Theory of the neuronal circuity of the brain and analytical thinking

ISBN 978-3-00-037458-6
ISBN 978-3-00-042153-2

Monograph of Dr. rer. nat. Andreas Heinrich Malczan

Part 2.7 The sequential storage of data in the cerebellum 

An important prerequisite for time memory is the ability to recapitulate two signals in the order in which they occur. Even in history lessons, it is often a question of: what was first, what was after, what was in between?

We have the ability to analyze perceived and learned signals with regard to their temporal occurrence. Therefore, in this section we will search for the material causes of this ability.

The reason for the sequential storage of data is that data is stored in the order in which it occurs. The first signal that can be imprinted is stored in the start group of the cerebellum cluster. The second memorizable signal becomes the inherent signal of the second Purkinje group. The third signal goes into the third group, and so on.

A signal to be detected runs along the moss fibres from the starting group to the second, third, fourth group, etc. If a group recognizes this signal as its own signal, its positive nuclear neuron reports the signal recognition to the thalamus. The Golgi cell at the end of the group interrupts the signal transmission along the moss fibres to the subsequent groups. The negative nuclear neuron of the recognizing group either inhibits the central distribution neuron for the secondary climbing fibres, so that no climbing fibre axon is active any more and thus no free Purkinj group stores the already recognized signal again. Or, in the case of sequential distribution of the climbing fibre signal, the distribution neuron of the recognizing Purkinje group is inhibited, so that the subsequent groups all do not receive a climbing fibre signal and thus cannot learn the current - already recognized - signal. All Purkinje groups in front of the recognizing group do not recognize their own signal, so their Purkinje cells together inhibit all the own signal detectors, so that they cannot be imprinted. As a result, no Purkinje group will be imprinted with this signal even if the climbing fibre signal is distributed sequentially.

If it is a signal which is not the start group's own signal, the start group will not recognize it. Thus the inhibition of the climbing fibre signal is not necessary. The predecessor group (here e.g. the start group) informs the successor group that the applied parallel fibre signal is not its embossing signal by generating a climbing fibre signal for the successor group. Now the successor group can analyze the current signal. If it recognizes it as its own signal, it will inhibit the climbing fiber signal in the olive and report the signal recognition to the thalamus. If it does not recognize the signal as its own signal, it will generate a climbing fiber signal for the successor group. This follower group also starts the analysis and suppresses the climbing fibre signal because it recognises its own signal and reports it to the thalamus. Or it does not recognize its own signal and generates a climbing fiber signal for the adjacent follower group.

This continues in case of non-recognition until the first free and unembossed Purkinj group is reached and the climbing fiber signal causes their embossing, so that the current parallel fiber signal becomes the inherent signal of this group. Or the number of Purkinj groups is exhausted, so that no new signal can be learned.

Thus the Purkinj groups form a chain to which a number of different eigen-signals belong. Each (imprintable) output signal of the cortex cluster passes through this chain until a Purkinj group recognizes this signal as its own signal or until this signal reaches a free and unimprinted Purkinj group and this group is imprinted with this signal or until the end of the chain is reached.

Either the imprintable cortex signal is already an inherent signal in the chain of Purkinj groups and is reported as recognized, or the next free Purkinj group is imprinted with it. This continues until all Purkinje cells in the chain are used up.

In view of these findings, we define two new terms for the creation of system-theoretical order.

Definition 2.6: Memory chain/signal chain of a cerebellum cluster

The signal neurons of a cortex cluster project onto a moss fiber population of a corresponding cerebellum cluster. This population supplies the parallel fibers with signals. The parallel fibers of a cluster form a parallel fiber population. The Purkinj groups supplied by this parallel fiber population form an ordered chain. This is called the memory chain of the Cerebellum cluster. The intrinsic signals associated with the imprinted Purkinj groups also form an ordered chain of signals, which we call the signal chain of the Cerebellum cluster.

Now the learning process of the cerebellum for complex signals should be emphatically specified from a systems theoretical point of view. The cerebellum learns complex signals that flow in from the cortex via the moss and parallel fibres, not from an inner drive, curiosity or of its own free will. The opposite is the case. The cerebellum learns the different (imprintable) signals by force. This is ensured by the algorithm of chaining the individual memory elements. Here, the Purkinje group is the smallest storage unit. It can store exactly one complex signal. And the first group in the chain learns its complex signal by force, because the average value of this complex signal is used to generate the primary climbing fiber signal. Thus this primary climbing fibre signal burns the first complex signal of the living being into the start group of the Purkinj groups. There is no freedom, no desire to learn and no curiosity. The first imprintable signal automatically lands in the first free group of Purkinje cells, i.e. in the start group.

And the next new (imprintable) signal is "burned in" in the neighbouring group. This is done fully automatically. Each new signal is stored within the memory chain in the next free Purkinje group. This storage of new signals is forced. Of course, the number of free Purkinj groups will not be infinite. And perhaps an intelligent deletion algorithm (sleep) will delete statistically insignificant signals.

But before we turn to this thought, the principle of the forced imprinting of Purkinje cells should be formulated as a new theorem of this theory.

Theorem 2.17: Theorem of forced imprinting of complex signals in the memory chain 

The Purkinj groups of a cerebellum cluster are forced by the incoming cortex output. Every complex signal that has not been imprinted so far, but is imprintable with regard to its intensity and duration, which differs sufficiently from the already imprinted own signals of the cerebellum cluster and which reaches the parallel fibres of the cerebellum cluster via the bridge cores from the cortex, leads to forced imprinting in the first free Purkinj group and thus becomes a new own signal of this group.

The exciting question at the current state of the author's theory is therefore no longer

The question is not how we learn something, but how we forget something.

This is because a finite chain of Purkinj groups will quickly exhaust its capacity in view of the many different complex signals that affect a living being on a single day alone. So how is this important forgetting organized? Perhaps this monograph will provide the answer. The principle seems to have already become clear to the author. 

But before that, it should be clarified how the Cerebellum learns temporal signal sequences and how a "video memory" mentioned above could work. Because the video memory is much better suited to "free up" memory space. Here the principle of compression becomes possible. In the auditory area, for example, (sequential) letter sequences, in which one Purkinje group is initially required per letter, are compressed by using, for example, "syllables" as new complex signals. If the sequence of letters is transformed into a sequence of syllables, only one Purkinje cell is also required per syllable. Thus the sequence of syllables is considerably shortened compared to the sequence of letters. If whole words are introduced as complex signals and only one Purkinje group is needed per word, the signal sequence is again shortened considerably (without loss of content). Storing whole phrases, even longer poems in one ride in a single Purkinje cell frees up previously used memory in an undreamt-of amount of memory. After all, a Purkinje cell has about 200,000 synapses, each representing a binary value. Thus a Purkinje cell could theoretically store a binary number with 200,000 dual digits.

Apparently, compression is the storage reserve of the cerebellum.

What is needed, therefore, is the development of a neural circuit that does exactly this fully automatically and that also uses the already recognized substructures of the (human) brain. The idea for this has been around for a long time, but the realization of the video memory delayed the project a bit. It should be mentioned here in anticipation that memory compensation (according to the author) is a phase of sleep. But more about that later.

First, the video memory, or more precisely the memory for time-sequential complex signal sequences, will be explained in more detail. As always it is a model. But the author tries to align this model to the known neural realities.

ISBN 978-3-00-037458-6
ISBN 978-3-00-042153-2

Monografie von Dr. rer. nat. Andreas Heinrich Malczan