Evolutionary way of thinking
Over 68.7% of human genes are responsible for the brain.
As a result of evolution, the nervous system of organisms consistently acquired new properties, listed below in the list.
The first three properties are quite simple, understandable and ordinary. With the latter, we constantly live and use, but still do not know how it works. As for the triple-point, yes, we do not know how to move from the conditioned reflex to thinking, as there are actually added a lot of new properties that are complex and separate, and in the sum are not necessarily sufficient for thinking.
The first property after the conditioned reflex, in my opinion, was the principle of dominant . This principle allows one to possess several models of behavior and reflexes, and at the same time to preserve the unity of actions of the whole organism, highlighting only one vector of behavior, inhibiting all the others. The above features of the dominant can be seen in the example of the embracing reflex in frogs, which occurs during mating as a result of hormonal influences. The lightest touch to the corns on the thumbs of the fore limbs of the male immediately causes an embracing reflex. Irritation of the skin with mechanical, chemical or electrical stimuli leads to an enhanced reflex. A high threshold, which is required to trigger defensive reactions, indicates concurrent inhibition of other behaviors.
')
The following properties of the nervous system make it impossible to assess the sequence of their occurrence, since each of them solves its specific problem. However, these are no longer reflex patterns of behavior and the first of them is obliged to the emergence of the need to develop conditional patterns of behavior. The reflex gives the only solution for controlling the body, the unconditional criterion for solving, but if a positive result of the solution to a problem depends on external factors, then the ordinary reflex loses its effectiveness. A conditional decision criterion is required. So, for example, a locomotive driver from the left window saw a red traffic light signal prohibiting travel, the brake lever is in the center of the cab to brake, it needs to be turned to the left. That is, if the driver was at the left window, then he should turn the lever towards himself, if he was right away.
It is possible that another property works according to the same principle as the development of a conditional decision criterion, but it is distinguished separately. This is a transfer of the decision criterion . The conditioned reflex unconditionally solves the problem, but due to the possibility of transferring the criterion of the solution, one can not wait for the conditioned stimulus, but (try) to create conditions equal to the conditioned stimulus. So observing that the wind sways the branches of a tree with fruits and as a result of which the fruits fall, one does not need to wait for the wind to blow the next time, in order to eat, it is enough to shake the branches yourself.
Well, the last thing that the conditioned reflex does not cope with is a combination of conditioned stimuli that are often found separately. If the acquired conditioned reflex uses two simultaneous conditioned stimuli, then they are extinguished separately without confirmation, and the next time when both stimuli appear simultaneously, the conditioned reflex will not work. Solves this problem and many other memory.
For ease of description, the behavior model is a ready-made solution, a ready-made mechanism on neurons that causes the body to perform actions. So below is the coarse structure of Fig. 1b controlling the muscles of the fin-wings of the angel clam Fig. 1a and which is the rhythm generator.
Fig. 1 Molusk Angel and the control structure of its fins-wings
Neurons "a" and "b" mutually suppress and produce a rhythm, "many" are motor neurons, and the neuron "c" is the controller.
Nature has implemented many models of behavior using neurons, and then I want to give you approximately a picture of the spectrum of the most elementary components that it uses. Taking only the properties of synapses and dendrites of a single neuron, you can make a signal coincidence detector, amplify or attenuate input signals, smooth out high-frequency input signals, perform logical operations on incoming signals and this is by the forces of only one neuron. The neuron itself can be a rhythm generator - this activity is called a pacemaker and is a translation of genetic memory. Using two neurons, nature made signal frequency multipliers, converting bursts of signals into a single signal, again a coincidence detector and a system for maintaining different rhythms.
Everything written below is solely my speculation about the location and the relation of memory to reflexes and the dominant principle.
Thinking how to come to the need to use memory in its simplest form, I remembered a wonderful indicative unconditioned reflex. An addictive reflex in the system has been called negative learning, which means that as it repeats, a stimulus loses its ability to cause the reaction that it previously caused. The addiction or extinction of the orienting reflex is associated with the formation of a “nervous stimulus model” - its multidimensional engram, which inhibits the activation system of the orienting reflex. A new property of the nervous system that can memorize nervous models of sensory information and highlight new, previously unseen models can be called memory.
