Nature of Perception

Written by : Mahmood Shamizi

Contents

Preface        …………………………………………………………………………………………………1

Nature of Perception      ……………………………………………………………………………….3

Abstract       ………………………………………………………………………………………………..3

Introduction      ……………………………………………………………………………………………4

Definition of Proposition Cell      …………………………………………………………………..5

Formation of Proposition Cell Structure and/or Semantics of Data   ………………….6

Competition of Proposition Cell     ………………………………………………………………..7

Installation of Proposition Cell     ………………………………………………………………….7

Sign      ………………………………………………………………………………………………………8

Meaning      ………………………………………………………………………………………………..8

Utilization      ……………………………………………………………………………………………..9

Types of Proposition Cell     …………………………………………………………………………9

Chronological Proposition Cell     ………………………………………………………………….9

Compositive Proposition Cell     ………………………………………………………………….10

Innate Proposition Cell     …………………………………………………………………………..10

Floating Proposition Cell     ………………………………………………………………………..11

Algorithm & Sensory Structure    ………………………………………………………………..11

Why and How to Develop it      …………………………………………………………………..12

Heuristic Procedure       ………………………………………………………………………………14

1. Final Cause ………………………………………………………………………………………14
2. Material Causes ………………………………………………………………………………..14
3. Formal Causes …………………………………………………………………………………..15
4. Efficient Causes ………………………………………………………………………………..15

Review in more general perspective and substaniation of compulsoriness

of evolutional path in mental construction of this system  ………………………………..15

Logical Feedback     …………………………………………………………………………………..16

Work      …………………………………………………………………………………………………..19

Control      ………………………………………………………………………………………………..20

Intricate Logical Feedback    ………………………………………………………………………20

What are the identification criteria of lower level in necessity for referral to the upper level? ….21

General and Special Intricate Feedback     …………………………………………………….21

Institutionalization of Innateness    ………………………………………………………………22

Conclusion     ……………………………………………………………………………………………23

References      ……………………………………………………………………………………………24

Appendix I      …………………………………………………………………………………………..25

Appendix II : Evolutionary and Combined Cells   …………………………………………25

Appendix III      …………………………………………………………………………………………26

Appendix IV     …………………………………………………………………………………………27

Appendix V     …………………………………………………………………………………………..27

Preface

   In this writing , you will face completely new concepts and perspective of the artificial intelligence in which how the real intelligence is constructed in a system will be explicated systematically and on the basis of a new phenomenon . This phenomenon was first presented in an article under the title of “ Non-Machine Vision” in the international conference on Machine Vision Image Processing & Applications M.V.I P2001 held in Birjand from 7-9 March 2001 by the writer and welcome by the masters of this technique, and the conference chairman acknowledged it through a special letter of appreciation . First this very article had been prepared under the title of Nature of Perception. Then to maintain the appropriateness with the subject of the above mentioned conference, the title was changed, and some points were omitted or added because of limitations in pages. Now the previous article is presented to the readers without any alterations.

     This paper explains how to prepare a ground for the growth of particular cells to the readers, these cells have several types and the ability to reproduce and show certain interactions to the input data and adjacent cells. These primitive cells have been designed to form different types of tissues like information maintenance, sense, perception of concepts and algorithms of the operations inside and outside the mind. This space has provided a ground for the opposition of concepts on the basis of dialectic principles and their hermeneutic growth in which the birth and death of each cell prepares a new substructure for the birth of a generation with new much clearer and more acceptable concepts. In this article, attempts have been made to clarify, as much as possible, some of the subjective and interdisciplinary concepts under discussion in philosophy and humanities, to reduce their hermeneutic aspect and add to the explicity of the experimental, revolutionary and objective definitions and make them accessible in engineering and cybernetics. These concepts are : meaning, concepts, sign, utilization, profitability, belief, surprise, incident, freedom, volition necessity, conjecture, confidence, probability, goal, will, idea, instinct, decision, skill, instrument, work, control. Of course, as you know, any article which is trying to change the reader’s view about the primitive concepts and axioms will face two problems simultaneously that seem to be interlocked and have created a barrier to the explanation of the new phenomenon in a way that it is necessary to make changes and revisions in the definitions and corrections in their prime concepts for the explication and clarification of that phenomenon, on one hand, because without agreement about the permit definitions between the writer and reader, explanation of the phenomenon seems not to be principal and fruitful. On the other hand, we known that if we want to present more precise definitions of the fundamental concepts, the surest way is to explain the phenomenon itself, but is it possible to clarify a phenomenon before changing the reader’s criteria? To pass through this paradox, these two problems must be solved at the same time; that is, the dear reader should review the article several times until the writter;s concepts and perspectives become clear from among the sentences little by little and all of a sudden the sentences are combined, and their logical coherence gets apparant. In brief, with consideration of the article type to realize the writer’s intention the reader’s creativity and interpretation power play a more basic role because the article cannot be proved on the basis of deductive inference and axioms geometrically, but it is a kind of inductive article, and because it intends to create changes in the basic concepts, the parties’ attempt is a necessary condition . Of course, because of limitations in the number of article pages and intensiveness of the materials, the satisfaction of  the majority of the readers may not be met, but it is enough to have a few people change their perspectives by concentrating on it. These can be initiators for starting the discussion about the phenomenon of intelligence including artificial or natural .

