You are on page 1of 186

АЛЕКСЕ́ Й АЛЕКСЕ́ ЕВИЧ ЛЕО ́ НТЬЕВ (1936—2004) ALEKSÉI ALEKSÉEVITCH LEONTIEV (1936—2004

)

DISCUSSÕES SOBRE LINGUAGEM E PSICOLOGIA
(ARTIGOS DA J.R.E.E.P 2003 E 2006)

IMPRESSO EM UMUARAMA 25 DE ABRIL DE 2013

CONTENTS

Journal of Russian and East European Psychology, vol. 41, nos. 3/4, May– June/July–August A.A. LEONTIEV AND T.V. RYABOVA (AKHUTINA) The Phase Structure of the Speech Act and the Nature of Plans. pp. 33– 38.

Journal of Russian and East European Psychology, vol. 44, no. 3, May– June 2006. A.A. LEONTIEV The Social and the Natural in Semiotics. pp. 6–16. Sign and Activity. pp. 17–29. “Units” and Levels of Activity. pp. 30–46. Personality, Culture, Language pp. 47–56. Sense as a Psychological Concept. pp. 57–69. The Psychological Structure of Meaning. pp. 70–82. What Are the Types of Speech Activity? pp. 83–86. Some Problems in the General Theory of Speech Activity. pp. 89–103.

Journal of Russian and East European Psychology, vol. 44, no. 4, July– August 2006, A.A. LEONTIEV Psycholinguistic Units and Speech Generation. pp. 7–88.

Journal of Russian and East European Psychology, vol. 41, nos. 3/4, May–June/July–August, pp. 33–38. © 2003 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2003 $9.50 + 0.00.

A.A. LEONTIEV AND T.V. RYABOVA (AKHUTINA)

The Phase Structure of the Speech Act and the Nature of Plans
The notion of the phase structure of the speech act—or to be more precise—the special structure of the “inner speech” stage in utterance production, belongs to L.S. Vygotsky. Vygotsky conceptualized the process of speech production, the progress from thought to word to external speech, as follows: “from the motive that engenders a thought, to the formulation of that thought, its mediation by the inner word, and then by the meanings of external words, and finally, by words themselves”1 Elsewhere he said, “Thought is an internally mediated process. It moves from a vague desire to the mediated formulation of meaning, or rather, not the formulation, but the fulfillment of the thought in the word.” And finally, “Thought is not something ready-made that needs to be expressed. Thought strives to fulfill some function or goal. This is achieved by moving from the sensation of a task—through construction of meaning—to the elaboration of the thought itself.”2 These ideas of Vygotsky may be used to derive the following

English translation © 2003 M.E. Sharpe, Inc., from the Russian text “Fazovaia struktura rechevogo akta i priroda planov,” Plany i modeli budushchego v rechi (materiali k obsuzhdeniiu) [Plans and Models of the Future in Speech (Material for Discussion)], pp. 27–32. Tbilisi: Metsniereba, 1970. Translated by Lydia Razran Stone.
33

34

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

sequence of stages (phases) in speech production: The process starts with (a) the motive. We have only slightly expanded Vygotsky’s concept when we speak of the motive not as an isolated factor but as the set of extralinguistic factors that give rise to the motivation for a speech act (in the broadest sense of the term).3 This motivation gives rise to the (b) speech intention. This stage corresponds to a “vague desire” or “a sensation of a task.” (Cf. the category of the “imagined situation” in D.N. Uznadze.) At this phase, the speaker has “an image of the result” (Miller et al.), but does not yet have a Plan of Action that must be performed to achieve this result. (Here we should probably provide some separate discussion of the dynamics of motivation. According to A.N. Leontiev, we should distinguish between motive and need. The need is objectified in the motive, and a motive is “the object that satisfies a particular need and that, being reflected in one form or another, controls his behavior.”4 During the “motive” stage we are dealing with a need but not a motive per se. The shift from need to motive is associated with the concept of speech intention.) Vygotsky calls this the phase of “thought.” We have called it the stage of “speech intention” because the use of the term “thought” to mean a particular stage in the speech-thinking process requires special discussion. First of all, we have to ask about the meaning of “thought” in analysis of the system of concepts. It is clear that Vygotsky uses this word in two senses. First, it is a process, and second, it is a particular stage in that process. But, does speech production include an independent stage in which thought exists separately and independently of all other stages in the speech act? Let us point out, first of all, that thought, clearly does not necessarily entail verbalization of this thought (cf. the work of A.N. Leontiev and E.V. Il’enkov). Next, there are a number of possible different ways in which the thinking process is realized in different psychological situations, even given that the thinking is verbal. (Cf. the concept of “vicarious perceptual acts” of V.P. Zinchenko, corresponding to the thinking component of perception.) Vygotsky

MAY–JUNE/JULY–AUGUST 2003

35

supposes that “today I saw a barefoot boy in a blue shirt run down the street” is a “thought.” (“I see all of this together in a single act of thought, but in speech I segment it into individual words.”) Evidently, there is some terminological inaccuracy here: “I” the speaker do not simply “see” a boy, I see him in a form that is already mediated. “Boy” is already a secondary image that carries a number of attributive (= predicative) characteristics that are still not verbalized in an objective language code: the boy I saw today; the boy → ran was in a blue shirt; the boy → was barealong the street; the boy → foot. At the start of the process, there might also exist, in addition to the visual image (“communication of events”), a secondary image that is already verbally represented, or a verbally represented concept (“communication of relationships”). Finally, at the start of the process, there also may be an indistinct emotion that is not directly verbalized, and so on. It is important to emphasize that all these cases, as do all others that are not discussed here, differ psychologically. In each, thought takes different psychological forms. Thus, we have considered two stages of the speech act that are pre-speech in the strict sense of the term. During these stages consideration of the future and planning occur in two forms: (1) that of stochastic (probabilistic) prediction (cf. the work of R.M. Frumkina) and (2) that of the “image of the result,” which is a function of the structure of the action as a whole, that is, the goal of the action. The next stage, and, in our opinion, a theoretically very important stage is (c) the inner program of the speech act. This corresponds to Vygotsky’s “thought mediated by inner speech.” Obviously, it also corresponds to the category of Miller et al., the “grammatical plan” and N.I. Zhinkin’s category, “plan.” Representation of the speech intention in the code of personal “senses” (to use Leontiev’s terminology developing Vygotsky’s understanding) occurs during this stage. These “senses” are represented by some subjective code units (resulting from internalization of objective external actions); it could be N.I. Zhinkin’s “code of images

36

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

and schemata.” (We distinguish among the inner program, inner speech, and inner pronunciation, talking to oneself.)5 Thus, what is usually called (by Vygotsky among others) inner speech, and what we call here inner programming, is precisely the tool that fulfills thought, the connecting link between the intention that gives rise to thought and the elaboration of the thought in an objective linguistic code. The transition to this code itself is a two-step process: first there is the transition from “senses” embodied in a subjective code to meanings of the “external” words of an actual language (i.e., a “translation” from a subjective code of “senses” to the objective code of language meanings, representation of the speech intention by the “meanings of external words”), and next: “the transformation of the grammar of thoughts into the grammar of words” (because “thought has a different structure from that of its verbal expression.”) Thus, we have stage (d): implementation of the inner program, which entails two relatively independent processes—semantic implementation and grammatical realization implementation. (See Leontiev’s Psycholinguistic Units, with respect to the possible interactions and inner structure of these processes.)6 In addition to the two processes in stage (d) that have been described, we can identify one more—the process of the acousticarticulatory and morphological implementation of the program (representation of thought in “external words” to use Vygotsky’s terms). This process must follow the selection of the utterance’s syntactic structure and directly precede the next stage, that is, stage (e)—the acoustic implementation of the utterance, or phonation. Strictly speaking, this last stage entails a process of motor programming (“motor plan”) , which is superimposed on the processes of semantic and grammatical implementation and depends on them. It is only through this motor programming and through the next process of acoustic and articulatory implementation that phonation per se occurs. During stages (c)–(e) of speech production, there are three processes that involve consideration of the future: (1) probabilistic

MAY–JUNE/JULY–AUGUST 2003

37

prediction in the selection of grammatical constructions, and the semantic and phonetic features of words; (2) constructive prediction (when I select and start to implement, let us say, a particular syntactic construction, I am predicting the future continuation of this construction; when I select an element I am not only selecting it but also predetermining quite a number of subsequent elements and the nature of their interactions; the same is true with regard to “grammatical obligations,” that is, the selection of morphemes); (3) programming, that is, the creation of a system of “key elements” that predetermines selection and decision making during subsequent stages of speech production. In analyzing these stages of speech production from the standpoint of the role of simultaneity and succession, we can see that the processes of inner programming (both grammatical and motor) and grammatical structuring (grammatical or syntactic actualization of the utterance) are processes that occur through successive synthesis involving the combination of elements. (Compare the data on the “natural” order of the components of an utterance in children’s agrammatical speech, the speech of the deaf, spontaneous mimic and hand speech, autonomous speech, and certain other cases, where the order “agent-attribute-patient-attribute-predicate-circumstance” [as in the sentence “The cat his black ear licks lazily”] is clearly fixed. Semantic elaboration [semantic realization], the translation of units in subjective code into units in the code of an external language, and acoustic-articulatory implementation are based on simultaneous synthesis and entail selecting elements from a paradigm. This interpretation is confirmed by data on impairments in various forms of aphasia.)7 In conclusion, we would like to relate the model presented in this report to statements by A.R. Luria and L.S. Tsvetkova. They identify all the basic stages in speech production that we have described, but they do not separate the stage of inner programming and the process of grammatical realization of the program, merging them in the concept of the “inner schema of the utterance,” or the “dynamic schema of the sentence.”

38

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Notes
1. L.S. Vygotsky, [Thought and Language], in [Selected Psychological Research] (Moscow: APN RSFSR, 1956), pp. 380–81. 2. L.S. Vygotsky, [From an unpublished work by L.S. Vygotsky: The Problem of Consciousness], in [Psychology of Grammar], ed. A.A. Leontiev and T.V.Akhutina (Moscow: Moskovskii Gosudarstvennyi Universitet), p. 190. 3. Cf. A.A. Leontiev, [The Extralinguistic Dependence of Speech Acts as a Problem in Language Teaching], Inostr. Ia. v Shkole, 1968, no. 2. 4. A.N. Leontiev. [Needs, Motives, and Consciousness] (Moscow, 1971), p. 5. 5. See A.A. Leontiev, [Psycholinguistic Units and the Production of Utterances] (Moscow: Nauka, 1969). 6. See ibid., with respect to the possible interactions and inner structure of these processes. 7. See T.V. Ryabova [Akhutina], [Mechanism of Speech Production Based on the Study of Aphasia,] in [Issues of Speech Production and the Teaching of Languages], ed. A.A. Leontiev and T.V. Ryabova [Akhutina] (Moscow, 1967), pp. 76– 94, and T.V. Ryabova [Akhutina] and A.S. Shtern, [To the Analysis of Grammatical Structuring], in [The Psychology of Grammar], ed. A.A. Leontiev and T.V. Ryabova [Akhutina] (Moscow: Moskovskii Gosudarstvennyi Universitet), pp. 78–105.

6

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 6–16. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440301

A.A. LEONTIEV

The Social and the Natural in Semiotics
It is not the mission of this work1 to communicate any new, previously unknown facts. The present work will be devoted to a review and rethinking of certain concepts and theoretical principles that are based on facts that are already well known. We will endeavor to demonstrate that, given a different methodological approach, the same causes can give rise to different effects. We will start with the concept of semiotics, referred to in the title of this article. In recent years, use of the term “semiotics” has come to serve as a sort of signal that an author in principle draws no qualitative distinction between the sign systems of animals and human language. In our opinion, it is useful to return to the original sense of this term and recall the words of Ferdinand de Saussure, who believed that semiotics or semiology are “part of social psychology and consequently of general psychology. . . . Defining the exact place of semiology is a task for psychology.”2 In keeping with Saussure’s view, semiotics, as we see it, is a discipline that studies the role of signs in the formation and functioning of the

English translation © 2006 M.E. Sharpe, Inc., from the Russian text “Sotsial’noe i estestvennoe v semiotike,” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 9–19. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov.
6

MAY–JUNE 2006

7

human psyche. In other words, semiotics is the branch of psychology that deals with behavior that is sign oriented in nature and guided by signs. The main distinction between the sign behavior of man and the communication-conditioned behavior of animals lies in the specifically psychological nature of the “human” sign, something that is completely alien to contemporary semiotics. We will permit ourselves to explain this distinction using the words of L.S. Vygotsky:
[U]ntil quite recently . . . it was presumed that . . . a sound in and of itself could be associated with any experience, with any content of mental life, and for this reason can convey or communicate this content or this experience to another person. At the same time . . . in order to convey a particular experience or content of consciousness to another person, there is no alternative to relating the content being conveyed to a specific class, and this . . . makes generalization absolutely essential. . . . Thus, the higher-order forms of psychological association that are characteristic of man are possible only as a result of the fact that man, through thought, reflects reality in generalized form.3

It would seem that the animal is also capable of generalization, inasmuch as it is capable of functional identification of things that are materially different and of drawing functional distinctions between similar signals. But does this mean that we can place an equal sign between the two? Can we conclude that a word has a fixed meaning for man just because it replaces something, because it is standing for [English in original] another, nonverbal signal? Consider, for example, the frequently encountered ideas of Charles Osgood. It is well known that the concept of meaning is usually imbued with this very understanding, while the “social character” of a word or the meaning of a word is understood by what is common in it for many people, as society’s or the social environment’s verbal usage “adhering” to an individual, as boundaries placed around variations in the usage of a particular word through association with other people. Hence, meaning ceases to be a social fact; its social aspect turns out to be extremely superficial, reduced to the commonality of principles according to which a word substitutes for nonverbal signals in different individuals.

8

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

It appears that there is a serious misunderstanding here, stemming from the fact that humanity is understood as a totality of individuals, of biological beings who live on their own in the biological world and only occasionally come together with other individuals for various purposes. This, however, does not correspond to reality: the social nature of a human being is part of what defines him as the biological species Homo sapiens. “All individual beings that we place into a single species belong to it specifically because they are connected by a certain number of properties common to all, properties inherited from a common ancestor.”4 The biological nature of every newborn animal reproduces changes that have accumulated throughout the history of the species. And, what is specific to the species and is realized in the individual is primarily morphological features, characteristics of the structure of the body of animals. Progressive development, evolution in the world of animals results in the improvement of biological adaptation of animals of a given species to the life conditions of that species. The pace of human development is not at all comparable with the pace of evolution in the world of animals. From the appearance of the first stone axe with a wooden handle, to man’s first flight into space, the horse, for example, barely had time to replace its three toes with a hoof. But, while evolution was progressing at such an amazing rate, human morphology essentially did not change; if you could dress a Cro-Magnon in European clothing and walk him through the streets of London, he would barely be given a second glance. It was not in the biological sphere that the evolution of the “human” species took place, but [in] some other sphere; it was not in the form of morphological changes that the accumulation of features and experiences belonging to the species took shape, but in some other form. The sphere was the social life of man, and the form was the preservation of the achievements of human activity within the sociohistorical experience of humanity. Now, the experience of the species is reflected not in changes to the structure of the human arm, for example, but in changes to the tool the arm uses, that is, the techniques and ways of using it that

MAY–JUNE 2006

9

are generalized in it, that are encompassed in it. “In the tools of labor, man acquires something akin to new organs that change his anatomical structure. Since the point in time where he attained the level at which he started using them, he has endowed the history of his development with a completely new feature. In the past, this history, like the history of all other animals, amounted to modifications in his natural organs; now it has turned into something that is primarily the history of advances in his artificial [functional] organs.”5 Another reason that the pace of human evolution is incommensurable with the pace of animal evolution is that man is never on his own in dealing with nature, as animals are (even Robinson Crusoe had centuries of human experience at his disposal on his island). Man’s relationship with nature is mediated by his relationship with society. He is always able to draw from the storehouse of social experience, and, therefore, does not need to experience everything firsthand; he is always one step ahead of nature, not letting himself be taken by surprise, while the animal is always one step behind nature. Man learns from his mistakes— and even more so from the achievements—of others. The animal learns from his own mistakes. But if we accept the thesis that the evolution of man is primarily the evolution of his artificial [organic] organs,6 then it becomes apparent that the subject of this evolution, and also of interaction with nature generally, is not the human individual, but the human species as a whole, the socium.* It is not the individual person who interacts with the biological environment, but human society overall; this is why within this society laws of evolution such as the law of natural selection, become invalid. It is not by mere chance that the principle of morality that rejects the legitimacy, the “naturalness” of natural selection—[the word] humanism—comes from the word human. This applies not only to the practical, labor-oriented activity of man that is mediated by the tools of labor, but to the theoretical,
*Knowledge of the society by every individual.—Ed.

10 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

predominantly cognitive activity that is mediated by what Vygotsky called “psychological tools,” that is, signs. “As a part of the process of behavior, the psychological tool also alters the entire course and structure of mental functions, just as the technical tool alters the process of natural adaptation determining the form of labor operations.”7 And, in exactly the same way that the tool—together with the work skills and abilities that lie hidden in it and are objectified in it—is introduced into the practical activity of the individual “from the outside” (a child is given a spoon and taught how to use it), the sign, the word—together with the means of using this sign that is objectified in it, its meaning—is introduced into verbal-cognitive activity “from the outside.” But this means that, in talking about man as a biological being and ignoring the fact that it is not the individual who serves as the subject in interactions with nature, with the reality that surrounds him, but the socium (or rather the individual as the representative of the socium, as the carrier of not only biological but also social characteristics), we create not only a philosophical, but a psychological, inaccuracy. At the same time, within the problem of the “innateness” of linguistic ability, we encounter just such a “biologized” understanding. We will briefly mention the view held by Eric Lenneberg on this question. Heredity, in his view, provides the individual with “language readiness,” in which we find the “latent language structure”; a person’s language acquisition is the process of actualizing, of transforming this latent structure into a real one. “Social conditions can be viewed as a type of trigger that sets off a reaction. Perhaps the best metaphor is the concept of resonance.”8 According to this understanding, where the individual is viewed as causa sui, the motive for his linguistic development is [an] internal need, with no role whatsoever for “arbitrary external factors,” such as, for example, the influence of mature adults on the child. It is illustrative that in Lenneberg’s book, which certainly holds a prominent place in the psychological literature on speech, everything associated with the acoustic-articulatory aspect of language is well argued [by] using biological, physioanatomical,

MAY–JUNE 2006

11

and psychopathologic data. But as, soon as Lenneberg reaches the structure of higher levels, for example, syntactic levels, he is forced to rely on disconnected observations and theoretical speculations. Between the author’s factual bases and methodological conclusions there is empty territory permitting rather diverse paths for discussion and interpretation of the original data. In any event, it is not valid to apply results from study of the “expressive plane” to the “content plane” of language, as Lenneberg does. So, according to Lenneberg, the individual contains within himself all the prerequisites for his subsequent linguistic development, and society serves only as his social environment, providing the conditions for this development, but not serving as its cause. This development has two levels. “There are two different levels relevant for language: in the formation of the latent structure and in the process of actualization from the latent structure to the realized structure.”9 Are these levels really so different for Lenneberg, Chomsky, Katz, McNeill, and other proponents of the theory of “innate ideas”? For them, human behavior is made up of two components: biological and another, which can be called “biosocial.” The biosocial component is viewed as a superstructure above the biological component— similar to behaviorism. True, behaviorism is sharply criticized, but primarily for the fact that its proponents apply simple models that have been developed for animals to more complex forms of human behavior; and, because behaviorists try to explain too much through conditioning, that is, through a factor that is external—if not social.10 The specific nature of human “social” behavior, including linguistic behavior, is based primarily on the principle of “rules,” again rooted in a biological predisposition. Any question about there also being fundamental distinctions in the psychophysiological organization of processes in man that are specifically social, as well as in processes that are not strictly social, and, correspondingly, in animals (such as cognition; consider Lenneberg’s characteristic claim: “it is clear that there is no formal distinction between the formation of concepts in man and the ability of animals to react to categories of stimuli”),11 as well as the question of

12 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

whether behavior that is certainly biological, in particular in animals, can also be based on a principle of rules—these questions are not raised by theoreticians of “innate ideas.” In general, it appears that within this theory, an actual psychological and physiological analysis has largely been replaced by a logical-philosophical analysis. Whatever the case may be, for its proponents, biological and “biosocial” components of behavior are fundamentally united; they do not envision physiological mechanisms that are specific to human behavior. At the same time, mechanisms that sustain uniquely human abilities exist, and, as they pertain to sense of pitch and certain other cases, have recently been investigated by A.N. Leontiev, who, following the physiologist A.A. Ukhtomskii, has advanced the concept of a “functional organ” formed during life as a result of a “specific activity of uniting different physiological mechanisms into a single functional system.” We will mention that this broader concept, advanced by P.K. Anokhin, presumes a “broad functional unity of variously localized structures and processes on the basis of attainment of a final (adaptive) effect.”12 One of the most important features of “functional organs” is their plasticity: “in fulfilling one and the same objective, they can have differing structures,”13 which provides for diverse possibilities in compensating for breakdowns in functions. These ideas of A.N. Leontiev stem back to Vygotsky’s well-known proposition that “in comparison with animals, the human brain possesses a new localization principle, as a result of which it became the human brain, the organ of human consciousness.”14 According to Vygotsky–Leontiev–Luria,15 uniquely human mental abilities are supported by just such mechanisms. These mechanisms take shape over the course of a person’s life in society (we will examine how that formation takes place below), and can certainly not be reduced to a mechanical actualization of readymade principles of development already situated within individuals. But, consequently, the alternative on which Lenneberg and his colleagues base their thinking (compare: “The individual does not serve as a passive means or channel through which information

MAY–JUNE 2006

13

is conveyed”)16 is no more correct than the opposition of “rationalistic” and empirical” viewpoints initiated by Chomsky. The alternative to “empiricism” is not necessarily “rationalism”: other approaches are possible, and there is one to which we subscribe. We just mentioned the formation of functional organs, contrasting it to the actualization of innate structures. At this point, it is important to underscore that the process of development of mental functions and abilities supported by “functional organs” differs at its very core both from the process of the unfolding of biologically inherited behavior, and from the process of acquiring individual experience. This process is carried out specifically in the form of the assimilation by each individual of a social and historical experience, of the collective knowledge of the socium.
Before the individual entering into life . . . is a world of objects embodying human abilities that came into existence through the process of the development of sociohistorical practice. . . . For the individual to discover the human aspect of the objects that surround him, he must carry out energetic activity in relation to them, activity that is consistent (although, of course, not identical) to the activity that is crystallized within them. Of course, this also applies to language. Another condition is that the relationships between the individual and the world of human objects must be mediated by his relationships with people; they must be included in the process of communication. . . . The individual, the child, is not simply thrown into the human world, but is introduced into this world by the people around him, and they guide him in this world.17

Therefore, society participates in the formation of human abilities more as an active force within this formation than merely as a “language environment.” The aforementioned mediated nature of uniquely human types of activity is the other side of the process of the active “socialization” of “natural/inborn” processes, as a result of the incorporation of objects and phenomena of the external world into these processes that objectify human abilities. If in the “natural” interrelations with the surrounding world that is characteristic of the animal the only regulator of these interrelations is its individual experience (the

14 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

experience of the species is already provided in ready-made form), in the “social” interrelations that are characteristic of man, his relationship with reality is regulated primarily by societal experience that has taken shape in the tool of labor or the psychological tool (sign) in the form of its function—in the form of those abilities and skills that can take shape and be realized through this tool. The “natural” processes that are mediated, “socialized” through the introduction of tools and signs are in and of themselves innate or at least can be actualized through a signal, but not through the formative influence of the external world; the introduction of “social elements” into these processes forces man to switch to a new way of enabling them physiologically. It appears that human activity, even at the early stages of its ontogenesis, has a psychophysiological character that does not lend itself in the least to interpretation popular to the theory of “innate ideas.” In the words of Vygotsky, “The child’s system of activity is determined at every given stage both by the extent of its organic development and the extent of its mastery of tools. Two distinct systems develop in tandem, forming, in essence, a third system, a new system of a special kind.”18 Turning again to Lenneberg’s book, we are told that there is a “substantial” difference between the formation of concepts in man and the generalization of stimuli in animals. According to Lenneberg, there is a display of “word tagging” in cognitive processes; therefore, “concepts are essentially superstructures above physical data, they are a means of ordering . . . sensory data.”19 But what has been said above about the role of tools and signs in the formation of uniquely human behavior to a great extent relates to the processes of perception and to other components of human cognition. This is expressed in that the experiential moment appears in cognitive processes not only in the external, sensory link of the perceiving system, as is also commonly presumed by psycholinguists;20 also, its main role is in the organization of this system’s effector link.21 The most important thing here is that language serves as a mediating link in the activity of cognition and immediately conditions this activity.

