legitimate. Anyone who thinks he can create new symbols or entities only by defining them will never say anything new. With a definition one brings ‘order’ – syntax – in things which are already known. This is useful, but not sufficient.

Since what I have to say is very unknown, and very new, I cannot draw the borders for you here right now. I will have to start at the other end, where all the little areas are, and reconstruct for you, bit by bit, the bigger areas and the scientific area which I have made up for myself. I will simply take you through my discoveries. That will be the style of this book. That is the lie I am presenting you. My actual discoveries went more from the opposite end – my book is split, just like the neural machinery that I will propose in it. It is important to mention this, because, as Jan Sleutels explains in his PhD thesis[^]

[^]Jan Sleutels. Real knowledge. The problem of content in neural epistemics. PhD thesis, University of Nijmegen, The Netherlands, 1994.

, one must realize that most things which I present in this book cannot be discovered by starting at the micro-scale end. Trying to discover the whole from the parts will usually only succeed by accident, and that is not a sound base for scientific inquiry.

I would, however, for those of you who have any knowledge about them, like to lay down some facts about the areas of knowledge which I use.

There are dozens of other scientific areas to be distinguished in my book. They are mainly located within the domains of brain physiology – but certainly not experiments on living animals – semiology, semiotics, and all kinds of psychology, and anthropology. I am especially fond of the more critical trends within the latter three, since they pick up the facts for which others are blind. My theory can only be better if it is able to explain both all and more than any concurrent theory can do – in this I am a true adherent of Karl Popper.

Formally it is important to take speech into account in order to distinguish its enormous actions and consequences from our non-linguistic, animal ways of thinking.

I think that I have now, in 2001, finally uncovered how our so-called short term memory works, and how it is implemented in our brains – but now, in 2021 I changed my mind again. I think this machinery, together with the rest of my machinery, is sufficient to explain how language can come into existence, and which languages can be comprehended by humans. From the latter predictions, among others, one could try to falsify my theory – or, as a positivist would erroneously say, one could try to prove my theory. I have not worked out the details, since I do not feel to have enough knowledge about syntax.

Earlier I spoke of temporal structures and their images. That is largely what Freud, seemingly without completely realizing it, speaks about when interpreting dreams. A dream is essentially a temporal ‘thing’ too, and when Freud speaks of “condensation” – “Verdichtung” in German – he is actually speaking about the non-temporal image which goes with this. You will see that this image is essential with regard to my explanation of consciousness.

Lacan interprets Freud’s theory linguistically, using semiologic theories of De Saussure. In semiology a distinction is made between signifier and signified – the symbol and that what it replaces. Lacan was well aware of the temporal aspects of speech, and of the importance of mathematics in understanding it. He didn’t really know how to use mathematics though – in my opinion.

Lacan makes a distinction between the “imagery” and the “symbolic.” To him these are separate systems. According to Lacan the imagery is incomprehensible – he wishes the scientists who try to uncover it good luck: I thank him – but the symbolic can be studied in itself. This would be the domain of psycho-analysis. At the same time Lacan did realize that things could break through this barrier separating the imagery and the symbolic. This is, for instance, necessary in order to create metaphors and to renew language.

Lacan’s separation between the “imagery” and the “symbolic” seems sensible to me, but I think that the ground of language is even more complex than the imagery is. Lacan knew from Freud that finding the ground of the imagery was beyond him. Psycho-analysis would never have existed if Freud would not have come to this conclusion. Freud too started studying neurons, just like me, and only because he saw that he could – in his time, we might now say – not learn to understand the human psyche this way, he tried something else, being psycho-analysis.

What is important to me is that Lacan realized that speech arrives from the interaction between the imagery and the symbolic. In order to understand speech, one has to understand this interaction, and the interconnections between the imagery and the symbolic.

The ‘lower’ parts of our brains perform, among others, essential stabilizing and controlling functions with regard to our ‘higher’ brains. In this respect the terms ‘lower’ and ‘higher’ are rather misleading. You’d sometimes better understand it exactly the other way round! This pyramid stands at its apex! The so-called ‘lower’ brains sometimes stand at the top, in a very fundamental way.

In a sense my story is about a relatively small part of our brains. I will, however, make the necessary links with some of the rest of it. Without this it would be difficult to comprehend the working of the phylogenetically younger part of our brains. Laying these connections won’t even be difficult qua principle. The problem is that it is impossible to know everything, even if we would know. I mean to say that it is not the quality of the facts about our brains which is difficult to understand, but the quantity of them. It’s the same as with our understanding of birds with regard to flying. It is much more easy to understand how and why an airplane flies than how and why a bird flies.

In electronic circuitry the things which one can calculate are generally not the difficult things. The principles of an amplifier or a radio are quite simple – although you can evidently make the story as difficult as you like by asking more questions. In practice the main difficulties stem from the irregularities, noise, and instabilities of the components which you have to work with – especially in a non-digital, analogue design. Amplification, or radio transmission, for instance, is itself peanuts – qua principle. In a real amplifier or radio transmitter, however, a lot of circuitry is necessary to overcome the irregularities of the electronic components you use.

The same goes for our mammal pair of brains. Like most things in this world, it is in the (very) end based on comprehensible principles – if only because evolution goes step by step. The problem is that a lot of extra circuitry is needed to keep everything stable and to avoid certain typical problems which arise in neural networks.

This said, let us not forget that there is a lot of psychic suffering too, so our brains do not always work that well. I will, now and than, use psychic illnesses to prove or understand the working of my thinking machine, but these phenomena themselves are, I think, separate areas which I should not delve into too deeply – however important they are for understanding the psychic illnesses – because otherwise I would go mad of mental exhaustion myself.

My theory complies with every fact that I know about humans and neural networks in every possible sense. I can’t say that I know everything, but in as far as ‘nature’ is concerned, I can, in general, say how ‘nature’ can be implemented in my machine. As far as ‘culture’ is concerned, I know, to put it simply, that I at least lay down most of the ground for a ‘system’ which can contain any cultural ‘system.’ This is not to say that there are not many problems to be solved, especially with regard to language and speech, on which culture heavily depends, but here too I can give you many hints.

There is another thing which makes it impossible to reconstruct our brains electronically, namely the gigantic number of neurons that it consists of – not to mention the, on average ten-thousand or so, synaptic connections per neuron. Further more, in living tissue useless neurons can die and others can come into existence.

All this said one must ask oneself: “How many feathers does an airplane. have?”