Nature of Perception

Written by : Mahmood Shamizi

Director of Development & Research Affairs of Iran Tractor Mfg. Industrial & Service Company

Abstract: In this article, a new system is presented in the artificial intelligence which is able to build concepts from the input data automatically and create statements for data processing. The ability of  this system to give a meaning to data is based on its profitability because that meaning used in the system is to achieve some goals which have been institutionalized  as tasks inside the system by the designer . This theory has been compiled on the basis of the dialectic principles, inspirations from the pragmatic philosophy and Aristotelian four causes, and the writer has made a sample robot which works accordingly and is able to expand its own mind .

   This system is able to discover and recognize the factors which impede its work and to create new operations in order to remove those impediments. This function is neither through trial and error about with reliance on statements it has extracted from the relationships existing in the environment and established in itself. Establishment of the statements by the system becomes applicable from the study of its own past observations. And according to the congruence principle, philosophical correspondence, the system is looking for differentiations and characteristics in the environment that have the ability to separate useful observations from useless ones. Discovery of these characteristics in the observations means redundancy in the excess date and their summarization, and it is , in fact, the birth of the framed concept and new proposition which has the ability to classify the observations that comes from abstraction of the information. Now without consideration of the verify and untruth of this statement, it can be said that a mutation has taken place in this system, and if the implementation of this concept is, in practice, not useful in detecting beneficial compounds from non-beneficial ones, it will not be accepted. And these changes will remain in the mind construction only when it has the ability to anticipate the conditions existing  in the environment before doing so; that  is , a mutation has taken place in a positive direction. But care must be taken that the mutation is not incidental and quite blindly in this regard but it profitability has been guessed to some extent before selection and is based something between science and ignorance or incident. Thus the system will do any discovery with an acceleration more than before and along with much more awareness. It is evident that such a system needs practice and periods of time to recognize the environment in order to reach a desirable efficiency and be able to duplicate its trained mind and reach mass production after this period.

Key Words : Sign, Meaning, Utilization, Semantic Net, Final Cause, Material Cause, Formal Cause, Efficient Cause, Pragmatic, Statement, Compositive Proposition Cell, Chronological Proposition Cell, Innate Proposition Cell, Logical Feedback, Heuristic Procedure, Formalization, Instrumentation .

Introduction:

    By putting the statements of mathematics and formal logic in a system, it can not be enabled to see and/or think at a desirable level because these statements have two weak points;

1. Because of dogmatic nature; that is their absoluteness, inflexibility and explicitness, these functions framed and shaped by a designer can not disintegrate n a course of time and have genetical history .

2. They are unable to carry the meanings of their signs and only show the type of the mathematical function and/or the logical relation between input and output (domain and range). For instance, if the letters of the formula F= mγ  are changed, except a simple multiplication function, nothing will remain for a physicist. In other words, the relationship between sign and meaning has been established only in the physicist’s mind not in the formula, and only a real intelligent system is able to establish a relationship between sign and meaning automatically.

  If the data are canalized artificially and compulsorily by the designer and directed to the related sign, system will do some machine-like actions and can not do more than that, so it will be dogmatic.

    And when the number of the concepts and statements needed to be processed become more, the designer will not be able to control and direct them any more, and various limitations related to design will be increasing progressively because they will encounter explosive compounds of them each. In order to remove these impediment a system must be manufactured which itself is able to build new concepts from the data or previous concepts .

The new concept is possible in two stages;

1. Proposition formation , 2- Leading suitable information into proposition

Definition of Proposition Cell:

These calls are a type of semantic net with an output branch called Axon and several input antennas called dendrite by which the existence of the state and/or a certain object is felt in the input and reported from its own output in a way that the output sign is a function of the order type or logical combination and/or numerical quantities of the inputs and also depends on the function type existing in the internal structure of the cell.

While a proposition cell exactly receives the environmental information through the cell input antennas is able to show the supplied quantity of a characteristic in the form of a phasic number in its output . This characteristic evaluates the profitability and non-profitability of the circumstances in the environment in terms of achieving a goal and sends it to the out put and shows, in other words, estimation of how much the circumstances are suitable and/or depend on the concept and/or counter-concept in its output. The output sign can appear in two forms in the system ;

1. It is directly used for system reactions
2. It is connected to a cell input antenna to participate in constituting a more complex concept : that is , it should have indirect relation with the application site so as to be used for the system reactions.

Formation of proposition cell structure and/or giving semantic to data :

Whenever a system, when required, proceeds to interpret and use raw data from the input, it has given semantic to those data. By interpretation it is meant searching, finding and extracting characteristics from among the data to existence of which not attention was paid nor evaluation nor manifestation were made already, but from now on the system paying attention to that characteristic will be able to have a new classification of those observations and evaluate the existence amount of that characteristic.

