Frame theory. Frame theory: the essence of the theory, main theses Social frame

Frames- these are the minimum information structures necessary to represent a class of objects, phenomena or processes. The frame reflects the basic properties of an object or phenomenon.

Frame is a network of nodes and relationships organized hierarchically, where the top nodes represent general concepts, and the bottom nodes represent more specific cases of these concepts.

The structure of a frame is written as a list of properties called in the frame slots.

Slots- these are some unfilled substructures of the frame, the filling of which leads to the fact that this frame is put in correspondence with a certain situation, phenomenon or object.

The term frame (from the English frame - “frame”, “frame”) was proposed in the 70s of the last century by one of the pioneers of AI, Marvin Minsky, to denote the structure of knowledge for the perception of spatial scenes. Like the semantic network, the frame model has a deep psychological basis.

A frame is understood as an abstract image or situation to represent a certain stereotype of perceptions.

In philosophy and psychology, the concept of an abstract image is widely used. For example, pronouncing the word “room” gives rise to a certain image of a room, i.e. “a living space with four walls, a floor, a ceiling, windows and a door.” Nothing can be excluded from the description - for example, by removing the windows, we will get a closet, not a room. However, in this description there are some unfilled attributes - “slots”, for example, the number of windows, ceiling height, etc. Slots are descriptions of particulars.

In frame theory, this image of a room is called a room frame. In addition, a formalized model for displaying an image is also called a frame.

There are sample frames, or prototypes, stored in the knowledge base, and instance frames, which are created to display real actual situations based on incoming data.

The frame model is to some extent universal, since it allows you to display the entire diversity of knowledge about the world through:

frame-structures , to designate objects and concepts (loan, pledge, bill);

role frames (manager, cashier, client);

frame-scenarios (bankruptcy, shareholders meeting, name day celebration);

frames-situations (alarm, accident, operating mode of the device), etc.

Frame structure can be represented using lists and tables.

FRAME NAME:

(1st slot name: 1st slot value),

(2nd slot name: 2nd slot value),

(Nth slot name: Nth slot value).

The same record can be presented in the form of a table, supplemented with two columns.

In the table, additional columns are intended to describe the type of slot and the possible attachment of special procedures to a particular slot, which is allowed in frame theory. The value of a slot can be the name of another frame, this is how networks of frames are formed ( Frame system- this is a hierarchical structure, the nodes of which are frames with a certain data structure).

Methods for a slot to obtain a value in an instance frame:

by default from the sample frame (Default value);

according to the formula specified in the slot;

clearly from the dialogue with the user;

attached procedure;

through inheritance of frame properties specified in the ACO slot.

The most important property of frame theory is the inheritance of properties, borrowed from the theory of semantic networks (used to reduce information redundancy in frame systems; it allows general (global) information for the system to be stored in a separate frame, and in all other frames to indicate only a link to the storage location of this information). In both frames and semantic networks, inheritance occurs according to AKO-connections (A-Kind-Of = This). The ACO slot points to a frame at a higher level of the hierarchy, from which they are implicitly inherited, i.e. the values ​​of similar slots are transferred.

Example. For example, in the network of frames in Fig. 1, the concept “student” inherits the properties of the frames “child” and “person”, which are at a higher level of the hierarchy. So, to the question: “Do students like sweets?” The answer is “yes”, since all children have this property, as indicated in the “child” frame. Inheritance of properties may be partial, since the age for students is not inherited from the child frame, since it is specified explicitly in its own frame. Specifying it explicitly overrides inheritance.

The main advantage of frames as a model for representing knowledge is the ability to reflect the conceptual basis of human memory organization, as well as flexibility and visibility.

Special knowledge representation languages ​​in frame networks FRL (Frame Representation Language), KRL, KAPPA and others make it possible to effectively build industrial ES.

Frame-oriented expert systems such as ANALYST, MODIS, TRISTAN, ALTERID are widely known.

Attached procedure. The slot value can be a procedural type program that runs on a message passed from another frame.

Types of procedures (launched when accessing the corresponding slot):

IF-NEEDED (if necessary) - runs if at the time of accessing the slot its value was not set;

IF-ADDED (if-added) - triggered when a value is substituted into the slot;

IF-REMOVED (if-removed) - runs when the slot value is erased;

Just like practice-oriented sociology, “frame theory” is not a holistic theoretical construction, but
is a set of concepts developing within the framework of sociology, psychology, cognitive science, cybernetics and linguistics. All these concepts are organized around the problem of contextualizing an action event (or utterance, i.e. a communication event). Actually, the term “frame” is a collective designation of context. But if in practice-oriented theories the starting point is the practical act itself, which already contains the conditions for its implementation or constitutes them in the process of implementation, then the focus of frame theory is the context of the practical act - a context endowed with a relatively autonomous, independent of specific practices existence.
The concept of “frame” today is most often used in two meanings, which with a certain degree of convention can be characterized as “linguistic” and “sociological”. According to the linguistic definition proposed in modern cognitive-oriented semantics, a frame is a cognitive structure, considered as a hierarchically structured system of knowledge about the signified [Tsurikova 2001: 145]. This structure is “the most generalized schematized representation of the basis of the value lt;...gt; the scheme of the image that underlies it” [Belyavskaya 1991: 83-84]. This definition goes back to Mervyn Minsky's work “Frames for the Representation of Knowledge” [Minsky 1979]. This is where the study of the problem of frames of human memory, thinking and communication in modern cognitive psychology begins [Solso 1996].
M. Minsky introduces the concept of a frame into the theory of artificial intelligence and interprets it as a static information structure that serves to represent stereotypical contexts [Minsky 1979: 26]. Soon after this, Charles Fillmore borrows ideas from Minsky's frame theory to construct the linguistic concept of frame semantics. He defines a frame as “a cognitive structure for schematizing experience” [Fillmore 1988: 24-25]. Thus, both in cybernetics and linguistics, the definition of “the elementary structure of communicative experience” is assigned to the frame.
The second conceptualization of the frame was proposed by G. Bateson and developed in detail by I. Goffman. For the purpose of this definition, a frame is interpreted as “the structural context of everyday interaction.” Thus, this concept becomes
psychological (G. Bateson), and then sociological (I. Hoffman) interpretation.
What do cybernetic, linguistic, psychological and sociological definitions of frame have in common? They all define a frame as: a structure that is stable and relatively static; cognitive education, the elements of which are
cognition (knowledge) and expectation (expectations); schema of representation, i.e. representing and meaning
form.
The fact that the category of frame is becoming so widespread in several disciplines at once is facilitated by the “cognitive revolution” of the 50s and 60s. In this sense, all frame theories have common roots: M. Minsky is inspired by the information research of Norbert Wiener, whose friend and associate for a long time was G. Bateson, who influenced the psychological (neurolinguistic programming) and sociological (frame analysis of I. Hoffman) directions in the theory frames.
We cannot pay enough attention here to the mutual influence of two lines of development of frame theory: cybernetic-linguistic and sociological-psychological. We are also forced to leave out interesting options for the development of the theory of frames in semiotics (among them we highlight an interesting line of interpretation of the concept of “frame” in the concept of Umberto Eco [Eco 2005]).
M. Minsky’s book “Frames for the Representation of Knowledge” and I. Goffman’s fundamental work “Frame Analysis: An Essay on the Organization of Everyday Experience” were published almost simultaneously (1974-1975). However, in Goffman's "Frame Analysis" there are no traces of the influence of M. Minsky (just as in the works of Minsky there is no influence of Goffman or Bateson). Six years after the publication of Frame Analysis, Erving Goffman noted: “All criticism must
to take into account differences in alternative conceptualizations of frames that have only recently received sufficient attention. For example, the work of Shank and Abelson in psychology, Minsky in artificial intelligence research, and Fillmore in linguistics. Indeed, today one could write a book focusing on the concept of “frame” and not touch upon “frame” in the sense that I or Bateson write about it.” (An example of such a book is the work of Deborah Tannen analyzed by Goffman).
We can conclude that both versions of frame theory (cybernetic-linguistic and sociological-psychological) are developing synchronously, but independently of each other.
The sociological conceptualization of the frame is based on the focus on studying everyday social interaction as a “contextual whole,” formulated in American pragmatism. “We never perceive or form judgments about objects and events in isolation,” wrote J. Dewey, “but only in connection with a contextual whole. This last is what is called a situation.” Situation is a category introduced to study the contextualization of human interaction. Proposed in the philosophy of pragmatism, it moves into the sociology of the Chicago School, where it is developed in detail by William Thomas, and then becomes one of the central concepts of symbolic interactionism.
Situations belong to classes of generalized situations. Moving from consideration of individual situational frameworks of the organization of everyday experience acquired in the process of interpersonal communication to expanded orders of interaction, the theorist observes the animation (in the words of I. Goffman - “layering”) of social contexts. This raises the question of the “ultimate context” or “context of contexts.” This question allows us to connect the problem of contextualization of practical action with the problem of the boundaries of social reality, which is, in this formulation of the question, the ultimate context of action.
W. James calls such an ultimate context a “universum”, and, interpreting it in the spirit of his own thesis about “plurality”
sticity of the universe,” describes the stratification of the universe into “subuniverses” - closed areas of contexts that obey their own logic of organization. Each such area has its own “mode of reality”. Actually, James owns the first definition of social reality in the categories of frame theory: “reality is the edging of our involvement.” The term “fringe”, in the form in which James uses it, can hardly claim the rigor of a scientific concept, but it accurately conveys the “frame metaphors” inherent in frame theory.
Everyday life, according to James, is one of such subuniverses (along with the worlds of religious experience, madness, scientific theorizing and eccentricity), its boundaries are the outer boundaries of the area of ​​​​all particular contexts of everyday experience. Thus, in pragmatist philosophy, another characteristic of a frame (context) is formulated: contexts are organized into meta-contexts (“frameworks”, “frame systems” or “frames of frames”). One of these metacontexts is everyday reality as a set of routine social interactions.
Contexts do not exist in isolation from each other; they are always in some relationship with each other. Not only does the meta-context contain the situational contexts of everyday actions, but these contexts themselves - “dinner party”, “business conversation”, “signing a document”, “outing with colleagues” - fit into each other, in the manner of a nesting doll. “A true philosopher,” James notes, “does not simply strive to place each object of his thought in its proper place in one or another subuniverse; he also tries to clarify the relations between subuniverses as elements of the totality that embraces them."
James's ideas about the plurality of realities are developed in the phenomenological sociology of A. Schutz. Schütz examines in detail the characteristics of “subuniverses” (in his definition, “finite areas of meaning” [Schütz 2003]). Finite areas of meaning
la exist as “contexts of contexts”, each of them can potentially be actualized, receiving the accent of reality. Schutz objects to James: “We are talking about areas of meaning, and not about subuniverses, because it is the meaning of our experience, and not the ontological structure of objects, that constitutes reality” [Schutz 2003: 17].
Shifting the focus of theoretical attention from the ontological structures of the metacontext to the experience of the acting actor is not accidental. This is a phenomenological revision of pragmatism. Although James is by no means inclined to ignore subjective experience (otherwise we would not be talking about the “edge of involvement”), he is much more interested in the own logic of the organization of subuniverses, inherent in them as closed areas of social experience. J.G. goes even further in this direction. Mead, who defined reality in the spirit of a thesis
A.N. Whitehead on “the organization of the prospects existing in nature.” The specific forms of experience inherent in each of the contexts are not considered as subjective, since perspectives do not belong to the subjects, but are woven into the very fabric of the world. Schütz follows Husserl’s reasoning: “any real unities are unities of meaning” [Schütz 2004: 123]. Consequently, the boundaries of the metacontext (or the “final domain of meaning”) are the boundaries of the possible understanding of the action. Don Quixote's actions - attacking windmills, fighting giants and searching for Dulcinea [Schütz 2003: 21] - are not without meaning; on the contrary, they are comprehended in the meta-context of his imaginary world.
How do meta-contexts and typical private contexts of action relate to each other? Schutz does not find an answer to this question. His attempts to describe the internal organization of metacontexts are limited to reproducing what was proposed by J.G. Mead's stratification of everyday life: from the “manipulative area”, the boundaries of which are constituted by the possibility of manipulating accessible objects, to the “world beyond reach” [Schutz 2003: 13]. This decision finds virtually no response in the future.
development of frame theory. Another idea, borrowed to answer this question from the phenomenological tradition, turns out to be more productive - the idea of ​​“brackets”, boundary signs separating episodes of experience [Goffman 2003a: 73]. Thus, the boundaries of contexts and metacontexts are distinguished as internal and external "brackets". Meta-context (frame of frames) sets the boundaries of the meaningfulness of particular contexts (frames) in which everyday events occur.
A similar interpretation of the relationship between internal and external contextualization is developed in the works of G. Bateson. It borrows the idea of ​​"brackets" and emphasizes the need to study discrete elements of everyday interaction. Hence the following constitutive feature of the frame: the frame is an explication of the discreteness of everyday life; behind frame theory is the “intuition of the discrete,” the idea of ​​the everyday world as having distinct internal and external boundaries.
“Intuition of the discrete” is opposed to “intuition of the continuum”: the tendency to interpret the object under study in the categories of the “flow” metaphor. In this metaphor, not only everyday activity can be comprehended (“the flow of symbolically mediated interaction” in G. Blumer or the “flow of practices” in G. Garfinkel), but also the experience of the everyday world itself, practical consciousness (“stream of consciousness” in the works of William James) . Contrasting discreteness with flow, the theory of frames endows the everyday reality under study with the properties of “elementary,” “collected,” and “discontinuous.”
The originality of Gregory Bateson's plan is partly determined by the interdisciplinarity of his ideas: Bateson tried to synthesize the ideas of phenomenology and pragmatism with the achievements of theoretical logic (the concept of logical types by B. Russell), linguistics (the Whorf-Sapir hypothesis of linguistic relativity) and the “cognitive revolution”. In his work “The Theory of Play and Fantasy” [Bateson 2000: 205-220], the term “frame” serves not so much to indicate the contextuality of the action or the boundaries of the meaningfulness of the perceived (imaginary), but to define the structural features of everyday communication. The most important of these features is the use of metacommunicative and metalinguistic messages.