If we again consider the evolutionary causes of the appearance of the properties of the nervous system, then the first, as is well known, property of the nervous system was a reflex. To make the behavior more complicated, new reflexes and behavior models were added, which used the same sensors and effectors, and that several of the behavior models would not work at the same time, a dominant mechanism appeared between the sensory information and the behavior models. The appearance of the orienting reflex required the appearance of a new property, which allows dividing sensory information into previously encountered and not encountered. This mechanism we call memory. It would be rational to use in the dominant mechanism better and processed sensory information, which is obtained as a result of the work of memory and is considered to be encountered. In my opinion, memory was used by the dominant mechanism at the level with sensory information, to control behavior. That is, the result of the memory operation for the dominant mechanism is represented as additional and better sensory information.
If we conditionally divide the neural network control system with such properties into its component parts, then, in my opinion, the following scheme will be obtained.
Fig. 2 Structural diagram of the neurosystem of control
Yes, memory as better sensory information is wonderful, but it can play a cruel joke from the moment it starts working with sequences. And because of the associative recognition will run ahead, which would be quite like nonsense in the sensory part of the system. Thus, memory recognition of a sequence of events will be perceived as a dominant mechanism, as if all these events occur simultaneously and now. This type of problem also includes the problem of differences in the image or copies of a different size from the real subject. So, for the first time when young children are confronted with a high-quality image of their sandals, they try to wear them, or sit on a toy chair, very similar to their home chair. It will take a department in the neural network system that will compare the reality and the result of the memory, as well as separate real events and objects from the intended ones.
Strangely enough, but as a result of finding a solution to such a problem, I came to the conclusion that this mechanism is in memory, and absolutely not separable from it. If we recall how we make up a pattern from components during a direct collision with the object or event being characterized, then in the process of recollection some more unobvious characteristics are added in the sum of the generators of the new pattern. A new pattern is created in the memory, which includes the characteristics “virtualizing” the pattern and the previous real pattern. A worthy example is the pattern consisting of the sensation in the mouth and the sense of taste from barberry candy or any other. Having imagined these sensations, you operate with a pattern of imaginary candy, since it still contains feelings of not having any food in the mouth and the usual taste from an empty mouth. If you start to chew any other candy, then it will be more difficult to remember the sensations of “Barbariska”, but you will learn it very quickly. This imaginary and virtual character allows the dominant mechanism to clearly separate the patterns responsible for reality and dreams. It is easy to see that we think or even rather dream with exclusively “virtual” patterns, and also completely unconsciously perfectly separate events and objects perceived by the sensory system from imaginary ones.
Virtual patterns can be classified into predicted, imaginary and simulated. Predicted include any patterns - activated as a result of working with sequences and not directly perceived. Simulated - are responsible for not completely real objects and events: specular reflections, images, sound and other imitations of reality, while such patterns have their own fully virtual patterns. Imaginary is already completely virtual patterns of their real counterparts, or pure imagination consisting of other virtual patterns.
The approach that a neuron pattern is unchanged - gives several consequences, namely: a neuron can have only one pattern, only one piece of information per neuron, the constancy of the pattern guarantees the invariance of the logical frame and the immutability of subsequent generated patterns, the memorization of all new information requires the addition of new neurons. But if you use the model of my hybrid neuron , where the pattern is stable and volatile associative links can be established, then most of the new information can be stored precisely through associations. That is, only rare new information from a constant sensory flow is really worthy of creating a new pattern, and for the overwhelming amount of new data, associative memorization is sufficient with a large library of already existing patterns. For example, for an apple pattern, adding new information in the form of adding new coloring colors (green, burgundy, light yellow), new tastes, sizes, names of varieties and so on does not require the creation of new patterns, it is enough to add new associative links with other patterns, each of which carries a unit of information. As an example of the prevalence of volumes of associative memory over memory patterns, it is possible to use associative techniques to memorize large amounts of information. However, this method of memorization does not give the full possibility of manipulating information, since it is still necessary to use the same techniques of associative memory access to extract memorized information.
As a result of evolution, the nervous system of organisms consistently acquired new properties, listed below in the list.