MAY–JUNE 2006

15

The process of verbal denotation in recognition . . . is not understood as a special process—separate from cognition—which then processes its product through thought, but as a process that is incorporated into the very activity of cognition. . . . After all, recognition, or the actual cognition of an object, demands the correlation of pre-information received with a reference, which in man is stored in generalizing systems that have a linguistic basis. Such references are not only the recipients of incoming pre-information, but they carry out the function of guiding recognition processes.22

All of this makes it evident that language does more than make the individual’s understanding of the external world easier, more precise, and faster. It serves as a force that indeed shapes this categorization, introducing a fundamentally new principle: “The child is compelled to reorganize his way of seeing and conceiving things so as to have the possibility of using language to signify what he knows”;23 but this reorganization itself is performed with the help of language. One hundred years ago, Steinthal said, “In order to think, one must be able to speak.”24 In order to perceive, it is also necessary to be able to speak, at least if we are talking about the human way of seeing things. After all, man sees things specifically as “social elements,” projecting onto them knowledge of their objective properties. For it to be possible to separate an object from the world around it, as a carrier of such objective traits, it has to be recognized; and, in order for it to be recognized, it has to be signified. This was well understood by the great Russian linguist and philosopher A.A. Potebnia—little known, unfortunately, outside his native land—when he was still a follower of Humboldt. We will conclude with his words. “Language is more than an external tool, and its function for cognition and action is closer to that of an organ in terms of its significance for man, like an eye or an ear. . . . The sum of acquired abilities and traditions always stands as an intermediary between a thing and the ability to perceive it.”25 Notes
1. This work is taken from the Russian text of a paper that was intended for presentation at the nineteenth International Congress of Psychology (London,

16 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

1969) as a part of the symposium, “Biological, Social and Linguistic Factors in Psycholinguistics.” 2. F. de Saussure, Kurs obshchei lingvistiki (translated from the French) (Moscow, 1933), p. 40. 3. L.S. Vygotskii [Vygotsky], Izbrannye psikhologicheskie issledovaniia (Moscow, 1956), pp. 50–51. 4. V.D. Komarov, Uchenie o vide u rastenii (Moscow-Leningrad, 1944), p. 207. 5. G.V. Plekhanov, K voprosu o razvitii monosticheskogo vzgliada na istoriiu. Izbrannye filosofskie proizvedeniia, vol. 3 (Moscow, 1956), p. 610. 6. This thesis appears to be generally accepted in anthropology and genetic psychology at present. See H. Pieron, “Le développement de la pensée conceptuelle et hominisation,” in Les processus de hominisation (Paris, 1958). 7. L.S. Vygotskii [Vygotsky], Razvitie vysshikh psikhologicheskikh funktsii (Moscow, 1960), p. 255. 8. E.H. Lenneberg, Biological Foundations of Language (New York [: Wiley], 1967), p. 378. 9. Ibid., p. 379. 10. N. Chomsky, “A Review of B.F. Skinner’s Verbal Behavior,” Language, 1959, vol. 35, no. 1. 11. Lenneberg, Biological Foundations, p. 332. 12. P.K. Anokhin, Biologiia i neirofiziologiia uslovnogo refleksa (Moscow 1968), p. 79. 13. A.N. Leont’ev [Leontiev], Problemy razvitiia psikhiki (Moscow, 1965), p. 206. 14. Vygotskii, Razvitie vysshikh psikhicheskikh funktsii, p. 393. 15. In addition to the works already mentioned, see in this regard the book by A.R. Luria, Mozg cheloveka i psikhicheskie protsessy (Moscow, 1963). 16. Lenneberg, Biological Foundations, p. 378. 17. Leont’ev, Problemy razvitiia psikhiki, pp. 185–86. 18. Vygotskii, Razvitie vysshikh psikhologicheskikh funktsii, p. 50. 19. Lenneberg, Biological Foundations, p. 333. 20. Charles Osgood is a typical example. Concerning his views on this question, see A.A. Leont ’ev [Leontiev], “Psikholingvistika i problema funktsional’nykh edinits rechi,” p. 6, and Voprosy teorii iazyka v sovremennoi zarubezhnoi lingvistike (Moscow, 1961). 21. See A.N. Leont’ev, “O mekhanizme chuvstvennogo otrazheniia,” Voprosy psikhologii, 1959, no. 2, as well as numerous works by V.P. Zinchenko et al. 22. A.N. Leont’ev and Iu.B. Gippenreiter, “O deiatel’nosti zritel’noi sistemy cheloveka,” in Psikhologicheskie issledovaniia (Moscow, 1968), p. 19. 23. J. Bruner, “An Overview,” in Studies in Cognitive Growth, ed. J.S. Braner et al. (New York, 1966), p. 323. 24. H. Steinthal, Abriss der Sprachwissenschaft, vol. 1 (Berlin, 1871). 25. A.A. Potebnia, Iz zapisok po teorii slovesnosti (Kharkov, 1905), pp. 643, 646.
To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 17–29. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440302

A.A. LEONTIEV

Sign and Activity
The goal of this article is to provide an analysis of a category of meaning outside the system of any particular science (and certainly not from the perspective of any particular scientific problem), but within a more general system, suitable to an integrated approach toward language, speech, and speech activity. The necessity of such an integrated approach is increasingly evident both on the theoretical level and in the framing and solving of applied problems. The external expression of this necessity is the birth of such international scientific disciplines as psycholinguistics, sociolinguistics, and ethnolinguistics, among others. The movement of scientific thought along the path of an integrated analysis of speech activity is hindered, however (among other difficulties), by the vagueness of a number of basic concepts, the logical consequence of which is the shifting of the interpretation of these concepts, provided within the framework of a specific science (usually linguistics), to a more general theoretical context, and from this stems the limitation of their treatment. This primarily effects the concept of “meaning,” a concept that is central not only to linguistics but also to psychology, logic, and semiotics. Only once
English translation © 2006 M.E. Sharpe, Inc., from the Russian text “Znak i deiatel’nost’,” in Iazyk i rechevaia deiatel’nost’’v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 33–45. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov. Notes renumbered for this edition.—Ed.
17

18 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

we have defined meaning as an object category of the science of man will we have the right to undertake its interpretation within the framework of a particular field, as something to which we apply scientific methods of investigation. Such an analytical path is all the more desirable when we address psycholinguistic problems associated with meaning. As is well known, in Soviet science, psycholinguistics from the very beginning takes the form of a theory of speech activity. It views speech as one of the types of activity (along with other types such as labor, cognitive, and mnemonic activity, etc.), and strives to apply the study of speech to those propositions and categories that have been developed within the general theory of activity, both in its social and psychological aspects. In this case, our task consists in applying an approach based on the perspective of activity theory to a more general circle of questions associated with the category of meaning, and discovering the factors involved in the emergence and mode of functioning of meaning within the system of human social activity. We encounter the concept of activity at the very start of our analysis of meaning, when we raise the question of the relationship between meaning and the sign. As works by Soviet philosophers show,1 the problem of the sign in its interpretation, from the perspective of the theory of reflection, is inseparable from the problem of the so-called ideal, or quasi-object. As is well known, the ideal object (quasiobject) arises in social activity as the transformed form of true connections and relations. These connections and relations are transferred onto a material object that is alien to them by its nature, or are taken into it, and are replaced by other relations that blend with the properties of this object, and serve as its properties and features. The apparent form of true relations takes their place; the direct reflection of content in form becomes impossible. An example is the monetary form, which is the form of transformed goods. For this reason, a tremendously important gnoseological problem arises: in

MAY–JUNE 2006 19

analyzing the quasi-object as a converted form* of true connections and relations, how is it possible to isolate in it what arises from its “substance,” from its own uniqueness, its features and properties, from what is “transferred” onto it and has been transformed in it. Language is the system of such ideal or quasi-objects—linguistic signs—where real relationships are replaced with their apparent form, where the real properties and relationships of the objects and phenomena of reality, actualized in activity involving these objects and phenomena, wind up being taken and moved into a new (linguistic) substance, and are filled with the materiality and properties of language. As in a number of other cases, here the materiality of quasi-objects prompts the emergence of fancies of consciousness: we often immediately correlate language with the objects and phenomena of the external world forgetting that between them there is no direct and unequivocal correlation, and that rigorous scientific analysis of the nature of any quasi-object demands an intermediate link, which was first introduced by Marx: the system of social activity. What is given to our consciousness, through immediate observation of and reflection on language, what in language presents itself to consciousness, hardly begins to cover the essence of language. For this reason, a one-sided semiotic and a one-sided linguistic approach to language, however subtle their analysis might be, are fundamentally incapable of discovering its essence. The concept of the quasi-object as the converted form of real
*Merab Mamardashvili (1970) introduced the concept of converted form to denote the processes of transition of some content from one substrate to another. The features of the content change in the course of this transition according to the properties of the substrate. An illustration can be borrowed from the psychology of art. When you try to transform a novel into a movie, even if you plan to maintain the content as close as possible to the original work, you can’t do it without some important changes. Indeed, the substrate, the film, imposes some limitations and offers some new possibilities.—Dmitry A. Leontiev.

20 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

relationships is inseparable from the Marxist interpretation of the concept of the ideal. From this perspective, the linguistic sign, as the quasi-object, is “the immediate body of the ideal image of an external thing.”2 Having its own sensory nature, the sign at the same time serves as a component part of a system of forms, and a means of external expression and a capturing of the ideal phenomena that are generally accepted, and whose meanings are generally agreed upon. And here again it is important to emphasize that the ideal itself “has immediate existence only as the form (means, an image) of activity of the social person. . . . The ideal can under no circumstances be equated with the state of the material found under an individual’s cranium. . . . The ideal is a special function of man, as the subject of social-labor activity.”3 The concept of the sign must also be introduced (as distinct from the concept of the quasi-object), as an implication of such an understanding of the ideal. If, in principle, the quasi-object has, as Marx said, its “material existence,” then being used as the “body” of an ideal image, in a certain sense, loses this “materiality.” According to Marx, in signs, “functional existence . . . so to speak, absorbs its material existence”:4 a thing in its material existence and functional properties “is transformed into a sign, that is, into an object that has no meaning in and of itself, but merely represents, expresses another object, with which it has nothing immediately in common, such as, for example, the name of a thing and the thing itself.”5 In light of the above, it is obvious that in the practice of scientific research the single term “sign” serves three different purposes. First, there is the sign as a thing or—as it applies to language—as a material linguistic “body” incorporated into the activity of man; in this sense, we will refer hereafter to the sign. Second, there is the sign as the equivalent of the real sign in everyday consciousness; this concept will be referred to as the sign image. Third, a sign is the product of the scientific conceptualization of the structure and functions of the objective sign—the model of the sign or the sign model. These three purposes, as a rule, have not been clearly distinguished or have not been distinguished at all in the course of analy-

MAY–JUNE 2006 21

sis, giving rise not only to terminological homonymy, but also to a fundamental confusion. Let us return now to the problem of meaning. From what was stated above it follows that the sign (in the sense of the term just mentioned) has a material side (its “body”) and that it has an ideal “weight” that is expressed and anchored in this “body.” The ideal aspect of the sign is not reducible to a subject’s subjective conception of the content of the sign image, but it also does not represent real objectivity or those real properties and features of objects and phenomena that stand behind the sign (the quasi-object). The paradox is that, existing before and beyond a particular sign, these properties may be regarded as meaning only after they have been “transformed,” that is, after we introduce the quasi-object with its own content characteristics: “extralinguistic” meaning does not exist, and at the same time “sign” meaning is not a simple copy of real connections and relationships. The ideal aspect of a sign is the result of transference, of “transformation,” in the Marxist sense, of connections and relationships of actual reality that take place in the process of activity. Objectively, the sign stands before the subject as a real sign with all that underlies it, including all its functional characteristics, which are determined by the special features of activity into which this sign is incorporated. But, subjectively, a sign is perceived as a sort of psychological formation in which actual social content of this sign is blended and transformed. The consciousness of the subject in this case remains a contemplating consciousness, and from his perspective the sign appears as a sign image, and meaning, as the form in which he fixes and experiences his own social experience, without assigning himself the task of penetrating its true roots and true nature. This is the path taken by most researchers regarding meaning, working not with the real sign, but with the sign image, and not reflecting, or only reflecting in part, those sign properties “in which the socially conditioned manner of its functioning is expressed, its ‘functional existence’”6 onto the corresponding sign model. Thus, several interconnected, but by no means identical, categories are correlated with what is intuitively understood as mean-

22 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ing. First, there is the system of connections and relationships between objects and phenomena of reality that exists beyond and before the individual sign; we will call this system the objective content of the sign. Second, the ideal “weight” of the sign, the ideal aspect of it, which is the converted form of the objective content, we will call the ideal content of the sign. Third, there is the social experience of the subject, projected onto the sign image, or, as we will refer to it, the subjective content of the sign (the sign image). Until now, stretching things somewhat, we have kept to the level of the isolated sign. Obviously, this is a mere convention: both signs objectively—in a person’s activity—and subjectively— in his consciousness—serve as an integrated system, as a sign system. The first question this raises is the following: to what extent do we have the right to talk about the existence of an objective-social system of signs? To put it another way, to what extent does the concept of a sign system correlate with the concepts of objective and ideational sign content that we introduced above? It is completely obvious that as it applies to objects and phenomena of actual reality—taken in the abstract, extra-activity existence—we have no basis for talking about a system in the sense that concerns us here. It arises only when these objects and phenomena are incorporated into activity that arises as a “system of content-based social connections,” subsequently transferred onto quasi-objects and “transformed” in them and as the structure of activity with these objects and phenomena. In the process of such a transference and transformation, this system converts into a system of quasi-objects in which the system itself undergoes a radical change. This happens primarily due to the fact that in and of themselves, taken in their own content-based (and formal) characteristics, quasiobjects cannot form a system. As it applies to linguistic signs and other quasi-objects—in which, according to Marx, “material existence” is absorbed by their “functional existence”—this thesis takes on a somewhat different appearance: it is as if the contentbased interconnections between these quasi-objects descend to a

MAY–JUNE 2006 23

lower “level” and become their formal connections; dislodged by the “system of content-based social connections” that are transferred onto them and transformed in them. These formal relations link up in this system to form a new, systemic, functional whole. The difference between activity that generates a “quasi-object” system and activity that generates an “object” system is that the former is for the most part cognitive activity, a reflective activity, while the latter is primarily an activity of social interaction. And the systematicity of linguistic signs is specifically that “equal effect” that permits them to hold—in the taken, converted form, of course—both “the system of content-based social connections,” and the system of operations that we can potentially realize with these signs in the activity of communication, correlating them with specific objects and phenomena; referring to them, and substituting them, generating the selection of the most appropriate signs (in particular, appellation), and combining them into a meaningful whole—an utterance. Correspondingly, it is possible to identify two sides, two aspects of the ideal content of the sign. One of them is the correlation of the ideal content with cognitive activity. The other is its correlation with activity of social interaction, with the use of signs for communication. The first dominates in those cases where we use signs in the process of communication. Neither aspect is a static component of content or abstract isolated units. It is as if sign content is “poured out” to the side where we “lean” our sign. The immediate reason for this is the incorporation of the sign into different systems, while the reason itself is rooted in the different nature of goals and objective problems that are solved in the process of activity, in the differences between problem situations that arise during that activity. The subjective content of the sign image is not identical with itself in different problem situations of sign usage. However this content might be modified for a speaker of a language, which remains, on the one hand, a “cognitive invariant,” which is dictated by the sign content, in correlation with the “system of contentbased social connections” fixed in the sign;7 and on the other hand,

24 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

its “communicative invariant,” the system of operations with this sign that is fixed in it and comprises the rules of its use within the framework of more complex communicative structures. The “cognitive invariant” of the subjective content of the sign image, as follows from what was stated above, is that in the content that stems from social activity is fixed in the sign, while “its communicative invariant” is what stems from activity that uses the sign. It appears that the former is closest to what is usually called a concept and the latter is exactly what is most often called meaning. In the most general sense:
behind linguistic meanings are hidden socially developed manners (operations) of action, in the process by which people come to know and change objective reality. In other words, in meanings something is represented—transformed and condensed in the material of language— the ideal form of existence of the objective world, its properties, connections, and relationships, discovered through the entirety of social practice. Therefore, meanings in and of themselves, that is, abstracted from their functioning in individual consciousness, are just as “unpsychological” as the socially known reality that underlies them.8

Because of this they develop in accordance with sociohistorical laws that are an outside individual consciousness. But, at the same time, reality is presented to human consciousness as signified reality. Meaning is a form of presentation of reality in consciousness.
In their second life, meanings become individualized and “subjectified,” but only in the sense that their movement within the system of relationships of society are no longer immediately contained in them. They enter into a different system of relationships, into another movement. But here is what is remarkable: at the same time they do no lose any of their sociohistorical nature, their objectivity.9

One of the most important features of the “second life” of meanings is their interrelatedness with sensory stimuli. In its role as the ideal content of the sign, meaning remains extrasensory, since, although the converted form of objective content presumes the material of the sign, it is taken as an extra-individual, abstract formation. But as soon as we switch to meaning as subjective content of the sign, it turns out that its existence in activity and its presen-

MAY–JUNE 2006 25

tation in the consciousness of the individual is inextricably tied to material (sensory-material) interrelatedness. Meanings do not exist for every one of us outside of the subjective reflection of materiality, for example, in the form of visual images or any image of perception. But at the same time it would be a mistake to think that such images precede meanings, and that meanings do nothing but “tag” (Lenneberg) cognitive processes. As numerous studies by Soviet psychologists demonstrate, uniquely human object perception is not possible without the participation of socially developed references primarily based in language, and “the process of verbal signification in recognition . . . is understood not as a separate process—isolated from perception—which then processes its product through thought, but as a process that is incorporated into the very activity of perception.”10 These references, which are stored in the visual system and are not possible without language (or some other means of social anchoring), nonetheless have a sensory nature. Experiments by V.P. Zinchenko, for example, showed that “names were assigned only after the collation and selection of references that correspond to images.”11 Here we are dealing with what M.S. Shekhter fortuitously labeled “secondary images,” that is, images forming as a result of generalization, usually mediated by language. We “see” a triangle, we recognize it because a generalized image of a triangle has been formed in our consciousness, but the image itself arises only as a consequence of an operation with immediate sensory data and on the basis of abstract features of any triangle that have been fixed in its linguistic form and reflected in the meaning of the word triangle. This materiality, this sensory nature of meaning, taken as the subjective content of the sign, is particularly clear in the process of meaning formation in child language acquisition. This is where one of the components of meaning comes from, one that is reducible neither to a cognitive nor a communicative invariant: the element in it that comes from psychological processes that stand “behind” the sign in various forms of its usage in activity, specifically, the extent and means of interrelatedness between content and