Thus, while observing the behavior of monkeys at the San Francisco Zoo, Bateson discovered and described the characteristic metacommunicative signs exchanged between individuals playing a game of fighting. The very possibility of the game exists only thanks to the message “this is a game.” This message is metacommunicative - it requires a view from outside the interaction, an indication of its context. “The signals,” writes Bateson, “exchanged in the context of the game, lt;...gt; are paradoxical twice: firstly, the playful bite does not mean what the bite it replaces would mean, and secondly, the bite itself is fictitious. Not only do playing animals not quite mean what they communicate, but the communication itself is about something that does not exist” [Bateson 2000:210].
The game world is organized according to its own rules: what is real in it is that which does not “really” exist. However, this reality is “enclosed in brackets”, and therefore the action of placing brackets is required - an action that asserts the boundaries of the context. In Bateson's terminology, such an action is a “communication of boundaries,” “persuasion,” “instructions,” an indication that all subsequent actions should be perceived differently from previous ones. The game, as a special context of action, requires a preliminary agreement: “This is a game.” “Any message,” G. Bateson summarizes, “explicitly or implicitly establishes a frame, by virtue of this very fact gives instructions to the recipient or contributes to his efforts to understand the messages contained in the frame” [Bateson 2000:215].
Thus, G. Bateson identifies the following constitutive characteristic of a frame: a frame is a metacommunicative formation, a “message about messages”, it is not reducible to other elements of communication and cannot be derived from them; The frame belongs neither to the content (the “activities in the frame”) nor to the environment (“the activities behind the frame”).
It was G. Bateson, who synthesized the achievements of cognitive science, communication studies and psychology, who opened (as paradoxical as it may sound) the prospect of the actual sociological conceptualization of the frame. In his work “The Theory of Play and Fantasy,” he proposes using two types of analogies to describe frame-analytic research: the analogy of a picture frame and the analogy
mathematical set. “The first step to defining a frame,” writes Bateson, “may be to say that it (the frame) is a class or limits a class (set) of messages (meaningful actions). Then the game of two individuals under certain circumstances will be defined as the set of all messages that they exchanged over a limited period of time lt;...gt; In a set-theoretic scheme, these messages would be represented by points, and the “set” might be delineated by a line separating them from other points representing non-game messages” [Bateson 2000: 214].
If on both sides of the “border” there are messages of the same “logical type” (here Bateson appeals to the categorical apparatus of Russell’s theory of logical types), then we are talking about a particular frame (“a game of two young orangutans”); if the boundaries of the context coincide with the boundaries of the logical type and, for example, the set of game messages is separated from the set of non-game messages, then we have an example of a meta-context (“game”). Episodes of the same game can be carried out in different frames (imitation of a chase, imitation of a fight, imitation of surrender), but they belong to the same frame system, the same meta-context. In other words, a metacontext is a frame that covers all frames belonging to the same logical type. These are James’s “subuniverses” and Schutz’s “finite regions of meaning.”
The difficulty here is that the analogy borrowed from set theory is overly abstract and invariably leads all subsequent arguments into the realm of formal logic. It does not provide an idea of ​​the frame as a “really existing” context. A frame is only as real as it is recognized by the participants in the interaction or by the analyst. Confirmation of the “recognizability” of a frame for Bateson is the presence of corresponding concepts in the dictionary: “game”, “film”, “work”, “interview” - they all refer to one or another set of contextually organized actions.
Another way to describe a frame metaphorically is to compare it to the frame of a painting. “If the analogy with a mathematical set is perhaps too abstract,” Bateson continues, “the analogy with a picture frame is too concrete. The concept we are trying to define is neither physical nor logical
skim. Rather, actual physical frames are added to physical pictures because it is easier for human beings to operate in a world where some of their psychological characteristics are exteriorized" [Ibid: 214].
Bateson writes about exteriorized psychological frames. But his logic of analysis is no more “psychological” than the theoretical constructs of James and Schutz. Moreover, James, Schutz and Bateson are united by a common formulation of the problem: if social reality is “multiple” and the meta-contexts of actions in it form closed worlds of meaning (“dream”, “theorizing”, “game”, “fantasy”, “everyday routine”) , then how is the unity of social experience possible, encompassing the experience of interaction in all these multiple contexts?
All theoretical constructs on which frame theory is based - regardless of whether they are produced by psychology, sociology or cognitive science - orient the researcher towards studying the complex machinery of social contexts and interaction frameworks. These frames are found in the spatiotemporal organization of interaction and in communication itself, but they are not part of practical action, just as the frame of a picture is not part of the space “inside” the picture, and quotation marks and parentheses are not part of the phrase placed in them. As Eviatar Zerubavel, a student of Erving Goffman and a representative of cognitive sociology, noted, “... a frame is characterized not by its content, but by the special way in which it transforms the meaning of this content.”
The above-described interpretation of the frame as a context of a practical act, relatively independent of its content, is consolidated in social theory and gives rise to the actual sociological study of the organization of contexts of everyday interactions. However, before we turn to frame analysis in sociology and try to reveal its theoretical perspectives, we should return to the original opposition of “frame” and “practice” as two alternative projects of studying everyday life. Now we know that the key to understanding this alternativeness should be sought in the problem of determining the context of everyday action.