The need for interpretation and semantic usually begins when the system has followed the sequence of the current events and routines in the environment and compared them with its past observations in order to achieve a goal , but it does not encounter usual and expected consequences for unknown reasons. The observation of the unexpected and non routine behaviors from the environment will create a contradictory feeling in the system that is indicative of the existence of the difference between the mental model structure of the system with the reality existing in the environment. The existence of this contradiction causes a system to start reproduce new cell and complete a tree of causality in it . The need for solving the contradiction will be severer, specially in the follow of the profitable goals . For example, when the system defeates as the result of the everyday operations, it can be said that this will be considered as a unusual and surprising behavior in the viewpoint of system, and it will automatically be looking for the cause of failure .

To do so, first, all the past observations done in connection with the same goal which has caused the sense of contradiction  are recalled from the memory. By observation , it is meant a continuous range of incidences recorded in the memory which have entered upon following the same goal and/or performing the related operations. Now the system is searching for a way to use these raw information to find a certain attribute to distinguish circumstances resulting in failure from those resulting in victory because  in this case , the system will obtain the capability to indetify and anticipate failure or discovery for each situation. That is why it must concentrate its studies on the initial parts of the observations; that  is on a part of the observation the results and consequences of which have not appeared yet and tries to find a characteristic with the most capability to discriminate. For this reason, all the recalled observations with successful consequences are separated with those with unsuccessful consequences and divided into two groups. Then a machine with different examinations and study tries to realize the cause of the consequences by finding the differential attribute between these two groups so that by measuring the amount of that characteristic the success or failure of each observation can be anticipated before it continues. In this case, it can be said that the system has succeeded in interpreting that information and discover the characteristic which accounts for the cause of contradictions, and this is the formation of  a new concept  in the system .

Proposition cell structure:

By automatically putting the discovered characteristic or concept  inside the shell of a cell and connecting its antennas to suitable information resources, a proposition cell is made. This cell receives  the data by means of its antennas and then passes them from the function or the cell internal concept in its output shows an interpretation of the environmental situation in the form of a phasic number. The methodology of concept and/or function discovery has been explained under the title of heuristic procedure in next pages.

Installation of A Proposition Cell:

After a proposition cell is made, the output will show that amount of the relative profitability of the situation existing in the environment. This ready interpretation must certainly be in connection with the application place it should be said that a complex effort will not be needed to identify the application place because it was already the same place which has announced the existence of contradiction and caused this new cell to be made and reproduced. The information leaving this proposition cell will enter a new conditional node in the causality tree of  the mental model. This new conditional node will be added to the previous concept as a new child. And if the previous concept had been known as a condition for achieving top goals, from now on the advent of the profitability state at the top or father level will be conditioned on the receipt of the profitability from the son, and conversely lack of the receipt of profitability will be known as an impediments to the goal achievement. Removing this impediment can be posited as a new executive will and case which requires the genesis of new operations. At this time a purposive and practical experiment and trial on part of the system will emerge and it will probably result in manufacture of a new algorithm. And this algorithm will be always be available to the said goal as an instrument so as to be used , if required, for converting non-profitable state to profitable state in the environment.

One of the important advantages of this system is the gradual growth and possibility of the evolutionary history of incomplete and inaccurate notion which is formed as a statistical proposition at first and then enables the system to throw its rope of beliefs over far horizons and form the initial semen of each concept not only in the frame of a strong reason but with weak conjectures, too and take it to the maturity and expliaty stage. This is the very hermeneutic genesis through which the concepts lessen their antropy little by little to obtain a perfect clearness. A group of the concepts created by the abstraction of the partial propositions are detached and cause general propositions  and pure logical and mathematical  orders, and a group is destroyed in the dialectic struggle before reaching maturity because need and utilization cause the survival of a concept, but the internal contradictions and so many distortions in the meaning has caused its disintegation. Thus the acquired science is obtained unconsciously from the environment (Appendix I )

Sign: A symbol that carries a message inside itself, like the proportion of sound meaning of the speech and/or shap and syntax of words to the meaning of the written text and/or any symbol and physical parameter receivable from the environment which carries information

Meaning: An interpretation which is attributed to a sign which, regardless of the physical appearance, carries a concept which is applicable .

Utilization: Using the meaning in order to anticipate the route of the environmental changes in a few moments later which is so-called “perception of the governing situation and atmosphere”. In case of the situation perception with non-profitable consequences and possibility to make changes in the system profitability, a suitable reaction will be shown. These changes can also be executed as operation inside the mind environment and cause its better performance in the future . In brief, it is the existence of utilization which gives meaning to a sign.

Types of Proposition Calls: since there are two types of data order in the nature (series-parallel), so in this system the birth of two types of proposition have been anticipated.

Chronological Proposition Cell:

 This is for registering and realizing the series information again which was flowed at least once in the past and caused the birth of the same cell. These cells are able to realize the advent sequence of different date and concepts from different proposition cells. These cells cause the perception of the temporal unity of information and observation and becomes a good cue to discover the existence of the relationship between cause and effect and/or a suitable date basis to study , analyse and extract cause and effect and generate logical statements.

Repetition of the receipt of a concept causes the high stability of proposition and/or deepening of the memory, and, on the other hand, the relationship of the output proposition with the utilization place and meaning explicity causes strength and lack of semantic causes disintegration of the proposition( Appendix II ). The formation and destruction of such statements in the natural intelligent systems are called Pavolf’s conditional reflexes.