26 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

its “sensory” aspect, the interrelatedness of subjective content with secondary images, with “visuality.” This aspect of subjective content can in certain cases (for example, in the child) take on an uncharacteristic significance: it is as if the subjective content of the sign image is projected onto sensory images that are related to the sign, and which becomes deformed to the extent of their limited (in comparison with the sign) psychological capabilities. For the subject, the sign seems to lose its ideal content, preserving only the part of it that is fixed in the and extracted from it. And since the sensory image, to a large extent, depends on the subject’s individual experience, the objective content of the sign is in a certain sense subjectivized, in a certain sense. A person begins to evaluate a sign in terms of his own individual experience, to give it those features that reflect, in essence, only the relationship of that person to the sensory image that represents for him represents a class of some real objects and phenomena. Below, when we speak of sensuous coloring of subjective content, this is what we refer to. This sensuous coloring is potentially greater in some signs than in others. The second component of meaning is that within the subjective content of the sign image, which comes from various levels of awareness and various levels of semantic explication of this content in the subject’s consciousness, in the speaker of the language. Undoubtedly, in the final analysis, both of these depend on factors that lie beyond individual consciousness. A person is aware of and explicates the content of a sign to the extent he needs. But the opposite direction is also critical—in certain situations the use of a sign is limited by the ability to explicate it (as happens, for example, with scientific terminology). We will call this aspect of subjective content its potential for explication, including the potential “depth” at which it is cognized. This can also differ for different signs. The third component of meaning is that within the subjective content of the sign image that derives from personal meaning and can be called semantic coloring of this content. Here, various forms of distortion are especially common; particularly characteristic is the substitution of objective (ideal) content with personal mean-

MAY–JUNE 2006 27

ing. The degree of semantic coloring, evidently, is largely tied to the degree of sensuous coloring and potential explicability of the sign: the greater the sensuous coloring and the less potential for explication, the greater the likelihood that the meaning of a sign will diverge from its ideal content. The fourth component of meaning is what can be called the sensuous coloring of the sign image’s subjective content. In the historical development of the system of linguistic meanings, all of these aspects of subjective content take on the status of factors that effect its change. Concerning the communicative invariant of subjective content, it can be presented in scientific analysis as a system of types of rules that set the boundaries of sign usage in the activity of communication. What are these rules? What operations with a sign are fixed in the sign image (however vaguely, as potential) and, consequently, must be viewed as forming the subjective content of this image? 1. Operations that are directly dictated by cognitive invariance, that is, signs’ cognitive-typological features that are brought into their usage. These are primarily rules that are warranted for a given sign concerning situational indication and substitution. There are types of signs (deictic signs) for which these operations almost exhaust the communicative invariant of their subjective content. 2. Interrelation and interchange operations among signs as elements of a sign system, that is, semantic elements, in the narrow sense of the word. As I.S. Narskii notes, they form “a sort of permissible circle of cases within which subject operations [that use signs— A.L.] correspond, despite all their individual differences—to a particular meaning.”12 Operations of this sort are realized in the mechanism of sign interchange, primarily in the rules for selecting semantic units for communication purposes. Specifically in this sense the psychological structure of meaning is determined by a system of interrelations and contrasts between words in the process of their use in activity. It is this “network of oppositions that, through interdiction, limits and directs the process of selecting appropriate meanings.”13 3. Operations that combine signs into quasi-objects (signs) of a

28 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

higher order, that is, within a sign’s semantics that is connected with the semantics of the utterance and represents compressed rules (that are attributed by us to the sign in question) of transition between sign and utterance.14 Operations of the second and third type can, in turn, be fixed in different ways in the sign. They can be content based, that is, they can enter the subjective content of the sign image of the speaker (or listener). For example, in isolating-type languages, rules for sign organization within an utterance are reduced to the organization of the corresponding semantic classes. But they can also be formal. In languages such as Russian, operations on formally grammatically marked classes of signs dominate the rules of utterance structuring. This characteristic is marginal for their subjective content and is relatively independent in relation to this content. In our previous analysis we purposely ignored, or at least avoided, the fact that a meaning of a sign appears not simply in the speech activity of a particular individual and in a particular situation (or, correspondingly, in a particular activity, which is not central here, as use of language in any nonspeech activity has as its necessary prerequisite actual or potential communicative use). The sign is a part and a condition of the processes of communication as one of the aspects of social interaction among people as members of a class or society overall. Contemporary psycholinguistics, as a rule, loses sight of that aspect of the problem, something that is associated with the treatment of communication itself by foreign (and Soviet, in some cases) science usually as interindividual communication aimed at conveying information.15 For this very reason, [in psycholinguistics today] speech is usually treated in the spirit of K. Bühler’s famous scheme,16 according to which the task of the speaker consists in conveying information about some objects and phenomena of the real world in a form allowing this information to be appropriately received by the listener. Be that as it may, only an approach based on the perspective of the psychology of communication can give us the key to correctly interpreting the nature of meaning and its interrelation with other philosophical and psychological categories. V.N. Voloshinov was correct when he wrote almost a half century ago, “Meaning is not

MAY–JUNE 2006 29

in the word or in the soul of the speaker, and not in the soul of the listener. Meaning is the effect of the interaction between the speaker and the listener on the material of the given sound complex. . . . Only the flow of speech communication sheds light of meaning on a word.”17 Notes
1. See E.V. Il’enkov [Ilyenkov], “Ideal’noe,” in Filosofskaia entsiklopediia, vol. 2 (Moscow, 1962); A. Poltoratskii and V. Shvyrev, Znak i deiatel’nost’ (Moscow, 1970); A.M. Korshunov, Teoriia otrazheniia i tvorchestvo (Moscow, 1971); M.K. Mamardashvili, “Forma prevrashchennaia,” in Filosofskaia entsiklopediia, vol. 5 (Moscow, 1970), and “Analiz soznaniia v rabotakh Marksa,” Voprosy filosofii, 1968, no. 6. 2. Il’enkov, “Ideal’noe,” p. 224. 3. Ibid., pp. 220–21. 4. K. Marks [Marx] and F. Engel’s [Engels], Sochineniia, vol. 23, p. 140. 5. Il’enkov, “Ideal’noe,” p. 224. 6. Korshunov, Teoriia otrazheniia i tvorchestvo, pp. 180–81. 7. This system is not always fully reflected in the subjective content of the sign. It would be more accurate to say that it is never adequately reflected in it. 8. A.N. Leont’ev, “Deitel’nost’ i soznanie,” Voprosy filosofii, 1972, no. 12, p. 134. 9. Ibid., p. 136. 10. A.N. Leont’ev and Iu.B. Gippenreiter, “O deiatel’nosti zritel’noi sistemy cheloveka,” in Psikhologicheskie issledovaniia (Moscow, 1968), p. 19. 11. V.P. Zinchenko, “Produktivnoe vospriiatie,” Voprosy psikhologii, 1971, no. 6, p. 40. 12. I.S. Narskii, “Kritika neopozitivistskikh kontseptsii znacheniia,” in Problema znacheniia v lingvistike i logike (Moscow, 1963), pp. 15–16. 13. A.A. Brudnii, “Znachenie slova i psikhologiia protivopostavlenii,” in Semanticheskaia struktura slova (Moscow, 1971), p. 22. 14. This is how we arrived at a system that on the surface coincides with the known differentiation between semantic, pragmatic, and syntactic meanings. However, our content-based interpretation of this differentiation is entirely different from its traditional interpretation. 15. See A.A. Leont’ev [Leontiev], Psikhologiia obshcheniia (Tartu, 1974). 16. K. Bühler, Sprachtheorie (Jena, 1934). 17. V.N. Voloshinov, Marksizm i filosofiia iazyka (Leningrad, 1929), p. 123.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

30 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 30–46. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440303

A.A. LEONTIEV

“Units” and Levels of Activity
The psychological structure of activity—its levels and its main units, or “formatives”—has been extensively analyzed in contemporary psychology, especially Soviet psychology. As early as 1935, in his Foundations of Psychology [Osnovy psikhologii] S.L. Rubinshtein introduced the following system of concepts: reaction—conscious action (or operation)—act (an action regulated by conscious relations); and in 1946, in Foundations of General Psychology [Osnovy obshchei psikhologii] the triad of “movement–action–activity.” But the most prevalent theory in our country and abroad was the theory of the internal structure of activity developed within the framework of the psychological school of L.S. Vygotsky and described in detail in the book by A.N. Leontiev, Problems of the Development of Mind [Problemy razvitiia psikhiki] (1959). This theory has undergone many restatements and interpretations and has been combined with various approaches and “adapted” to a variety of specific studies. Throughout, even within the framework of activity theory itself, an ambiguous understanding of the units and levels of activity organization can be seen. P.Ia. Galperin’s concept of “action” can be cited as an example

English translation © 2006 M.E. Sharpe, Inc., from the Russian text “‘Edinitsy’ i urovni deiatel’nosti,” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 66– 82. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov.
30

MAY–JUNE 2006 31

(e.g., Galperin, 1976), an understanding that, like certain other aspects of his theory, demands analysis beyond the scope of this article. It appears that the very possibility of alternative solutions to the problem of units and levels of activity given the sameness of the original theoretical postures reflects the open-ended, preliminary nature of the proposed theory, the existence from the very start of “reserves” within it for further development and clarification, and its fundamentally antidogmatic nature. At the same time, however, activity theory imposes certain methodological and theoretical limits on the diversity of possible interpretations of the structure of activity—something that is not always fully understood by some researchers. This makes it imperative that the problem of the structural levels of activity, as an object of special theoretical investigation, be given particular attention. Inasmuch as the theory of A.N. Leontiev served as a starting point for many psychologists and philosophers working on the question of the structure of activity, it would be wise to clarify his understanding of the problem of units and levels that interests us here. First and foremost, attention should be paid to the fact that Problems of the Development of Mind was not written all at once: works from various years, reflecting the evolution of activity theory, are collected in this book. The chronological structure of he book and subsequent evolution of its author’s views (see the monograph Activity. Consciousness. Personality [Deiatel’nost’. Soznanie. Lichnost’]) are often overlooked in reference to its various propositions. Before characterizing the understanding of the problem of activity units and levels in article, there are several things that should be said about the very concepts of “level” and “unit” in psychology. It is no secret that the concept of level in activity theory is genetically tied to the concept of level in the work of N.A. Bernstein, formulated in 1935. In Bernstein, this concept is dynamic, system-activity-oriented; levels are interpreted as a way of realizing sensory synthesis, a way “that is best suited for solving a particular problem given the quality and makeup of its contributing afferentations and their synthesis” (Bernstein, 1966, p. 97).

32 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

According to Bernstein, one and the same movement can be supported by different physiological organizations; but such an organization always has multiple levels. The concept of the unit is particularly complicated. As is well known, A.N. Leontiev does not provide an explicit definition of it; as a rule, he puts the term “unit” within quotation marks, and in so doing, “determines” it. And this is justified: after all, as it applies to his point of view, the concept of unit has little applicability to activity, action, or operation, since it presumes their discrete nature. In other words, the concept of the unit is better suited to the model of Miller, Pribram, and Gallanter, for example (the “TOTE unit”). In A.N. Leontiev’s conception, the only thing that can be called a “unit” in the strict sense is activity (an activity act). As A.N. Leontiev sees it, the structure of activity takes the following form. At its basis lies the concept of action, of process, the object and motive of which are not the same. Next, there is the concept of operation. “Psychologically . . . the merging of separate, individual actions into unified actions is their transformation into operations” (Leontiev, 1972, p. 298). Another sort of operation is born out of the simple adaptation of action to the conditions of its execution. (For the sake of brevity, we will call operations of the first type conscious operations or “C-operations,” and operations of the second type, which have a different relationship with consciousness, will be called adaptive operations, or “A-operations.”) Finally, we have the introduction of the concept of activity as an action that has acquired an independent motive. In this case, and only in this case, we are dealing with a conscious motive. We should note that awareness of a motive is not elemental, but it demands a certain special process of reflection of the relationship between a particular activity’s motive to the motive of the broader activity. All of these tenets of the related theory of activity are often cited. Unfortunately, it is not uncommon for these citations to depict the structure of activity as being closed; concepts relating to the psychological nature of consciousness are given only an explanatory role. In fact, the most important feature of this conception is constituted in the fact that within it, the structure of activity

MAY–JUNE 2006 33

and the structure of consciousness are interchangeable concepts— they are tied to one another in a unified, integrated system. The fact that an analysis of the structure of activity usually precedes an analysis of the structure of consciousness, which is determined genetically by the primate. But genetically, consciousness cannot be understood in any other way than as a product of activity. Functionally, they are interconnected: activity is “directed by consciousness,” and at the same time, in a certain sense, it is activity that directs consciousness. It is particularly important, therefore, to devote particular attention to the problem of the connection between the structure of activity and the structure of consciousness. From the very beginning, A.N. Leontiev emphasizes that the appearance of a differentiated internal structure in activity is the consequence of the emergence of collective labor activity (1972, p. 273). It is possible when, and only then that man subjectively reflects the real or potential connection between his actions and the attainment of the overall end result. This is what makes it possible for a person to carry out separate actions that would not appear to be effective if taken in isolation, outside of collective activity. “Together with the birth of an action,” A.N. Leontiev writes, “with this main ‘unit’ of human activity, there arises the main social (by its nature) ‘unit’ of the human psyche, the rationale for a person regarding what he is directing his activity toward” (ibid., p. 274). At the same time, the possibility of awareness appears, of presentation of the material world, as a result of which awareness in the true sense emerges, as a reflection of reality through meanings. The genesis, the development, and the functioning of consciousness are products of a level of development of the forms and functions of activity. “Along with a change in the structure of a person’s activity there is also a change in the internal structure of his consciousness” (ibid., p. 186). How does this happen? Any mental reflection is always “biased.” But it features what has objective ties, relationships, interactions, what enters into social consciousness and

34 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

is fixed in language, and what depends on the relationship of a given subject to the reflected object. This is the origin of the differentiation of signification and personal meaning that is so often analyzed by various authors. What interests us now is meaning as the specific relationship that arises in a subject’s activity between what motivates him to act and what his action is directed toward, that is, the relationship between motive and goal. The relationship between signification and meaning is the relationship between the main “formatives” of the internal structure of human consciousness. We could put it even more categorically: this relationship is its main “formative.” The development of production dictates the emergence of a system of coordinated actions, that is, of complex action; and this signifies—on the level of consciousness—the most important step: the move from a conscious goal to a conceived condition of action, the appearance of levels of awareness. On the other hand, the division of labor and production specializations give birth to a “shift of motives onto goals,” and the transformation of action into activity. New motives and new needs are born, and from here we get the subsequent qualitative differentiation of awareness. Another exceptionally important step is the transition to truly internal mental processes, the emergence of a theoretical phase of practical activity. Internal speech actions appear, which, accordingly, form the general law of the shift of motives, internal activities and internal operations. Like activity, consciousness is not merely the sum total of its elements; it has its own structure, its internal integrity, its logic. And if human life is a system of activities that alternate with one another and coexist or conflict with one another, then consciousness unites, supports their creation, their variation, their development, their hierarchy. So it is not the element-by-element connections of “units” of consciousness with “units of activity” that is most important, but, first, the system-forming role of consciousness in relation to the entirety of activities; and second, in the double-sided interdependence between the dynamic of the internal structure of consciousness and the dynamic structure of activity.

MAY–JUNE 2006 35

Let us look at how the ideas described above are treated in the book Activity. Consciousness. Personality (Leontiev, 1975, 1977). What is emphasized here is primarily the nonadditive, molar nature of activity. It is “a system with its own structure, its own internal transitions and transformations, its own development, . . . incorporated into the system of social relationships” (ibid., p. 82). After all, in society, it is not merely a matter of man encountering external conditions to which he must adjust his activity; these societal conditions themselves encompass motives and goals for his activity, its means and ways, and through this “society generates the activity of the individuals who comprise it” (ibid., p. 83, emphasis added). What directs the process of activity is, primarily, the object itself, the material world, and, secondarily, its image as a subjective product of activity that fixes, stabilizes, and encompasses the material content. The conscious image is understood here as the ideal measure, reified in activity; human consciousness plays an essential role in the movement of activity. Thus, along with “consciousness-image” the concept of “consciousnessactivity” is introduced; and overall, consciousness is defined as the internal movement of its formative structures, movement incorporated into the overall movement of activity (ibid., p. 157). It is emphasized again and again: actions are not isolated, “separate entities” within the makeup of activity: uniquely human activity exists in no other way than in the form of actions or chains of actions (ibid., p. 104). One and the same process serves as activity in its relation to motive, and as actions or a chain of actions in subordination to a goal. Thus, action is neither a component nor a unit of activity—it is specifically its “formative,” its moment. He goes on to analyze in greater detail the relationship between motives and goals. The concept of the “motive-goal” is introduced, that is, the motive serving the role of the “overall goal” (the goal of activity and not of action), and the idea of the “zone of goals,” the delineation of which is entirely dependent on the motive. The selection of a specific goal and the process of goal formation is connected with “the approbation of goals through action” (ibid., p. 106). Along with this line of thought, the concept of the two

36 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

aspects of action is introduced. “Besides its intentional aspect (what must be achieved), action also has its operational aspect (such as, the means by which it will be achieved)” (ibid., p. 107). This leads to a somewhat different definition of operation—it is the quality that forms actions. While the genesis of action is in the exchange of activities, in the interrelations between “the collective subject” (Marx) and the individual subject, the genesis of operation is found in the interrelation among actions of a subject (one and the same), their incorporation into one another. The question is raised as to the decomposition of activity into units smaller than operations (here no quotation marks are placed around the word units). The concept of the functional block, proposed by V.P. Zinchenko, is given as an example. But this is a transition to the analysis of the intracerebral processes that implement activity. Finally, the concept of personality as the internal element of activity is introduced. It is specifically, and only as a result of the hierarchy of an individual’s separate activities, which realize his essentially social relationship with the world, that he takes on a special quality—he becomes a personality. A new step in the analysis here is reflected in the fact that—while it was the concept of a system of actions that took center stage in the examination of activity—in the analysis of personality, the most important aspect is the concept of hierarchical connections between activities, the hierarchy of their motives. These connections, however, are in no way assigned to the individual as something that takes shape outside of activity or over activity. The development, the expansion of the circle of activities itself by necessity leads to their connection into “nodes,” and from here to the formation of a new level of consciousness—the consciousness of personality. It has been necessary to repeat certain well-known propositions about the conception of activity in order to show that, on the basis of its internal logic, this conception is widely open to further development. It is open both “downward” and “upward.” It is open “downward” because it demands investigation of the intracerebral (psychophysiological) processes and structures generated by the

MAY–JUNE 2006 37

phylogenetic and ontogenetic development of material activity, that is, studying them as dependent on activity and at the same time conditioning the possibility of implementing activity. It is open “upward” in that it demands attention to concepts and categories of a more global nature than the concept of activity (as a unit), and first and foremost to concepts of the system and hierarchy of activities. But, naturally, this attention presumes the study of the interrelations between the structure of activity and the structure of consciousness and, then—with the concept of personality— including an analysis of the structure of activity in a broader context. Even from the cursory representation of the state of inquiry into the structure of activity provided above, it is evident that many questions essentially remain unexplored and have been posed only in general terms. Therefore, broad possibilities open up for a variety of solutions, which have indeed been proposed by a number of authors. Without pretending to offer a summary that is by any means complete, we will pause to examine only two ideas of this issue that appear to raise the most serious questions. Let us first turn to the propositions put forth by E.G. Iudin in his articles from 1976– 77. Perhaps the most important points here are this author’s understanding concerning the methodological status of activity theory and the resulting dilemma he sees: does the threefold structure of activity pertain to an analysis and explanation or to the actual object of study? Iudin believes that in activity theory, the only true psychological object is the level of action, while two other levels carry out a more “clarifying role”: activity, as a means of integrating psychology into a social-philosophical context, and operation, which integrates psychology into neurophysiology (1976, p. 75). Iudin believes that these levels are only an “explanatory schema,” unconsciously understood, also as a “schema of the object.” Assuming from the beginning that concepts identified within the system of activity are “units of analysis,” Iudin further notes further that in the trinomial structure of activity, such categories as motive,

38 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

goal, and condition occupy their own special place and that, evidently, they “must form a special category of units” (ibid., p. 77). In another article (1976a), Iudin expresses the opinion that the trinomial scheme, sufficient as an explanatory scheme, demands special verification as the object of study. Furthermore, he reproaches A.N. Leontiev for turning directly to social phenomena (the division of labor, etc.) in explaining psychological phenomena. The author generally believes that one is not justified in defining consciousness and personality “solely through activity” (1977, p. 36), and calls for the creation of a psychological taxonomy. While in the psychological theory discussed above, the nonadditive nature of activity is emphasized and viewed as a developing system that is characterized by the movement of its internal formatives, of its moments, and by their transformations, what stands out in the works of Iudin is the concept of units, their categories, their connections as independent entities, their taxonomy. Consequently, the main proposition of the conception criticized by Iudin has been replaced. There is another problem in the works of this author. While he claims to hold a “neutral” position—he refuses to answer the question whether activity theory is the characterization of an object or a tool of analysis—in actuality, Iudin clearly understands it as a theoretical construct. This is the basis for the “socialization” reproach concerning some of its concepts. If activity theory is a system of units of analysis, then this analysis should be undertaken within the framework of one particular science, in this case, psychology. But, if we are going to look at activity not simply as a theoretical construct, but as a methodological category, it becomes obvious that, in principle, it is impossible to construct a system of concepts of activity theory that would be “self-sufficient,” that could describe a system of activity as such, in isolation from the “big” system in which it is contained, of which it is a part. For this reason, Iudin’s reproaches in the “socialization” of certain concepts of activity theory appear to us to lack any foundation. In essence, we have absolutely no methodological bases for the dilemma Iudin proposes. “Units of analysis” do not have their own