LOGIC, METHODOLOGY AND PHILOSOPHY OF SCIENCE

THE IMPORTANCE OF FRAME THEORY IN MODERN SCIENCE

Modern frame theory is used to structure and understand knowledge systems in various fields of science. The formulation of the basic principles of the theory occurred in parallel and independently in different sciences, so it is difficult to answer the question of which area is the source of the formation of the “frame approach”. The need to structure knowledge exists in any science, so the concept of a frame is used as a tool. The provisions of frame theory describe the mechanisms of constructing and perceiving reality, therefore their importance is great in the study of subjects of psychology, sociology, and linguistics.

Key words: frame, frameology, diagram, structure, frame theory, knowledge, reality.

The history of science of the 20th century is characterized by the development of many directions in various fields of knowledge. The picture of the world is changing again - the concept of four-dimensional space-time allows us to consider reality as a relative “quantity”. In physics, scientists are increasingly moving away from studying the visible world, going beyond the boundaries of the upper layers of the atmosphere and the boundaries of the molecular nucleus, studying the organization of near-Earth space and developing nanotechnology. The bold assumptions of science fiction writers are realized in human space flights, advances in robotics, and the creation, or attempts to create, artificial intelligence. The computer industry is developing at an amazingly fast pace, which affects not only the further scientific process, but also the structure and organization of society.

New trends are emerging in psychology aimed at studying the deep mechanisms of the psyche that control human behavior - Gestalt psychology, cognitive psychology. Practical sociology is becoming increasingly important, which, with the help of very specific techniques, reveals patterns in social connections. In linguistics, there is an interaction between different approaches to the study of the structure and perception of language. Understanding of language is considered inextricably linked with the social situation, discourse and psychological experience of a person. A characteristic feature of the development of scientific thought in the twentieth century is the emergence of many interdisciplinary connections and border sciences, as a result of which special methods and terms previously inherent in one scientific field penetrate into another and acquire a different meaning. This process of borrowing terminology occurs largely due to the fact that the terms themselves do not yet have a clearly formulated definition and can describe completely different subjects of study.

S.I. Nekrasov N. S. Molchanova

Moscow State Technical University of Civil Aviation

[email protected]

SCIENTIFIC GUIDELINES

For example, initially, in the most general case, the term “frame” denotes a structure containing some information. This understanding is the basis for explaining many cases of its application in various fields. So, in programming, a frame means a certain area of ​​​​data, which an ordinary PC user calls a “window” or “page”. In psychology, a frame is an ideal image of reality recorded in a person’s memory. In sociology - a certain model of behavior. Examples of frames can be found in linguistics (paradigm), in bibliography (description of publication, footnote form at the bottom of the page). In philosophy, various realities are correlated with the frame: science, everyday life, mythology and religion, etc.

As such, the science of frameology does not exist, because Frame theories are developed in many special scientific fields, but the subject of research is incomparable phenomena - the relationships of individuals in society, communication in the animal world and the human world, mental processes in humans, symbolic organization of communication, scientific activity, processes in cyber reality, features of language understanding. Thus, it is believed that the theory of frames was first developed by the American scientist Marvin Minsky. In his work “Frames for the Representation of Knowledge” (1979), he considers the frame from the point of view of a person’s perception of new conditions of reality and thinking in the process of adapting to an unfamiliar situation. It is necessary to make a reservation that the author developed the theory of frames within the framework of theoretical research in the field of artificial intelligence, i.e. how knowledge should be represented in a machine model of the real world. The Minsky frame represents some information on the basis of which a person makes predictions and also correlates his behavior. Therefore, Minsky calls a frame a certain set of data structured in a certain way, in which a stereotypical situation is encoded. At the same time, he does not distinguish between human thinking and artificial intelligence, believing that the human thinking process is based on the presence in his memory of a large set of various frames, from which, if necessary, the appropriate one is selected. Artificial intelligence is built in a similar way. The boundaries of the frames, which contain information about ideal situations, are presented in the form of questions. The answers to these questions complete the frame of a specific situation in given circumstances and a given period of time. Because There are many options for specific situations, then knowledge is structured into an extensive system of frames.

Developed in the 1970s by M. Minsky, the theory of frames for representing knowledge about the surrounding reality is reminiscent of the semantic field theory of J. Trier, L. Weisgerber and others in linguistics, formulated back in the 1930s. In accordance with this theory, when studying language, isolated semantic areas (fields) of vocabulary are identified on the basis of various relations of the linguistic units included in them (synonymous, generic, “part-whole”, etc.).

The main idea of ​​Trier and Weisgerber's structural semantics is that an adequate description of the meaning of a word depends on the lexical units included in the same semantic field of a given language, and on its (the word's) semantic relations with other members of the field. Thus, the content and scope of meanings of words included in the semantic field of designating intervals within a day in the Russian language differs from the content and scope of the same lexical units in the German language. In many situations where the word "morning" is used in Russian, native speakers of German would use the word Vormittag "pre-noon time", which is included in the semantic field along with the lexical units "morning", "afternoon", "evening" and "night" . Thus, in order to accurately determine the boundaries of the meaning of the words “morning” or “evening”, it is necessary to know which segments of the time of day exist in a given language as independent entities, i.e. understand the volume and content of the entire semantic field.

Semantic fields can contain identical linguistic units with multiple meanings, intersect and be associated with other semantic fields, thus creating a language system. At the same time, according to the neo-Hum-Boldtian linguo-philosophical concept developed by Weisgerber, the conceptual systems underlying specific languages ​​are unique. In structural semantics and the related semantic field theory, individual concepts, as well as the meanings of the lexical units denoting them, are considered not as objectively existing, extra-linguistic phenomena, but as “spiritual objects” created by language that structure reality.

S.I. Nekrasov, N.S. Molchanov. The Importance of Frame Theory

However, in linguistics, the term “frame” was first used only in the 1980s by Charles Fillmore within the framework of frame semantics to define a system of concepts of language, the understanding and use of which is based on the totality of human knowledge and experience in a given period of time. If Trier and Weisgerber talked about uniting linguistic units into semantic fields regardless of the social situation, then Fillmore focuses on the specifics of specific situations in which certain words arise, are used and change. The Fillmore frame is a kind of semantic scheme of the world, the elements of which are described by certain lexical units. Thus, in English, the meanings of the words “shore” and “coast” (translated into Russian as “shore”) differ in relation to the boundary between land and water, viewed from the land side or from the water side. The statement “We will soon reach the coast” implies a journey by land, and “We will soon reach the shore” - a journey by water. Capturing such nuances is due to a person’s ability to recognize ways of schematizing the world with the help of words1.

A huge variety of such frame schemes are formed depending on cultural traditions, areas of human activity, specific “closed” situations, motivations of the speaker, etc. A separate frame, as part of the surrounding reality, structures and clarifies the meaning of a word that will be used in communication in a certain place and at a certain point in time, and a word spoken at another point in time and in another place activates another frame, i.e. that part of experience and knowledge that lies beyond verbal interaction.

Although the use and understanding of language in a social context became the subject of study by linguists in the second half of the twentieth century, already in the 1930s F. Bartlett defined the concept of “mental schema”, a schema represented in memory. This concept later entered cognitive psychology. The process of perceiving reality through speech communication is determined by knowledge structured in memory in the form of meanings of linguistic units, images, and ideas.

Bartlett's scheme is the same Minsky frame, the same structure containing knowledge about the surrounding reality, determining human behavior and modifying ideas depending on specific conditions. Bartlett's scheme is the same Philmore frame, the same semantic field associated with experience recorded in memory. In the process of communication, memory, stimulated by words or context, affects corresponding frames, like Minsky's artificial intelligence, frames that interact with each other2. Thus, in the last century, scientists from different countries at different times and in different scientific fields developed, if not the same, then very similar theories.

Thus, in the philosophy of science, T. Kuhn’s theory of scientific revolutions also contains elements of the theory of frames, namely the concept of “paradigm” he uses, which is understood as a set of basic theoretical views, classical models of research, methodological tools that are recognized by all members of the scientific community. Kuhn's scientific paradigm is a part of scientific experience, a number of theories, concepts and views that define and structure the worldview of a scientist at a given stage of the development of science; in turn, the scientist formulates theories and develops approaches within the framework of the same scientific paradigm, using the knowledge presented in it. Consequently, the scientific paradigm can be considered as a frame that determines the perception and formation of scientific reality.

William James in his book “Principles of Psychology” says that there are probably many realities, each of which implies its own special and separate style of existence: the world of physical things, the world of science, the world of ideal relationships, the worlds of mythology and religion, the worlds of individual beliefs and etc. In this case, reality can be considered as an “edge”3, i.e. endowing an object or abstraction with certain properties (relevance, size, color, etc.) or conclusion

1 Kolesnik N.V. Frame semantics of Ch. Fillmore // Language, consciousness, communication: Collection. articles / Rep. ed. V. V. Krasnykh, A. I. Izotov. M.: MAKS Press, 2002. Issue. 22. P. 63.