Compositive Proposition Cell: This is for registering  and realizing the parallel information entered from several input data whether continuous or discrete; that is, inside the cell, a topological space with the dimensions equal to the number of the cell antenna is formed. Any possible arrangement of the input composition can be known like a point in that space. This topological space has been divided to integrated limits ( Renge) with profitability and non-profitability, consequently the input of any composition will be valuable. Usually a discrete output is generated.  From another perspective, these cells can be interpreted as convertors of the continuous signals, (Polerization) (Appendix III).

Innate Proposition Cell: These cells can be of the chronological or compositive type, and the only difference of these cells is that they have been programmed and created already by a designer and can not be changed by the system, that is, their meaning has not been changed, and the profitability or non-profitability of the arrived date has been identified before the system start by the designer. For example, sweet taste can not be seen non-profitable and/or blue colour can not be seen red. All the innate values defined by the designer are called breast knowledge. It is these very values that prepare a ground based on which the system is able to interpret and define the profitability and non-profitability of other data and  place them in new cells . The innate proposition cells cause sensations upon the system birth when it is started and enable it to give meaning to and interpret the situation and atmosphere in which they lie and show innate reaction on the basis of these reactions. These cells indicate that the mind can not be a blank board upon birth but a half-written board and an ambiguous model contaminated with errors from the environment has been institutionalized in it.

Floating Proposition Cell: This is able to realize a certain characteristic from the region focused on . And the characteristic established in it can be either compositive or chronological type.  By giving address to this cell, it will be able to zoom its camera on any cell of the group members.

And in case of the desired characteristic existence in a part its address is reported and recorded in a member. This cell plays a fundamental role in extraction of the aspect common between concepts and induction and discovery of the logical orders and formation of the abstract and visual concepts.

Sensory & Algorithm Structure: As said earlier, the innate proposition cells are considered as basis and prime buds for the growth of the mind configuration which have been anticipated and institutionalized already by the designer. Their contents can not be changed and created.

The reproduction and binding of the new cells are carried out on them through a certain process and in accordance with the regulations and cause the development of the   sensory

configuration and semantic of the new data and that of simple innate algorithm which is harmonious with Freudian view. This process is explained in the following example, if the output of an innate proposition cell is supposed A, it can be lied in one of the value (A+) and/or anti-value (A-) situations. By value it is meant the same profitability and by anti-value it is meant the same non-profitability .If it is deviating from the value (A-), the system will start the innate algorithm and stimulate it as a work instrument (α). This instrument causes changes in the environment and consequentially in the concept of A and converts it from A- to A+ . But if the work does not go in the normal way the following stage will begin .

Cause of Development & How To Do it :

Whenever the instrument (α) stimulation does not create a desirable result because of any unpredicted change in the environment; that is, it does not take A- to A+, it can be said that the inefficiency of an instrument (driver) had been observed and a  contradiction   has  been  created  in  the system which denotes the insufficiency of the algorithm  and lack of adjustment of the internal statements of the system with the real conditions of the environment.

This contradiction causes a start of the machine which builds statements, and a new proposition cell like B will be made to show the pre-requisites required before implementation of the instrument (α) in  itself; that is, the stimulation of the instrument (α) depends on the receipt of the pre-requisites from  B. Thus it is seen that contradiction has caused the development of the system .

(Appendix IV).

The existence and inexistence of these pre-requisites in the data resources are  sensed by the antennas of the proposition cells which can show the profitability and/or  readiness of  the environmental conditions to stimulate the instrument (α).

Now if the environmental conditions are not suitable for work, no action must performed , and the system will wait for the improvement of the conditions. This property indicates the power of the system to evaluated feasibility as well as the phenomenon of hunting; that is, proceeding  to  do  an  action  in  the right

moment. On the other hand, the system will not remain idle and will try to find a new tool, by another machine, to change the concept B from the B- to B+ and will try the Hillclimbig available tools .

 We call this found tool (b). Now, if the tool(b), under special circumstances, could not change B- to B+, another expansion similar to the above stages will be made .

For example;

Let us suppose that conversion of A- to A+ (fruit eating ) is possible with the innate tool (α), and A has found a pre-requisite like B ( fruit touching ) and B has found a tool called (b) ( fruit picking), but when fruit is not available, the action of picking (b) will not cause it touching (B) and because of this contradiction, the algorithm making machine will start and present such a statement.

  If D + and C +   then   >  B + ; that means, if you go onto a table (C+ ) and raise you hand (D+ ), the action of fruit  picking will successful. This is a hypothesis which the proposition making machine has taken into account and presented to the tool (b) as pre-requisite . If, in practice, such an interpretation and idea comes out true, it  will be accepted as a truth and will remain in the sensory structure. (like natural choice in evolution theory ) .4-51

  Therefore , from the connection of the  minor statements, a pyramid-like sensory structure along with the operation algorithm will emerge. In this structure, each target (if found), will have its own tool with it; in other words, the stages and the sequence of the necessary operation will be apparent beneath itself for achieving each target ( Appendix V)

Heuristic Procedure:

The procedure of the production of a proposition by a proposition making machine takes place on the basis of the four causes of Aristotle. Of course, his words can only be assumed about the intelligent systems. He says, “four causes are necessary for anything to be made.”