- Unconditioned reflex.
- Addictive
- Conditioned reflex.
- ...
- Thinking (Profit)
The first three properties are quite simple, understandable and ordinary. With the latter, we constantly live and use, but still do not know how it works. As for the triple-point, yes, we do not know how to move from the conditioned reflex to thinking, as there are actually added a lot of new properties that are complex and separate, and in the sum are not necessarily sufficient for thinking.
The first property after the conditioned reflex, in my opinion, was the principle of dominant . This principle allows one to possess several models of behavior and reflexes, and at the same time to preserve the unity of actions of the whole organism, highlighting only one vector of behavior, inhibiting all the others. The above features of the dominant can be seen in the example of the embracing reflex in frogs, which occurs during mating as a result of hormonal influences. The lightest touch to the corns on the thumbs of the fore limbs of the male immediately causes an embracing reflex. Irritation of the skin with mechanical, chemical or electrical stimuli leads to an enhanced reflex. A high threshold, which is required to trigger defensive reactions, indicates concurrent inhibition of other behaviors.
')
The following properties of the nervous system make it impossible to assess the sequence of their occurrence, since each of them solves its specific problem. However, these are no longer reflex patterns of behavior and the first of them is obliged to the emergence of the need to develop conditional patterns of behavior. The reflex gives the only solution for controlling the body, the unconditional criterion for solving, but if a positive result of the solution to a problem depends on external factors, then the ordinary reflex loses its effectiveness. A conditional decision criterion is required. So, for example, a locomotive driver from the left window saw a red traffic light signal prohibiting travel, the brake lever is in the center of the cab to brake, it needs to be turned to the left. That is, if the driver was at the left window, then he should turn the lever towards himself, if he was right away.
It is possible that another property works according to the same principle as the development of a conditional decision criterion, but it is distinguished separately. This is a transfer of the decision criterion . The conditioned reflex unconditionally solves the problem, but due to the possibility of transferring the criterion of the solution, one can not wait for the conditioned stimulus, but (try) to create conditions equal to the conditioned stimulus. So observing that the wind sways the branches of a tree with fruits and as a result of which the fruits fall, one does not need to wait for the wind to blow the next time, in order to eat, it is enough to shake the branches yourself.
Well, the last thing that the conditioned reflex does not cope with is a combination of conditioned stimuli that are often found separately. If the acquired conditioned reflex uses two simultaneous conditioned stimuli, then they are extinguished separately without confirmation, and the next time when both stimuli appear simultaneously, the conditioned reflex will not work. Solves this problem and many other memory.
Behavior patterns
For ease of description, the behavior model is a ready-made solution, a ready-made mechanism on neurons that causes the body to perform actions. So below is the coarse structure of Fig. 1b controlling the muscles of the fin-wings of the angel clam Fig. 1a and which is the rhythm generator.
Fig. 1 Molusk Angel and the control structure of its fins-wings
Neurons "a" and "b" mutually suppress and produce a rhythm, "many" are motor neurons, and the neuron "c" is the controller.
Nature has implemented many models of behavior using neurons, and then I want to give you approximately a picture of the spectrum of the most elementary components that it uses. Taking only the properties of synapses and dendrites of a single neuron, you can make a signal coincidence detector, amplify or attenuate input signals, smooth out high-frequency input signals, perform logical operations on incoming signals and this is by the forces of only one neuron. The neuron itself can be a rhythm generator - this activity is called a pacemaker and is a translation of genetic memory. Using two neurons, nature made signal frequency multipliers, converting bursts of signals into a single signal, again a coincidence detector and a system for maintaining different rhythms.
Memory
Everything written below is solely my speculation about the location and the relation of memory to reflexes and the dominant principle.
Thinking how to come to the need to use memory in its simplest form, I remembered a wonderful indicative unconditioned reflex. An addictive reflex in the system has been called negative learning, which means that as it repeats, a stimulus loses its ability to cause the reaction that it previously caused. The addiction or extinction of the orienting reflex is associated with the formation of a “nervous stimulus model” - its multidimensional engram, which inhibits the activation system of the orienting reflex. A new property of the nervous system that can memorize nervous models of sensory information and highlight new, previously unseen models can be called memory.
