Home Pulpitis Why the use of models affects the limits of applicability. Subjective aspects of the application of mathematical modeling of military operations in the work of military command and control bodies

Pulpitis

Why the use of models affects the limits of applicability. Subjective aspects of the application of mathematical modeling of military operations in the work of military command and control bodies

MILITARY THOUGHT No. 10/2011, pp. 49-53

ColonelO.V. TIKHANYCHEV ,

Candidate of Technical Sciences

TIKHANYCHEV Oleg Vasilievich was born on October 30, 1965 in the city of Shuya, Ivanovo region. Graduated from the Kazan Higher Military Command and Engineering School (1988), Mikhailovsky Artillery Academy (1997). Served as a platoon commander, deputy battery commander in the GSVG and North Caucasus Military District. Since 1997 - in the 27th Central Research Institute of the Ministry of Defense of the Russian Federation in positions research fellow, head of department, leading researcher of the research department.

In 2005 he defended his thesis for Candidate of Technical Sciences. Author of more than 100 scientific papers. Professor at the Academy of Military Sciences.

ANNOTATION. The experience of developing mathematical models for automated control systems and the use of prototypes of mathematical model programs at operational training events is analyzed. The need to improve the procedure for developing mathematical models in order to reduce the influence of subjective factors on the effectiveness of their application is substantiated.

KEYWORDS: mathematical modeling, procedure for developing models, experience in operational training activities, objective and subjective factors, improving the organization of model development.

SUMMARY. The author analyzes the experience of developing mathematical models for automated control systems and application of software prototypes of mathematical models for operational training activities. The necessity of improving the procedure of development of mathematical models is reasoned to reduce the influence of human factors on the effectiveness of their application.

KEYWORDS: mathematical simulation, modeling procedure, experience of operational training activities, objective and human factors, improving the modeling organization.

IN MODERN CONDITIONS priority direction of reforming the Armed Forces Russian Federation is to increase the efficiency of their use, including through the automation of command and control of troops (forces). Automation of command and control of troops (forces) is understood as the process of equipping headquarters, command posts and combat complexes with electronic means. computer technology and their use in the work of governing bodies.

The intellectual component of the complex of automation tools for the automated troop control system (ATCS) is software, which is divided into general, system-wide and special. Special software (SPO) of automated control systems consists of calculation, information problems and mathematical models. The latter play a significant role in the process of planning operations (combat actions) and command and control of troops (forces), providing forecasting of the development of the situation and a comparative assessment of the effectiveness of decisions made.

The article “Modeling Armed Confrontation: Development Prospects” examined a number of important aspects of the application of mathematical modeling in military affairs. But subjective factors remained behind the scenes, although in practice they have a significant impact on the use of mathematical modeling in the process of organizing operations (combat operations). Subjective reasons The limited use of mathematical modeling in the practical work of headquarters has not received adequate coverage in subsequent publications related to mathematical modeling. Thus, in the article “Problems of automation of intellectual decision support for combined arms commanders at the tactical level” it is noted that mathematical models should be the most important component of automated control systems, but they have not found wide application in the process of making a decision to fight and managing it. Why this happened is not specified. The disadvantages of existing models and objective technological factors that hinder the use of mathematical modeling are mainly considered. Subjective reasons are mentioned in passing.

At the same time V military field, characterized by fierce confrontation and high personal responsibility of the decision maker, the presence of a subjective factor is not just inevitable, but also a natural phenomenon. In conditions of incomplete information, experienced commanders (chiefs) are able to formulate the right decisions on an intuitive level. At the same time, they usually proceed from their subjective ideas about the importance various criteria optimality and effectiveness of possible alternatives to decisions made. This is what often gives rise to subjective rejection of the results of mathematical modeling, which can ultimately lead to serious errors in planning and combat control.

Thus, the presence of subjective factors hindering the use of mathematical modeling in military affairs - real fact, requiring reflection and taking appropriate measures.

What specifically determines cases of subjective rejection of the use of mathematical modeling by officials of military command and control agencies (MCA)? There are many reasons, and they manifest themselves both at the development stages and at the stage of using mathematical models.

The main reasons for the rejection of any innovation, as psychologists say, are a lack of understanding of its essence, ignorance of its features and the inability to apply it.

The existing procedure for using open source software implies that the official - user of the automated control system reliably knows the limitations and assumptions adopted during the development of open source software, and the limits of applicability of mathematical models from the open source software. It is within these boundaries that inspections and tests of open source software elements are carried out to confirm its performance and adequacy. This fully applies to mathematical models as an integral part of open source software. Theoretically, officials of the administrative authority who use SPO components in their practical activities must understand the limits of applicability of the mathematical model by carefully studying the operational documentation for the components of the software. Understand, remember and always be guided by them. Unfortunately, this ideal situation is not always realized in practice, primarily due to the imperfection of the organization of the learning process officials OVU work on automation equipment.

Another problem is the problem of dividing responsibility for decisions made between the user of the model and the developer of its mathematical apparatus. If in technical systems The division of responsibility for operating errors between the developer and the user is prescribed in the relevant GOSTs and technical regulations, but for software there are no such documents yet. The high degree of responsibility of officials of the educational institution for the results of their activities, coupled with an uncertain understanding of the limits of applicability of models, gives rise to certain concerns among officials when using mathematical modeling in planning practice real operations(combat). Without solving this problem, it is impossible to ensure the full use of mathematical modeling in the practice of operating a device.

Significantly influences the introduction of mathematical modeling into practice OVU is the irrationality of the interface layout of mathematical models created by industry. IN At present, insufficient attention is paid to this aspect when developing programs. Engineering psychology and ergonomics do not add optimism: they deal primarily with operator operating modes and workplace equipment, but not with the quality of program interfaces.

At the same time, with the development information technologies, increasing the capabilities of computer technology, people are increasingly becoming the link slowing down decision-making in automated control systems. And the reason here is the program interface, which slows down both the process of entering initial data and the analysis of modeling results. After all, the interface is the main element of communication between the user and the program. Often, it is the convenience of the interface that determines whether the user will turn to the program at critical moments and whether he will be able to quickly carry out calculations and analyze their results.

It’s bad that creative and “piecemeal” work on creating program interfaces and developing approaches to unify them, which can only be performed by a specialist with a broad operational and technical outlook, does not relate to scientific activity at all. At the same time, the lack of unified approaches to the interface implementation of mathematical models and information and calculation problems significantly reduces their user properties and makes it difficult for officials to master and implement them into the activities of educational institutions.

In accordance with the governing documents, two categories of developers take part in the creation of interfaces for models and tasks from the automated control system software: employees of the National Research University of the Ministry of Defense, leading military-scientific support for the creation of automated control systems, and software developers at industrial enterprises. All of them are at least experts in the use of computer technology. But these skills can also play a negative role. The specialist unconsciously creates a model interface “for himself”, and not for a staff officer who works under severe time pressure and is a specialist in the military field. And the logic of a programmer is often different from the logic of an ordinary person. No wonder they joke that normal person believes that there are 1000 bytes in a kilobyte, and the programmer is sure that there are 1024 grams in a kilogram. As a result of these differences, the simplicity of the interface during development is often sacrificed for the sake of some additional qualities and capabilities that seem necessary to the programmer. As a consequence, there are difficulties in mastering the interfaces of models and tasks by officials of the educational institution, and a reluctance to work with them when solving practical problems.