1. Final cause 2.Material Cause 3.Formal Cause       4.Efficient Cause.

He gives this example that a silver sugar box is made

1. according to the need and application for keeping sugar in it ( final cause)
2. The existence of silver metal is needed to be shaped into a sugar box (material cause)
3. An idea or shape must be formed in the mind which seems useful( formal cause ).
4. There must be silversmith who can turn the material cause to the formal cause. (efficient cause) 8-7,12,13

In the system , too, presented in this article, the above mentioned stages are seen. The existence of failure shows a need for making a concept because inefficiency of the tool implies a contradiction between the environmental behavior and inner model of the system which creates a surprise pulse.3-88

89,90 , 1-22

1. Where a contradiction is created, a pulse will be produced which , in fact, sets forth a question and needs an answer which will be produced by the proposition making machine and sent to that place. This machine will consider that concept reaching a profitability status as a target ( final cause) A+.
2. This stage is to find an answer to this question” what parameter change has caused change in the environment and changed its behavior?” or “for what reasons failure has taken place?” One can guess that one of the same parameters which have caused changes in the environment upon failure has probably caused it. 2-42

Consequently, by referring to the memory and studying the  past observations, that is the contents of the chronological statements, one can find the factors effective in the target and have access to the date sources. This searching method is based on the contiguous principle. Thus the cause of failure-identification (material cause)-will be found. Now it the found parameter(s) are restored to the normal conditions, no failure will probably be resulted . But  what is the normal condition and where is its border line? The answer to this question must be given below.

3. Formalization: By this stage it is meant the determination of a acceptable form from among possible forms and shapes or universal compositions of the material cause sometimes instead of one form, group of them may be acceptable. Sometimes, the forms will be of a continuous spectrum with relative acceptability in terms of profitability, but a border must be demarcated for isolating profitability from non profitability. The width of the border line denotes the limits for the phasic space operation , and the two regions outside the line denote the existence of Aristotelian dual logic in the every concept. And among the composition, there will be a composition with  maximum acceptability which is called the ideal form or new desire (subject) .In this stage, the idea formation has taken place but the limits and quantity of the design coefficients have not been determined yet ( formal cause B+). This idea has not actually been implemented and changed to object and is like a hypothesis in the methodology of recognition . (Appendix II) 8-180
4. Instrumentation: In this stage, this question “By what means can turn B- to B+ is set forth. By searching, a suitable tool and testing it  the answer will be found. After finding  its place, it will be recorded for ever, and not search and test will be needed  in the future ( material cause) (b)

Now, after the above stages and practical test, the correctness of the theorem can be accepted; that is, actions can be taken to make a proposition cell. From now on, operations based on knowledge can be controlled, and it can be said that knowledge is power .2-34   4-62, 64   7-323   9-76

Review In More General Perspective & Substantiation of Compulsories of Evolutional Path in Mental Structure of This System :

Before reviewing, it is necessary to explain a  new concept called “Logical Feedback”.

Logical Feedback: There may exist various types of  feedback in systems with some of which we are familiar in the control engineering for example;

1. Positive Feedback Loop in which the system input is fed by the output through the same phase .
2. Negative Feedback Loop in which the system input is fed by the output with the phase difference 2Π.
3. By adding a comparator and reference, a new feedback is introduced to the automatic control engineering the aim of which is to control the value of the parameter which can return it to the level specified in the reference .

4. Figure Feedback : This type of feedback is used in the expert systems in which the designer does not mean to control the value but to maintain a cretain figure and quality . To do this, it is sufficient to establish and control certain mathematical relationships between the parameters. That is to say that a group of the certain compositions among universal compositions of those parameters has the desired figure and value in the designer’s opinion. I both of these feedbacks, the passage of the error signal means a deviation outside the system with an internal criterion, and the system struggles to change the outside environment and conform it with the internal criterion.

5. Logical Feedback: In this feedback, the error signal does not make any changes in the internal criterion and does not have any activities outside on the contrary to the above feedback. This feedback is modulated on the figure feedback and will work when the expert system, after trial and error, does not succeed in changing the environment and the signals received from it to transfer to the internal criterion, and the logical error feedback signal will appear and enter a upper machine which has the ability the make changes in the algorithm of the lower machine ( expert system) and will get a new algorithm issued to replace the previous one. Now after the new algorithm having been run by the lower system, if the signals received from the environment do not conform with the criterion it can be deducted that the algorithm designing machine ( the same proposition cell making machine in this article ) has been  successful in its task, and the sensory structure is more compatible with the environment behavior and found the ability to anticipate the environment behavior, but if the result were not desirable, the system must make a newer algorithm so that conformity can be maintained. As we saw, in this feedback, the error signal has caused changes in the structure of internal logical criterion.