If we again consider the evolutionary causes of the appearance of the properties of the nervous system, then the first, as is well known, property of the nervous system was a reflex. To make the behavior more complicated, new reflexes and behavior models were added, which used the same sensors and effectors, and that several of the behavior models would not work at the same time, a dominant mechanism appeared between the sensory information and the behavior models. The appearance of the orienting reflex required the appearance of a new property, which allows dividing sensory information into previously encountered and not encountered. This mechanism we call memory. It would be rational to use in the dominant mechanism better and processed sensory information, which is obtained as a result of the work of memory and is considered to be encountered. In my opinion, memory was used by the dominant mechanism at the level with sensory information, to control behavior. That is, the result of the memory operation for the dominant mechanism is represented as additional and better sensory information.
If we conditionally divide the neural network control system with such properties into its component parts, then, in my opinion, the following scheme will be obtained.
Fig. 2 Structural diagram of the neurosystem of control
Yes, memory as better sensory information is wonderful, but it can play a cruel joke from the moment it starts working with sequences. And because of the associative recognition will run ahead, which would be quite like nonsense in the sensory part of the system. Thus, memory recognition of a sequence of events will be perceived as a dominant mechanism, as if all these events occur simultaneously and now. This type of problem also includes the problem of differences in the image or copies of a different size from the real subject. So, for the first time when young children are confronted with a high-quality image of their sandals, they try to wear them, or sit on a toy chair, very similar to their home chair. It will take a department in the neural network system that will compare the reality and the result of the memory, as well as separate real events and objects from the intended ones.
Strangely enough, but as a result of finding a solution to such a problem, I came to the conclusion that this mechanism is in memory, and absolutely not separable from it. If we recall how we make up a pattern from components during a direct collision with the object or event being characterized, then in the process of recollection some more unobvious characteristics are added in the sum of the generators of the new pattern. A new pattern is created in the memory, which includes the characteristics “virtualizing” the pattern and the previous real pattern. A worthy example is the pattern consisting of the sensation in the mouth and the sense of taste from barberry candy or any other. Having imagined these sensations, you operate with a pattern of imaginary candy, since it still contains feelings of not having any food in the mouth and the usual taste from an empty mouth. If you start to chew any other candy, then it will be more difficult to remember the sensations of “Barbariska”, but you will learn it very quickly. This imaginary and virtual character allows the dominant mechanism to clearly separate the patterns responsible for reality and dreams. It is easy to see that we think or even rather dream with exclusively “virtual” patterns, and also completely unconsciously perfectly separate events and objects perceived by the sensory system from imaginary ones.
Virtual patterns can be classified into predicted, imaginary and simulated. Predicted include any patterns - activated as a result of working with sequences and not directly perceived. Simulated - are responsible for not completely real objects and events: specular reflections, images, sound and other imitations of reality, while such patterns have their own fully virtual patterns. Imaginary is already completely virtual patterns of their real counterparts, or pure imagination consisting of other virtual patterns.
Not patterns uniform memory is strong
The approach that a neuron pattern is unchanged - gives several consequences, namely: a neuron can have only one pattern, only one piece of information per neuron, the constancy of the pattern guarantees the invariance of the logical frame and the immutability of subsequent generated patterns, the memorization of all new information requires the addition of new neurons. But if you use the model of my hybrid neuron , where the pattern is stable and volatile associative links can be established, then most of the new information can be stored precisely through associations. That is, only rare new information from a constant sensory flow is really worthy of creating a new pattern, and for the overwhelming amount of new data, associative memorization is sufficient with a large library of already existing patterns. For example, for an apple pattern, adding new information in the form of adding new coloring colors (green, burgundy, light yellow), new tastes, sizes, names of varieties and so on does not require the creation of new patterns, it is enough to add new associative links with other patterns, each of which carries a unit of information. As an example of the prevalence of volumes of associative memory over memory patterns, it is possible to use associative techniques to memorize large amounts of information. However, this method of memorization does not give the full possibility of manipulating information, since it is still necessary to use the same techniques of associative memory access to extract memorized information.
Source: https://habr.com/ru/post/127866/
All Articles