The negative impact of this factor can be eliminated only by changing the existing procedure for developing SMPO, ensuring closer participation V process of developing the end user mathematical model. For this purpose, it is advisable to introduce a mandatory stage (stages) of trial operation of open source software elements V mock-up execution with the involvement of officials of the OVU. Based on the results of the stage, it is necessary to provide for the refinement of SPO elements V parts of the program interface organization. By the way, global experience in software development shows that any technology used (cascade, spiral or breadboard) necessarily contains a prototyping stage, based on the results of which the software, including its interface part, is finalized.

It is also important personal attitude of each official to the results of mathematical modeling. This attitude can be expressed in a general distrust of the results obtained using an unknown mathematical apparatus, and is formed during “communication” with models. The last one deserves special attention.

It’s no secret that sometimes ODU officials, dissatisfied with the modeling results, try to different ways correct them. A user (operator) who knows the model well can “play” various factors to influence the results V the right side. When he becomes a decision maker, he has the opinion that the model can show any result, if only there is a desire. This opinion is deeply erroneous and arises from ignorance of the features of mathematical modeling. Yes, the simulation result can be slightly adjusted by changing any of the initial conditions for organizing the actions of opposing groups, which fall into the category of uncertain and selected by the operator within established boundaries. But it is impossible to falsify the results without changing the original data, especially if the model is used for comparative analysis options for the use of troops (forces), all other things being equal. The results themselves may change, but the model will still show the correct trend in the situation.

An approach To the resolution of this situation, in our opinion, is the same - involvement of officials in the development of mathematical apparatus, which is embedded in the SMPO created to automate them activities. First of all, this relates to the formalization of the simulated process and the formation of a system of tolerances and restrictions.

Involving officials of the educational institution in the development of SMPO, in particular for describing the apparatus of mathematical models, is not an easy path. This requires certain efforts from the customer and industry, not only technical, but also organizational, and sometimes educational plan. But the practical experience of such work available at the 27th Central Research Institute of the Ministry of Defense testifies to the effectiveness of this method. The development of a number of methods for operational calculations together with officers of the internal affairs department showed that subsequently software tools that implement the jointly created mathematical apparatus are perceived by officials much better. Knowledge of the mathematical apparatus used in software and the limits of its applicability ensures confidence in the modeling results.

Thus, the analysis of subjective factors that interfere with the use of mathematical modeling in the practical work of educational institutions shows that the existing shortcomings are systemic. They do not depend on the specific open source software developer and the approach he has chosen to create automated control system open source software: functional, structural or process. To eliminate them, it is necessary to change the procedure for both creating mathematical models, introducing mandatory stages that provide for the participation of future users of the models in their development, and the procedure for preparing officials of the educational institution to work with them.

Besides, It is worth dwelling on one more subjective factor of distrust in mathematical modeling, arising in cases where industry representatives unreasonably often modify mathematical models or try to implement them where there is no objective need for this.

Analysis foreign experience shows that the most acceptable is the gradual increase in the capabilities of mathematical models through their modernization without radically altering the mathematical “core” and, of course, the use of mathematical modeling for planning operations (combat actions) only where it is really necessary, where there are conditions for this. Unfortunately, with us everything often happens exactly the opposite. Unreasonably frequent modification of models, the extension of mathematical modeling to areas where it is not applicable (for example, to the “battalion - company (battery) - platoon" level), subjectively reduces confidence in the process of using models when planning military operations, and discredits the very idea of mathematical modeling.

So, in order to reduce the negative impact of subjective factors on the use of mathematical modeling in the practice of the control unit, it is necessary to increase the knowledge and skills of SMPO users and overcome the reluctance of developers to take into account their requirements (to overcome the automated control system under the firm guidance of the customer, with the help of the control system and organizations providing military-scientific support works).

To do this you need:

improving the procedure for developing mathematical models, including in the development process the mandatory stages of prototyping and testing of models in the educational institution; change in attitude (increased attention) to the creation of software interfaces for mathematical models from the automated control system;

adjustment of guidelines that define the content of the stages of development of mathematical models;

optimization of the process of training officials who use mathematical models as part of open source software for control point automation kits.

The implementation of these measures will allow mathematical modeling to take its rightful and appropriate place in the process of organizing operations (combat actions) and command and control of troops (forces).

Military Thought. 2009. No. 7. P. 12-20.

Military Thought. 2009. No. 9. P. 43-53.

Foreign military review. 2006. No. 6. P. 17-23; 2008. No. 11. P. 27-32.

To comment you must register on the site.

Victor Kuligin

Disclosure of content and specification of concepts should be based on one or another specific model of the mutual connection of concepts. The model, objectively reflecting a certain aspect of the connection, has limits of applicability, beyond which its use leads to false conclusions, but within the limits of its applicability it must have not only imagery, clarity and specificity, but also have heuristic value.

The variety of manifestations of cause-and-effect relationships in the material world has led to the existence of several models of cause-and-effect relationships. Historically, any model of these relationships can be reduced to one of two main types of models or a combination of them.

a) Models based on a time approach (evolutionary models). Here the main attention is focused on the temporal side of cause-and-effect relationships. One event – “cause” – gives rise to another event – “effect”, which lags behind the cause in time (lags). Lag is a hallmark of the evolutionary approach. Cause and effect are interdependent. However, reference to the generation of an effect by a cause (genesis), although legal, is introduced into the definition of a cause-and-effect relationship as if from the outside, from the outside. It captures the external side of this connection without deeply capturing the essence.

The evolutionary approach was developed by F. Bacon, J. Mill and others. The extreme polar point of the evolutionary approach was the position of Hume. Hume ignored genesis, denying the objective nature of causality, and reduced causality to the simple regularity of events.

b) Models based on the concept of “interaction” (structural or dialectical models). We will find out the meaning of the names later. The main focus here is on interaction as the source of cause-and-effect relationships. The interaction itself acts as a cause. Much attention Kant took this approach, but the dialectical approach to causality acquired its clearest form in the works of Hegel. Of the modern Soviet philosophers, this approach was developed by G.A. Svechnikov, who sought to give a materialistic interpretation of one of the structural models of cause-and-effect relationships.

Existing and currently used models reveal the mechanism of cause-and-effect relationships in different ways, which leads to disagreements and creates the basis for philosophical discussions. The intensity of the discussion and the polar nature of the points of view indicate their relevance.

Let us highlight some of the issues being discussed.

a) The problem of simultaneity of cause and effect. This is the main problem. Are cause and effect simultaneous or separated by an interval of time? If cause and effect are simultaneous, then why does the cause give rise to the effect, and not vice versa? If cause and effect are not simultaneous, can there be a “pure” cause, i.e. a cause without an effect that has not yet occurred, and a “pure” effect, when the action of the cause has ended, but the effect is still ongoing? What happens in the interval between cause and effect, if they are separated in time, etc.?

b) The problem of unambiguity of cause-and-effect relationships. Does the same cause give rise to the same effect, or can one cause give rise to any effect from several potential ones? Can the same effect be produced by any of several causes?

c) The problem of the reverse influence of an effect on its cause.

d) The problem of connecting cause, occasion and conditions. Can, under certain circumstances, cause and condition change roles: the cause becomes a condition, and the condition becomes a cause? What is the objective relationship and distinctive features of cause, occasion and condition?