2 Digoeva E. Language and science of the twentieth century / / http://www.darial-online.ru/1999_5/digoeva.shtml.

3 Schutz A. On the plurality of realities // Sociological Review. Volume 3. No. 2. 2003. P. 3.

16 SCIENTIFIC GUIDES No. 16(71) 2009

meaning in the frame of a special style of existence. A person, in the process of his activity, constantly acts within the framework of one or another reality, necessarily following the rules and norms that determine its structure.

By choosing an object of research, a scientist enters an already structured world of scientific contemplation, transmitted to him by the historical tradition of his science. From that moment on, he is included in a discourse that includes the results obtained by others, the problems raised by others, the solutions proposed by others, the methods developed by others. This theoretical foundation of a particular science is the final area of ​​meaning (in A. Schutz’s terms), the frame within which specific problems should be developed. Thus, a scientist must always remember that the subject of his research is inextricably linked with the components of a certain scientific field.

In sociology, almost simultaneously with M. Minsky, the frame was considered by Erving Goffman in his essay “Frame Analysis” (1974). However, in Goffman’s frame appears only as a tool or method for describing human social behavior in the process of interpersonal communication; Goffman’s frame fits into sociology and does not pretend to exist as a separate science of “frameology” or theory. Goffman's use of this term follows Gregory Bateson, who in the 1950s described the fighting behavior of great apes. Bateson proposed that primates imitating a fight exchange metacommunicative signals with the help of which they define the situation as either “real reality” or “make-believe reality.” These signals, transforming empirical reality, contribute to the process of interaction associated with understanding the content of a particular situation. Bateson uses the term “frame” to define the structural features of communication, namely the use of metacommunicative and metalinguistic messages5.

Goffman calls frames “definitions of a situation based on the principles of organization and involvement in events that govern events”6. This concept is close to the concept of a mental schema introduced by F. Bartlett, which determines an individual’s reaction to a certain situation, and the concept of frame semantics by Charles Fillmore, which explains the understanding of linguistic expressions by indicating typical situations of word use.

Goffman deploys a system of microsociological descriptions and frames that are constantly changing under the influence of social reality. Frame systems are not defined as strict algorithms of perception; reality is constantly framed, i.e. when correlating the real situation with the ideal image enclosed in a frame, the boundaries of this chosen frame are deformed in accordance with subjective perception, external circumstances and conditions, in accordance with the style of existence of the reality in which social interaction occurs. But unlike G. Bateson, who believed framing to be a psychological process, Goffman considers it as a property of the organization of events and perception. Goffman's frame serves to contextualize social interaction and indicate its structural characteristics, thereby acquiring a certain duality. A frame is both a “matrix of possible events”, which is based on the “arrangement of roles” of interacting subjects, and a “scheme of interpretation”, which is part of any perception7.

The fact that M. Minsky and I. Goffman simultaneously and independently develop “their” essentially similar theories of frames speaks of possible continuity not only between the sciences of the same field (only humanitarian or only technical), but also between sciences of different cycles - humanitarian and technical. However, artificial intelligence is a narrow field of study, although based on

4 Ibid. P. 29.

5 Vakhshtain V. “Practice” vs. “frame”: alternative projects for studying the everyday world // Sociological Review. Volume 7. No. 1. 2008. P. 82.

6 Batygin G. S. Continuum of frames: the sociological theory of Erving Goffman // Goffman I. Analysis of frames: an essay on the organization of everyday experience: Trans. from English / Ed. G.S. Batygin and L.A. Kozlova; entry article by G.S. Batygina. M.: Institute of Sociology RAS, 2003. P. 43.

7 Vakhshtain V. Sociology of everyday life: from “practice” to “frame” / Sociological Review. Volume 5. No. 1. 2006. P. 71.

S.I. Nekrasov, N.S. Molchanov. The importance of frame theory...

knowledge about psychological and social processes, but strictly aimed at creating a machine model of the real world. Sociology, the purpose of which is to identify patterns and mechanisms of interpersonal interaction in various social groups, in this regard provides more opportunities for the use and development of scientific terminology and the principles of “frameology”. M. Minsky has many assumptions and reservations regarding the application of frame theory: in each specific case, his theory is supplemented either by structural components, or organizational characteristics, or options associated with circumstances, intentions of role subjects, and psychological characteristics. Therefore, Minsky's frame theory is often called only a “frame approach” to studying the problem of knowledge representation.

The importance of frame theory in its modern multifaceted formulation is difficult to overestimate. Despite the differences in terminology, the initial provisions of the theory have been formed by scientists since the beginning of the twentieth century in a variety of scientific fields. The basic principles of frame theory can be applied in any field of science to systematize knowledge about various objects of reality based on a person’s psychological, social and linguistic experience. The universality of frame theory is based on the fact that any knowledge must be classified, structured, schematized, and put into form. The history of science represents a variety of views and knowledge, formed into trends, directions, schools, concepts, which have their own characteristics and boundaries that separate some theories from others. Any knowledge is formulated into a theory within the framework of a certain picture of the world, containing scientific and philosophical ideas about reality, which fragment and structure its very [picture of the world] content. The theory of frames is most consistently revealed in sociology and cognitive psychology, because here the frame is considered not only as a static structure of knowledge that determines the relationship between the elements of the system, but also as a scheme of cultural experience that changes in the process of perception.

1. Kolesnik N.V. Frame semantics of Ch. Fillmore // Language, consciousness, communication: Collection. articles / Rep. ed. V. V. Krasnykh, A. I. Izotov. M.: MAKS Press, 2002. Issue. 22.

2. Digoeva E. Language and science of the twentieth century // http://www.darial-online.ru/1999_5/digoeva.shtml.

3. Schutz A. On the plurality of realities // Sociological Review. T. 3. No. 2. 2003.

4. Vakhshtain V. “Practice” vs. “frame”: alternative projects for studying the everyday world // Sociological Review. Volume 7. No. 1. 2008.

5. Batygin G. S. Continuum of frames: the sociological theory of Erving Goffman // Goffman I. Analysis of frames: an essay on the organization of everyday experience: Trans. from English / Ed. G.S. Batygin and L.A. Kozlova; entry article by G.S. Batygina. M.: Institute of Sociology RAS, 2003.

6. Vakhshtain V. Sociology of everyday life: from “practice” to “frame” / Sociological Review. Volume 5. No. 1. 2006.

Bibliography

ROLE OF THE FRAME THEORY IN CONTEMPORARY SCIENCE

Moscow state technical university of civil aircraft

[email protected]

S. I. Nekrasov N. S. Molchanova

The frame theory is used for structuring and interpretation of knowledge system in different scientific fields. The main ideas of the theory were formulated in different sciences separately. Therefore it is difficult to determine in what field the “frame-work” was formed first. The notion of “frame” is used as an instrument due to the necessity to structure scientific knowledge and cultural experience. The ideas of the frame theory describe the processes of design and perception of the social reality. Therefore they are important in scientific research of psychology, sociology, linguistics.

Key words: Frame, framology, scheme, structure, frame theory, knowledge, reality.

Relational frame theory: a review of the controversy

Amy C Gross and Eric J Fox

Although Skinner's book Verbal Behavior (1957) was published more than 50 years ago, research into the behavioral analysis of language and perception has developed slowly. In recent years, a new approach to language learning known as relational frame theory has attracted considerable attention and generated much research and debate. It can be difficult to fully appreciate the controversy surrounding relational frame theory, partly because of the complexity of the theory itself and partly because the debate has continued for several years in several publications. The purpose of this article is to provide a brief overview of relational frame theory and summarize the key points of the debate.

Keywords: relational frame theory

The first comprehensive paper presenting relational frame theory as an approach to the study of human language in behavior analysis was presented at the annual Association for Behavior Analysis in 1985 (Hayes & Brownstein). Since then, relational frame theory has generated a great deal of research, discussion, and debate. It was not uncommon to witness heated debates about relational frame theory during symposia and conference sessions on the behavioral approach. Numerous articles have been published criticizing this approach, usually dating back to the first large-scale study of the book format (). While behavior analysts were accustomed to harsh criticism from outsiders from those who did not share their worldview (e.g., Chomsky, 1959), such intense controversy was not typical of behavior theory within the field. It is probably true to say that relational frame theory has become one of the most controversial, hotly contested topics in contemporary behavior analysis.

Much of the discussion surrounding relational frame theory seems to stem from two original sources. First, this approach to language learning differs significantly from that proposed by the “founding father” of our discipline, Skinner (1957). Although relational frame theory is in some respects an extension of Skinner's views on verbal behavior (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2000), its proponents also directly criticize key components of Skinner's analysis. Given Skinner's importance and relevance to the development of behavioral psychology, it is not surprising that a theory that approaches such an important topic as verbal behavior from a different perspective may be met with fear, disdain, and even disdain.

Second, if relational frame theory is correct, it has major implications for the future development of the science of human behavior (Hayes & Berens, 2004). The change in properties of stimuli examined in the literature on stimulus relations suggests that stimuli can influence behavior solely through their participation in verbal relations with other events.