MAY–JUNE 2006 39

existence, independent of the object of study; a descriptive system cannot be opposed to the system of an object, about which L.K. Naumenko has written very clearly (1968, pp. 143–61). If we refuse to grant “units of analysis” a separate existence—and it seems that this is the only solution to the problem—then a special taxonomy of units of activity, and the introduction of a new series of units even more so, leads to a simple duplication of the problem. And such a special taxonomy seems hardly possible: the structure of activity as viewed within general psychological theory is inseparable from the structure of consciousness and from personality. Figuratively speaking, this is not so much a theory of activity as it is a theory of “activity–consciousness–personality.” The problem of units has been raised in the research of V.P. Zinchenko, which is of particular interest both in terms of its affinity to A.N. Leontiev’s ideas concerning activity, and in terms of its richness. According to Zinchenko, motives, goals, and conditions are components of activity. They are closely associated with three types of units. Each of these units, in turn, is a system of interconnected units from the preceding level; in a unified activity act, its organization is realized “in the unification of functionally defined processes (elements) that are subject to one and the same motive” (Zinchenko and Gordon, 1976, p. 83). In this connection, according to Zinchenko, there are problems within activity theory that demand an extension of its categorical framework. For example, in certain specific types of activity not all potentially possible properties of actions and operations are evident, but only those essential to the given conditions of activity, to the given goal that has been set. For this reason the question of the structure of new units of activity should be raised, “in particular those that are based in a functional dependency, together with the relations between separate elements (actions, operations), and their properties in the structure of the activity as a whole” (ibid., p. 101). This is where we derive the category of “functional structure.” Analyzing operations, Zinchenko proposes breaking them down into even smaller units. He considers the functional block to be

40 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

such a unit, although it may not have “a direct application for behavior” (i.e., it may be just an element of an operation), but can have such an application “when it is essentially the same as what is known in activity theory as an operation” (ibid., p. 113). In an article by V.P. Zinchenko and V.M. Munipov, we see an interpretation of operations that differs somewhat from A.N. Leontiev’s: operations are determined “not by the conditions, as such, under which an act is carried out, but only by what is functionally significant within the conditions” (1976, p. 51). Conditions consist in the material properties of reality, which differ from functional ones. Consequently, according to Zinchenko, there is a special level of analysis of operations that is determined by the material properties of the situation: this is the level of functional blocks. A clear conception of the levels of activity takes shape: microanalysis (blocks and subblocks) permit activity to be imbued with a particular material content, since only more elementary units are directly connected with reality. Therefore, the materiality of activity ascends from the bottom to the top, starting with elementary units, while its conceptualization goes from the top to the bottom. “It is their meeting that gives rise to activity” (ibid., p. 53). In analyzing the works of Zinchenko, it is easy to see certain parallels between his views and those of Iudin. Whether explicitly or not, Zinchenko nonetheless often treats the structure of activity as an explanatory model, although elsewhere this structure is understood as a property of the object, that is, of activity itself. But what is particularly striking is Zinchenko’s additive understanding of the very “unit” of activity. For him this unit has no quotation marks and at times it is even simply an element. This is the source of the three categories of analytical units of activity— operational, cognitive, and personality-based (Zinchenko, 1977, p. 23)—and also the source of the qualification of an action and an operation as representing “operational” units of activity. This is the very idea of separateness, of the additive nature of activity units, which, in our view, represents the idea of a “functional structure”: this idea would hardly be justifiable given any other understanding.

MAY–JUNE 2006 41

In essence, any structure of activity is “functional” in the way Zinchenko looks at it: it has no other structure. The matter becomes more complicated when it includes an operation. As we recall, Iudin believes an operation is a more physiological than a psychological concept. We will immediately state that this is totally incorrect as it concerns conscious or Coperations; here, we undoubtedly have a psychological concept. (We will note that A.N. Leontiev, who introduced a distinction between C-operations and A-operations in Problems of the Development of Mind, later [in 1974] seems to have forgotten about the latter and deals only with C-operations.) Concerning operations of the second type—adaptive, or A-operations—and their relationship to C-operations, the question remains open. One point is beyond doubt: the overall class of “operations” features its own relationships of incorporation and subordination (in particular, C-operations can incorporate A-operations). For this reason, in our own works analyzing speech activity, it became necessary to introduce the concept of “macro-operations” (e.g., the transformation of the sentence) and “micro-operations” (e.g., the choice of a word). But even the latter follows a more complex psychological course than Zinchenko’s functional blocks (see A.A. Leontiev, 1974). As it appears to us, transcending from “bottom to top,” Zinchenko switches from functional blocks to A-operations when the functional blocks “can be transferred into new behavior” and to C-operations when there is no such direct outlet. In any event, he is aware of the fact that between operations and functional blocks (and even more so in the case of subblocks) there is a qualitative transition to new reality. This reality for him is a psychophysiological reality, on the one hand, and a “material” reality, on the other. It seems that in this case it would be unjustified to talk about a “material content.” The materiality of activity is something else; it characterizes the level of activity and the level of action. In general, elementary units are hardly closer to reality—they are closer to “nonhuman,” dead reality. Here there seems to be a logical jump from the structure of a particular kind of activity, a structure that

42 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

has an orienting nature, to the structure of any activity. But this is another question, one about the hierarchy of activities themselves. In light of what has been described above, it is possible to identify several main problems associated with the structural level of activity that have not been sufficiently clarified in the literature or that have been given contradictory interpretations. The first of these problems is a certain ambiguity, and, in any case, a lack of refinement of two critical concepts: sodezhanie deiatel’nosti [content of activity] and aktivnost’ [activity]. This leaves room for a very varied interpretation of these two concepts, up to the structural characteristics of the former being attributed to deiatel’nost’ and the latter being subordinate to aktivnost’ as species is to genus. We imagine a more precise interpretation of aktivnost’ as a process that implements a subject’s personality (super-activity) properties, and, consequently, something that is beyond the bounds of a “separate activity” within a system of activities (Petrovskii, 1975). The “system of activities” itself is yet another unresolved problem in activity theory, or at least a part of this problem. As we have seen, in his 1975 monograph, A.N. Leontiev raised the question of the system of activities as being a characteristic of personality. But much is unclear here. First, the interrelations between motives (the hierarchy of activities) is not the same as a system of activities—they are separated in the text, but the author went no further. There are also inconsistencies in the treatment of the interrelation between the hierarchy and the system of activities, on the one hand, and personality and consciousness, on the other. While the individual activity is the basic unit of analysis of consciousness, at the same time, it is consciousness that “holds together” activity. While personality is characterized by hierarchical relationships among activities, at the same time, it is a product of the reflection by consciousness of the connections between activities and their hierarchies. V.A. Petrovskii generally talks about personality as a “system of activities” (1975, p. 37). It strikes us that this problem has many sides. One aspect, the system of activities, is viewed from the perspective of personal-

MAY–JUNE 2006 43

ity: the simultaneous “bundle” of a subject’s potential activities interrelated in various ways, for which it is possible to identify the primary aspect. This train of thought can be found in the works of A.N. Leontiev from the early 1940s. It is indeed personality that directly determines this system and is simultaneously determined by it. But it cannot, of course, be reduced to a system of activities and equated with it: this is another category, and, evidently, if we want to introduce into a system of activity levels a new, highest level, we must seek a “unit” for it. One way or another, however, the transition to this highest level, like the transition from operations to functional blocks, marks the transition to a new quality. In this regard, we prefer to talk about a personality system of activities, or about a P-system. The second aspect is a hierarchy (or rather hierarchization) of activities in accordance with the degree to which they are “conceptualized,” the degree to which they are endowed with personal significance, which is directly linked to the hierarchy of motivesgoals. Following this path, we arrive at different levels of activities, as we arrived earlier at different levels of operations. These activities can exist in tandem (or can conflict), or they can be incorporated into one another (although, here conflict is also possible); in this case one and the same system of actions plays a dual role for a subject—as the achievement of a short-term and longterm goal, which can be understood as the satisfaction of an immediate or remote motive. Here it is particularly helpful to distinguish between meaning-formative motives and stimulusmotives (Leontiev, 1977, p. 202). It is useful in this case to speak about a system of activities interrelated with consciousness (Csystem). The third aspect is essential, and is associated with the existence of types of activity (such as perceptive activity), which by their very essence usually demand “incorporation” into another activity. This fact leads some researchers to deny the status of activity in such processes, and others, on the contrary, to proclaim such activities to be the only object of psychological study. Here we encounter an entire web of questions, the first of which is the

44 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

concept of orienting activity: is this an activity, and how does it enter into the overall system of activity, or the A-system? From this perspective, which is narrowly activity-focused, we examine the third system of activities, or the A-system. Finally, the systemization of activities is at least threefold. It is absolutely essential and urgent to discover the nature, the specific dynamic process and interaction of these systems. The third problem, which has been partially developed by O.K. Tikhomirov and V.P. Zinchenko, is associated with the difference between the intentional and the operational aspects of actions, and, consequently, with the dual orientation of the structure of activity, which seems to dissolve into two substructures: “activity–action” and “action–operation.” The ellipses here signify the open nature of these substructures “upward” and “downward” respectively. If we present the structure of activity in this way, organically entered into the model are, first, the concept of a “system of activities” (another question is whether we are dealing with a Psystem, a C-system or an A-system; it is possible that an A-system is primarily correlated with the structure of interrelations between intentional and operational aspects); and second, the concept of the functional block. Under these circumstances, the concept of operation breaks down into various parts. It is not the confluence of conceptualization and materiality that gives “birth” to activity, but the confluence of intentionality and operationality. The “dynamic paradigm” of activity separated out by A.G. Asmolov and V.A. Petrovskii (1978) correlates perfectly with the second, the operational substructure; and we therefore consider Asmolov’s idea about the levels of sets (attitudes) to be correct, but we feel that their hierarchy has a somewhat different character—all types of sets in a certain sense are operational (see Asmolov, 1977). Consequently, one of the main questions that must be resolved is the dynamic relationship between the intentional and operational aspects at different levels of the structure of activity. It is within this understanding, and not within the understanding of the development of “the conception of the aktivnost’ of person-

MAY–JUNE 2006 45

ality” where it is possible and necessary to overcome the “rough schemes interpreting the principle of the materiality of activity” (Smolian and Solntseva, 1977, p. 121). The fourth problem was identified by A.N. Leontiev himself. This is the problem of the relationship between psychology and psychophysiology, the transition between the two in the analysis of activity. What is meant here are “those psychophysiological functions that realize activity, that in part compose its natural prerequisites and place certain limitations on its course, and in part are reshaped within it and are even generated by it” (Leontiev, 1974, p. 9). Naturally, we have not exhausted the problems that arise at this stage in the development of activity theory. Indeed, this was not our objective. We merely attempted to demonstrate that the system of concepts describing the structure of activity is neither rigid nor closed, and that this system “works” not with elements or autonomous processes, but with “units” of a completely different way. And—of particular importance, in our view—the system of levels and units of activity is not only and not merely opened “upward” and “downward,” it is not comparable with the structure of consciousness, and even less so is it contrasted to it, but is one with it. Finally, the system of levels and units of activity is internally multifaceted, dialectic, and dually oriented. These features regarding the understanding of the structure of activity—which are not always clearly conceived, even by psychologists who base their work on activity theory—are the true prerequisites for the further development and refinement of this theory and its conceptual framework. References
Asmolov, A.G. 1977. “Deiatel’nost’ i urovni ustanovok.” Vestnik Moskovskogo universiteta, ser. 14. Psikhologiia, no. 1. Asmolov, A.G., and B.A. Petrovskii. 1978. “O dinamicheskom podkhode k psikhologicheskomu analizu deiatel’nosti.” Voprosy psikhologii, no. 1. Bernshtein [Bernstein], N.A. 1966. Ocherki po fiziologii dvizhenii i fiziologii aktivnosti. Moscow. Gal’perin [Galperin], P.Ia. 1976. Vvedenie v psikhologiiu. Moscow.

46 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Iudin, E.G. 1976. “Poniatie deiatel’nosti kak metodologicheskaia problema.” In Metodologicheskie problemy issledovaniia deiatel’nosti. Ergonomika-10. Moscow. ———. 1976a. “Deiatel’nost’ kak ob”iasnitel’nyi printsip i kak predmet nauchnogo izucheniia.” Voprosy filosofii, no. 5. ———. 1977. “Deiatel’nost’ i sistemnost’.” In Sistemnye issledovaniia. Ezhegodnik 1976. Moscow: Nauka. Leont’ev [Leontiev], A.A. 1974. “Psikhologicheskie osnovy obucheniia russkomu iazyku kak inostrannomu.” Russkii iazyk za rubezhom, no. 4. Leont’ev [Leontiev], A.N. 1964. “Myshlenie.” Voprosy filosofii, no. 4. ———. 1972. Problemy razvitiia psikhiki. Moscow (1st ed. 1959). ———. 1974. “Obshchee poniatie o deiatelnosti.” In Osnovy teorii rechevoi deiatel’nosti, Moscow: Nauka. ———. 1977 [1975]. Deiatel’nost’. Soznanie. Lichnost’. 2d ed. Moscow: Politizdat. Naumenko, L.K. 1968. Monizm kak printsip dialekticheskoi logiki. Alma-Ata. Petrovskii, V.A. 1975. “K psikhologii aktivnosti lichnosti.” Voprosy psikhologii, no. 3. Rubinshtein, S.L. 1935. Osnovy psikhologii. Moscow-Leningrad. ———. 1940. Osnovy obshchei psikhologii. Moscow. Smolian, G.L., and G.N. Soltseva. 1977. “Psikhologicheskie faktory optimizatsii trudovoi deitel’nosti.” Voprosy filosofii, no. 6. Tikhomirov, O.K. 1977. “Poniatie ‘tsel’ i ‘tseleobrazovanie’ v psikhologii.” In Psikhologicheskie mekhanizmy tseleobrazovaniia. Moscow: Nauka. Zinchenko, V.P. 1977. “Printsipy analiza funktsional’noi struktury poznavatel’noi i ispolnitel’noi deiatel’nosti.” In Deitel’nosti i psikhicheskie protsessy. Moscow. Zinchenko, V.P., and V.M. Munipov. 1976. “Ergonomika i problemy komleksnogo podkhoda k izucheniiu trudovoi deiatel’nosti.” In Metodologicheskie problemy issledovaniia deiatel’nosti. Ergonomika-10. Moscow.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 47–56. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440304

A.A. LEONTIEV

Personality, Culture, Language
To the memory of two great philosophers whom I was fortunate enough to know personally: Evald Ilyenkov and Meraba Mamardashvili.

What is personality? The concepts referred to in the title of this article can be defined— and have been defined—in a multitude of different ways. It would be impossible to introduce all of these definitions here and discuss their strengths and weaknesses. We will therefore attempt, without giving definitions or categories, to simply explain the possible interpretations we accept—and why. We will begin with personality [lichnost’]. Our interpretation of this category is rooted in the last works of L.S. Vygotsky associated with “height” psychology. In a manuscript from 1929, “The Concrete Psychology of Man” [Konkretnaia psikhologiia cheloveka] (Vygotsky, 1986), Vygotsky expressed an original view of personality as a psychological category that is primary in regard to activity and consciousness. It is not only activity and

English translation © 2006 M.E. Sharpe, Inc., from the Russian text “Lichnost’, kul’tura, iazyk,” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 119–28. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov.
47

48 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

consciousness, but, first and foremost, personality that is socially (culturally) determined—its essence is social. In many recent works devoted to the ontogenesis of language and to the human mind overall, there are a variety of interpretations of the process of interiorization. In the majority of these works, however, this process is reduced to the interiorization of cultural facts or (and) culturally determined operations. In a well-known book by M. Cole and S. Scribner (1976), the “role of culture in mental development” is understood only as the influence of culture on cognitive processes. In a study by P. Tul’viste (1987), categorization is understood as being within human experience that varies and develops under the influence of a particular ethnic culture. Something similar (with the exception of their view of values as the result of cultural determination) can be found in the works of J. Bruner, even in his article about Vygotsky (1985). Of course, there is no doubt that “the world is a world of symbols in the sense that it consists of conceptually organized, rule-based systems of knowledge about what exists, how to achieve goals, and what should serve as an object of evaluation” (Bruner, 1985, p. 32]. However, can one believe that this is only a world of symbols, confined to concepts, goals, rules, and so on? In jotting down notes about a personal conversation he had with the famous writer, V.F. Tendriakov, A.N. Leontiev once wrote, “I find (have) my ‘I’ not in myself (it is others who see it in me), but outside myself, existing in a conversation partner, in a loved one, in nature, and also in the computer, in the System” (1983, p. 241). Here “System” means the social structure, removed from a specific person—people can only personify the “System.” Another thought of Leontiev: “a return to the construction of the image of the external, multidimensional world in the consciousness of the individual, of the world as it is, in which we live, in which we act, but in which our abstractions do not ‘reside’ in and of themselves” (ibid., p. 255). At the same time, for Bruner, Cole, and many other psychologists, that is exactly what the world is—abstractions “residing” in it—only a conceptual construct—it is a theoretical world.

MAY–JUNE 2006 49

I know of only three major thinkers who understand the world another way: L.S. Vygotsky, A.N. Leontiev, and M.M. Bakhtin. Let us hear what Bakhtin has to say. “The world, where the act truly transpires and is performed, is the unified and unique world that is concretely experienced: seen, heard, felt, and conceived. . . . The recognition of my unique involvement in it, of my non-alibi, guarantees reality the unified uniqueness of this world” (Bakhtin, 1985, p. 511). If I do not put myself in this position, if I do not perform the act in a specific time and space, guided by specific emotional-volitional and motivational factors that are effective, and, at times, form here and now, then this world “disintegrates into the abstract-general, the onlypotentially-possible moments and relations that can be reduced to that same only-potentially-possible, abstract-general unity” (ibid., p. 512). This is the world in which our abstractions “reside”; however, we live in another, real world. Bakhtin goes on to talk about “large” and “small” experience.
With a small experience, there is one perceiver (everything else is an object of perception), one free subject (everything else represents something dead), one thing living and open (everything else is dead and closed), and one who speaks (everything else is dumbly silent). With a large experience everything is alive, everything speaks— this is experience that is deeply and essentially dialogic. The world’s thought is about me, the thinker; it is actually I who am objective in a subjective world. (Ibid., pp. 519–20)

My personality is the process (and the result) of placing myself in this key position in the world of a “larger experience.” It is an interiorized “one and only” world in its interrelations, with my perception and my act, with my motives and my will, with my social experience and my values. The process of perception only reflects a part of the process of internalization: it forms an orienting basis for my activity in the world. These processes of perception are subordinate to personality, which defines and regulates them. But, what is psychological personality? The answer can be found in the works of Vygotsky: it is the dynamic meaningful system

50 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

that incorporates motivational, volitional, and emotional processes, the properties of action and the properties of thought. All of these components have different relationships to one another and can form different “alloys.” “In the process of social life . . . there arise new systems, new alloys of mental functions, unities of a higher order, which rule particular laws, particular interdependencies, and special forms of connection and movement” (Vygotsky, 1984, p. 328). In the 1929 manuscript cited above, we find a somewhat different answer to the question raised: personality is fundamentally dialogic, it is always a drama, and not a simple (perhaps even contradictory) process or system of processes. For example, a person’s activity is determined by different social roles “played out” by that person. Dramatic conflict may arise between my role as a judge (“I must condemn him”) and myself as a human being (“I understand him”) (Vygotsky, 1986). To summarize: personality is the process of a person’s constant self-definition in the real world that regulates the processes of perception, act, experiences, and so on. Personality is primary in relation to activity and consciousness. Once the leader of the Georgian school of set theory, A.S. Prangishvili, once asked me, half in jest, “What is more important— our set or your activity?” I answered, as I recall, “If we say that personality is more important, then I do not think we will have any disagreements.” Only much later did I realize that this was no joke. What is culture? The understanding of personality described above greatly narrows the spectrum of possible definitions of culture. We cannot accept any of the interpretations where culture is synonymous with “a world of symbols,” and is interpreted as the aggregate or even a system of facts of culture that exist in a certain “social space” outside and apart from a particular person. R. Brislin provided the simplest of all definitions: “‘Culture’ relates to those aspects of society in which all of its members

MAY–JUNE 2006 51

participate and that they all possess it and pass it down to the next generation. ‘Personality’ correlates with the unique combination of features . . . that differentiates individuals within a given society” (Brislin, 1981, pp. 51–52). Actually, this definition is deeper than it may appear at first glance. If personality is a self-definition in the “large,” real world, then culture is a generalization of the same type as personality itself. According to Hegel and Marx, the ideal essence of external things “immediately exists only as the form (manner, image) of an activity of a social person” (Ilyenkov, 1964, p. 220). Correspondingly, culture is a system of ideal phenomena that has its own real existence in the process of social activity of human beings. A more consistent interpretation of the concept of “culture” in this sense is provided in the books and articles of the Armenian philosopher and ethnographer, E. Markarian (1969, 1973, 1983). For him, culture is the specific means of human activity. Cultural phenomena are any nonbiological means that make human activity goal oriented, allowing for the actualization of this activity in various situations and spheres of its application (Markarian et al. 1983, pp. 3, 4). Culture is always a dynamic unity of two currents: (a) the negotiation of existing standards and stereotypes, and (b) the standardization and stereotyping of innovations accepted by society (Markarian, 1969, p. 50). These ideas closely parallel Vygotsky’s thoughts regarding the dialogic essence of human personality. Not all social experiences or phenomena are cultural. Culture is function, but not substance: man as a social subject “is in control of himself” in a certain way that can and should be described in cultural terms. The same thing is true about social structure, social groups, and so on. There are at least three levels of culturally determined features of activity: (a) culturally determined features of individual mental processes and operations, in particular, categorization; (b) sociocultural norms, social roles, meanings overall that are associated with the consciousness of a given person; and (c) manners of behavior that are cultural in nature and are determined by particular features of one’s personality. In traditional

52 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Adygei society, there are, according to B.Kh. Bgazhnokov, two constructive principles that regulate decision making in various social situations. One is “honor” (namys), the other is “adygeism,” that is, the complex of personality traits attributed to a “true Adygei.” Expressed metaphorically, culture is a type of indicator of the optimal way of acting in the world and of understanding the world, and an indicator of the boundaries that influence the selection of experience in this optimal way. What is language? Let us start with the concept of the quasi-object, or the ideal object, first developed in the course of Hegelian ideas by K. Marx, and further refined by the Russian philosophers E.V. Ilyenkov and M.K. Mamardashvili. This ideal object serves in social activity as a conversion of real connections and relations. These connections and relations, carried out in the process of activity, are transferred and projected onto a material object, which is, by nature, alien to them. They are reflected in this material object and somehow push aside those properties that were previously inherent and that reflect its functioning as a material object. The object begins to play the role of a substitute for the connections and relations that are transformed in it, but the object is not a direct and immediate reflection of these connections, relations, and properties. A typical example is money, in which materiality is completely subordinate to function, and which is a conversion of economic relations in society. It is not surprising that in one of Marx’s manuscripts, money is compared with the sign, and in another manuscript, Marx writes, “Logic is the money of the spirit.” From here an extremely important epistemological problem associated with the analysis of the quasi-object as converted into real connections and relations: how can we separate what is associated in the quasi-object with its “substance,” its own properties, and its qualities, from what is carried over onto it and reconstructed within it? (Ilyenkov, 1964; Mamardashvili, 1970).