Various feedbacks can be categorized in to three groups in terms of the passage of their error signals;

• Quantity Feedback
• Figure Feedback
• Logical Feedback

1. In the quantity feedback, the error signal informs whether the quantity is too much, too little or enough and causes changes and correction in the value.
2. In the figure feedback, the error signal informs of a deviation in the environment produced in comparison with the internal criterion. The deviation is interms of the form of the combination of the quantities not their quantity. For example, the identification of the alphabet is done without attention to their dimensions.
3. In the logical feedback, there is no news about the quantity of deviation from the criterion nor that from the criterion and reference, but the error signal is resulted from the victory or failure in a certain target, and the signal meaning reports the truth or falseness of the inner structure of the system itself; that is, is the function of the proposition and logic inside the system, regardless of what it is, able to succeed in data processing. Consequently, transformation and restoration of logic structure and correction of the internal criterion of the lower system becomes necessary and it is evident that such a signal can not carry the quantities of deviation. That is, this is a critique against the inner structure.

Type of                        Meaning of                             Type of         Ability of System

Feedback    Sensor      Error Signal                           System          to apply error signal

Quantity      Quantity   Shows Deviation in Quantity    Command-    Able to correct

                                    oriented                                   quality

Quality        Quality     Shows Deviation in Quality      action-           Skilled in Conformation

                                    oriented                                   with quality

Logical        Quiddity   Informs of problem in              Reason-         able to change inter-

                   logical model structure                             oriented         pretation of data & logical

                   of internal criterion                                                        structure & to conform

                                                                                   with environment

If we assume the corrective actions by the upper machine on the inner structure equivalent to positive and/or negative mutation I the evolution theory, these corrections in the concepts and algorithm will remain in the sensory structure only when they are confirmed by the logical error signals ( principle of natural choice), and they are confirmed provided their profitability in achieving the target has been tested in practice, consequently, the concepts and algorithms will be driven towards becoming logical and compatible with the governing relationships in the environment .

    Since the survival of a new concept in this system depends on its compatibility with the observations, the emergence of a mental model, even inaccurate, but compatible with the environment seems necessary, and it will be the critiques against this mental model and inaccurate subjects that cause the correction of its mental models and subjects , and the system itself will be responsible for identifying the parameter effective in the victory and failure, which is the same concept of cause discovery and determining the optimal quantity of profitability, but in the quantity and the quality feedback, the designer himself is responsible for identifying and selecting a parameter as a reference and the quantity of the reference, and the system only has to correct the environment and produce object and tries to conform it with the pattern and does not have any ability to discover the cause. And finally the skill of the expert system can be increased  trough some methods, but it is not able to make and produce new subjects.

Care should be taken that the nature and function of such a system is fundamentally different from the current systems philosophically because this system guesses the profitability of the situation in the environment before any action and trial and error and tries to  render it applicable only after the target is specified. In other words, the target is created before action which indicates the priority of idea to work. In this system,  searching for a correct way, on the contrary to the current systems, is not based only on  action, trial and error and blind test but based on the birth of an idea, which has been created as a result of analysis and reasoning, though not clear, by the system itself, before starting to do any actions and tests. And , in fact, the critique of the system against itself about the mechanism of idea generation, analysis algorithm, thinking and/or algorithm  of the operations will be acceptable.

Also we should notice that other current systems, are controllers; namely, they have been designed to maintain a predictable situation or some situations. In the event that this system tries to do different new tasks, because of its ability to create new ideas, they have not been designed to maintain a certain and predictable situations. And in case a control is needed during the task performance, the system it self will be capable of creating it.

Work: Purposeful performance of an action and/or a range of operations which only starts to create some changes in the environment in order to reach a new stable and valuable situation, and it is not necessary for the system to have an activity and control so as to keep and maintain the generated value because the environment will keep that action as a memory outside the system. Work is a range of operations which is established as a pre-requisite to be implemented in the next situations of the environment by the system. For example, to release a stone is some work which can be done by a system while it does not have any controls on its execution. A seemingly simple operation like throwing a stone can be analysed into tens of work each of which is used at a particular time and released then.

Control: A continuous activity by the system to protect a certain situation. And whenever that activity is stopped, the desired situation, too,  will be destroyed . The real intelligent systems must not be introduced as the control systems because they have the ability to perform a completely new task and regulate the identification of the priority and posteriority in the operations is based not on trial and error and skill acquisition but on the discovery of the cause priority to effect according to its beliefs from the environment. Thus, these systems have freedom of volition because if the reactions and algorithm had been anticipated and written by the designer the freedom of this system in selecting the way does not have a meaning any longer. And even if a way is chosen randomly, again it can not be either considered a part of intentional and voluntary tasks of the system nor called a decision along with awareness and freedom. But if a way based on the discretion of the  system itself is chosen, neither the designer nor probability can have a role in it, but it will be  decision based on the system knowledge. And it, only in this case, can be considered an international decision based on the awareness of the system itself.

Intricate Logical Complex:

Another logical feedback loop can be mounted at the top of the algorithm correcting (proposition making) machine so that the algorithm correcting machine can be considered as the lower machine, and its operations can be criticized to make some corrections in the algorithm correcting  machine itself. In this case, the higher machine will be seeking for these change in the lower machine so that machine will incur less failure. Thus it must try to identify the factors effective in the failure of the lower machine and introduce them to it, so to the state of those parameters can be paid attention. Therefore, the upper machine will get experiences in presenting solutions in different states, and must obtain the sensory ability to anticipate consequences of each change it creates in the algorithm of the lower machine, and choose the most suitable changes in the mechanism and reaction logic of the lower machine by identifying the failure factors and taking into account the conditions. Thus the lower machine will increase its skill and will not have a fixed reaction logic but a dynamic and flexible state which is controlled from a top layer. But in the case that the top layes do not work, the system will not remain without reactions and will show primary reactions. Apart from this logical feedback which is a common type there is another special type of intricate concept that will be discussed later.