The solution to these problems depends on the chosen model, i.e. to a large extent, on what content will be included in the initial categories of “cause” and “effect”. The definitional nature of many difficulties is manifested, for example, in the fact that there is no single answer to the question of what should be understood by “cause.” Some researchers think of a cause as a material object, others as a phenomenon, others as a change in state, others as an interaction, etc.

Attempts to go beyond the model representation and give a general, universal definition of the cause-and-effect relationship do not lead to a solution to the problem. An example is the following definition: “Causation is such genetic link phenomena in which one phenomenon, called the cause, in the presence of certain conditions, inevitably generates, causes, brings to life another phenomenon, called the effect.” This definition is formally valid for most models, but without relying on the model, it cannot solve the problems posed (for example, the problem of simultaneity) and therefore has limited theoretical-cognitive value.

When solving the problems mentioned above, most authors tend to proceed from the modern physical picture of the world and, as a rule, pay somewhat less attention to epistemology. Meanwhile, in our opinion, there are two problems here that are important: the problem of removing elements of anthropomorphism from the concept of causality and the problem of non-causal connections in natural science. The essence of the first problem is that causality as an objective philosophical category must have an objective character, independent of the cognizing subject and his activity. The essence of the second problem: should we recognize causal connections in natural science as universal and universal, or should we consider that such connections are limited in nature and that there are connections of a non-causal type that deny causality and limit the limits of applicability of the principle of causality? We believe that the principle of causality is universal and objective and its application knows no restrictions.

So, two types of models, objectively reflecting some important aspects and features of cause-effect relationships, are to a certain extent in contradiction, since they solve the problems of simultaneity, unambiguity, etc. in different ways, but at the same time, objectively reflecting some aspects of cause-effect relationships , they must be in mutual connection. Our first task is to identify this connection and refine the models.

Limit of applicability of models

Let us try to establish the limit of applicability of evolutionary type models. Causal chains that satisfy evolutionary models tend to have the property of transitivity. If event A is the cause of event B (B is a consequence of A), if, in turn, event B is the cause of event C, then event A is the cause of event C. If A → B and B → C, then A → C. Thus In this way, the simplest cause-and-effect chains are formed. Event B may act as a cause in one case, and as a consequence in another. This pattern was noted by F. Engels: “... cause and effect are representations that have meaning, as such, only when applied to a given individual case: but as soon as we consider this individual case in general connection with the entire world as a whole, these representations converge and intertwine in the representation of universal interaction, in which causes and effects constantly change places; what is a cause here or now becomes an effect there or then and vice versa” (vol. 20, p. 22).

The transitivity property allows us to carry out detailed analysis causal chain. It consists of dividing the final chain into simpler cause-and-effect links. If A, then A → B 1, B 1 → B 2,..., B n → C. But does a finite cause-and-effect chain have the property of infinite divisibility? Can the number of links in a finite chain N tend to infinity?

Based on the law of the transition of quantitative changes into qualitative ones, it can be argued that when dividing the final cause-and-effect chain, we will be faced with such content of individual links in the chain that further division will become meaningless. Note that infinite divisibility, which denies the law of the transition of quantitative changes into qualitative ones, Hegel called “bad infinity”

The transition of quantitative changes into qualitative ones occurs, for example, when dividing a piece of graphite. When molecules are separated until a monoatomic gas is formed, the chemical composition does not change. Further division of a substance without changing it chemical composition is no longer possible, since the next stage is the splitting of carbon atoms. Here, from a physicochemical point of view, quantitative changes lead to qualitative ones.

The above statement by F. Engels clearly shows the idea that the basis of cause-and-effect relationships is not spontaneous expression of will, not the whim of chance and not the divine finger, but universal interaction. In nature there is no spontaneous emergence and destruction of movement, there are mutual transitions of one form of motion of matter to others, from one material object to another, and these transitions cannot occur otherwise than through the interaction of material objects. Such transitions, caused by interaction, give rise to new phenomena, changing the state of interacting objects.

Interaction is universal and forms the basis of causation. As Hegel rightly noted, “interaction is a causal relation posited in its full development.” F. Engels formulated this idea even more clearly: “Interaction is the first thing that appears to us when we consider moving matter as a whole from the point of view of modern natural science... Thus, natural science confirms that... that interaction is a true causa finalis of things. We cannot go further than the knowledge of this interaction precisely because behind it there is nothing more to know” (vol. 20, p. 546).

Since interaction is the basis of causality, let us consider the interaction of two material objects, the diagram of which is shown in Fig. 1. This example does not violate the generality of reasoning, since the interaction of several objects is reduced to paired interactions and can be considered in a similar way.

It is easy to see that during interaction both objects simultaneously influence each other (reciprocity of action). In this case, the state of each of the interacting objects changes. No interaction - no change of state. Therefore, a change in the state of any one of the interacting objects can be considered as a partial consequence of the cause - interaction. The change in the states of all objects in their totality will be full investigation.

It is obvious that such a cause-and-effect model of the elementary link of the evolutionary model belongs to the class of structural (dialectical). It should be emphasized that this model does not reduce to the approach developed by G.A. Svechnikov, since under investigation G.A. Svechnikov, according to V.G. Ivanov, understood “... a change in one or all interacting objects or a change in the nature of the interaction itself, up to its collapse or transformation.” As for the change of states, this is a change in G.A. Svechnikov classified it as a non-causal type of connection.

So, we have established that evolutionary models as an elementary, primary care contain a structural (dialectical) model based on interaction and change of states. Somewhat later we will return to the analysis of the mutual connection of these models and the study of the properties of the evolutionary model. Here we would like to note that, in full accordance with the point of view of F. Engels, the change of phenomena in evolutionary models reflecting objective reality occurs not due to the simple regularity of events (as in D. Hume), but due to the conditionality generated by interaction (genesis ). Therefore, although references to generation (genesis) are introduced into the definition of cause-and-effect relationships in evolutionary models, they reflect the objective nature of these relationships and have legal basis.

Fig. 2. Structural (dialectical) model of causality

Let's return to the structural model. In its structure and meaning, it perfectly agrees with the first law of dialectics - the law of unity and struggle of opposites, if interpreted:

– unity – as the existence of objects in their mutual connection (interaction);

– opposites – as mutually exclusive tendencies and characteristics of states caused by interaction;

– struggle – as interaction;

– development – as a change in the state of each of the interacting material objects.

Therefore, a structural model that relies on interaction as a cause can also be called a dialectical model of causality. From the analogy of the structural model and the first law of dialectics, it follows that causality acts as a reflection of objective dialectical contradictions in nature itself, in contrast to the subjective dialectical contradictions that arise in the human mind. The structural model of causality is a reflection of the objective dialectics of nature.