MAY–JUNE 2006 53

The quasi-object has its own “material existence.” But, being used as an “immediate body of an ideal image of an external thing” (ibid., p. 224), this quasi-object, can then be converted into a sign. In signs, according to Marx, “functional existence . . . so to speak, absorbs their material existence” (1960, p. 140). The sign is an “object that in and of itself has no meaning, but only represents and expresses another object with which it has nothing immediately in common” (Ilyenkov, 1964, p. 224). Of course, this last citation does not strike us, at first glance, as something fundamentally new in comparison with the traditional understanding. However, if language is understood as a system of such ideal objects, of linguistic signs in which a transformation replaces real connections and relations, we must not forget that between linguistic signs and quasi-objects overall, on the one hand, and phenomena of the external world, on the other, there is no direct and immediate correlation. We must not forget that a truly scientific analysis of the nature of the quasi-object demands that we introduce a mediating link, as Marx has done: this link is the system of social activity (Mamardashvili, 1968). In language, what is presented to consciousness does not begin to fully encompass the essence of linguistic meaning. The meaning that represents the ideal aspect of the sign is the result of the transfer and transformation (in the Marxist sense) of connections and relations of actual reality that occur in activity. Consequently, the classical “triangle” of semantic relations, stemming back to Ogden and Richards, do not appear to be complete to us (see Leontiev, 1975, 1976). Most linguists associate meaning only with the linguistic sign. But in reality there are at least three types of meanings: (a) linguistic (verbal) meanings; (b) meanings of images (images of perception, of memory, or imagination, that is, material meanings; and (c) meanings of operational components of human activity as immediate properties of this activity, for example, social roles as meanings. We will examine the second type of meanings, in particular, in the works of Vygotsky (“the meaning of things”), as well as the works of A.N. Leontiev in the article, “Perception and

54 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Activity” [Vospriiatie i deiatel’nost’], and his last manuscript, “Image of the World” [Obraz mira]. Similar ideas can be found in the works of S.L. Rubinshtein, D.N. Uznadze, and many others, including Western psychologists. In what way is the meaning of the word “table” associated with the meaning of an actual table that presents itself to our consciousness as an image of one type or another, or that we attribute to such an image? As a philosopher, I might answer this way: this image is by no means a sign, but a type of ideal object projected onto a real object, and subjectively merged with it into a unified whole (Vygotsky’s object “duplication”). As a linguist, I might answer: the word “meaning” has two meanings—a narrower one (actual linguistic meaning) and a broad one (a linguistic plus an objective meaning). In some situations, we mistakenly equate linguistic meaning with material meaning, and vice versa. But as a psychologist, I cannot give a definitive answer to just how these two types of meanings are connected: as strange as it may seem, this problem has yet to be studied seriously. In any event, in a normal individual, operations with linguistic objects (words) and corresponding presentations of objects were essentially identical in an (unpublished) experiment by L.A. Dergachev. It is paradoxical but true: a multitude of practical problems associated with teaching (and not only foreign-language language teaching) require the development of a comprehensive theory of meaning, for example, the problem of illustratability or the problem of the so-called visual semantization of a foreign word. We will attempt to summarize the positions stated above: language is the entire system of meanings, including both linguistic and material meanings, that reflect the qualities and properties of the real world. Exactly how they reflect it can be described in terms of “transformation” and “transference,” or as vehicles, which require a consistent distinction between real objects, quasi-objects, and signs. But, in talking about meanings, are we talking truly (and only) about meanings? In 1947, A.N. Leontiev first made reference to the fact that “sense” is a broader concept than “meaning.” Meaning is a type of

MAY–JUNE 2006 55

core of personal sense. In a manuscript discovered after his death, we find the same idea about the image of the world, but formulated in a different way. In all of his works, Leontiev used the concept of a “sense field,” which he borrowed from Vygotsky. In this field, objective meanings exist and function. These meanings, which represent more “sense” than “meaning,” are potentially and actually “built into” human activity, and reflect the motives, experiences, and values of real people; they also form a unified and unique “larger” world, which is interiorized. Between the “larger” world and myself as a personality, there is a constant dialogue. And language (in the broad sense) is the means of this dialogue. One of the most important premises of the Cartesian approach to scientific thinking is the clear distinction between an internal (mental) world and the World. We ask ourselves: how is the World reflected in “me,” in my internal world? However, in asking this question we are not able to understand that “I” is also a part of the World, and that this World exists only under the condition of my existence and my activity in it. I am an inseparable, integral part of this World. Otherwise, it is a different World. It is difficult to continue this reasoning today, at the very least because, while we understand the problem overall, we lack the answers to a multitude of specific questions. And, not least of all because, while working on this article, I arrived at a point where I realized with horror that a discussion of even the most fundamental problems associated with the synthesis of approaches formulated above demands an entire book, and not a brief article. So, for better or worse, it will be necessary to stop. Nevertheless, this article is necessary for me in order to enable those readers who are inclined to agree with its main theses to make further, independent steps in the proposed direction. References
Bakhtin, M.M. 1985. Literaturno-kriticheskie stat’i. Moscow. Brislin, R.W. 1981. Cross Cultural Encounters. New York[: Pergamon]. Bruner, J. 1985. “Vygotsky: A Historical and Conceptual Perspective.” In

56 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Culture, Communication, and Cognition: Vygotskian Perspectives, ed. J.V. Wertsch. Cambridge[: Cambridge University Press]. Cole, M., and S. Scribner. 1976. Kul’tura i myshelenie. Moscow. [Culture and Thought (New York: Wiley, 1974).] Il’enkov [Ilyenkov], E.V. 1962. “Ideal’noe.” In Filosofskaia entsiklopediia, vol. 2. Moscow. Leont’ev [Leontiev], A.A. 1975. “Znak i deiatel’nost’.” Voprosy filosofii, no. 10. ———. 1976. “Iazyk kak sotsial’noe iavlenie.” Izv. AN SSSR. Ser. lit. i iaz., no. 4. Leont’ev [Leontiev], A.N. 1983. “Iz dnevnikovykh zapisei.” In Izbr. psikhol. proizv., vol. 2. Moscow. Mamardashvili, M.K. 1968. “Analiz soznaniia v rabotakh K. Marksa.” Voprosy filosofii, no. 6. ———. 1970. “Forma prevrashchennaia.” In Filosofskaia entsiklopediia, vol. 5. Moscow. Markarian, E.S. 1969. Ocherki teorii kul’tury. Yerevan. ———. 1973. O genizise chelovecheskoi deiatel’nosti i kul’tury. Yerevan. Markarian, E.S. et al. 1983. Kul’tura zhizneobespecheniia i etnos. Yerevan. Marks [Marx], K. 1960. Das Kapital [Kapital], vol. 1. Moscow. Tul’viste, P. 1987. Kul’turno-istoricheskoe razvitie verbal’nogo myshleniia. Tartu. Vygotskii [Vygotsky], L.S. 1984. “K voprosu o psikhologii tvorchestva aktera.” In Sobr. soch., vol. 6. Moscow. ———. 1986. “Konkretnaia psikhologiia cheloveka.” Vestnik Moskovskogo Universiteta, ser. 14, Psikhologiia, no. 1.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 57–69. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440305

A.A. LEONTIEV

Sense as a Psychological Concept
In the title of this article, the term “sense” is used differently from the way it has generally been used in logic and philosophy since the time of Gottlob Frege. Nonetheless, replacing it with some other term is not possible, as in psychology—in particular in the Soviet Vygotskian school of psychology—it is as universally accepted as it is in logic, or almost so. For Frege, meaning is something designated by a “proper name,” a “sign,” that is, referent, while a sense of the name (Sinn) is the information it contains, and unambiguously characerizes the object or the path in which a name signifies an object. The classical example of expressions with one meaning but different senses is morning star and evening star. Furthermore, Frege distinguishes the idea between “meaning” and “sense” representations, the internal image of an object “arising from memories of sensory impressions that a person previously had. The representation is subjective: it is often permeated with emotions, the clarity of its constituent parts differs and is contantly changing; even in one and the same person, representations connected to one and the same sense are different at different times; one person’s representation is

English translation © 2006 M.E. Sharpe, Inc., from the Russian text “Smysl kak psikhologicheskoe poniatie,” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 141–52. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov.
57

58 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

not another’s.”1 If an idea is subjective, then sense is objective. It “can be the common property of many people, and, consequently, is not a part or mode of an individual mind; for one can hardly doubt that humanity has a storehouse of thoughts that it conveys from one generation to the next.”2 The modern version of Frege’s theory is most clearly presented in a well-known book by A. Church.3 Here we do not have the concept of the “representation,” but sense is informally defined as something that is assimilated when a concept is understood. Here, Church supplies a special footnote to emphasize that the concept of sense is devoid of any psychological implication. This footnote is quite indicative: what the true logician nowadays fears most is committing the sin of psychology. (As can be seen above, Frege had no such fear.) This fear is rooted in the commonly held objective of contemporary formal logic of operating with forms of knowledge , and not forms of thinking . This feature can be best characterized by words stated eighty years ago by the great Russian linguist A.A. Potebnia: “it (logic) is a hypothetical science. It says: if a thought is given, then the relationship between its elements must be of such-and-such a nature; otherwise, the thought is not logical. But logic says nothing about how we arrived at such a thought. . . . For example, in making assertions, logic does not examine the process of stating, but from its one-sided point of view evaluates the results of the completed process.”4 But let us return to Frege and ask ourselves the following question. What parameters does he apply in contrasting sense and representation? The answer is clear. A representation is a subjective category, a psychological category, because for Frege it is individual and weak; sense is a logical-objective category, because it “can be the common property of many people.” What can be “common property” can therefore not be a “part and mode of an individual mind.” In a word, Frege, like his modern followers such as Church, bases his thinking on the premise that the psychological = the individual = the subjective, while the logical = the common (social) = the objective. It is an understandable position, but not one that can claim to be

MAY–JUNE 2006 59

unique and mandatory. It would be sufficient to find a psychological theory in which the equations prove to be wrong, whereby the system of categories proposed by Frege would demand a significant reevaluation. Just such a psychological theory was developed by the Vygotsky school.5 We will examine several if its aspects that are fundamental to a psychological description of the concept of thought. According to this theory, man is not in relationships of adaptation with surrounding reality, like an animal, but of active mastery, of influencing reality. This becomes possible for him as a result of the human ability to foresee his own actions and to consciously plan them. And this ability, in turn, is conditioned by the fact that every uniquely human activity—practical, labor activity and theoretical, cognitive activity—is mediated by auxiliary means that are socially developed and preserved in the “collective memory” of society. In practical activity this is tools; in theoretical activity, this is signs, including linguistic signs. In being incorporated into human activity, tools and signs are not automatically “over-added to the mix.” They alter the very structure of activity, as they force man to form new, more complex connections in his mind, connections that permit new, higherorder forms of behavior. Thus, the human psyche, in incorporating tools and signs into activity, takes on a new quality, and does not merely undergo a quantitative change. Language and labor, or rather, labor and language restructure the human psyche down to its foundation, rather than basically refashioning it. Humanity’s sociohistorical experience, its material and spiritual culture, are man’s “essential powers” (Wesenskräfte) expressed in mediated form, in the “form of the existence” of human abilities (Marx). Every individual person “appropriates” (aneignet) these mediated abilities and properties, organically merging them with neurophysiological conditions that they have inherited genetically.6 The spiritual and mental development of the individual is the active process of appropriating sociohistorical experience in the course of his practical and theoretical activity; he does not “discover” the world for himself as a result of insight or anything analogous to it;

60 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

but over a long period, in a painstaking and orderly manner, molding the formation of his psychic awareness, erecting level after level, building them up to tear them down; and, finally, reaching a height from which the boundless horizons of human knowledge and human ability opens up before him, which is now accessible to him as well. So, humanity’s intrinsic “storehouse of thoughts” is indeed a “part or mode of an individual mind.” Furthermore, the “mind,” the individual psyche, can exist only by retrieving information from this storehouse. But the relationship here is two-sided—after all, the “storehouse” itself has real existence only within millions of individual psyches.7 This is the viewpoint held by L.S. Vygotsky and his school on the nature of human mental activity. With respect to its specific structural nature, it is determined first and foremost by the fact that, from the very start, there is a conscious goal in activity. Activity is planned and organized—consciously or unconsciously—in precisely a way such that the goal is achieved using optimal means and minimal expenditure of time and energy.8 In addition to a goal, the activity act is characterized by a specific motive; one and the same (on the surface) activity can be carried out as the result of different motives, driven by different needs. The attainment of a goal is the satisfaction of a need; with the attainment of the goal, the activity act is accomplished. We will emphasize once again a point made above: an individual’s assimilation of social values takes place in the course of activity. The child begins to use a spoon in its characteristic function not because he has some abstract knowledge of a spoon. He is simply presented with the necessity of eating his porridge on his own, and we make available an appropriate means to achieve this and give him a minimal idea of operations with these means (it is not surprising that any child, speaking any language, when he forgets the name of the spoon will say, “the thing for eating porridge” or something of that nature). The same is true for linguistic signs. This side of the problem was thoroughly studied by Vygotsky himself and has been repeatedly

MAY–JUNE 2006 61

dealt with since then.9 Using linguistic signs to the extent possible at each level of his/her development, the child gradually masters the appropriate rules of usage for these signs. At first, these are the same signs as those used by adults only in the case of objective reference, but the way of referring and the nature of the interconnection of signs with one another is different. The usage of signs has the opposite effect on the psychophysiological organization of linguistic ability,10 so the possibility for more complex forms of activity emerges, and so on; ultimately, not only object reference, but the subjects’ correlation per se, the rules for using a particular sign come to resemble the rules generally accepted in the given society or a given linguistic community. One and the same word with identical reference to objects and phenomena of the external world “means” different things to the child at different ages and different stages of development. In Frege’s terminology, the meaning of a word remains the same, but its sense develops as a means in which the name signifies the object development. In the psychological conception of Vygotsky and his school, the terminology used for these two aspects is not clearly differentiated; Vygotsky talks about the “development of concepts,” although the term “concept” is hardly appropriate for the genetic aspect of meaning. Below, we will first of all preserve Frege’s term “meaning,” and second, we will talk about “signification.” So, in entering into the activity of the individual (the child), the word, with its objective, essentially extrapsychological meaning, acquires a different, gradually developing signification that approximates the generally accepted one. We should note two aspects of this phenomenon. First, this becomes possible only as a result of the fact that the word has correspondence to the referent, as if replacing it by activity; in this sense, the mastery of meaning is the most important way—one might say the determining way—in which individual behavior is mediated through social experience. Second, the meaning of a word is by no means confined to the fact of its correspondence to a given object or class of objects (phenomena) of reality, although such a correspondence (object reference) forms its basis; in other words, our concept of “meaning”

62 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

does not correspond with Frege’s, despite identical terminology. If for Frege meaning is denotatum, for us it is something closer to the potential possibility of a word’s reference to a given referent or class of referents, realized through various “significations,” and depending primarily on the place of the given referent within activity. Based on the above, it can be concluded that meaning is a socially codified form of social experience assimilated by each individual. This codification is its constituting feature, which is expressed in the fact that the corresponding referent can be consciously perceived. What is meant by this “can be perceived? First, it signifies that the referent, that is, the fragment of reality reflected in meaning, can be an object of actual perception, toward which the conscious activity of the subject is directed. Second, it signifies that what we are aware of in objects or phenomena of reality turns out to be common to any person speaking the given language and living in the given society, who has attained a certain level of mental development. Third, it signifies that we have at our disposal some kind of objectivized psychological foundation on which we project, so to speak, our knowledge about reality. More often than not this foundation is the word.11 If we look at meaning as a fact of the psyche, as Frege’s “representation,” and not as “meaning,” we notice something paradoxical. While meaning is objective, it never appears to be an objective phenomenon for each individual. It is not a matter of individual variations in the acquisition of a given meaning, the differences “in sensory impressions that a person had previously,” and so on; the point is that in acquiring social experience that has been captured in meanings, the individual incorporates this experience into a system of his living relationships, into a system of his activity. And this is primarily expressed in the fact that any content encapsulated in meaning is perceived by a person in different ways depending on the motive of the corresponding activity. This was expressed well by I. Verhaar in a talk at the ninth International Congress of Linguists. “Thoughts that come with the word prison

MAY–JUNE 2006 63

will be different for an architect, a lawyer, a prisoner, or an official reviewing a verdict. But the meaning that is ‘put into’ the word prison is absolutely the same for all these people.”12 So if we attempt to define “sense” in our understanding, then it will be most accurate to talk about it in terms of an analogue of meaning in a concrete activity. But this is not simply an individualpsychological aspect of meaning and even less, of its affective coloring.
At the start of his life, a person usually behaves as if life lasts forever. But then something changes in his life or, perhaps, his life approaches its end, and the same person now counts his remaining years, even months. He hurries to realize some of his intentions and abandons others. You could say that his awareness of death has been altered. But has his meaning changed or expanded, has his awareness of the very concept, the “meaning” of death been altered? No. The sense has changed for him. . . . [And then] indeed in the first case the idea of death can be extremely affective for the subject, and in the second case, to the contrary, it cannot elicit any strong emotion at all.13

In order to take the next step in our reasoning, we will have to look at the relationship between meaning and sense from the perspective of the historical development of human consciousness. And, the first fundamental fact that we must encounter is the dependence of forms and manner of a person’s reflection of objective reality on the particular features of the society in which he lives. As it pertains to capitalist society, it has been described by Karl Marx as the fact of “self-alienation.” If we examine the collective activity of primitive hunters, we see that the motive is in agreement with this activity, or one could even say it coincides with the objective result. It is stimulated by each person’s share in the overall catch, but this catch is at the same time the result of the activity of the primitive collective and each of its members. Things are different in a class society, in particular (most keenly felt), in capitalist society. Here the motive for labor activity by a worker does not coincide with its result, the objective content of activity does not coincide with the subjective content, the sense of his labor does not coincide with its meaning.

64 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

The worker produces for himself neither the silk that he weaves nor the gold that he extracts from the mine, nor the palace that he builds. For himself he produces a wage, but the silk, the gold, the palace are transformed for him into a specific number of vital means, perhaps into a cotton jacket, into a copper coin, into a place to live in some basement. . . . The sense of his twelve-hour labor does not consist in the fact that he weaves, spins, drills, and so on, but in the fact that it is a way of earning a wage that makes it possible for him to eat, to go to a tavern, to sleep.14

A paradoxical phenomenon emerges, the very one that is characterized by Marx as “self-alienation” of the worker, alienation from his own essence. On the one hand, he is a creator, a producer of material value, the bearer of knowledge and skills, ensuring the life and development of society; on the other hand, for him the content of his labor is secondary, he is working not in order to produce, but in order to live. It is as if he is split in two in his relationship to labor, and this cannot fail to be reflected in his consciousness and his perception of material and social reality, the world of objects, and the world of human relations. It is not unusual for this bifurcation to take on hideous forms that, perhaps, are more obviously demonstrated in internal tendencies that are common to all capitalist society. “The glazier rejoices in the hailstorm that could break all the panes of glass,” wrote Fourier. This phenomenon is also vividly reflected in speech activity, giving rise to a sharp divergence between the meaning of words and their sense. Look at the (extremely telling) words that a contemporary Spanish writer, who has nothing at all to do with Marxism, puts in the mouth of his protagonist, a proletarian living on the outskirts of town.
Uptown folk have taken over the language of people on the outskirts. Words used to be coins—real or counterfeit. Now there is nothing but counterfeit in circulation. The words “bread,” “justice,” and “man” have all lost their original meaning. They have become empty sounds, instruments of lies. . . . Uptown folk have taken words, deprived them of their living flesh and transplanted them into their barren soil. The truth cannot roll from their tongue, just as grass cannot grow through the asphalt of their streets. . . . For them, the word “bread” does not

MAY–JUNE 2006 65

mean bread and the word “man” does not mean man. Each word has become a mousetrap for them, every phrase is a snare. . . . And the people from the outskirts have to keep quiet. Their tongues no longer obey them. (Juan Goytisolo, Undertow)

Let us attempt to analyze this tirade. Of course, Goytisolo’s protagonist does not actually believe that the meaning (in our understanding of this term) of the word “bread” is different for “uptown folk” and “people on the outskirts.” The word continues to signify bread. But the bread itself is perceived in different ways. For one person it is a means of survival, a means of satisfying a natural need, and, at the same time, it is a social value that is clearly recognized and easily measurable in terms of the amount of labor expended for it. There is a reason why peasants in a Russian village consider it impossible to throw away even a crust of bread, not to mention a whole slice, and sternly punish children if they violate this prohibition. For another person, bread is not perceived as having a social value; he pays for it in a store with an impersonal coin that has been easily earned, not with the sweat of labor, working at a lathe or plowing the furrows, and the attitude toward it is fundamentally different. The sense of bread, and, consequently, the perception of the word “bread” differs; the worker, the novel’s protagonist, intuitively senses this. The root of this difference is the social structure of capitalist society, reflected in the motivation of labor and other human activities, coloring their perception of such seemingly neutral words as “bread” or “person.” This does not mean that the difference in motives for an activity, and therefore the difference in the sense of a referent or the word that corresponds to it necessarily depends on class differences. But in a class society (most clearly in capitalist society), the objective meaning of the referent (or word) and its sense always diverge to some degree, since the “personal” interest, the “personal” motive and the overall interest of society diverge. A most important step forward was taken by socialist society in comparison with capitalist society, and this step was the liquidation of the socioeconomic basis for alienation.