Research on stimulus equivalence has shown that stimulus properties generally transfer from one element of the equivalent class to another (see, for example, Augustson & Dougher, 1997; Dougher, Augustson, Markham, Greenway, & Wulfert, 1994; Dymond & Barnes, 1994 ; ). Research on other derived stimulus relationships has shown that the behavioral properties of a stimulus can also be modified by its relationships with other stimuli (Dougher, Hamilton, Fink, & Harrington, 2007; Dymond & Barnes, 1995; Roche & Barnes, 1997; Roche, Barnes-Holmes, Smeets, Barnes-Holmes, & McGeady, 2000).

For example, if someone infers the arbitrary relation “A is greater than B,” and B then becomes a behaviorally conditioned reinforcer, then even without further training, A may acquire more reward properties for the individual (for example, the person will choose A but not B, or will work harder to get A but not B). Such changes in stimulus properties mean that our ability to predict and influence human behavior through the power of our speech will be greatly hampered if we rely solely on analysis of the immediate consequences of behavior. We also need to consider the person's verbal behavior. Such analyzes often require the development of new experimental techniques and may result in the use of interventions that are likely to seem foreign and irritating to many behavior analysts. For example, the best-known extension of relational frame theory, acceptance and commitment therapy (Hayes, Strosahl, & Wilson, 1999), often uses mindfulness meditation and practice exercises that may seem outside the scope of behavioral psychology.

However, relational frame theory has brought important philosophical, conceptual, and empirical questions to the forefront of our research field. Although relational frame theory in general and its controversy in particular have received much attention, it appears that the theory is still not well understood by many behavior analysts. This is partly due to the complexity of the theory. Although the core tenet of relational frame theory is relatively simple (the relationship between stimuli is a learned operant), fully understanding research and analysis of the theory can be challenging and requires familiarity with its general principles, ideas, and terms.

Even experienced behavior analysts need to learn some new terms to understand relational frame theory. Another problem is that criticism of the theory and responses to these criticisms have been scattered across various journals for several years. Without first conducting an extensive review of the literature, it may be difficult to fully appreciate the nature of the debate surrounding relational frame theory, given the excellent arguments made on both sides. To help remedy this situation, in this article we have set ourselves the task of presenting a general overview of relational frame theory, summarizing the main criticisms of this theory that we have found in the literature, and also presenting responses to these criticisms by adherents of the theory.

What is relational frame theory?

Relational frame theory is a behavioral analytic explanation of the processes of human speech and perception. It is similar in principle to Skinner's theory, and differs from most cognitive and linguistic approaches to language learning in that it “views verbal events as a process rather than an outcome” (Hayes, Fox, et al., 2001, p. 22) . Relational frame theory differs fundamentally from Skinner's approach in the way it defines and explains verbal events and processes.

A challenge to Skinner's theory of verbal behavior (1957)

Informal conversations with behavior analysts sometimes reveal their overconfidence in their perception of the topic of verbal behavior. Many people seem to believe that, thanks to Skinner's (1957) book, the topic has already been addressed and studied to a sufficient extent. Unfortunately, the impact of Skinner's analysis on research and practical application has been limited. Although there is a growing body of research based on Skinner's theory of verbal operants (Sautter & LeBlanc, 2006), the volume and scope of these studies leave much to be desired. Many researchers focus on teaching children with developmental disabilities (Sautter & LeBlanc) basic verbal operants (primarily mand and tact). Moreover, Dymond, O'Hora, Whelan, and O'Donovan (2006) found that between 1984 and 2004, “Verbal Behavior was the most frequently cited topic in non-empirical articles” (p. 81).

Relational frame theorists have argued that the limited influence of Skinner's theory of verbal behavior may be due to the way Skinner defined verbal behavior and verbal stimuli. Skinner defined verbal behavior as behavior that is reinforced by another organism that has been taught by others to provide similar reinforcement. Hayes, Blackledge, and Barnes-Holmes (2001) argue that this definition is too broad because many behaviors are inherently socially mediated, so the definition fails to distinguish verbal behavior from virtually any other social behavior. Skinner noted that this definition also included the behavior of non-human animals in the experimental chamber; animal and experimenter constitute “a small but very real verbal community” (p. 108). Such a general definition raises the question: why do we need a separate definition or interpretation of verbal behavior? Moreover, its use does not lead to any visible improvements in the techniques we can use to study verbal behavior (). Relational frame theorists have argued that Skinner's definition of verbal behavior is nonfunctional because it includes the history of behavior of another organism (e.g., the listener) as a distinctive feature (Hayes, Blackledge, & Barnes-Holmes, 2001).

In all other areas of behavior analysis, behavior is defined depending on its function, and not on the history of behavior of another organism. Considering the behavior of another organism in determining Skinner's verbal behavior, proponents of relational frame theory argue that Skinner placed the researcher in a special position: in order to classify the behavior of the organism of interest (the speaker), he needed to study the history of another organism (the listener). This requirement, some argue, can lead to both conceptual and methodological confusion when one attempts to conduct research on verbal behavior. However, some other analysts do not consider Skinner's definition of verbal behavior to be problematic in this regard (see, for example, Leigland, 1997; Palmer, 2008).

In addition, relational frame theory researchers and others have criticized Skinner's definition of a verbal stimulus (Hayes, Blackledge, & Barnes-Holmes, 2001; Hayes & Hayes, 1992). Again, the main complaint was that Skinner's definition was non-functional. For Skinner, a verbal stimulus is simply a product of verbal behavior. This is a unique way of classifying stimuli in behavior analysis because such classification is based on the source of the stimuli rather than their function for the organism of interest (Hayes, Blackledge, & Barnes-Holmes, 2001).

Behavior analysts characterize other stimuli (including reward, punishment, and discriminative stimuli) by their functions rather than by what produces them. Some may see this as a minor quibble, but the meaning and coherence of Skinner's taxonomy of verbal operants suffers without a functional definition of the verbal stimulus. In order to correctly classify any (except mand) response as an example of one of Skinner's verbal operants, it is necessary to first determine whether the behavior-controlling stimulus is verbal. For example, imagine that a girl in the forest hears “cuckoo” and says “clock.” Is this reaction a tact or an intraverbal? Using Skinner's definitions, this depends not on the function of the stimulus, but on whether it is a product of verbal behavior. To categorize the girl's response in this case, we must consider the key variable to be the source of the stimulus, not its function. If the sound "peek-a-boo" was, for example, the result of her brother's verbal behavior, then the girl's reaction would be intraverbal. However, if the peek-a-boo sound was made by a bird (i.e., was not a product of verbal behavior), then the girl's response would be tactful. How we categorize a girl's verbal behavior depends on the source of the controlling stimulus, despite the fact that the girl responds the same way in both situations to the same perceived stimuli. This method of classification is akin to classifying stimuli according to their topographic properties, and is clearly not functional.

Perspective on the relationships between derived stimuli

Research on stimulus equivalence and derived relationships has provided behavior analysts with new insights into the treatment of verbal and rule-governed behavior (Hayes, Barnes-Holmes, & Roche, 2001). We now know that most people are able to voluntarily establish relationships between different stimulus-events without direct training or instruction. For example, a person who has been taught that A = B and A = C, when presented with new stimuli, will most likely be able to infer the relation B = C. This relationship between B and C is arbitrary because the stimuli are not physically identical (not equivalent). It is also derivative because it was not taught directly. Many relations other than equivalence can be inferred in this way: for example, comparison and contrast relations (Dymond & Barnes, 1995; Green, Stromer, & Mackay, 1993; Roche & Barnes, 1996). Moreover, humans are capable of constructing highly complex “relational networks” using only a few directly learned relationships (Steele & Hayes, 1991). A description of the nature of reactions that are a derivative of the interaction of stimuli is beyond the scope of this article; however, other researchers have discussed this issue (e.g., Carr, Wilkinson, Blackman, & McIlvane, 2000 ; Devany, Hayes, & Nelson, 1986 ; Hayes, Fox, et al., 2001, p. 28). Moreover, the behavioral properties of a stimulus can change depending on its derived relationships with other stimuli. For example, if A is a conditional reward for an individual, and he independently deduces that B is opposite to A, then B will most likely acquire the properties of a conditional punishment. This effect has been demonstrated with a variety of stimulus properties, including conditioned reward properties (Dymond & Barnes, 1995; Hayes, Brownstein, Devany, Kohlenberg, & Shelby, 1987; Hayes, Kohlenberg, & Hayes, 1991; Roche & Barnes, 1997) distinctiveness properties (Hayes et al., 1987), conditioned emotional responses Dougher et al., 1994), extinction properties (Dougher et al.), and self-determination (Dymond & Barnes, 1994).

Derived relationships between stimuli can generate a useful model for the analysis of speech and other complex behaviors in humans. The arbitrary nature of the relations between stimuli corresponds to the symbolism or referentiality of language, in which words and their referents usually share only a few formal properties (for example, the word "fox" is nothing like a real fox), but people often respond to them as if they were are equivalent and share many psychological characteristics (Sidman & Tailby, 1982).