What is the criterion to identify the lower layer when it is necessary to refer to its higher layer?

Whenever a lower cell faces status of non-definiteness in measuring and estimating the provided quantity of the input concept; that is, it lies on the border line of value and anti-value in a way that there is not confidence in the truth of the estimation; in other words, whenever the perceptional status of the cell lies in the doubtful region it will refer to the layer higher than itself before giving out put so that the top layer will have access to the more processed date and more sophisticated concepts and will supply each of the doubtful regions with definitions of the value and anti-value quantity by using a more common logic and take the lower layer out of suspension. It is seen that the function logic and values of the lower layers are influenced temporarily by the top layer, and value and anti-value will be defined to it afresh.

Common & Special Intricate Feedback :

The common intricate technique has been used in all the sensible creatures like animals and insects, but only in human being the special intricate technique has appeared and made him intellectual.

In the common intricate technique, there are some operations outside the system under the supervision of the top layers and cause them to evolve , but the special intricate technique controls the manner of the internal operations of the system through another similar mechanism in addition to this mechanism. In fact, it proceeds to create necessities and equipment of thinking and turn its evolutionary laws from statistic state to dynamic state. To put it other way, in the common technique, the best operation logics were selected, but in the special technique, the best logics will be selected for problem solution, correct thinking and reasoning methods.

In addition, as a result of the addition of the higher feedback layers, specific capabilities will emerge in the system, and it is to discover and obtain features and laws governing how to control the lower layers. Hence, when we go towards the top layers, we will observe more common, abstract and inclusive concepts and nomic proposition so that we will reach the layers with capability of discovering formal logic.  10-311   6-94

The more common laws are discovered, the less the probability of the falseness of the new statements issued by the proposition making machine. In other words, the probability of  choosing positive mutations rather than the negative mutations will increase. And this shows the evolution any path of the system towards becoming meta-experimental. Also the discovery of the reasoning routines causes the issuance of the more complex and logical commands by the system , and it is only in this case that the system has made decisions based on its knowledge, and they have not been built in by the designer. Therefore the way for compulsory evolution has been traveled. Having the ability of each element of the system to reproduce shows the  Fractal of  its structure. 4-56

Institutionalization of Innateness:

The type of the concept generated inside each proposition cell depends on the type of the value defined n the previous cell and extremely on the type of the value institutionalized in the innate cell because only the concept which will be involved in supplying that target will develop. Therefore, type of the concepts formed inside each mental structure depends on the type of the criterion postulated in the primary cell as well as the logic flowing in that environment .

For example, if entering much more pressure on the joints or body surface of the robot is institutionalized as an anti-value inside an innate cell, a particular algorithmic structure which is trying to prevent the entry of those much more pressure will be generated and consequently provides the system with the institution of self-protection .

Thus we can institutionalized any desirable criterion in the innate cells so that the system can evolve that way and possess different features , and each innate cell can cause  the growth and evolution of the machines with different institutions. |For example, if we introduce the assignment of meaning as a value for innate cell, the system will possess the attribute of curiosity, and finding inexpensive ways compared to the previous ways to a cell is considered as a value , a system with the attribute of economy will be created .

Conclusion :

As seen earlier , as a result of predominant compositions during the encounter of the system with the complex conditions, they can not be anticipated in advance and formulated upon designing and nor a general method can be built in the analyse , infer and identity them, but the system must provide the ability to assign meaning, interpretation from the objects to the extent which meets the needs and conform with its objectives and complexity level of the concepts. The mechanism of this action is equivalent to thinking as knowledge theory. The meaning assignment and recognition happens in a system when it can infer the governing and current relationship in the environment gradually . Therefore, in this article, a new method for identifying the environment conditions by the system has been presented which is based neither on trial and error and blind search nor by using the statements built in the system by the designer which have been acquired from the environment by the human knowledge and directed to the system, but a new  and practical method for How to create the ability to make the statements has been made by the system itself, and how to design a system with real knowledge was expressed and it was shown that a system that has the key to acquire knowledge and recognize the environment and has the ability to repair the mental tissues and develop logic in them can be created.