Let's consider an example illustrating the application of a structural model of cause-and-effect relationships. Such examples, which are explained using this model, can be found quite a lot in the natural sciences (physics, chemistry, etc.), since the concept of “interaction” is fundamental in natural science.

Let us take as an example an elastic collision of two balls: a moving ball A and a stationary ball B. Before the collision, the state of each ball was determined by a set of attributes Ca and Cb (momentum, kinetic energy, etc.). After the collision (interaction), the states of these balls changed. Let us denote the new states C"a and C"b. The reason for the change in states (Ca → C"a and Cb → C"b) was the interaction of the balls (collision); the consequence of this collision was a change in the state of each ball.

As already mentioned, the evolutionary model in in this case is of little use, since we are not dealing with a causal chain, but with an elementary cause-and-effect link, the structure of which cannot be reduced to an evolutionary model. To show this, let us illustrate this example with an explanation from the position of the evolutionary model: “Before the collision, ball A was at rest, so the cause of its movement is ball B, which hit it.” Here ball B is the cause, and the movement of ball A is the effect. But from the same positions, the following explanation can be given: “Before the collision, ball B was moving uniformly along a straight path. If it weren’t for ball A, then the nature of the movement of ball B would not have changed.” Here the cause is already ball A, and the effect is the state of ball B. The above example shows:

a) a certain subjectivity that arises when applying the evolutionary model beyond the limits of its applicability: the cause can be either ball A or ball B; this situation is due to the fact that the evolutionary model picks out one particular branch of the consequence and is limited to its interpretation;

b) a typical epistemological error. In the above explanations from the position of the evolutionary model, one of the material objects of the same type acts as an “active” principle, and the other as a “passive” principle. It turns out that one of the balls is endowed (in comparison with the other) with “activity”, “will”, “desire”, like a person. Therefore, it is only thanks to this “will” that we have a causal relationship. Such an epistemological error is determined not only by the model of causality, but also by the imagery inherent in living human speech, and the typical psychological transfer of properties characteristic of complex causality (we will talk about it below) to a simple cause-and-effect link. And such errors are very typical when using an evolutionary model beyond the limits of its applicability. They appear in some definitions of causation. For example: “So, causation is defined as such an effect of one object on another, in which a change in the first object (cause) precedes a change in another object and in a necessary, unambiguous way gives rise to a change in another object (effect).” It is difficult to agree with this definition, since it is not at all clear why, during interaction (mutual action!), objects should not be deformed simultaneously, but one after another? Which object should deform first and which should deform second (priority problem)?

Model qualities

Let us now consider what qualities the structural model of causality contains. Let us note the following among them: objectivity, universality, consistency, unambiguity.

The objectivity of causality is manifested in the fact that interaction acts as an objective cause in relation to which interacting objects are equal. There is no room for anthropomorphic interpretation here. Universality is due to the fact that the basis of causality is always interaction. Causality is universal, just as interaction itself is universal. Consistency is due to the fact that, although cause and effect (interaction and change of states) coincide in time, they reflect different aspects of the cause-and-effect relationship. Interaction presupposes a spatial connection of objects, a change in state - a connection between the states of each of the interacting objects in time.

In addition, the structural model establishes an unambiguous relationship in cause-and-effect relationships, regardless of the method of mathematical description of the interaction. Moreover, the structural model, being objective and universal, does not impose restrictions on the nature of interactions in natural science. Within the framework of this model, instantaneous long- or short-range action and interaction with any finite velocities are valid. The appearance of such a limitation in determining cause-and-effect relationships would be a typical metaphysical dogma, once and for all postulating the nature of the interaction of any systems, imposing a natural philosophical framework on physics and other sciences on the part of philosophy, or it would limit the limits of applicability of the model so much that the benefits of such a model would be very modest.

Here it would be appropriate to dwell on issues related to the finiteness of the speed of propagation of interactions. Let's look at an example. Let there be two stationary charges. If one of the charges begins to move with acceleration, then the electromagnetic wave will approach the second charge with a delay. Doesn't this example contradict the structural model and, in particular, the property of reciprocity of action, since with such interaction the charges are in an unequal position? No, it doesn't contradict. This example does not describe a simple interaction, but a complex causal chain in which three different links can be distinguished.

Due to the generality and breadth of its laws, physics has always influenced the development of philosophy and has itself been influenced by it. While discovering new achievements, physics did not abandon philosophical questions: about matter, about motion, about the objectivity of phenomena, about space and time, about causality and necessity in nature. The development of atomism led E. Rutherford to the discovery atomic nucleus and to...

Determining boundaries through feasibility and cost assessments

The applicability limits for the models are determined based on the implementation constraints identified in the previous section. As already mentioned, each of them affects one of the main limiting factors (or both at once) - economic efficiency (increasing implementation costs) or feasibility (decreasing the significance of the results obtained for the company).

The purpose of this section is to formulate recommendations for which companies a particular model is applicable to. Obviously, the applicability of the model strongly depends on individual conditions - the strategic priorities of the company, the characteristics of its structure and management style, financial resources, and the like. However, it seems possible to determine the initial approximate boundaries by solving the following subproblems (determining more precise boundaries may be the subject of future practical research):

· Identification of potential conflicts of company goals and limitations at this level

· Determination of points of occurrence of additional implementation costs for certain models (through already identified constraint factors)

· Approximate cost estimates where possible

Recommendations regarding the first task are already contained in the formulation of the corresponding constraint, which arises at the level of the goal “Selecting a partner for interaction” and extends to the Schillo and Computational Trust and Reputation Models. The company's goals must include the purpose of the model being implemented. For the above example of goal and model, a conflict is obvious in the situation of a monopoly market for the supplier - the consumer company cannot choose a partner for delivery using models, since there is only one option. To clarify the presence of this relationship, a company may need to decompose its goals using a goal tree, an object widely used in BPM.

During the analysis of the classification developed in the previous section and the literature on reputation models and special cases of their implementation, the following points of occurrence of additional costs were identified:

Collection of data on the reputation of counterparties. Occurs in the “Input Data” constraint, at the model level. What is taken into account here is the final value of reputation, which can be calculated internally (by implementing a model with a corresponding purpose) or acquired from relevant service providers. In the first case, there are costs for implementing two models instead of one, however, the potential benefits may be greater due to the functionality of the model for calculating reputation (the solution, therefore, depends on the set of objectives that the company needs to achieve using reputation). In the second case, costs are formed from the price of using tools for extracting the necessary data. Much here depends on the company's business environment. For companies operating within reputation systems (for example, sellers on EBay), it is possible to use the API of these systems, which are often already “hard-wired” necessary functions(as, for example, in the Yandex Market Content API) and the use of which is relatively cheap. You should also not lose sight of the costs of paying for the time of employees using the API, or automating these processes. In the case when the reputation of agents is not calculated centrally, the problem arises of extracting it from unstructured data, such as reviews (from different sources, different formats - for example, video reviews on YouTube, which are also a form of feedback), messages in intracorporate networks. Tools that solve these problems are more expensive - and their price increases the more data sources they are able to process. Very few companies have the resources to develop products of appropriate complexity, which also affects the cost. In addition, in the case of analyzing internal data (for example, corporate correspondence), the company must have the necessary data (generate it), and therefore the technology for storing it. If this condition is not met, new restrictions arise that significantly increase the cost of implementation and affect the feasibility. A comparison of different reputation data collection tools is shown in the table below:

Table 6. Comparison of reputation data extraction tools

Tool name	Price per month usage, thousand rubles
Reputation systems API
	For free
Yandex Market Content API	Free/20 (for those who do not sell on Ya-m)
Tools for extracting reputation from unstructured data
Sidorin Lab (sidorinlab.ru)
Brandspotter (brandspotter.ru)
Brand Analytics (br-analytics.ru)	150-515 (depending on the depth of the retrospective)
Semantic Force (semanticforce.net)
SAP HANA, Event Steam Processing based on Hadoop	From 370 (only the cost of the license per month is taken into account)

As can be seen from the table, most tools that analyze external data are affordable even for small companies (for example, small online stores; the average monthly profit of an e-commerce enterprise here is considered to be 750 thousand rubles, as in). The really expensive solutions involve analyzing large amounts of data generated by companies that can afford the cost. It is also worth noting that most of the inexpensive solutions are focused on working with the company’s reputation in its external environment (in the market, in the public space). Thus, when solving problems of personnel management (see applications of the organizational approach, Chapter 2, Fig. 8), where you need to analyze objects in internal environment companies, only expensive solutions are left to choose from.

Collect hard-to-find input data. Such data includes the input data of the “Reputation from the point of view of consumers” model, namely data on the cost structure of competitors. There are two ways to obtain them: accept approximate data (for example, accept your cost structure) or buy data from relevant service providers. The first case is suitable for companies in markets that are homogeneous in terms of products and sellers, close to perfect competition, but even there this precondition can lead to a serious decrease in the quality of the result obtained. The solution may be to use the model output as an argument to a decision function, which will take into account various factors with weights. The second case relates to the analysis of the competitive environment, which is part of the range of marketing analysis services that are widespread in the market. Examples of costs for such services are listed in the table below.

Even though the quality of information may directly depend on cost, competitive analysis services are available to a wide range of companies. It is worth noting, however, that the more dynamic the market, the lower the barriers to entry, the faster the number of competitors and their diversity grows - and the more often it is necessary to conduct competitive analysis, the higher its cost in terms of the period.

Ensuring data quality. If the input data for the model is difficult to access, there is another way - to use approximate data. For example, in the case of competitors’ specific variable costs, it seems possible to use the costs of the company implementing it in the model. In order to avoid the negative effects of data inaccuracy, it is enough to use several data sources (which is not a problem, since in most potential cases of introducing models, according to the author of this work, since the mechanisms for making appropriate decisions / solving problems / achieving goals are obvious , exist in companies without reputation being involved). In addition, weights can be assigned to these sources depending on the reliability of the data used. This path, however, comes with additional costs for the decision maker or for automating the process. Also, many models (for example, Sporas) require protection against unfair transactions and assessments. This can be resolved by implementing a certified reputation or OERM methods. For example, such methods include promptly responding to negative reviews or creating an artificial positive background in ratings/reviews. The costs associated with OERM methods are comparable to the costs of collecting reputation data - the deeper the analysis / more data about the company, the more expensive the services. Certified reputation is usually implemented at the reputation system level - as is the case with TripAdvisor - so all a company can do here is to choose the right system or model for which the level of protection will be acceptable.

Computational complexity. Occurs at the level of models, in the corresponding limitation. Of the models considered, the ones that use reflection are the most relevant to him - these are “Supplier and Intermediary”, “Reputation from the Point of View of Consumers”, “Model of Firms Competing in the Market”. In the calculations made there, phantom agents are used - agents that exist only in the minds of other agents (including phantom ones, determined by the rank of reflection). Additional calculations require additional power. Due to the variety of services for the provision of such capacities, as well as the requirements for them that are not directly related to the situation under consideration (for example, equipment sizes, virtuality, data security requirements), it is difficult to give a cost estimate of costs. Only one thing can be said for sure - the more agents or the higher the rank of reflection, the more complex the calculations in the model. Thus, reflexive models are best suited for companies operating in a market with a small number of players (oligopoly).

Cost of changes. If we turn to the constraints that arise at the process level (potentially covering all models), we can see that almost all of them are associated with changes in the company - in processes, connections between them, and various internal structures. These changes are more difficult to implement, the larger the company itself - accordingly, the larger the company, the more expensive it is to implement reputation models. For an accurate assessment, audit data from a large number of companies is required (to estimate the cost of possible changes) and data on practical implementation cases (for clarification and subsequent generalization). All of this could be areas for further research.

results

The result in this section is a list of models with their corresponding application recommendations. According to the results of the analysis, the main aspects here turned out to be the cost of necessary services for the company, its internal structure and parameters of the external environment.

0. All models - The larger the company, the more difficult it is for it to make changes in its internal structure, the less applicable the models are for it.

1. SPORAS - it is necessary to extract information to calculate reputation. Well applicable for companies within reputation systems; for others, costs arise proportional to the volume of data required for processing. Requires many prerequisites for technical implementation; to ensure them, it can be implemented together with other models (for example, certified reputation models)

2. Schillo - requires specific input data, the cost is proportional to the number of players in the market. For oligopolies or niche markets. In addition, the rating scale is binary, which leads to inaccuracy of data; correction of the solution may be required.

3. E-bay model. Simple summation - information needs to be extracted to calculate reputation. Well applicable for companies within reputation systems; for others, costs arise proportional to the volume of data required for processing.

4. Computational model of trust and reputation - information must be extracted to calculate reputation. Well applicable for companies within reputation systems; for others, costs arise proportional to the volume of data required for processing. In the case of monopolies among partners, application is inappropriate. In addition, the rating scale is binary, which leads to inaccuracy of data; correction of the solution may be required.

5. Model of firms competing in a market - best suited for oligopoly or niche markets. The more players, the less applicable.

6. Reputation from a consumer perspective (non-dynamic) - best suited for oligopoly or niche markets. The more players, the less applicable, because uses reflection and requires specific input data, which is more expensive the more players there are.

7. Reputation from a consumer perspective (with dynamics) - best suited for oligopoly or niche markets. The more players, the less applicable.

8. ReMSA - it is necessary to extract information to calculate reputation. Moderately applicable for companies within reputation systems, as it takes into account data that may not be collected within the system. For other companies, costs arise proportional to the volume of data required for processing.

9. Certified reputation model for Trip advisor - for companies within reputation systems or other networks with an established mechanism for counterparties to evaluate each other. For other business conditions (for example, when counterparties evaluate each other in free form) it is less applicable.