The phenomenon of the emergence of complex networks of relationships between stimuli following direct training on only some of these relationships may explain the remarkable generativity of language (Barnes-Holmes, Hayes, Dymond, & O'Hora, 2001). Indeed, there is growing empirical evidence that derived stimulus relations are associated with language development. Verbal humans readily demonstrate the occurrence of derived stimulus relations, but humans with verbal impairments and nonhuman animals do not demonstrate these behaviors convincingly and unambiguously (Barnes, McCullagh, & Keenan, 1990; Devany et al., 1986; Dugdale & Lowe, 2000; Hayes, 1989; Sidman & Tailby, 1982). In addition, researchers have shown that the ability to establish derived stimulus relationships correlates with cognitive and verbal skills (Barnes et al., 1990; Devany et al., 1986; O'Hora, Peláez, & Barnes-Holmes, 2005; O'Hora et al. al., 2008). It arises in early childhood but develops gradually and at about the same time as language skills (Lipkens, Hayes, & Hayes, 1993). And finally, derived relations produce the so-called. priming effect, evoked potential difference, and patterns of neural activity that resemble patterns that occur during semantic processing (Barnes-Holmes et al., 2005).

Derived stimulus relations pose a problem for behavior analysts because it is impossible to derive results from a strict conditioning paradigm; therefore the relationship is often called derivative or emergent. According to relational frame theory, derived stimulus relations represent a special form of relational response. Most living organisms, from insects to birds to mammals, can learn to respond to certain physical properties of stimuli, for example by choosing the brightest or longest of two stimuli (a phenomenon commonly known as “transposition”; for a review, see Reese, 1968). . Relational responding of this kind has some interesting characteristics. First: stimulus relationships are two-way, and all relational responses reflect this. If A is somehow connected with B, then B is necessarily connected in the same way with A. When an organism chooses a stimulus because it is, for example, longer than another, it simultaneously rejects the second stimulus because it is shorter than the first. The relationship between two stimuli in one direction (from A to B) entails a relationship between stimuli in the opposite direction (from B to A). In relational frame theory, this property is called "mutual correspondence" (in equivalent relations, this property is often called "symmetry"). The second characteristic of relational responding is that relationships between stimuli can be connected in such a way that new relationships emerge. When ordering three objects by size, for example, responses to two kinds of stimulus relations may reveal a third kind. For example, you might place a watermelon to the left of an apple because it is larger than an apple, and a cherry to the right of an apple because it is smaller than an apple. By responding to these two relations between stimuli (watermelon is larger than an apple and cherry is smaller than an apple), a third relation appears: watermelon is larger than a cherry. This relationship between stimuli occurs when two other stimuli are combined, and can be determined without directly comparing watermelon to cherry.

The way in which two-way stimulus relations can combine to give rise to new ones is called “combinatorial correspondence” (in equivalent relations this property is often called “transitivity”). Proponents of the theory of relational frames argue that the early language training that most people undergo makes it possible to subsequently monitor arbitrary contextual and social cues, rather than simply the formal properties of stimuli, to control relational responses. For example, in the presence of a contextual cue “X is greater than Y,” a person may learn to respond as if X is indeed greater than Y (for example, by pointing to X after asking “Which is greater?”). Moreover, because mutual correspondence is a property of relational responding, in this context a response demonstrating that Y is less than X is likely to be reinforced (e.g., pointing to Y after asking “Which is less?”). In addition, since the property of relational response is also combinatorial correspondence, as we said, if “Y is greater than Z,” then we can combine the two indicated relations and, thus, answer that X is also greater than Z. All these relations between X, Y and Z are arbitrary because they are not based on physical properties, and some of them are derived (rather than directly learned) due to the characteristics of relational response - bidirectionality and combinatorial properties. From the perspective of relational frame theory, people are able to establish derived relationships between stimuli because they learn voluntary relational responses to environmental stimuli based on contextual cues.

Different types of relationships also have different response models, and the term “relational frame” (literally “relationship frame”) is used to describe a universal model of arbitrary relational response, which has the properties of mutual correspondence, combinatorial correspondence and transformation of stimulus properties.

Contextual cues define both the relevant relationships and properties that will be transformed into a relational frame. The metaphor “frame” is used to emphasize the idea that this type of response can include any stimulus-event, even new ones, just as any picture can be framed” (Hayes, Fox, et al., 2001 , p. 34) and corresponds to Skinner's term “autoclitic frame” (1957). A number of relational frames have been identified and studied, including agreement, contrast, distinction, comparison, hierarchy, and deictic other-perspective frames (Hayes, Barnes-Holmes, & Roche, 2001). Although researchers use the noun “relational frame” for convenience, it is important to remember that relational frames (and relational networks) describe behavioral repertoires, not hypothetical or assumed mental structures or information constructs. Specifically, relational frames refer to situationally context-controlled patterns of relational responding that people acquire due to the consequences (e.g., reinforcement) provided by the verbal and social community.

According to relational frame theory, voluntary relational responding is the basis of human language and perception; therefore, the definition of “verbal behavior” is “the placing of events in a relational frame” (Hayes, Fox, et al., 2001, p. 43). Accordingly, the definition of “verbal stimuli” is “stimuli that influence behavior by being part of a relational frame” (Hayes, Fox, et al., 2001, p. 44). The history of the behavior of the organism itself, and not the history of another organism or the listener, is the basis for the emergence of the functions of verbal stimuli. In relational frame theory analysis, both the speaker and the listener are involved in verbal behavior. The speaker - because he produces stimuli based on events contained in a relational frame, and the listener - because he reacts, focusing on these events.

Relational frame theory as an approach to the study of verbal behavior is leading to an increase in empirical research, applications, and the development of conceptual analysis. It serves as the theoretical basis for the dissemination of a type of psychotherapy known as “acceptance and commitment therapy” (Hayes et al., 1999). Moreover, the insights from relational frame theory have been used to study areas and topics such as psychological development, rule following, logical argumentation, persuasion and rhetoric, problem solving, social behavior, prejudice and stigma, learning from another perspective, sexual attraction and even religion and spirituality (see Hayes, Barnes-Holmes, & Roche, 2001).

Criticism of relational frame theory and response to criticism

Novelty of relational frame theory

One criticism of relational frame theory is that it is not truly post-Skinnerian because it shares many of the same fundamental principles as the Skinnerian approach (Osborne, 2003). Osborne argued that relational frame theory is simply an extension of Skinner's work on verbal behavior because, like Skinner's analysis, it describes complex human behavior using a limited set of behavioral laws. He noted that the fundamental laws of verbal and nonverbal behavior in humans and animals are unchanged. Thus, there is no paradigm hidden in relational frame theory” (p. 22). Adherents of relational frame theory hold the view that it is a new theory. These researchers presumably use the fundamental laws of behavior proposed by Skinner (1957); however, they go beyond Skinner's analysis of verbal behavior. However, relational frame theory is post-Skinnerian because of its approach to human language and perception (Hayes, Barnes-Holmes, & Roche, 2001, 2003).

New behavioral laws or new theory

Although Malott (2003) viewed relational responding as an important component of human speech, he was hesitant to call it a new law of behavior. Rather, he argued that fundamental behavioral processes can explain relational responding. Malott believed that due to stimulus control, processes of amplification and generalization, a complex behavioral chain is formed for the relational frame. Although he did not claim to have identified the correct fundamental law, he argued in discussions for a molecular approach to explaining human speech rather than the molar approach proposed by relational frame theorists. Similarly, Salzinger (2003) reported his concerns about the departure from once established laws of behavior. Indeed, he argued that it is rule-controlled behavior that can explain relational responding rather than that relational frame theory explains rule-controlled behavior. Accordingly, he concluded that relational frames may simply be a form of behavior under the control of rules, rather than the act of placing events into a relational frame.

Burgos (2003) argued that relational frame theory is not new because it builds on other theories. In particular, he made the following argument: relational frame theory is a combination of set theory and symbolic logic applied to verbal behavior. A detailed explanation of each of these theories is beyond the scope of this article; however, below we will briefly explain how relational frame theory can be a combination of set theory and symbolic logic. Relational frame theory is based on stimulus equivalence relations, and Burgos noted that the approach uses the language of set theory. Regarding symbolic logic, he argued that the authors of relational frame theory note that correspondence can be viewed from the point of view of symbolic logic. Burgos stated that additional ideas from relational frame theory, not just "mutual correspondence", are equivalent to concepts from symbolic logic. He also explained that many mathematical and philosophical theories combine set theory and symbolic logic, and are therefore very convincing. He concluded that the theory of relational frames, like these theories, has great potential and can be widely applied, since it is a combination of set theory and symbolic logic as applied to verbal behavior.

Relational frame theorists have rejected Malott's (2003) attempt to explain verbal behavior in terms of verbal chaining. Malott's analysis requires the use of vocal repertoire, and Hayes et al. (2003) cited several studies in which relational responding developed in the absence of vocal repertoires. In addition, relational frame theorists have pointed to the existence of several other possible molecular mediating agents of speech and perception, but have noted a lack of empirical evidence. Therefore, it does not appear that a behavioral chain or other molecular mediating events are sufficient to explain the complex process of speech and perception (Hayes et al., 2003).

The explanation offered by relational frame theory differs from Sidman's equivalence explanation in four important ways. First, although Sidman offered one of the earliest behavioral explanations of stimulus equivalence, his approach remains primarily descriptive. Indeed, he noted: “My theory was aimed not so much at explaining equivalence relations as at formulating a descriptive system—a consistent, consistent, and economical way of defining an observed phenomenon” (Sidman, 1994, p. 536). An accurate and consistent description of an empirical phenomenon is, of course, important, but it is not sufficient for a functional, behavioral explanation. Additionally, Sidman's explanation is a description of the behavioral phenomenon known as "stimulus equivalence," while relational frame theory is a behavioral explanation of how exactly this and other phenomena occur.