66 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Whether or not class stratification of society is directly reflected in the particular features of sense, as a rule, sense is tied to the social structure of society in one way or another. Sense is not individual, since people’s relationship with objective and social reality is not individual; this relationship is always colored by a particular set of group interests. And it would be strange if it were any other way, since a person is not Robinson Crusoe living on an island, surrounded by a boundless sea of unfamiliar individualities. And no matter how hard he might try at times—in the person of philosophers such as [José] Ortega y Gasset—to publicly close his eyes to his social essence, it remains with him. So, sense, like meaning, is a form of social influence on the individual, a form of social experience that is required by the individual. But unlike meaning, it is not in a codified form. As a rule, this sense does not exist as something separate from meaning for the person individualizing it. On the contrary, it seems that a person immediately perceives the referent (or word) in its objective meaning. But all objects of human reality, like all words of human language, are seen by each of us as if through the prism of our “personal” (and, in practice, social) interest. And it takes a special effort of analytical thought to be able to rise above this interest to see the separateness of sense and meaning. *** We have covered a long distance over the course of a few pages. Having started by contrasting meaning and sense on a one-sided logical basis (Frege), we made our way step by step toward a fundamentally different point of view on their interrelations. Along the way we had to call on the findings of philosophy, psychology, logic, and linguistics. As a result, we have arrived at a certain understanding of sense, one that is only conditionally labeled “psychological” in the title of this article because, while psychological, at the same time, it extends beyond the boundaries of psychology in its traditional understanding. This understanding could be called semiotic if

MAY–JUNE 2006 67

contemporary semiotics—having forgotten the warning of one of its prophets, Ferdinand de Saussure—did not deliberately limit itself to a purely phenomenalistic view of the structure of sign systems. It could also have been called semantic—recalling the definition of semantics by Hayakawa as the study of the relationships between language, thought, and behavior—if only the understanding of the interconnection and relative significance of these three components among proponents of “general semantics” had not been diametrically opposed to ours. One way or another, the concept of sense outlined above and first proposed in 1944 by A.N. Leontiev is not, as the concept of meaning, the domain of a single science or a component of a single model of reality. Without demanding each time a fundamental change in the perspective on the objective essence of sense, on the objective significance of this phenomenon for the true reality of people in society, the concept of sense can also be used, and to a certain extent is already used, in a variety of aspects of scientific research by a number of sciences: philosophy, psychology, pedagogy, sociology, ethnography, linguistics, and finally, psycholinguistics. In serving within these sciences in its various aspects, we repeat, it does not lose its integrity and qualitative distinctness. Is it justified to advance such “global” notions in an age of scientific differentiation and a multiplicity of scientific models? Yes, if we view this differentiation, this multiplicity as a step toward integrating them into a new Science of Man; and, in particular, (as our next task) toward the creation of a “general theory of the position of language in social life,” about which D. Hymes dreamed not long ago.15 And this dream is not unfounded: observing the evolution of contemporary, including American, psychology, the evolution of contemporary linguistics, and so on, one can clearly see a tendency toward synthesis taking on increasing significance and increasingly coloring the system of ideas and concepts of these sciences. The opposite tendencies, no matter how strong they might be now, belong to yesterday’s science. One can agree or disagree with the theoretical dogma of one

68 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

school or another. But the history of science, the history of ideas is following its course, and it is essential that we be able to understand in time where this course is leading us. Notes
1. B.V. Biriukov, “Teoriia smysla Gotloba Frege,” in Primenenie logiki v nauke i tekhnike (Moscow: Academy of Sciences of the USSR, 1960). 2. G. Frege, “Uber Sinn und Bedeutung,” Zeitschrift für Philosophie und philosophische Kritik, 1892, vol. 100, p. 20. 3. A. Church, Vvedenie v matematicheskuiu logiku, vol. 1 (Moscow, 1960). 4. A.A. Potebnia, Iz zapisok po russkoi grammatike (Moscow: Uchpedgiz, 1958), p. 70. 5. In our subsequent discussion of the views of the Vygotsky school on the problem of sense we primarily have in mind A.N. Leontiev, A.R. Luria, and their followers. As far as Vygotsky himself is concerned, he shared the position of F. Paulhan, from whom he also borrowed the term “sense” (smysl) (see F. Paulhan, La double fonction du langage [Paris, 1929]); in Vygotsky, the distinction between meaning (znachenie) and sense (smysl) was generally equivalent to the distinction between connotative and denotative meanings that is generally accepted in linguistics. 6. See A.N. Leontiev and A.A. Leontiev, “Social and Individual in the Language,” Language and Speech, 1959, vol. 2, p. 4. 7. Or, as the same thought was expressed by F. Engels, “Human thought exists only as the individual thinking of many billions of current, past, and future people” (K. Marks [Marx] and F. Engel’s [Engels] Sobr. soch., vol. 20, p. 87). 8. The physiological basis for optimization of activity is provided in the works of N.A. Bernstein, primarily in the book published after his death, Ocherki po fiziologii dvizheniia i fiziologii aktivnosti (Moscow: Meditsina, 1966). 9. Here we will cite, in addition to Vygotsky’s famous monograph, our own book, The Word in Speech Activity [Slovo v rechevoi deiatel’nosti] (Moscow: Nauka, 1965), where the corresponding literature is listed. 10. Here it would be fitting to introduce, following the examples of P.K. Anokhin and A.N. Leontiev, the concept of the “functional system,” but that would take us too far afield. 11. But other forms of fixing meanings are also possible: “in the form of a skill, as a generalized ‘mode of action,’ of norms of behavior, and so on” (A.N. Leont’ev [Leontiev], Problemy razvitiia psikhiki, 2d ed. [Moscow, 1965], p. 289). Naturally, in talking about awareness, we did not have in mind awareness of meaning, and even less so, signification. Both of these are optional; for us, they are epiphenomena. On the other hand, the potential to be an object of awareness also plays a certain role in cases where the corresponding denotatum, reflected in meaning, does not immediately serve as the goal of an activity. 12. J.M.W. Verhaar, “Speech, Language, and Inner Form,” Proceedings of

MAY–JUNE 2006 69

the Ninth International Congress of Linguists (The Hague, 1964), p. 749. 13. A.N. Leont’ev [Leontiev], “Psikhologicheskie voprosy soznatel’nosti ucheniia,” Izvestiia APN RSFSR, no. 7 (Moscow, 1946), p. 28. 14. Marx and Engels, Soch., vol. 6, p. 432. Here and subsequently we rely in part on the reasoning expressed by A.N. Leontiev in his book Problems in the Development of Mind [Problemy razvitiia psikhiki] (Moscow, 1965), p. 315 f. 15. D. Hymes, “Review of New Directions in the Study of Language, ed. E.N. Lenneberg,” Contemporary Psychology, 1965, vol. 10, no. 12, p. 549.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

70 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 70–82. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440306

A.A. LEONTIEV

The Psychological Structure of Meaning
The very concept of the “psychological structure” of meaning already presumes the idea that it is made up of components. At the same time, of course, from the very start we discard all variants that equate meaning with a simple sum of substantial semantic elements. Just as a phoneme is not the sum of acoustic-articulatory differential features, meaning is not the sum of semantic components. Is it, perhaps, possible to understand meaning as a system of semantic components understood in the above sense? Such an understanding is rather widespread in contemporary linguistics in association with the ideas of the “semantic field,” and analogous concepts. But even if this understanding is acceptable, it is only as an abstract linguistic concept, that is, in a model of language; if we strive to discover an objective semantic systemicity on in the process of speech activity, then here, as the Rumanian investigator T. Slama-Cazacu1 has shown, we are dealing exclusively with what she has called “dynamic structurization of significations:” in other words, a system of psychological interconnections between words
English translation © 2006 M.E. Sharpe, Inc., from the Russian text “Psikhologicheskaia struktura znacheniia,” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 152–63. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov. Notes renumbered for this edition.—Ed.
70

MAY–JUNE 2006 71

(or, rather, their equivalents within a person’s linguistic abilities; at this point we will not analyze what these equivalents are) that is complex, labile, and constantly changing, and are influenced by various external and internal factors. It could be argued from a psychological (or psycholinguistic) perspective that meaning is actually a system of processes, a system of actions, but not a system of substantial elements. Accordingly, in L.S. Vygotsky’s definition, the meaning of a word is the “unity of generalization and association, communication and thinking.”2 It should be noted that association and generalization are understood by Vygotsky as processes that preserve their nature as such in the given definition as well. In other words, meaning is a dynamic unity, a kind of result of association and generalization. Meaning as a psychological phenomenon is not a thing, but a process, not a system or aggregate of things, but a dynamic hierarchy of processes. Here, evidently, the stated viewpoint should be clearly contrasted with all types of psychological theories of meaning advanced within the framework of behaviorism, as well as with certain neopositivist and pragmatist interpretations. According to these theories, meaning is limited to the understood reaction of an organism to a given word, to a given verbal stimulus. Meaning is a type of pseudonym for reaction—real or potential—to a corresponding word.3 Such an understanding appears to be unacceptable to us, inasmuch as we do not share the theoretical and methodological basis of behaviorism. Here again we will cite Vygotsky, who said in a paper that was not published during his lifetime: “Meaning is not the sum of all psychological operations that stand behind a word. Meaning is something significantly more definite—it is the internal structure of the sign operation. It is what lies between the thought and the word.” And, somewhat earlier on he wrote, “Meaning is the path from thought to word” (what is meant by word here, as in other works by Vygotsky, is its phasic aspect, something we would call the plan of expression).4 In other words, meaning can and, from our perspective, must, be defined first and foremost through the generation of speech as a specific method of moving from what we want to express to a

72 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

specific linguistic expression. Consequently, its psychological structure should be sought in the internal structure of the hierarchy of processes of the psychophysiological generation of a speech utterance. Here, one can anticipate objections. Is it really possible to fit meaning fully and completely into the structure of speech generation? It is well known how much a specific utterance—including the semantic component—depends on different types of situational, contextual, and other factors of a more or less random nature, on how unstable its generation is. At the same time, meaning, after all, is not something labile, plastic, dynamic. In a certain sense it is constant, permanent, static. It has been, so to speak, “lexiconized.” This is truly the case, and it is impossible to argue with this point of view. It is necessary, however, to clearly define concepts, which we are now attempting to do. Let us start by distinguishing two aspects of psychophysiological speech generation that manifest themselves at any level of this generation. Generally speaking, these two aspects are not related to speech; they are two sides of any organized human behavior that has any degree of complexity, two inseparable links in any activity, which was clearly shown by N.A. Bernstein.5 We have in mind the distinction between the mechanism (and, correspondingly, the process) of accomplishment of the activity, and the mechanism (or process) that controls this accomplishment. In order to demonstrate the difference between these two aspects we will cite research done by L.A. Chistovich showing that the phoneme is not the unit of generation as such, of what we just called the accomplishment of the activity: these units are acoustic-articulatory, differential elements (like I.A. Baudouin de Courtenay’s kinakem), syllables, syntagmas. But the phoneme is present in speech generation as a control unit.6 Until now we have been talking about the accomplishment of the activity; and here, at this level, there is no “dictionary” meaning. In our view, it appears only at the control level. A “dictionary” meaning is psychologically determined through a control mechanism only.

MAY–JUNE 2006 73

The second essential clarification in this case concerns the distinction between meaning as an objective psychological (or mental) phenomenon and meaning as a subjective phenomenon that arises in a language speaker’s awareness of his own speech. It is clear that even if a speaker of a language is not aware of the independence of meaning, its “separateness” from the signified object, its volume, and so on, does not fundamentally change anything in how he will use the corresponding word. The subjective relationship of the speaker to meaning is not what should interest us now—for us it is an epiphenomenon. In connection with what was just said, it is essential to emphasize that awareness itself can differ. Any child can separate out the consonants in a word, such as the K and Sh in the word kasha. But he is only aware of the phonemic makeup of a word—in most cases—when he starts to learn to read.7 An analogous distinction between separating out and being aware can be drawn in our case as well. Meanwhile, this separating out (a spontaneous process of separating reference points within the flow of speech) concerns this very link of the carrying out of activity, while the process of actual awareness is associated with the control link. We will make a small digression concerning the philosophical aspect of the question. What is referred to as meaning in most of our philosophical and gnoseological works corresponds precisely to meaning in the control link. For example, I.S. Narskii defines meaning in the following way: it “forms a permissible circle of cases, within which the operations of a subject, despite all their individual differences, correspond to a given meaning.”8 The understanding of meaning as a certain ideal equivalent of signs that are mutually interchangeable in the process of semiosis in the collective monograph Problems of Thinking in Contemporary Science [Problemy myshleniia v sovremmennoi nauke] comes close to this. According to the authors of this monograph, meaning is a sort of ideal “social substance” not immediately manifested in the material form of signs and realized only in the mechanism of interchange. This “social substance” of signs, “being the basis of their equivalency, cannot be . . . the external object itself, but just its thought reflection.”9

74 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Given such an approach, “thought reflection,” the psychological equivalent of linguistic meaning or meaning as a psychological phenomenon, is realized in the mechanism of the interchange of words. Consequently, the psychological structure of meaning is, first of all, a system of distinctive features of meaning, correlated with different types of interrelations among words in the process of real speech activity, a system of semantic components viewed not as an abstract linguistic concept, but within the dynamic of communication, in the entirety of linguistic, psychological, and social conditionality of the usage of a word. Or, to put it more simply: the psychological structure of meaning is determined by a system of correlations and contrasting of words in the process of their usage in activity, and not in the process of their comparison as units of a lexicon. Looking at it from the psychological perspective, what is this system of correlation and contrasting of words? It is a system of associative connections between words. The psychological structure of meaning, as indicated above, is its associative structure. There are at least two pieces of evidence that can be introduced in defense of this thesis. One of them is the result of B.V. Zeigarnik’s observations of the phenomenon of “disrupted thought” or “disrupted speech” among schizophrenics. Here, one and the same phenomenon of “disrupted speech,” according to Zeigarnik, can appear both in the form of semantic breakdown of speech (“I was an honest person, I wanted to look in the kitchen, I have a clock on which there is an attorney . . .”—Zeigarnik’s example:
The strength of the cone acts powerfully As a high-altitude law For elevation, like the horizon, In carry-up style flight

from the letter of a schizophrenic to the Academy of Science Division of Language and Philology in 1915), and in the form of a breakdown in the system of associations: “In less severe cases, when the ‘disruption of thought’ does not appear in spontaneous speech, experimental psychological research manages to detect a

MAY–JUNE 2006 75

rich and multidimensional character in patients’ associations, the domination of speech patterns featuring connections based on sound consonance.”10 In this regard, that is, if we allow that associative characteristics correspond psychologically to semantic characteristics (incidentally, H. Hörmann11 defines meaning as “knowledge of associations,” which is very close to our understanding), it is easy to interpret many phenomena known from experimental psychology, in particular, the phenomenon of semantic accumulation in the well-known experiment by Howes and Osgood,12 as well as the phenomenon of the phonetic merging of words in aphasics, such as the tractor plows → traclows [in Russian, traktor pashet → trashet—Ed.] when there is a disruption in the grammatical structuring of the utterance. The second piece of evidence for the thesis regarding the associative nature of semantic components of the word can be found in the works of the American psycholinguist J. Deese. Deese took several dozen words occurring most often in the Kent-Rosanoff associative norms, in reaction to the stimulus “butterfly,” and used them as stimuli in an associative experiment. After compiling a distribution matrix of answers and processing it in a certain way, Deese was able to study the mutually associative relations between words that objectively belong to the same semantic (associative) group (after the word “butterfly” was chosen, all subsequent operations, with the exception of separating syntagmatic answers from paradigmatic answers, there was no dependency on the meaning of the stimulus words and the reaction words). After submitting the results to a factorial analysis, Deese came up with a system of six factors. Factor 1 is realized in words that signify animals (moth, insect, bird, fly, beetle, bee, etc.). Factor 2 is realized in words that obviously stand in opposition to this feature (color, flower, blue, yellow, sunlight, garden, sky, nature, summer, spring), but its semantic “face” was not very clear. Factor 3 allows human beings to be separated into “positive” and “negative.” Factor 4 separates “nonliving” things (summer, sunlight, garden, flower, spring—blue, sky, yellow, color). In the “music” group, the first factor includes the words: music,

76 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

opera, orchestra, piano, song, symphony, instrument; the second includes tone, sound, noise, loud, listen, ear. In the group of “aggressive words,” one of the factors clearly describes weapons: gun, pistol, bullet, explosives, as well as explosion, shooting, and so on.13 Analyzing Deese’s results, it is easy to see that the factors turned out to be insufficiently defined where they described a collection of words that were clearly insufficient for a semantic grouping. From this perspective it would be interesting to conduct a comprehensive investigation of a more significant group, using more experimental material (Deese used only fifty subjects, which is clearly not enough for a statistically reliable data set). Such an investigation could be built around the Dictionary of Associative Norms of the Russian Language [Slovar’ assotsiativnykh norm russkogo iazyka], which is currently being compiled in the Russian Language Methodological Research Center [Nauchno-metodicheskii tsentr russkogo iazyka] at Moscow State University.14 In any event, Deese’s data clearly demonstrate the possibility of identifying factors conceptually interpreted as semantic components of meanings of experimentally studied words on the basis of formal processing of data from associative experiments. Let us move on to the second half of this article, where certain hypotheses will be advanced relating to the real psychophysiological content of meaning, interpreted by Vygotsky as “the path from thought to word.”15 Working with such an understanding, it is possible to view semantic features of meaning as the semantic criteria for selecting a word from long-term memory. The overall idea about such selective criteria, like the concept of long-term memory itself, belongs to D.E. Broadbent. According to Broadbent, there are at least three factors relevant to “long-term” memory: the semantic (associative) closeness of words, their sound, and their subjective probability characteristic.16 The idea that a probability principle is used in the speech mechanism belongs to B. Mandel’brot.17 The most recent works on memory, in particular publications by P.B. Nevel’skii, make it absolutely clear that this principle is

MAY–JUNE 2006 77

used exclusively in the link that we call (using Nevel’skii’s term) “permanent memory” (this concept is not identical to the concept of long-term memory, but we are not able to analyze the differences here). We used the term “subjective probability characteristic.” This term was chosen for a reason. It is associated with the fact that speech mechanisms do not use an a priori, let us say, linguistic probability, the frequency of a word in the texts of a given language, and so forth, but a value that reflects the organization of words according to the frequency in the activity of a particular speaker of the language and—most of all—in his particular activity (although the fundamental organization remains unified). This subjective probability organization, as it pertains to operating with nonspeech stimuli, has been extensively studied in the Soviet Union and the United States in recent times.18 Without reviewing all of the currently available materials concerning the probability organization of speech memory, we will focus on several aspects of greatest relevance to us. In their time, G.A. Miller and P. Nicely19 noted that in the perception of a word under white-noise conditions, its individual sound (acoustic-articulatory) features are perceived independently of one another and the thresholds of recognition are different for different features. So, the number of syllables can be determined even given a very high level of noise. In this connection, H. Savin20 advanced the hypothesis that, having a particular set of word features, we can conduct a search among a small group of the most frequent words that have these features; with an increase in the number of features, the power of the set from which the search was conducted increased correspondingly, and so on. Thus, the high or low recognition threshold is correlated with the power of the subset in which words that shared specific features were included; if there were few such words and words could be confused based on these features, then the threshold was lower, and vice versa. In other words, we produce something like movement in a field of sound features of a word, and in moving from one feature to another, we change the statistical parameters correspondingly.