This is the same knowledge by which animals can regulate their behaviors without engineering calculations and using physical formulae, mathematical and logical formal statements. For instance, they can control their body gravitation well facing different conditions by receiving and analyzing the environmental information .  2-48,18  5-122   9-65

References :

• PHILOSOPHICAL ESSAYS
  Bertrand Russell, O.M., F.R.S
  First published in 1910
  Revised edition (C) George Allen & Unwin Ltd., London,1966
  Translated by Dr. Mir Shamseddin Adib Soltani, Amir Kabir Publications-Tehran (1984)

140093796. General Recongition of Science
           Bronovosky Jacob Translated by Dr.Mohammad Alipour Abdollah, Astan Ghods Razavi, Mashad (1991)
140093797. Image of World In Modern Physics
           Plank Max ,Translated By Morteza Saber, Pocket Book Stock Company- Tehran (1969)
140093798. The Theory of Knowledge
           Fort, M.K.
           Translated By Farhand Nomani and Manouchehr Sanahian, Amir Kabir Publications-Tehran (1976)
140093799. The Philosophy of Sciences
           Capaldi Nicholas,
           ISBN;964-435-039-1
           The Historical Development of Scientific Concepts and their Philosophical Implicants
           Translated By Ali Haghi, Sorush Publication – Tehran (1998)
140093800. Cybernetic Today
           M. Makarov ,Translated By Fatemehsadat Rafiei, Roz Publications-Tehran (1971)
140093801. Modern Hermeneutics
           Selected Essays
           Nietzche, Friedrich Wilhelm
           Translated and Edited by :
           Babak Ahmadi, Mehran Mohajer & Mohammad Nabavi
           ISBN 964-305-416-0
           Markaz Publications-Tehran (1998)
140093802. The Philosophy of Technology
           Heidegger and Others
           ISBN;964-303-411-x
           Translated By Shapour Etemad, Markaz Publications-Tehran(1998)
140093803. An Introduction to Philosophy
           Hermes John & Bacler Jastos,
           Translated by Amir Jalaleddin Alam , Soroush Publications-Tehran (1996)
140093804. An Introduction to Philosophy of Science
           Rudolph Carnap
           Translated into Persian By Yusef Afifi
           Editor Martin Gardner
           Second edition : Sprin 1994
           Niloufar Publications-Tehran (1994)

Appendix I

Concepts have crookednesses in some respects, during to genesis history as follows;

• Logical correctness: The cell synopses have not been connected to the completely exact places, so logically they generate meanings close to reality .
• Impurity: Sometimes the feature and criterion built in the cell for realization and isolation of the input compositions have wrong and reverse result on some of the compositions; that is, the realizing feature does not have the same correctness on all the compositions, and causes error occasionally.
• Ambiguity: In some of the compositions , it is unable to identify and respond.(Theorem is not widely accepted).
• Doubt: There is no precise borders for isolating profitable and unprofitable compositions, and in some of the certain compositions, possible behavior can be anticipated. ( phase logic)
• Accuracy: This occurs when the output is in the form of scalar number
  When the exact quantity of the output can not be guessed.
  B. When the input composition can not have the sufficient ability of isolation and  resolution for use.

1. The accuracy of the output is different from different compositions.

The above complications can also be found in scientific theorem and/or artistic works There are techniques to remove each of them, but they can not be discussed in details depending on the article subject .

Appendix II :

Cells Being Generated & Combined :

These cells , in fact, carry statistical statements and in their output, they do not show precision in meaning or an exact isolation of profitability from non-profitability  is  impossible  around  the border line and produces and ambiguous response in the output. The system tries to make the width the range of doubtful an ambiguous compositions narrower as much as possible  and  approach  them  to the Aristotelian dual discrete logic and pure concepts. Therefore, the input status of such cells will fall into one of these three states; possible , impossible or on the phaseic band.8-121

To explain the above cases, an example has been shown in one-dimension space.

• Range of profitable compositions and/or limit of meaning A+
• Range of non-profitable compositions and/or limit of anti-meaning A-
• Range of border line and/or limit of doubt and ambiguity ( phase band)

Appendix III :

For example, the some of five proposition cells are presented below geometrically in two-dimension space . Each has two dendrites called M and N as two scalar continuous inputs and can show the appropriateness or non-appropriateness of the environmental conditions in their outputs based on the input composition. This space is, in fact, able to create boolean logic logics from the continuous quantities of the input and change them to discrete quantities in their output or convert multi–dimension concepts to one dimension and do the meaning  purification .

Such topological space , in addition to the presentation of the bull logics and mathematical functions, they are able to present more complex logics which can not be expressed through mathematical and logical function because the shape of the profitable and non-profitable ranges have been formed as a result of real experimentation and observations and depend on the relationships among the parameters in the environment.

The quantities from M and N will show a coordinate in two-dimentional space, and consequently the value related to it will appear in the Axon output. And in the case that the dendrites are a lot in a cell, a topological space will be created which has discrete output.

Appendix IV:

The evolution of mind can take place in two directions;

1. Mental model or ability to interpret data and make concepts and/or changes in world view .

2.To develop tool making or skill of how to execute operations, there is a certain method in each of them for identifying the necessity of development which can not be included in this article. Finally, care should be taken that development in each of the above forms will bring about changes in the system behavior.

Appendix V:

It should be noted that wherever in the mind there is a need for making a concept if the similar shape of that concept has been used in another pyramid shape structure, the system can use them and it is enough to be junction between that place and place of need. The use of one pyramid by another pyramid is shown below .

One of the other advantages of this structure is its ability to localize the shapes because it has a Fractal Structure the quantum of which consist of the layer of concept and tool; that is, when there appears a contradiction in each of the concepts, growth and development, too, will take place in the same place and relate with the source of data .

END