Secondly, the theory of relational frames from a scientific point of view is a more conservative way of explaining equivalence, since the process of explanation does not require the derivation of new behavioral laws (proponents of the theory propose a new behavioral law at the inference stage, but this is due to the unusual influence of the transformation of stimulus properties; Hayes, Fox, et al., 2001). On the other hand, Sidman argues that equivalence may be a fundamental property of a stimulus and that its mathematical explanation involves “new—previously unknown—behavioral variables or theoretical principles” (1994, p. 537).

Third, the terms used in relational frame theory to write equivalence and other derived stimulus relations are more inclusive than the terms adapted by Sidman and other equivalence researchers. Relational frame theory introduced these new terms to make it possible to talk about all kinds of stimulus relations (e.g., greater than, before-after, opposite, darker than). Sidman's terminology, borrowed from mathematical set theory and applied to equivalence relations, does not work for other types of stimulus relations except equivalence. Terminology with a broader scope was needed because relational frame theory researchers study many derived stimulus relations other than equivalence.

Fourth, as discussed above, the study and application of relational frame theory focuses on a broader range of stimulus relations, not just equivalence. Stimuli can be related to each other in a variety of ways and to varying degrees, and, presumably, each of these types of relationships between stimuli can be under the voluntary control of the situation. Although the issue of stimulus equivalence has generated the most excitement in the behavioral science community because of its apparent connection to word meaning and vocabulary development, it is difficult to find explanations for more complex lexical items such as sentences based solely on equivalence relations. By analyzing stimulus relationships other than equivalence, relational frame theorists have been able to provide compelling explanations for complex phenomena related to human language such as sentences, rules, spirituality, morality, etc. (see Hayes, Barnes-Holmes, & Roche, 2001). Moreover, the complex response patterns that are observed when learning multiple types of stimulus relations (see, e.g., Steele & Hayes, 1991), as well as the accompanying changes in stimulus properties, were predicted by relational frame theory rather than by Sidman's theory of stimulus equivalence.

Clarity of concept

Some authors have criticized relational frame theory for unclear or incomplete explanations of concepts. Malott went so far as to claim that “the difficulty of understanding their analyzes may be so great that they fail to convince anyone” (2003, p. 17). Another author stated that the analysis becomes less precise because the topics that relational frame theory examines are too broad” (Tonneau, 2004).

Certain misunderstandings have arisen regarding the description of higher order operants, the process of transfer of stimulus properties, relational frames as a class of responses, and the new law (or principle) of behavior. Galizio (2003a) believed that calling a relational frame a higher order operant is inaccurate. He argued that relational frame theory does not define a higher order operant; Accordingly, it makes no sense to call a relational frame that way. In addition, he wondered how the transformation of stimulus properties begins. He argued that the analysis proposed by the theory's proponents did not include the process of initial development of the transformation of stimulus properties, thereby making the description of this phenomenon unclear. Palmer (2004b) criticized the theory of relational frames for its vague explanation: whether relational frames are a class of reactions or part of an individual's history, due to which a class of reactions is formed. When theorists describe the relational frame as both an outcome and a process, the distinction becomes blurred (Palmer, 2004b). Finally, Palmer (2004a) suggested that if proponents of the theory claim that relational responses are a higher order operant and that transformation of stimulus properties occurs without direct learning, they must derive a new law or principle. He stated that “the establishment of a new law” (Palmer, 2004a, p. 231) is necessary for a complete explanation of relational operants.

In response to Palmer's (2004a) criticism that the theory lacks new laws and principles, proponents have argued that a new behavioral law is not necessary to address the idea of ​​relational operants. Hayes and Barnes-Holmes (2004) noted that relational responding is the result of differential reinforcement, a process that has been discussed in detail in behavioral analysis. They also stated that other function-oriented behavior analysts have no difficulty accepting the idea of ​​relational operants without deriving a new behavioral principle or law and that it is therefore unclear why Palmer had difficulty with this approach .

Variety of research areas

Relational frame theorists agree with Palmer (2004a) that people likely have hidden behaviors in relational frames, but they do not consider such behaviors to mediate events. Thus, the researchers did not limit the scope of application of the theory, but studied the relationship between hidden and observable reactions, using some variants of methods and procedures - such as the “methodology for assessing relational response”, “methodology for pairing respondents”, “development studies”, “method talk aloud", "response latency", "semantic priming" and implicit association testing procedures, as well as recording of evoked potentials. Through these experimental interventions, researchers develop empirical explanations for underlying forms of relational responding in search of explanations more precise than Palmer's (Hayes & Barnes-Holmes, 2004) mediating events. Moreover, in response to criticisms of mediating events, Hayes and Barnes-Holmes noted that mediating events are not central to the study of human language and perception. Differing opinions about the significance of mediating events are most likely due to unresolvable philosophical differences between researchers who are concerned with this issue and those who are less focused on it.

conclusions

Relational frame theorists are exploring an area that most behavior analytic researchers have not previously considered. Since these attempts began, other researchers and theorists have been quick to condemn their activities. They criticized the novelty of relational frame theory, the clarity of the concepts introduced, and the scope of the study. Proponents of the theory responded to all criticisms, despite the fact that the origins of these disputes may lie more in the field of philosophy than logic or empiricism.

Despite this, even critics agree that researchers are studying an important subject area (Malott, 2003: McIlvane, 2003; Spradlin, 2003). Behavior analysts have long ignored areas such as human language and perception, and relational frame theorists have ambitiously taken up the study of these complex and important topics (McIlvane). Although critics may not agree with all of the ideas in relational frame theory, they have recognized that the theory is useful in the study of language and perception (Galizio, 2003a; Malott; McIlvane). Applied and experimental behavior analysts are encouraged to consider relational frame theory when conducting research—especially when verbal behavior is an important independent or dependent variable. Moreover, the wide range of issues that relational frame theory addresses and the methods it employs may make the field of behavior analysis more attractive to those who have long considered behaviorism "dead" and irrelevant. We hope that the intense debate and controversy generated by relational frame theory will contribute to the development of behavioral analysis, as well as a deeper understanding of the complexity and behavioral significance of human language.

Translation by Anastasia Galitskaya

Original article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2779078/

Frame theory refers to psychological concepts related to understanding what we see and hear. These modes of perception are interpreted from a consistent point of view, and conceptual modeling is carried out on their basis.

To recognize the fact that given information in these areas has a single meaning, human memory must be able to associate this information with special conceptual objects. Otherwise, the information will not be able to be systematized.

Frame theory is based on the perception of facts by comparing information received from the outside with specific elements and meanings, as well as with frames, definitions for each conceptual object in our memory.

The structure that represents these frames is called a frame. Since there are some analogies between various conceptual objects, a hierarchical structure with classifying and generalizing properties of the “abstract - concrete” type is formed.

Complex objects are equivalent to frame networks. Each frame is complemented by associated factors and procedures that ensure that requests are made to each other.

Frame theory is more of a goal-setting theory than an outcome theory.

Mark Minsky, in 1975, in his work “Framework for Representing Knowledge in the Phychology of Computer Vision” described the basics of frame theory.

The essence of the theory: when a person finds himself in a new situation, he recalls in his memory a basic structure called a frame. A frame is a unit of knowledge representation, memorized in the past, the details of which can be changed according to the current situation. Each frame can be supplemented with various information, which may relate to how to use this frame, the consequences of this application, and the actions that need to be taken if the forecast does not come true. Each frame can be viewed as a network consisting of several vertices and relationships.

The top level is a fact about the state of an object that is considered to be true.

The next levels are a set of terminal slots that must be filled with specific values ​​and data.

Each slot specifies a condition that must be met when establishing a correspondence between values ​​(the slot either sets it itself or a smaller component of the frame is created.)

A simple condition is called a label and can contain requirements, for example, that the match must be set by the user, that the described value is sufficiently complete, that there is an indicator of the special component of the frames.

Complex conditions indicate relationships between facts corresponding to multiple terminals. By connecting many frames that are relationships, you can build a frame system. The most important result of this construction is the ability to transform frames in one system. Frames are used to cost-effectively carry out various calculations and image processing.

When analyzing a visible object, different frames of the native system describe it from different angles of view, and the transformation from one frame to another shows the result of the transition from one observation point to another. For non-visual frames, differences between frames in the same system indicate some action, cause-and-effect relationship, or change in conceptual perspective, etc. In the same system, different frames may share terminals. This is a serious point to pay attention to, since it makes it possible to link information obtained from different points of view.

Rice. 2. Connection of the frame with additional information

The main features of frame theory are manifested in the ability to predict and include other processes. Multiple terminals of the same frame are usually predefined with default values. Therefore, even when detailed information regarding a location is not given, the frame will still be quite informative. This method can be widely used for presenting general information, analyzing many similar problems, practicing techniques for solving problems using logical methods, and it can also serve as a useful means of disseminating frame systems. Since the default value is not hardcoded to the terminal, it can easily be replaced with new information, for example to better suit the relevant situation. Therefore, the default value can be used as a variable, either in special cases of inference from facts or in generally accepted cases. In addition, it can be used instead of logic limiters.