78 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

In this connection, it would be interesting to inquire into the types of features that are being used in this process. Unfortunately, there has been very little work in this area. We will mention the exceptional article by R. Brown and D. McNeill about the “tip of the tongue” phenomenon and the publication by O. Spreen and colleagues.21 What both of these works address is not the question of the hierarchy of words used, but only their nomenclature, and even this is addressed to a limited degree. So this question remains open. But, we are primarily interested in the fundamental aspect of the problem. It is possible to advance the hypothesis that not only phonetic (acoustic-articulatory) features, but also semantic features can serve as a base of reference for features (or as a base of reference for features of word selection). That is, we “move” not only in a field of sound features, but also in a field of semantic features; if a word is characterized by a particular feature, then a search for that feature takes place within a particular subset of words that possess this feature. Only such an assumption reasonably explains the mechanism of selecting word utterances that are based on the latter’s semantic or semantic-associative characteristics. On the other hand, it is obvious that we will allow for such a path only for the generation of speech; perception of speech demands another method. But modern psychology and physiology of speech, based on the fundamental unity between the generation link and the perception link (“analysis through synthesis”), absolutely allows for a difference in their concrete organization. If this hypothesis is true, in a normal situation, the search evidently takes place along at least two22 independent, but coordinated, pathways, on the basis of a hierarchy of acoustic-articulatory features, and on the basis of a hierarchy of semantic features, with an emphasis on belonging, in most cases, to the semantic features. However, in some cases—specifically in children—in adults given the effect of a different type of stimulant, in schizophrenics, and so on—the coordination breaks down. The sharp ascendancy of a search for sound arises, and in the end the random selection of a word within the end subset. The opposite phenomenon (ascendancy

MAY–JUNE 2006 79

of a semantic search) results in the well-known phenomenon of “a loss of words” (“I know what word I want, I just can’t remember it”). Finally, the ongoing account of intermediate search results in accordance with this principle in a search using a different principle (motivated search), results in the equally well-known phenomenon of “sound symbolism.” D. Norman23 observed that word selection criteria, in long-term memory, are determined by a subject’s strategies specific to the given situation. Everything depends on the task that faces him: he will act one way if, let us say, he has to give a correct answer, and another way if a mistake is allowed, but it is important to keep reaction time to a minimum. By this means, we introduce the heuristic principle into our problem, and the place it occupies in the human mind overall, as it turns out, suggests that it can be applied more broadly than was done by Norman. In particular, it can be assumed that this principle can regulate the primary selection of one or another series of features (phonetic or semantic) used as a foundation for the search: that is, what a specific situation or class of situations depends on, whether or not we are searching for a word using semantic or phonetic criteria (not to mention how these criteria are tied to probabilities). Evidently, it is this heuristic path that a subject follows in the situation described by Goldiamond and Hawkins:24 if a word is not recognized, the subject in these experiments uses a special strategy eliciting the word-reaction more or less arbitrarily from a subset of the most frequent words. In this connection, we can also examine R. Abelson’s data, which shows that given a change of the utterance there is a group of verbs (help, need, use, buy, produce, etc.), whose meaning is particularly easy to generalize and apply to a new utterance.25 A problem arises in light of what has been stated above: can we believe that a word “lies,” or “is entered” somewhere in the brain “lexicon,” and the pathways and means of search described here are aimed only at finding such a record? We believe otherwise. In our opinion, a word is entered in the form of the search for this word. Operating by means of corresponding features, we are already

80 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

“reading the entry” in the lexicon. And it hardly makes sense to search somewhere in the nerve cells for an engram of a word sound form, or of some kind of “impression” with a label attached to it. So, the word is its search. If we have “lost” a needed word, this means that the word, as such, never was: we did not reach the end of our search. What is at the “end,” if not an engram? Evidently, simply a signal that it is the end of the search. What kind of signal, we do not know. In brief, “words are not preserved in memory as words, but as sets of features. When words are used, they are not reproduced by memory, but rather are reconstructed from the features that comprise these words.”26 The psychological aspect of a word’s meaning, as has been stated above, is not a thing, but a process or—taken more broadly, as the psychological equivalent of a “dictionary meaning”—both a thing and a process, but by no means merely a thing. This leads to serious conclusions regarding the structure of the linguistic sign27 that have not been touched upon here. In general, by adopting this point of view, it evidently becomes necessary to take a fresh look at a number of questions concerning linguistic semantics and linguistics overall. Notes
1. T. Slama-Cazacu, “La ‘structuration’ dynamique des significations,” Mélanges linguistiques publiés a l’occasion du VIII Congrès International des linguistes à Oslo, Bucharest, 1957. 2. L.S. Vygotskii [Vygotsky], “Myshlenie i rech’,” in Izbrannye psikhologicheskie issledovaniia (Moscow, 1956), p. 52. 3. See A.A. Leont’ev [Leontiev], Slovo v rechevoi deitel’nosti (Moscow, 1965), pp. 180–81. 4. From unpublished works of L.S. Vygotskii [Vygotsky], “Psikhologiia grammatiki” (Moscow, 1968), p. 187. 5. See N.A. Bernshtein [Bernstein], Ocherki po fiziologii dvizhenii i fiziologii aktivnosti (Moscow, 1966). 6. V.A. Kozhevnikova and L.A. Chistovich, ed., Rech’, artikuliatsiia i vospriiatie (Moscow-Leningrad, 1965). 7. Leont’ev, Slovo v rechevoi deiatel’nosti, p. 128 f. 8. I.S. Narskii, “Kritika neopozitivistskikh kontseptsii znacheniia,” in Problema znacheniia v lingvistike i logike (Moscow, 1963), p. 15.

MAY–JUNE 2006 81

9. Problemy myshleniia v sovremennoi nauke (Moscow, 1964), p. 82. 10. B.V. Zeigarnik, “Narusheniia myshleniia u psikhicheski bol’nykh,” dissertation abstract (Moscow, 1959), p. 20. 11. H. Hörmann, Psychologie der Sprache (Berlin, 1967), p. 227. 12. D. Howes and C. Osgood, “On the Combination of Associative Probabilities in Linguistic Context,” American Journal of Psychology, 1954, vol. 67. 13. See J. Deese, “On the Structure of Associative Meaning,” Psychology Review, 1962, vol. 69; and Deese, “The Psychological Structure of Meaning,” in The Structure of Association in Language and Thought (Baltimore, 1965), ch. 4. 14. See “Aktual’nye problemy psikhologii rechi i psikhologii obucheniia iazyku,” Slovar’ assotsiativnykh norm russkogo iazyka (Moscow, 1970). 15. L.S. Vygotskii, “Psikhologiia grammatiki,” p. 187. 16. See D.E. Broadbent, “Recent Analyses of Short-Term Memory,” Eighteenth International Congress of Psychology, symposium 21. Short-Term and Long-Term Memory [Kratkovremennaia i dolgovremennaia pamiat’], Moscow, 1966. 17. B. Mandel’brot, “O rekurrentnom kodirovanii, ogranichivaiushchem vliianie pomekh,” in Teoriia peredachi soobshchenii (Moscow, 1957). See literature on this question, for example, R.M. Frumkin and A.P. Vasilevich, “Veroiatnost’ slova i vospriiatie rechi,” in Voprosy porozhdeniia rechi i obucheniia iazyku (Moscow, 1967). “The hypothesis is that it is presumed that a word is ‘recorded’ in memory together with its probability” (R.M. Frumkin, “Problemy vospriiatiia slov v zavisimosti ikh veroiatnostei,” in Problemy iazykoznaniia [Moscow, 1967], p. 93). 18. See A.N. Leont’ev [Leontiev] and E.P. Krinchik, “Pererabotka informatsii chelovekom v situatsii vybora,” in Inzhenernaia psikhologiia (Moscow, 1964). 19. G.A. Miller and P. Nicely, “Analysis of Perceptual Confusions Among Some English Consonants,” JASA, 1955, vol. 27, no. 2. 20. H.R. Savin, “Word-Frequency Effect and Errors in the Perception of Speech,” JASA, 1963, vol. 35, no. 2. Savin’s hypothesis was proved in experiment: L.L. Havens and W.E. Foote, “The Effect of Competition on Visual Duration Thresholds and Its Independence of Stimulus Frequency,” Journal of Experimental Psychology, 1963, vol. 65, no. 1. 21. R. Brown and D. McNeill, “The ‘Tip of the Tongue’ Phenomenon,” JVLVB, 1996, vol. 5; O. Spreen, I.G. Borkowski, and A.L. Benton, “Auditory Word Recognition as a Function of Meaning, Abstractness and Phonetic Structure,” JVLVB, 1957, vol. 6. 22. Perhaps there are even more pathways. According to M. Anisfeld and M. Knapp, “each word consists of a series of features or attributes that uniquely characterizes it and distinguishes it from all other words in the lexicon. These features . . . incorporate semantic, syntactic, phonological, and—for the literate—orthographic aspects” (M. Anisfeld and M. Knapp, “Association, Synonymity, and Directionality in False Recognition,” Journal of Experimental Psychology, 1968, vol. 77, no. 2, p. 178). 23. D.A. Norman, “Memory and Decisions,” Eighteenth Congress of Psychology, symposium 22. Memory and Activity [Pamiat’ i deiatel’nost’], Moscow, 1966. Compare the work by Krinchik and Leontiev cited in note 18.

82 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

24. J. Goldiamond and W. Hawkins, “Vexierversuch. The Relationship between Word Frequency and Recognition Obtained in the Absence of Stimulus Words,” Journal of Experimental Psychology, 1958, vol. 56. 25. R.P. Abelson, “Heuristic Processes in the Human Application of Verbal Structures in New Situations,” Eighteenth International Congress of Psychology, symposium 25. Heuristic Processes in Cognitive Activity [Evristicheskie protsessy v myslitel’noi deiatel’nosti], Moscow, 1966. 26. Anisfeld and Knapp, “Association, Synonymity,” p. 178. See also D. Broadbent, “Perceptual and Response Factors in the Organization of Speech,” in Disorders of Language (London, 1964); H.S. Yavuz and W.A. Bousfield, “Recall of Connotative Meaning,” Psychological Reports, 1959, vol. 5, pp. 319–20. 27. See A.A. Leont’ev [Leontiev], “Iazykovoi znak kak problema psikhologii,” materials for the conference on Language as a Special Sort of Sign System [Iazyk kak znakovaia sistema osobogo roda], Moscow, 1967.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

Journal of Russian and East European Psychology, vol. 44, no. 3, May–June 2006, pp. 83–86. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440307

A.A. LEONTIEV

What Are the Types of Speech Activity?
It is important to understand that the types of speech activity taught in the mother tongue differ from methods of teaching foreign languages. We attribute these ideas to the renowned linguist and educator, Lev Vladimirovich Shcherba. In essence, it is both a methodological and a psychological concept. After all, learning to read, write, and speak is essentially the formation of specific verbal skills and the speech or communicative-speech abilities that are based on them (here is meant the use of skills for solving various specific problems, primarily of a communicative nature). The various types of speech skills and speech abilities are indeed types of speech activities. Four main types of speech activities are usually identified: reading, auditory articulation (listening; they share the designation of receptive types of speech activity); spoken language, and writing (together with written speech; the last two types of speech activity are usually designated as productive). The concept of types of speech activity in the methodology of a native language allows for a clearer understanding of the psychoEnglish translation © 2006 M.E. Sharpe, Inc., from the Russian text “Chto takoe vidy rechevoi deiatel’nosti?” in Iazyk i rechevaia deiatel’nost’ v obshchei i pedagogicheskoi psikhologii (Moscow and Voronezh: IPO MODEK, 2001), pp. 385–88. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov.
83

84 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

logical principles underpinning the formation of corresponding skills and abilities. It is logical to expect that methodological techniques, types of exercises, and so on, must relate to the structure and formation of corresponding psychological mechanisms, which are always complex and contain many levels. In practice, the need to support the formation of separate psychological operations and their complexities must be considered with the interaction of different types of speech activity, their mutual intertwining, especially regarding the solution of complex communicative problems. For example, underestimating the work in forming phonemic hearing gives rise to a number of mistakes in writing. The great Vygotsky wrote in one of his works: “It is not thought that thinks, but a person.” In just the same way, it is not the hand that writes, not the tongue that speaks, and not the ear that listens. A person, as an integral subject of psychological activity, as a personality, uses his speech skills (in the broad sense of the word) and abilities in life to solve the problems that confront him. And it is necessary to be literate not so much to receive a high school diploma as to become a full-fledged person among other people, to fully realize oneself. Any normal child can speak and hear (but not listen!). Nonetheless, in methodology we talk about the “development of speech.” What is it that is developed? What is developed is the ability to define the communicative problem (what is it I want to achieve?), to be aware of one’s speech, to make it volitional, and to be able to adequately select linguistic or speech means for the attainment of the goals at hand. But, in addition to the simplest speech skills (which are in fact extremely complex) and equally “simple” communicative abilities (for example, dialogic speech or “primer” reading), we teach more complex (to say the least!) abilities based on the same skills. It is these complex abilities of spoken language that are combined in the School 2100 system in the discipline of Rhetoric. The child who is able to communicate well with his parents or children his own age must learn from scratch how, for instance, to answer during school classes (the simplest form of

MAY–JUNE 2006 85

public speaking). To be able to hear what your conversation partner is saying is different from listening to and understanding what your teacher is saying during a lesson and being able to correctly react to it. The case of auditory articulation (listening) is no less complicated. It seems to many that this is something that does not need to be taught. Indeed, in many children, the listening ability forms spontaneously, without our active participation. However, years will pass, and it will turn out that during a lesson in one of the higher grades, one such child is not able to isolate the important points of what the teacher is saying and to memorize them, or even to concentrate on the process of listening at all. For some reason this is often called laziness or inattention, and the child himself is blamed. But he is the last one to blame: this is the result of insufficient education. The Chinese sage Confucius taught that if something does not turn out right, the worthy man blames himself and the fool blames someone else. Our teachers and methodologists would do well to take Confucius’s commandment to heart. As far as writing is concerned, to teach writing is something entirely different from teaching written speech, the written exposition of one’s own thoughts. Teaching someone to memorize and express someone else’s thoughts cannot be considered an adequate preparation for writing. Indeed, graduates of nonspecialized high schools are not able, for the most part, to write a commercial, scientific, or literary text. But they will need many of these abilities in institutions of higher education, at work, and simply in life. To teach written language is to open up an entirely new horizon of communicative possibilities for a pupil. It is hardly correct to “work on reading” or “writing,” or on spoken language and listening separately and in a disconnected way. In the pupil’s mind they are inseparable, and they merge together in any event, no matter how we might teach them. Would it not be better from the very beginning to construct integrated reading and writing lessons? This is exactly what we do in our system. And there are two more misconceptions. The first is that teach-

86 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ing reading, writing, and so on, must take place during special classes of reading or writing. The second misconception is that reading and writing fulfill a propaedeutic function only as they relate to a native language and literature (and generally speaking to the educational area of philology). In fact, teaching reading, writing, spoken language, and listening can and should be conducted in all lessons, even including mathematics. After all, a mathematical text is a special kind of text with its own language. Often even students in the higher classes cannot solve a problem only because they do not know how to correctly read and understand it, to make sense of it. The same is true for biology, chemistry, history, and all other subjects. In general, a thorough education in the types of speech activities in one’s native language is a condition for the successful study of the activity of a pupil within any educational discipline. It is therefore difficult to overestimate an integrated instructional approach to varying types of speech activity in the elementary school, as it relates to the future of a child. A pupil’s formal literacy is not the primary task of the elementary school (although it would be ridiculous to deny its necessity). Here the foundation of the child’s functional literacy must be established, as well as the capability for proficient spoken and written communication. And, at the same time, this is a tool of a child’s self-development, which creates a favorable environment for the development of the child’s “self-education” and socialization, and his/her overall values, relating to cognitive growth, personality and aesthetics.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

JULY–AUGUST 2006

89

Journal of Russian and East European Psychology, vol. 44, no. 4, July–August 2006, pp. 89–103. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO10610405440302

A.A. LEONTIEV

Some Problems in the General Theory of Speech Activity
For Vygotsky, the concept of the sign grows out of a specific theoretical interpretation of an activity that is distinctly human. It is a stimulus impacting the organism just as other stimuli do, however it is not a stimulus-object, but a stimulus-tool: one can discern a
dual relationship existing between the behavior and the external phenomenon: the external phenomenon (the stimulus) can at times play the role of the object at which a behavioral act is directed in order to resolve a problem facing a person (to memorize, to compare, to choose, to evaluate, to weigh something, etc.), and at other times it plays the role of the tool by means of which we direct and realize psychological operations (memorization, comparison, choice, etc.) essential to solving a problem.1

But this definition of L.S. Vygotsky touches on only one aspect of the sign, one that could conditionally—in the case of the language sign—be called the “communicative” aspect. But just as important—and philosophically, immeasurably more important—is another aspect, one associated with the above-mentioned role of language in the process of awareness and reflection of the external world as an ideal form in the mind of man. Here, it is most important to point out that the sign can genetically be traced back to the material object or phenomenon, to the “thing.” This “thing” (or “symbol,” in the terminology of E.V. Ilyenkov), in its capacity
English translation © 2006 M.E. Sharpe, Inc., from the Russian text excerpted from A.A. Leont’ev, Slovo v rechevoi deiatel’nosti. Nekotorye problemy obshchei teorii rechevoi deiatel’nosti, 2d ed. (Moscow: URSS, 2003), pp. 24–41. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov. Notes renumbered for this edition.—Ed.
89

90 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

as a mediating element of activity, takes on particular properties. The meaning of the symbol
always remains outside its immediately perceived guise, within other sensory-perceived entities and is revealed only through the entire system of relationships of other entities things to this given entity, or, the other way around, of the sensory-perceived entity to all other entities. If it is actually taken out of this system, this sensory-perceived entity loses its role, the meaning of the symbol. This shows that its existence and function as a symbol belongs not to the entity as such, but merely to the system within which it resides. For its existence as a symbol, the bodily, sensory-perceived encasement, the “body” of the symbol . . . is something absolutely incidental, fleeting, temporary. The “functional existence” of such an entity completely engulfs, in the words of Marx, its “material existence.” . . . And if this has taken place, then the material body of this entity will resonate with its function. As a result, the symbol will be transformed into a sign, that is, into an object that in and of itself has no meaning, but merely represents, expresses another object with which it has nothing immediately in common, such as, for instance, the name of the entity and the entity itself. . . . The function of the symbol is actually constituted as an immediate body of an ideal image of an external entity.2

So, if we are to introduce a concept of a sign operation, it is evident that in the genetic sense, this is an operation with two objects, but only in the genetic sense. However, if we examine the operation of a sign not only in its “cultural historical” or phylogenetic aspect, and not only in its ontogenetic aspect, but simultaneously, as something given, then it becomes clear that the stimulus-tool, or the sign, is not necessarily a real, material object. After all, in the final analysis a real object functioning as a stimulus-tool presumes the existence of a certain sensory, mental equivalent of this object, that is, it does not presume the realization in activity of all of the properties intrinsic to it as an object, as a thing, but only those objective properties corresponding to the given activity, its inclusion in an activity not as a “thing” in its totality, but as an aggregate of properties drawn from this object-entity. If, in a given instance, we are not concerned with the properties of a stimulus-tool that do not influence the fundamental structure of a particular sign operation and the structure of the activity overall, that is, the nature of the unity of operations into one dynamic whole, then it is evident that we can also envision a sign operation in which the role of stimulus-tool will be played not by an “entity,” and not even by a sensory-perceived object (such as, for instance, the auditory encasement of a word); indeed, in cases where a sign operation is directed toward self-communication, that is, when it serves

JULY–AUGUST 2006

91

exclusively as a means of self-regulation of behavior, the function of the sign, of the stimulus-tool, can be carried out by a mental equivalent of the sign not expressed in an immediate material form, such as the representation of a word (although the very fact of the emergence of such an equivalent presumes its secondary nature, its dependence on the actual object or objects or, to be more precise, on the genetic conditionality of its full-fledged, external activity). In this connection, it is useful to distinguish two aspects of the concept of the sign. In some cases we should talk about the virtual sign, while in others we should talk about the real sign.3 A virtual sign refers to certain features of activity divorced from specific sign operations and attributed to the corresponding material object that is fixed in a sign form; it concerns activities objectified in a sign. A real sign is an element of a specific sign operation. The key to resolving the question of the “materiality” of language rests, in our view, in distinguishing between virtual and real signs. It is completely obvious that a real sign is material or, at least, has the potential to be material— for instance, a word can always be pronounced. But a virtual sign is fundamentally nonmaterial. It is always and an abstraction, only inasmuch as we view it as a mediating link within activity (but compare below the two types of sign operations). As we can see, the distinction between virtual and real signs corresponds to two typical research situations in which we usually apply the concept of the sign. The demands of psychological research create the necessity of understanding the real sign; the psychologist, in particular the psychologist studying the mechanism of interiorization of external activity, is specifically interested in the structure and course of particular sign operations. The concept of the virtual sign is, perhaps, closer to a linguistic interpretation. It arises as a result of the necessity of finding a fixed point of departure in modeling various aspects of verbal activity. The problem is, as we will demonstrate below, that any study of language involves modeling it to attain certain objectives. A certain set of extralinguistic attributes is always involved in this process, introduced into the model without any methodological basis, that is, without a “prelinguistic identity.” They are selected arbitrarily for each given model. However, in cases where we perform global modeling of all three fundamental components of verbal activity, that is, when we create a system of mutually intersecting models, our opportunities for choosing the structure of these models, and the bases of prelinguistic identity supporting them are limited. The concept of the virtual sign is suited to the intersection of models of different components of speech activity. Let us take a closer look at the problem of the structure of the linguistic sign.