Table 7. Visualization of the limits of applicability

Relates. rep. systems	Many counterparties	Add. ed. on support quality Dan.
Simple sum/average	Outside/Inside
	Outside/Inside
	Outside/Inside
	Outside/Inside
Calc. trust and reputation model	Outside/Inside
Firms competing in the market	Outside/Inside
Reputation in the eyes of consumers (stat.)	Outside/Inside
Reputation in the eyes of consumers (din.)	Outside/Inside
Certificate rep. for TripAdvisor	Outside/Inside

Designations:

Green - good applicability

Yellow - applicable with restrictions/costs

Red - applicable with significant restrictions/costs

b) Models based on the concept of “interaction” (structural or dialectical models). We will find out the meaning of the names later. The main focus here is on interaction as the source of cause-and-effect relationships. The interaction itself acts as a cause. Kant paid much attention to this approach, but the dialectical approach to causality acquired its clearest form in the works of Hegel. Of the modern Soviet philosophers, this approach was developed by G.A. Svechnikov, who sought to give a materialistic interpretation of one of the structural models of cause-and-effect relationships.

Let us highlight some of the issues being discussed.

c) The problem of the reverse influence of an effect on its cause.

Attempts to go beyond the model representation and give a general, universal definition of the cause-and-effect relationship do not lead to a solution to the problem. As an example, we can cite the following definition: “Causality is such a genetic connection of phenomena in which one phenomenon, called the cause, in the presence of certain conditions inevitably generates, causes, brings to life another phenomenon, called the effect.” This definition is formally valid for most models, but without relying on the model, it cannot solve the problems posed (for example, the problem of simultaneity) and therefore has limited theoretical-cognitive value.

Limit of applicability of models

The transitivity property allows for a detailed analysis of the causal chain. It consists of dividing the final chain into simpler cause-and-effect links. If A, then A → B 1, B 1 → B 2,..., B n → C. But does a finite cause-and-effect chain have the property of infinite divisibility? Can the number of links in a finite chain N tend to infinity?

The transition of quantitative changes into qualitative ones occurs, for example, when dividing a piece of graphite. When molecules are separated until a monoatomic gas is formed, the chemical composition does not change. Further division of a substance without changing its chemical composition is no longer possible, since the next stage is the splitting of carbon atoms. Here, from a physicochemical point of view, quantitative changes lead to qualitative ones.

It is easy to see that during interaction both objects simultaneously influence each other (reciprocity of action). In this case, the state of each of the interacting objects changes. No interaction - no change of state. Therefore, a change in the state of any one of the interacting objects can be considered as a partial consequence of the cause - interaction. A change in the states of all objects in their totality will constitute a complete consequence.

So, we have established that evolutionary models, as an elementary, primary link, contain a structural (dialectical) model based on the interaction and change of states. Somewhat later we will return to the analysis of the mutual connection of these models and the study of the properties of the evolutionary model. Here we would like to note that, in full accordance with the point of view of F. Engels, the change of phenomena in evolutionary models reflecting objective reality occurs not due to the simple regularity of events (as in D. Hume), but due to the conditionality generated by interaction (genesis ). Therefore, although references to generation (genesis) are introduced into the definition of cause-and-effect relationships in evolutionary models, they reflect the objective nature of these relationships and have a legal basis.

Fig. 2. Structural (dialectical) model of causality

Let's return to the structural model. In its structure and meaning, it perfectly agrees with the first law of dialectics - the law of unity and struggle of opposites, if interpreted:

– unity-as the existence of objects in their mutual connection (interaction);

– opposites– as mutually exclusive trends and characteristics of states caused by interaction;

– fight– as an interaction;

– development– as a change in the state of each of the interacting material objects.

As already mentioned, the evolutionary model is of little use in this case, since we are not dealing with a causal chain, but with an elementary cause-and-effect link, the structure of which cannot be reduced to the evolutionary model. To show this, let us illustrate this example with an explanation from the position of the evolutionary model: “Before the collision, ball A was at rest, so the cause of its movement is ball B, which hit it.” Here ball B is the cause, and the movement of ball A is the effect. But from the same positions, the following explanation can be given: “Before the collision, ball B was moving uniformly along a straight path. If it weren’t for ball A, then the nature of the movement of ball B would not have changed.” Here the cause is already ball A, and the effect is the state of ball B. The above example shows:

Model qualities

Let us now consider what qualities the structural model of causality contains. Let us note the following among them: objectivity, universality, consistency, unambiguity.

Objectivity causality is manifested in the fact that interaction acts as an objective cause in relation to which interacting objects are equal. There is no room for anthropomorphic interpretation here. Versatility due to the fact that the basis of causality always lies interaction. Causality is universal, just as interaction itself is universal. Consistency is due to the fact that, although cause and effect (interaction and change of states) coincide in time, they reflect different sides cause-and-effect relationships. Interaction presupposes a spatial connection of objects, a change in state - a connection between the states of each of the interacting objects in time.

In addition, the structural model establishes unambiguous connection in cause-and-effect relationships, regardless of the method of mathematical description of the interaction. Moreover, the structural model, being objective and universal, does not impose restrictions on the nature of interactions in natural science. Within the framework of this model, instantaneous long- or short-range action and interaction with any finite velocities are valid. The appearance of such a limitation in determining cause-and-effect relationships would be a typical metaphysical dogma, once and for all postulating the nature of the interaction of any systems, imposing a natural philosophical framework on physics and other sciences on the part of philosophy, or it would limit the limits of applicability of the model so much that the benefits of such a model would be very modest.

1. The interaction of the first charge with an object, which causes its acceleration. The result of this interaction is a change in the state of the source that influenced the charge, and in particular the loss of part of the energy by this source, a change in the state of the first charge (acceleration) and the appearance electromagnetic wave, which was emitted by the first charge during its accelerated motion.

2. The process of propagation of an electromagnetic wave emitted by the first charge.

3. The process of interaction of the second charge with an electromagnetic wave. The result of the interaction is the acceleration of the second charge, the scattering of the primary electromagnetic wave and the emission of an electromagnetic wave by the second charge.

In this example, we have two different interactions, each of which fits into the structural model of causation. Thus, the structural model is in excellent agreement with both classical and relativistic theories, and the finite speed of propagation of interactions is not fundamentally necessary for the structural model of causality.

Regarding the structural model of causality, we note that the decay reactions and do not contradict it. synthesis of objects. In this case, a relatively stable connection between objects is either destroyed as a special type of interaction, or such a connection is formed as a result of interaction.

Since quantum theories (as well as classical ones) widely use the categories “interaction” and “state,” the structural model is fundamentally applicable in this area of natural science. The difficulties that are sometimes encountered are due, in our opinion, to the fact that, although they have a well-developed mathematical formalism, quantum theories are not yet fully developed and refined in terms of conceptual interpretation.

Mario Bunge writes, for example, about the interpretation of the f-function:
“Some attribute the function ψ to some individual system, others to some actual or potential statistical ensemble of identical systems, others consider the ψ-function as a measure of our information, or the degree of confidence regarding some individual complex consisting of a macrosystem and an instrument, or, finally , simply as a catalog of measurements made on many identically prepared microsystems.” Such a variety of options for interpreting the ψ-function makes it difficult to strictly causally interpret the phenomena of the microworld.

This is one of the indications that quantum theories are in the stage of formation and development and have not reached the level of internal completeness characteristic of classical theories.