Frame systems are associated with information retrieval networks. If the candidate frame does not match the current problem, in other words, if the terminal match does not quite meet the cue condition, that network specifies another frame. Using this cross-frame structure, knowledge can be represented regarding facts, similarities, and other information useful for understanding. When a certain frame is selected as a unit of representation of a certain state, then during the coordination process such values ​​are substituted into all terminals of each frame so that the conditions in the appropriate places are met. This negotiation process is driven by information that complements the frame (this information relates to the processing of unforeseen situations), as well as the purpose of the system at the current moment in time. Further, if the approval process is unsuccessful, the use of the information obtained is of some interest.

This presentation may seem difficult to understand, so we will explain it with a simple example.

In Fig. Figure 2 shows the connection of the frame with additional information. A frame of this type describes only one conceptual object, and additional information is specific information for each frame. What is important in this example is that there is some physical object called a room, some conceptual object called a state within that room, and that the additional information concerns not only these objects, but also the specific manipulation of them.

In Fig. Figure 3 shows an example of a frame representation of a compound arch. In this case, the frame is presented as a semantic network. However, since a frame consists of many terminal slots, the relationships between them are described using their values. Conditions that limit the values ​​are also added to the corresponding slot. In addition, since parts A, B and C are represented by other frames, the arch frame slot is supplemented with pointers of these frames.

In Fig. Figure 3 shows an example of a frame system regarding the viewing angles of an arch. In this case, the front view of the arch is taken as the first angle of view. If we rotate it 90° clockwise, as shown at position 2, only parts A and B will be visible. When we view an object, for example from visual angle 2, then recalling visual angle frame 1 from memory, we can assume that this the object is an arch. In this way, the fact that part C is not visible is compensated; if, when rotating this object by 90 o counterclockwise, its appearance coincides with the frame of visual angle 1, the assumption was correct. Otherwise, another frame is called and compared. In this case, a similarity indicator can be used.

The main properties of frames are discussed below.

(1) Basic type. By effectively using a frame system like the one shown in Fig. 4, a quick understanding of the essence of a given object and its state can be achieved, but memory is required to remember various positions in the form of frames. Therefore, only the most important objects of a given subject are remembered in the form of basic frames, on the basis of which frames for new states are built. In this case, each frame contains a slot equipped with a substructure pointer, which allows different frames to share the same parts. This substructure does not change when the viewing angle changes. Thanks to this property, it is possible to present and use information obtained at different times and from different areas as independent information associated with a given point of view (i.e. as knowledge).

(2) Matching process. The process by which the correct frame selection is verified is called the matching process. Typically this process is carried out in accordance with the current goal and the information (values) contained in a given frame. In other words, the frame contains conditions that constrain the slot's values, and the target is used to determine which of these conditions is relevant in a given situation.

As a result, the frame matching process is carried out as follows:

1) first, using assumption and intuition, a certain basic frame is selected; with the help of knowledge based on identified features, relevance, or with the help of subframes assumed to be the most relevant, a given frame itself confirms or does not confirm its relevance. This determines which slot constraint should be used for matching based on the current target. When confirmed, the matching process is completed.

Otherwise:

2) if there is a slot in a given frame in which an error has occurred regarding, for example, a condition for consistency with default information, then information is needed to ensure that the appropriate value is assigned to this slot. For example, if in the process of matching the frame “viewing angle 1” shown in Fig. 4, only parts A and B are specified as information, then for a given frame information regarding part C is required. Assigning the required information to a given slot must not contradict slot restrictions and expectations. If the match fails even under the following circumstances (as an example):

Field of view on the right

Part C should be visible; if you can’t see it, turn right 90°,

Part A is supported by part B,

There is no connection between B and A, then

3) as a final step, control is transferred to another appropriate frame from this system. If this frame is not suitable, then control is transferred to the corresponding frame from another frame system. This process uses transformation information (for example, a change in viewing angle, a change in location, etc.). If this comparison ends in failure, then this means that the problem has no solution.

The matching process discussed above only concerned one frame, but in complex visual or text objects, a similar process should be introduced into the complete processing process.

(3) Hierarchical structure. A frame usually corresponds to a representation of a general concept with a classificatory hierarchical structure; as shown in fig. 5. The peculiarity of such a hierarchical structure is that the attribute information contained in the top-level frame is shared by all lower-level frames associated with it. For example, the attribute “animal moves” is common to both birds and the canary, which is at the lowest level.

If human memory had a similar structure, it would be possible to organize and remember similar concepts, avoid unnecessary complexity regarding attribute information, and add new concepts or knowledge to appropriate positions in this hierarchy (learning). At the same time, it would be easier to detect inconsistencies in knowledge and manage the sequence, and if one could understand that this is a bird even without using specific knowledge (for example, that this is a “chicken”), then the flexibility of the system would noticeably increase, for example, it would be possible to obtain conclusions based on knowledge about birds and animals. (It should be noted that such a hierarchical structure is characteristic not only for representing knowledge with frames. For example, in a knowledge representation language that combines frames and production systems, sometimes a part of such a structure is used as a database of the production system. In this case, such a system can also be called a frame system , however, although it is much more powerful than the production system, in its versatility it is inferior to the universal frame system.)

(4) Interframe networks. Memorizing a conceptual object that has a classification hierarchical structure is easily explained by a frame model, however, as mentioned above, if the “matching process” is unsuccessful, there is a need to search for a frame similar to the previous one. Such a search, carried out using difference pointers, is possible by connecting frames that describe objects with slight differences with these pointers and forming a network of similar frames.

In Fig. Figure 6 shows an example of the so-called Winston network. If, when matching the “chair” frame, it turns out that the object is too large and does not have a back, then using the difference pointer, the “table” frame is searched. If you find that the object is too wide and there is no back, you can effectively search for the “bench” frame using another difference marker.

In this example, we looked at difference pointers, but semantic networks can be built using all kinds of other pointers and use them to perform various inferences with high efficiency. (In such cases, the frame system should be considered to include a semantic network.)

(5) Default value. When a person considers something and thinks about what it would mean, or mentally imagines something and thinks about what it would mean, the process can be thought of as the distribution of specific values ​​among the terminal slots of a frame. Moreover, in the case of mental representation, the limits regarding the distribution of these values ​​are wide. In such cases, the assumed value is called the default value. For example, when reading the sentence “Taro took the ball,” the reader’s imagination draws not an abstract ball, but a very concrete one, for example, a tennis or golf ball. In addition, this ball must have certain attributes in his mind, for example, size, color and mass, the values ​​of which are implied by default. Undoubtedly, these meanings are evoked associatively on the basis of the reader’s personal experience. Such default values ​​are loosely associated with slots and then they are gradually replaced with reliable information.

This is based on the following assumption: “Until the terminal values ​​are allocated, no decision is made to store them in long-term memory. Until this point, the frame stores a default value, albeit loosely related to it.” Inferences derived from default values ​​are called default inferences. Based on these findings, you can continue to draw conclusions and fill in the deficiencies of the given information. This feature usually enhances the capabilities of the system, but at the same time there is a danger of drawing incorrect conclusions based on erroneous beliefs.

The default output serves a very important function in image or speech recognition. For example, if only part of the image is visible, then by replacing the remaining parts with default values, the complete image can be depicted. In the same way, using the default value, you can restore the meaning of the context from which individual sentences were snatched. As an example, consider the following text:

Hanako is invited to Taro's birthday party.

She believed that he liked to wind up the spinning top.

She headed to her room and shook her piggy bank.

The piggy bank didn't make any sound.

From such a text we can easily reconstruct the full content. To do this, it is necessary to understand in general what is said in the given sentences. You can easily understand that the main topic of the first sentence is a certain birthday. By focusing around the birthday frame and proceeding in a similar manner, you can assign values ​​to slots in an appropriate way and build a chain of frames:

birthday ® gift ® spinning top ® no spinning top ® buy ® money ® piggy bank ®

® shake ® no sound ® no money ® contact mother ® etc.

Here, highlighted places or default values ​​establish the connection between frames; they are not present in the given sentences. When this inference method is used, it is assumed that there is an efficient inter-frame network and daemons in addition to the default values.

(6) Relationships “abstract - concrete” and “whole - part”. The hierarchical structure discussed above is based on the “abstract-concrete” relationship, but in addition to this type of structure, there are others based on the “whole-part” relationship.

The “abstract-concrete” relationship is typical. that at the upper levels, as shown in Fig. 5, abstract objects (concepts) are located, and concrete objects are located at the lower levels, and the objects of the lower levels inherit the attributes of the objects of the upper levels. These relationships are also called IS-A or KIND-OF relationships. Such names are explained by the spelling forms “canary IS-A bird” (a canary is a bird) and “canary KIND-OF bird” (a canary is a type of bird).

Another whole-part relationship concerns structured objects and shows that a lower-level object is part of a higher-level object. For example, a wall is a structural element of an auditorium, but it is not an auditorium, so the “wall” object does not inherit the “audience” attribute, which is a top-level object. Rather, attribute inheritance is performed based on an IS-A relationship of the type “body - wall - auditorium wall - auditorium wall A”. Note that in Minsky's frame theory both types of these relationships were considered, but their separation, like the one given here, was not made.

Only the “abstract - concrete” relationship has received practical application in frame systems. However, sometimes it is necessary to describe and manipulate a structured object, for example in CAD, so in such cases it is necessary to handle whole-part relationships. In this case, the components of the system are described by IS-A relations, and the structure - by PART-OF relations. However, in PART-OF relationships, attribute inheritance cannot be used, which is why the advantages of the frame knowledge representation model are not visible. In such cases, other methods are necessary.