But the problems of the formation of quantum theories are evidenced not only by the interpretation of the ψ-function. Although relativistic mechanics and electrodynamics at first glance appear to be complete theories, a deeper analysis shows that for a number of reasons these theories also did not avoid contradictions and internal difficulties. For example, in electrodynamics there is the problem of electromagnetic mass, the problem of the reaction of charge radiation, etc. Failures in attempts to resolve these problems within the framework of the theories themselves in the past and the rapid development of theories of the microworld gave rise to the hope that the development of quantum theories will help eliminate the difficulties. Until then, they must be perceived as an inevitable “evil” that one has to put up with one way or another and expect success from quantum theories.

At the same time, quantum theories themselves faced many problems and contradictions. It is interesting to note that some of these difficulties are of a “classical” nature, i.e. inherited from classical theories and is due to their internal incompleteness. It turns out to be a “vicious circle”: we assign the resolution of the contradictions of classical theories to quantum theories, and the difficulties of quantum theories are determined by the contradictions of classical ones.

Over time, hope in the ability of quantum theories to eliminate contradictions and difficulties in classical theories began to fade, but until now interest in resolving the contradictions of classical theories within their own framework still remains in the background.

Thus, the difficulties that sometimes arise when explaining the phenomena of the microworld from the position of causality have an objective origin and are explained by the peculiarities of the formation of quantum theories, but they are not fundamental, prohibiting or limiting the application of the principle of causality in the microworld, in particular the application of the structural model of causality.

Causality and interaction are always interrelated. If interaction has the properties of universality, universality and objectivity, then cause-and-effect connections and relationships are equally universal, universal and objective. Therefore, in principle, one cannot agree with Bohm’s statements that when describing the phenomena of the microworld, one can in some cases rely on philosophical indeterminism, in others one can adhere to the principle of causality. We consider V.Ya.’s idea to be deeply erroneous. Perminov that “the concept of complementarity indicates path reconciliation(our italics – VC.) determinism and indeterminism”, regardless of whether this thought relates to the philosophy of natural science or to a specific natural scientific theory. The way to reconcile the materialist point of view with the position of modern agnosticism on this issue is eclecticism, there is a denial of objective dialectics. IN AND. Lenin emphasized that “the question of causality is especially important for determining the philosophical line of one or another new “ism” ...” (vol. 18, p. 157). And the path to the formation of quantum theories lies not through denial or limitation, but through the affirmation of causality in the microworld.

Two sides of scientific theories

The structure of scientific theories of natural science and the functions of scientific theories are directly or indirectly related to the causal explanation of the phenomena of the material world. If we turn to the structural model of causality, we can identify two characteristic points, two important aspects that are in one way or another connected with the functions of scientific theories.

The first concerns the description of causal relationships and answers the question: how, in what sequence? It corresponds to any branch of private consequence that connects conditioned states. It gives not only a description of the transition of an object from one state to another, but describes and covers the entire causal chain as a sequence of related and conditioned states, without going deeply into the essence, into the source of changes in the states of the chain links.

The second side answers the question: why, for what reason? On the contrary, it splits the cause-and-effect chain into separate elementary links and provides an explanation for the change in state, based on interaction. This is the explanatory side.

These two sides are directly related to the two important functions scientific theory: explanatory and descriptive. Since the principle of causality has been and will be the basis of any natural science theory, the theory will always perform these two functions: description and explanation.

However, this is not the only way in which the methodological function of the principle of causality is manifested. The internal structuring of the theory itself is also related to this principle. Let's take, for example, classical mechanics with its three traditional branches: kinematics, dynamics and statics. In kinematics, force interactions are not considered, but there is a description (physical and mathematical) of the types of motion of material points and material objects. Interaction is implied, but it fades into the background, leaving priority to the description of complex related movements through the characteristics of their states. Of course, this fact cannot serve as a reason for classifying kinematics as a non-causal method of description, since kinematics reflects the evolutionary side of cause-and-effect relationships connecting various states.

Dynamics is a theoretical section that includes a complete cause-and-effect description and explanation, based on a structural model of cause-and-effect relationships. In this sense, kinematics can be considered a subfield of dynamics.

Of particular interest from the point of view of causality is statics, in which the chains of consequences are degenerate (absent), and we are dealing only with connections and interactions of a static nature. In contrast to the phenomena of objective reality, where there are no absolutely stable systems, static problems are an idealization or a limiting case, permissible in private scientific theories. But the principle of causality is also valid here, since it is impossible not only to solve static problems, but also to understand the essence of statics without applying the “principle of virtual displacements” or related principles. “Virtual displacements” are directly related to changes in states in the vicinity of the equilibrium state, i.e. ultimately with cause and effect relationships.

Let us now consider electrodynamics. Sometimes it is identified only with Maxwell's equations. This is incorrect because Maxwell's equations describe the behavior of waves (emission, propagation, diffraction, etc.) under given boundary and initial conditions. They do not include a description of the interaction as a reciprocal action. The principle of causality is introduced along with boundary and initial conditions (retarded potentials). This is a kind of “kinematics” of wave processes, if such a comparison is permissible. “Dynamics”, and with it causality, is introduced by the Lorentz equation of motion, which takes into account the reaction of charge radiation. It is the connection between Maxwell's equations and the Lorentz equation of motion that provides a fairly complete cause-and-effect description of the phenomena of electromagnetism. Similar examples could be continued. But the above are enough to make sure that causality and its structural model are reflected in the structure and functions of scientific theories.

If at the beginning of our work we went from an evolutionary model of causality to a structural one, now we have to go back from the structural model to the evolutionary one. This is necessary to correctly assess the mutual relationship and distinctive features evolutionary model.

Already in an unbranched linear cause-and-effect chain, we are forced to abandon a complete description of all cause-and-effect relationships, i.e. We do not take into account some particular consequences. The structural model allows unbranched linear cause-and-effect chains to be reduced to two main types.

a) Object causal chain. It is formed when we select a material object and monitor the change in its state over time. An example would be observations of the state of a Brownian particle, or the evolution spaceship, or the propagation of an electromagnetic wave from the transmitter antenna to the receiver antenna.

b) Information causal chain. Appears when we monitor not the state of a material object, but some informing phenomenon, which, in the process of interactions of various material objects, is connected sequentially in time with various objects. An example would be the transmission of oral information using a relay race, etc.

All linear, unbranched causal chains can be reduced to one of these two types or a combination of them. Such chains are described using an evolutionary model of causality. In an evolutionary description, interaction remains in the background, and a material object or an indicator of its condition comes to the fore. Because of this, the main attention is focused on describing the sequence of events over time. Therefore, this model is called evolutionary.

A linear, unbranched causal chain is relatively easy to analyze by reducing it to a set of elementary links and analyzing them through a structural model. But such an analysis is not always possible.

There are complex causal networks in which simple cause-and-effect chains intersect, branch, and intersect again. This leads to the fact that the use of a structural model makes the analysis cumbersome and sometimes technically impossible.

In addition, we are often interested not in the internal process itself and the description of internal cause-and-effect relationships, but in the initial impact and its final result. This situation often occurs in behavior analysis. complex systems(biological, cybernetic, etc.). In such cases, detail internal processes in their entirety turns out to be redundant, unnecessary for practical purposes, cluttering up the analysis. All this led to a number of features when describing cause-and-effect relationships using evolutionary models. Let us list these features.

1. In the evolutionary description of the causal network, the complete causal network is coarsened. The main chains are highlighted, and the unimportant ones are cut off and ignored. This greatly simplifies the description, but such simplification is achieved at the cost of losing some information, at the cost of losing the unambiguity of the description.

2. In order to maintain unambiguity and bring the description closer to objective reality, cut off branches and causal chains are replaced by a set of conditions. The completeness, unambiguity and objectivity of the cause-and-effect description and analysis depend on how correctly the main causal chain is identified and how fully the conditions compensating for coarsening are taken into account.

3. The choice of one or another cause-and-effect chain as the main one is determined largely by the researcher’s goals, i.e. what phenomena he wants to analyze the connection between. Exactly target setting forces us to look for the main cause-and-effect chains, and replace the cut-off ones with conditions. This leads to the fact that with some settings main role execute some chains, while others are replaced by conditions. With other settings, these chains may become conditions, and the role of the main ones will be played by those that were previously secondary. Thus, the causes and conditions change roles.

Conditions play important role, linking objective cause and effect. Under different conditions affecting the main causal chain, the consequences will be different. Conditions seem to create the channel along which the chain flows historical events or the development of phenomena over time. Therefore, to identify deep, essential cause-and-effect relationships, a thorough analysis is required, taking into account the influence of all external and internal factors, all conditions influencing the development of the main causal chain, and assessment of the degree of influence.

4. The evolutionary description focuses not on interaction, but on the connection of events or phenomena in time. Therefore, the content of the concepts “cause” and “effect” changes, and this is very important to take into account. If in the structural model interaction acts as a true causa finalis - the final cause, then in the evolutionary model - the active cause (causa activa) becomes a phenomenon or event.

The investigation also changes its content. Instead of connecting the states of a material object during its interaction with another, some event or phenomenon acts as a consequence, closing the cause-and-effect chain. Because of this, the cause in the evolutionary model always precedes the effect.

5. In the above sense, cause and effect in the evolutionary model can act as single-quality phenomena that close the cause-and-effect chain on both sides. The consequence of one chain can be the cause and beginning of another chain, following the first in time. This circumstance determines the transitivity property of evolutionary models of causality.

Here we have touched only on the main features and distinctive features of the evolutionary model.

Conclusion

The structural model of causality can be successfully used for relatively simple causal chains and systems. In real practice, we also have to deal with complex systems. The question of a cause-and-effect description of the behavior of complex systems is almost always based on the evolutionary model of causality.

So, we examined two types of models that reflect cause-and-effect relationships in nature, analyzed the mutual relationship of these models, the limits of their applicability and some features. The manifestation of causality in nature is diverse both in form and content. It is likely that these models do not exhaust the entire arsenal of forms of cause-effect relationships. But no matter how diverse these forms are, causality will always have the properties of objectivity, universality and universality. Because of this, the principle of causality has performed and will always perform the most important ideological and methodological functions in modern natural science and philosophy of natural science. The variety of forms of manifestation of cause-and-effect relationships cannot serve as a reason for abandoning the materialistic principle of causality or statements about its limited applicability.

Information sources:

Svechnikov G.A. Causality and connection of states in physics. M., 1971.
Svechnikov G.A. Dialectical-materialistic concept of causality // Modern determinism: Laws of nature. M., 1973.
Tyukhtin V.S. Reflection, systems, cybernetics. M., 1972
Uemov A.I., Ostapenko S.V. Causality and time // Modern determinism: Laws of nature.
Orudzhev Z.M., Akhundov M.D. Temporal structure of causal connection // Philosophy. Sciences. 1969. No. 6.
Zharov A.M. Temporal relationship of cause and effect and uncertainty. 1984. No. 3.
Kuznetsov I.V. Selected works on the methodology of physics. M., 1975.
Materialistic dialectics: In 5 volumes. T. 1: Objective dialectics / Under the general. ed. F.V. Konstantinov and V.G. Marakhova; Rep. ed. F.F. Väkkerev. M., 1981.
Naletov N.3. Causality and theory of knowledge. M., 1975.
Hegel G.W.F. Encyclopedia of Philosophical Sciences: In 3 volumes. T. 1: The Science of Logic. M., 1974.
Starzhinsky V.P. The concept of “state” and its methodological role in physics. Minsk, 1979.
Ivanov V.G. Causality and determinism. L., 1974.
Materialistic dialectics. T. 1. P. 213.
Bunge M. Philosophy of Physics. M., 1975. P. 99.
Bohm D. Causality and randomness in modern physics. M., 1959.
Perminov V.Ya. The problem of causality in philosophy and natural science. M., 1979. P. 209.
Nikitin E.P. Explanation is the function of science. M., 1970.

Kuligin V.A. Causality and interaction in physics. Collection of Voronezh State University: “Determinism in modern science.” Voronezh, 1987.

19. Modeling as a method of cognition. Types of models. Adequacy, limits of applicability of models. Examples of using models in the study of biological systems.

Modeling- this is a method in which the study of some complex object (process, phenomenon) is replaced by the study of its model. The main stages of modeling can be summarized as follows:

1. Primary collection of information. The researcher must obtain as much information as possible about the various characteristics of a real object: its properties, the processes occurring in it, patterns of behavior under various external conditions.

2. Formulation of the problem. The purpose of the research, its main objectives are formulated, and what new knowledge the researcher wants to obtain as a result of the research is determined. This stage is often one of the most important and time-consuming.

3. Justification of the main assumptions. In other words, the real object is simplified, characteristics (item 1) that are not significant for the purposes of the study are isolated and can be neglected.

4. Creation of a model, its research.

5. Checking the adequacy of the model real object. Indication of the limits of applicability of the model.

Thus, the model, as it were, coordinates the real object with the purpose of the study: on the one hand, it simplifies the object, making it possible to conduct research, but on the other hand, it preserves the main thing that interests the researcher. In biophysics, biology and medicine, physical, biological, and mathematical models are often used. Analogue modeling is also common.

Physical model has a physical nature, often the same as the object being studied. For example, the flow of blood through vessels is modeled by the movement of fluid through pipes (rigid or elastic).

Biological models are biological objects convenient for experimental research, on which the properties and patterns of biophysical processes in real complex objects are studied. For example, the patterns of occurrence and propagation of action potentials in nerve fibers were studied only after finding such a successful biological model as the giant squid axon.

Mathematical models- description of processes in a real object using mathematical equations, usually differential. Computers are now widely used to implement mathematical models.

If the processes in the model have a different physical nature than the original, but are described by the same mathematical apparatus (usually the same differential equations), then such a model is called analog. Typically used as an analog model electric. For example, an analog model of the vascular system is an electrical circuit of resistances, capacitances and inductances.

Basic requirements that the model must meet.

1. Adequacy - the model must reproduce the patterns of the phenomena being studied with a given degree of accuracy.

2. The limits of applicability of the model must be established, that is, the conditions under which the selected model is adequate to the object being studied must be clearly defined, since no model provides an exhaustive description of the object.