92 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Let us imagine that we have been given a set of objects (or their equivalents) serving the function of the stimulus-tool in activity. Let us also imagine that within this set there are two subsets, M′ and M″ . Any object in subset M′ can be substituted with another object from the same subset without changing the content of the sign operation and structure of activity as a whole. However, if that same object is substituted with an object from subset M″ (or the other way around), the content of the sign operation and/or the structure of activity as a whole will change. It is obvious that to simplify the analysis we can express all objects from each subset using a “conventional” virtual sign or, more precisely, using a certain model. The model will not reflect those properties, those peculiar features of each of the objects m′ 1, m′ 2, m′ 3 m″ 1, m″ 2, m″ 3, . . . , that do not effect their functioning as signs, i.e., that are irrelevant to sign function. Then we will be able—instead of set M of real objects—to operate in the study of activity using two conditional, generalized objects, two models—[m′ ] and [m″ ]—each of which will replace the series of objects m′ 1 . . . and m″1 . . . and correspond to a virtual sign. How will these models differ from one another? Each of them can incorporate, or, rather, combine really different objects if they are used in the same way in activity. Consequently, the material nature of the substituted object is irrelevant. The objective content of the activity, the function of the given object in activity is another matter. It is primarily in the objective content of the sign activity that our models differ from one another. To be more precise, in the description above, they differ only in terms of sign function, only in the objective content of the activity of the sign. In reality, however, such a simplified scheme, evidently, is never encountered in sign activity. First of all, the sign is never an isolated given, but is always part of a system of signs. Second, the sign is essentially polyfunctional; it is characteristic of lower order signs that they take on more complex functions, incorporating themselves in more complex sign systems. The third and final point is central—the sign, like sign operation in general, can serve not only as the precondition and form of the realization of a particular activity, but can also be the object of activity. In other words, the sign can serve as the goal of the action, and consequently4 can be conscious, can be “activated” in consciousness. Without, for now, delving into an analysis of this aspect of the question, let us limit ourselves to the statement that the most typical action performed on language signs is the juxtapositioning of signs (words) in terms of their acoustic form and corresponding acoustic analysis of words, that is, the identification of “diacritical elements” in them,5 which for now can be conditionally equated with phonemes. Keeping this aspect in mind, it is useful to

JULY–AUGUST 2006

93

introduce at least two components of the sign model into the analysis: these are its semantic component, reflecting the function of the sign as a stimulustool, that is, its place in activity that corresponds to it, and the phasic component, reflecting the essential characteristics of the material “body” of the sign that become evident in the process of comparing signs, and the change that accompanies change in the function of the sign.6 The progressive development of practical, production activity is expressed (a) in the change of function of the tools used in this activity, particularly machinery; and (b) in the increased complexity of the construction of these tools (machinery), which leads to the possibility of change in their function; and, as a result, creates the opportunity for functional change on a new level. An analogous process takes place in theoretical sign activity, in particular, speech activity. Corresponding to the change in the function of the tool there is a change in the functional content, the functional load of the language sign (given the identity of its language status); and corresponding to the increased complexity of the construction of the tool, there is a change in the language status of the given sign; that is, first and foremost a change in the morphological and phonological nature of the word or class of words. Of course, such changes in the linguistic status of a language sign occur in steps and unevenly, but this says nothing about the process of development of sign activity in general, which lays the ground for these changes; and A.A. Potebnia was right in comparing them “with footprints in the sand; they can be followed, but that does not mean that they contain the foot itself; and the word does not contain the meaning itself, just a footprint of the meaning.”7 The reason behind the irregularity and step-by-step changes in the linguistic status of the language sign is that the sign never enters into activity on its own, as an independent, isolated unit; it can be viewed in this capacity only if we deliberately limit ourselves to an analysis of sign activity that is separate and removed from the overall context of activity. If we step out of this framework—and in examining any evolution of the sign, we must step out of it—then it turns out that along with external factors rooted in the particular features of the activity, the sign is also subject to the action of internal, purely linguistic factors, and that it impacts a number of differently directed vectors and finds itself situated in different types of interconnections with other signs.8 In other words, a “sign system” or “system of signs” is the way we designate a special type of dependence in the functioning and development of signs—specifically an interdependence of signs that are analogous in terms of their place and function in activity. Such a dependency can come in many forms; for example, the functioning

94 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

and development of any given sign is guided by a whole array of differing interconnections with other signs. When we talk about a language system or rather about the aggregate of language subsystems, what we have in mind is one or another form of an objective interaction of the functioning and development of signs within a sign system. What do we mean here by “forms of interaction”? Primarily, they are the factors that determine the development of the semantic aspect of the sign in different directions or the need to create a new sign in order to capture and express new content. They are also the factors that specifically determine what the newly created sign will be. They are the factors that condition the morphological articulation within a sentence; that is, the factors that in the final analysis determine the distribution of signs classifying them in terms of their syntactical functions. They are the factors that determine the acoustic (and the overall phonetic) makeup of a given sign-word. Finally, they are the factors that determine any variability in the sound of a given word in connection with the varying conditions of communication. It should be said that such a view of the nature of systematization is nothing new. It relates back to the well-known philosophical thesis about the identity of connection and interaction inasmuch as “entities are mutually connected, implying that they influence each other.”9 In the history of linguistics (independent of the general philosophical understanding of system and connections) there have been several attempts to comprehend the concept of the system as a certain constant form of interaction of elements (W. von Humboldt,10 Baudouin de Courtenay, and others). In recent years such attempts have been increasingly tied to the correlative concept of “system-norm.” In the view of E. Coseriu,11 a language’s “system” is a system of language invariants or an aggregate of language phenomena that serves in defining a specific function in language (usually the function of distinguishing), and can take the form of a set network of oppositions (structures). A “norm” in language12 is an aggregate of the language phenomena that does not carry an immediate distinguishing function in language, taking the form of common and generally accepted (traditional) realizations. We will introduce specific examples from the field of phonetics (phonology). As is well-known, the correlation of consonants by hardness and softness does not apply to sibilants or the letter ts in standard Russian (system). However, some sibilants (zh, sh) and ts, wherever they are placed, sound like hard consonants [zhir] (fat), [shes’t’] (six), [tsep’] (chain), etc.; others that do not correlate to them (sh’, zh’, ch’, according to R.I. Avanesov’s transcription system), whatever their placement, are realized as soft: [ish’u] (I

JULY–AUGUST 2006

95

search), [dazh’:a] (even), [noch’] (night), and so forth (norm).13 In the Russian language there is a correlation based on whether consonants are voiced or voiceless, which serves a functional role, that is, it serves to distinguish (system). However, at the end of a word (and in certain other positions), this correlation is neutralized, that is, there is a devoicing of voiced consonants (norm). Because the phenomenon of neutralization of phonemic oppositions (as with any positional alternation) should be considered as a reflection of the norm and not system, it is evident that a change of positions, being common to all speakers, is not imperative by nature; if, instead of the normal l’ek] we pronounce [v’esh’:ij ol’eg], this does not interfere with [v’esh’:’j Λ our understanding, although it will leave a somewhat strange impression.14 We find phonological correlations in the same identical form in the speech of any native speaker of a given language. Here there can be no question of “correctness” or “incorrectness.” At the same time, there are a number of phenomena relating to the phonetic norm—the nature of a vowel reduction or the degree of assimilation, for instance—that vary from speaker to speaker. [RΛd’ilsa] or [rΛd’ils’a] (was born), [vΛnz’it] or [vΛn’z’it] (plunge, thrust), [l’isa] or [l’iesa] (fox)—these distinctions are not essential to communication, they do not carry communicative weight. This is why in linguistic practice the problem of normalization arises, a problem that is essentially nonlinguistic and is usually decided either statistically or by turning to some “language authority.” While the phenomena of norms do not create any functional oppositions— it represents a system of identification and not of opposition—one can talk about a certain functional weight involved in phenomena of norms. This weight involves the opposition of one norm to another norm, of one system of realization to another system of realization. The opposition of different degrees of reduction can serve to distinguish styles of speech; on the one hand, the opposition of one norm to another can reflect social differentiation of a speech community.15 To get somewhat ahead of ourselves, we will say that the concept of norms permits us to move from the category of linguistic standards to the category of linguistic processes. On the other hand, if we take into consideration that the variability of system phenomena is limited by individual psychology, it can be said that the category of system itself forms a transition between the category of language standards and the category of language ability. As it pertains to phonetics, the concept of norm as a product of averaging the most typical phonetic features of individual speech (or norms as “aggregates of constant sound elements independent of their function”)16 is not Coseriu’s discovery. This concept is implicit in many works on phonetics

96 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

and phonology, especially those by linguists belonging to the Prague School.17 In developing the views of I.A. Baudouin de Courtenay, G. Ulashin came close to the idea of norms in opposing a “subjective system based on . . . a subjective acoustic-articulatory unity” and an “objective system based on an internal functional unity.”18 Finally, the concept of norms is encountered in the works of adherents to the “phonometric” school, which we will not be able to address.19 However, Coseriu was the only one to systematically delineate the phenomena of system and norms,20 identifying, instead of two traditional branches of the study of the sounds of speech (phonetics and phonology), three independent disciplines—allophonetics, normophonetics, and phonology. Allophonetics deals with individual speaking, normophonetics—with the overall language “system” of realization, and phonology—with the functional system or structure. Developing his idea further, Coseriu contrasted norm and system using yet another important parameter: a norm is “an aggregate of already realized forms and models that should be used as such,” while a system is “a mechanism allowing for a departure from what is already realized, that is, ‘a system of possibilities.’”21 From this interpretation of system, which stems directly from Humboldt’s definition of language form, it can be concluded that:
Usually a system is viewed as a given, and change, as a problem. However, strictly speaking, it would be more logical for system and change to exchange places: the “becoming” of a language element precedes its “realized state.” . . . Activity that creates language is itself a system; . . . the thing “due to which language is language” is . . . the language activity creating language and preserving it as a tradition.22

A.W. de Groot arrived at a similar interpretation somewhat earlier. In the history of words, in his opinion, one does not see a tendency toward orderly change, as is usually thought, but a tendency toward a preservation of the sound of a word that is undermined by various (primarily social and psychophysiological) factors. For him, systematization represents the boundaries and trends in the disturbance of the development of identity. In this connection, de Groot promotes a concept of “norm” that is close to Coseriu’s concept of norm.23 And even before 1931, analogous ideas could be traced back to the views of I.A. Baudouin de Courtenay, which were expressed by E.D. Polivanov.24 It is natural, however, that the linguist does not deal directly with a system in this sense of the word, with system as interaction. What he calls a “system” is a certain abstraction, a skeleton comprised of imagined elements (units) that models—albeit in a very simplified form—actual internalized relations. This is how the majority of contemporary linguists understand the system

JULY–AUGUST 2006

97

of language. As a rule, however, they are not clearly aware that, in staying within the framework of language, they are not dealing with an actual systematization, but with a model or complex of models. For this reason, within contemporary linguistics, the concept of a system of language (and equally, a structure of language) is not simply ambiguous, but in the majority of cases, has not been given any explicit definition or interpretation whatsoever, remaining essentially an intuitive concept.25 What is the content that most linguists assign to the terms “system” and “structure” at present? It seems that we can identify at least three approaches to interpreting the concept of system, assuming we stay within the bounds of a statistical understanding (a synchronic understanding). The first of these approaches attempts to ascertain a concept of system at the level, so to speak, of language (of a linguistic standard) and nothing more. This viewpoint is the most widespread, and it is not surprising that the sessions devoted to this question at the Erfurt Symposium arrived at a definition of system as “an (internal) articulation, inherent in the material, of (mutually dependent) elements or linguistic devices that can be formally identified through either a paradigmatic or syntagmatic analysis, or both.”26 Within that approach, two trends can be identified. One of them applies the concept of system both to paradigmatic and syntagmatic relations, at times introducing different terms for these two types (the London School: system in paradigmatics, structure in syntagmatics), while the other adheres to an exceptionally traditional “purely paradigmatic” understanding. The second approach reduces the concept of language system to an “individual language system.” This viewpoint is represented, in particular, by E. Buyssens, who states outright that “in each person there exists a language system. . . . The language of a people is the aggregate of diverse individual systems and, in particular, what is common to all individual systems: the language itself is not a system.”27 For the third approach, a language system, as formulated by R. Mikusch, is “no more, no less—a construct of the scholar that is absolutely the same as the periodic system of elements or the classification of birds in zoology.”28 Along with these three approaches, we can divide all linguists into two groups depending on whether or not they recognize a unity in the system of language (E. Benveniste, A.A. Reformatskii) or view it as a complex of subsystems (R. Jakobson, M.M. Gukhman). Some authors (V. Skalichka, N.N. Korotkov) attempt to find a compromise solution to this dilemma, interpreting the relations within each subsystem as being paradigmatic par excellence and connecting separate subsystems into a syntagmatic plan. As should be evident from the above, we do not share any of the views on

98 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

system described here. In other words, for us, a language system is neither a text structure nor an “individual” language system of speech behavior, nor a pure construct: it is the actual form of interaction among the elements of speech activity that can be interpreted by using various models. It is essential that the units or elements forming the system be very clearly distinguished from the system itself, which is the form of organization of these units. It is obvious that, in recognizing any mode of existence as a system and asserting its being in a particular form, we are in no way obliged to apply this same assertion to its units. In other words: systematic development of the phonetic aspect of language is not, in principle, tied to any particular phonological theory or to any particular understanding of the phoneme. In general, the problem of the unit is secondary in comparison with the problem of the system of language; the language system is objective. It exists before the linguist and is independent of him. The system’s nomenclature, the complex of units, albeit organized accordingly, are constructs of the linguist. The language system is conditioned by the subject of study; the units to a great extent depend on the method of study, although, of course, they are not only determined by it. To make sense of this, we will follow the reasoning of a linguist studying and describing a specific language. The first thing that the linguist must deal with in this case, something that he absolutely must have before him, is the object of study itself, that is, the concrete language. We cannot, however, directly observe the language in question—we can only observe specific examples of speech, the verbal activity of the language community. It would be even more precise to say that we do not have before us he verbal activity of a language community, but always the verbal activity of a specific speaker of that language. In studying this activity, we more or less explicitly exclude all the features that are not characteristic of the verbal activity of other speakers of that language. And, finally, even more to the point, what is happening could be formulated in the following way: the linguist not only consciously or unconsciously ignores everything individual in the verbal activity of his informant, but he also ignores what is not needed for the specific task of describing language. It can also work the other way: depending on a specific task, he may focus attention on certain features of verbal activity that, given a different approach, would not be taken into account (just compare the analysis of Russo-Japanese phonological parallels and the discrepancies of E.D. Polivanov).29 The second thing that the linguist who is describing a specific language

JULY–AUGUST 2006

99

has “on hand” is a linguistic tradition, a system of units and a method of analysis (a system of scientific theories) that have been developed by science at the moment that the linguist starts his work. And finally, there is a third component that shapes any linguistic description. That is the specific objective of the research. This objective determines precisely which of the available units and methods will be used in each specific instance. Returning to our linguist, we must state that before him, as a “given” in describing language, there always arise the verbal activity of a specific speaker of the language and a system of units and methods of analysis. Thus, the linguist does not find himself in a situation that can be depicted by using an arrow uniting two points. Language Knowledge of language

• • His situation can be more precisely described as shown in Figure 1. In other words, our knowledge of language is mediated by a system of units and methods of analysis. Language is realized in speech activity as a multifaceted phenomenon, so that in speech activity we can observe not only what has been represented in our diagram as a1a3a6a8, but also a2a4a5a 7, which is “not needed” by the linguist (at least at the given moment). What forces us to choose just a part of the multitude of language data, in this case, a1a3a6a8, is the system of scientific theories “imposed” on us. But what do we end up with once we have “filtered” verbal activity through this system of scientific theories? It is something far removed from knowledge of language, which is why this term has been enclosed in brackets in Figure 1. For now, it is just a model of language. For it to become knowledge of language we would need to confirm the validity of this model of language reality. One way to confirm such a model would be to conduct a linguistic experiment that could be constructed in any number of ways. Having been confirmed and verified, the model could be placed at the foundation of knowledge about language; however, it is not the same thing as our knowledge about language, which represents an abstraction from all possible models of a given language. Until now we have placed our linguist in an essentially linguistic situation. However, a no less characteristic situation for him can be called metalinguistic. While in a linguistic situation we have speech activity and a system of scientific theories as a given, which, as a result, provide us with a model, at this point the circumstances are reversed. Knowing the nature of the needed model, having it as a given and having speech activity also as a given, we accordingly reshape the system of units and methods of linguistic

100 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Figure 1. [Language, System, and Speech Activity]
[Knowledge of language]
b1 b3 b5 b8

System

a1

a2

a3

a4

a5

a6

a7

a8

Speech activity

Language

analysis and the system of scientific theories, something that happened, for instance, during the development of descriptive linguistics. It is interesting to look back on how, over the history of linguistics, a gradual differentiation occurred and an awareness of the distinction between separate components of the linguistic and metalinguistic situation arose. The first period in the history of the science of grammar is the logicalgrammatical period, the period “of rational grammars.” It is characterized by a disregard for the specific task of description, a disregard reflected in the complete absence of a metalinguistic context. The system of scientific theories here is a system of theories as in the most extensive sense of the word: it represents a priori traditional schemes, subjected to no proofs whatsoever. Based on this schema, one finds the results of transferring the approach in logic to language research, which results in substituting a linguistic model with a model of logic. This is the source of the one-sidedness and narrowness of the understanding of the very subject of linguistics. A classical example of such a construct is the famous General and Rational Grammar by Lancelot and Arnauld (the Port-Royal Grammar, 1660). R.O. Shor characterizes it in the following way:
General laws of the structure of speech are defined by the authors of the Port-Royal Grammar—in keeping with the views on language held by the rational philosophy of the seventeenth century—as laws of formal logic; the sentence is equated with judgment, a word is defined as a sign of understanding; and all grammatical definitions are given accordingly. For the authors of the Port-Royal Grammar, in language, everything is subordinate

JULY–AUGUST 2006

101

to logic and expediency. They . . . conceive language as an external mechanism obedient to the will of scholars who set down laws to govern it in their grammars.30

The second period is completely dominated by the gigantic figure of Wilhelm von Humboldt, who was able to completely overturn the old approach.31 The history of the scientific study of language actually begins with Humboldt, since he was the first to clearly proclaim the ontological independence of language, to introduce the concept of linguistic forms,32 and to declare that language categories are not fundamental replacements for thought categories and that language is a “formative organ of thought”—an understanding that, in the end, is shared by all modern science. Humboldt was the first to construct an actual linguistic model and the first to clearly determine the object of linguistic study. The third period can conditionally be called the psychological period, and it included the neogrammarians and Steinthal, who stood at the fountainhead of neogrammarianism. During this period, the system for describing language was brought as close as possible to data of immediate linguistic perception; either consciously or unconsciously, language was equated with linguistic ability. The problem of method during this period was simply eliminated—the basis for constructing a model of language was the intuition of a speaker of that language. Many vestiges of this period have been preserved in modern linguistics and the worst of them is intuitionism—the uncontrolled introduction to linguistics of the intuition, the linguistic “feel,” of the speaker. We are far from thinking that it is possible to construct a language’s grammar on a purely rational or axiomatic basis; but it is clear that intuition in the study of language must be assigned a limited and strictly defined place. The next and fourth period—the sociological period—is represented by Meillet and the Geneva School, Shcherba and Polivanov, Sapir and Marr. They are typified by a different understanding of the subject of linguistics as well as an explicit separation of linguistic standards (language) and linguistic ability. After this period there is a certain turning point. Previously the problem of method—if it was raised at all—was implicit; a linguist did not consciously place himself in a metalinguistic situation. But now, with the emergence of structural linguistics, this entire circle of problems became explicit. Structural linguistics—descriptive linguistics in particular—is a modeling approach to linguistics par excellence, and in their best works, proponents of descriptivism explicitly take into account all three of the components mentioned above—the nature of the object, the system of methods and the

102 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

units of analysis, and the specific task of modeling that is connected to it. We will go more deeply into the concept of the model and modeling in general and as they apply to language in particular. Notes
[Bibliographic information is incomplete in the original.—Ed.] 1. L.S. Vygotskii [Vygotsky], Instrumental’nyi metod v psikhologii, pp. 226– 227. See, for example, A.N. Leont’ev [Leontiev], Razvitie pamiati; V.V. Davydov, Analiz stroeniia scheta, and others for experimental research of the role of “stimulustools” in the psychological life of modern man. 2. E. Il’enkov [Ilyenkov], Ideal’noe, p. 224. 3. These terms were proposed by A.N. Leontiev. A similar distinction (virtual and actual signs) is made by E. Coseriu (Determinación y entorno, p. 34). 4. “In order that there be awareness of a perceived content it must occupy the structural place of the immediate goal of action in the activity of the subject” (A.N. Leont’ev, Psikhologicheskie voprosy soznatl’nosti ucheniia, p. 11). 5. K. Bühler, Sprachtheorie, p. 33. 6. The above relates to the virtual sign. However, as with any concept developed for a class of real objects or phenomena, it can be attributed as well to specific phenomena or objects belonging to a given class or, as in this case, used in examining specific sign operations, that is, real signs. It was specifically during this type of examination of verbal activity that the understandings of phasic and semantic components of the sign were proposed. This was done by L.S. Vygotskii in his book Myshlenie i rech’. 7. A.A. Potebnia, Iz lektsii po teorii slovesnosti, pp. 133–34. 8. This aspect of the question has been extensively developed by linguists of the Geneva School. Compare, for instance, the concepts of “grammatical transpositions” and “semantic transpositions” in S. Karcevski (“Système du verbe russe” and “Du dualisme asymétrique”); and H. Frei, “Ramification des signes dans la mémoire.” 9. F. Engel’s [Engels], Dialektika prirody, p. 392. For a discussion of the “dynamic” nature of the concept of connections compare I.I. Novinskii, Poniatiia sviazi v marksistskoi filosofii. 10. W. von Humboldt, Über die Verschiedenheit, pp. lviii–lix, and elsewhere. 11. See E. Koseriu [Coseriu], Sinkhroniia, diakhroniia i istoriia; E. Coseriu, Sistema, norma y habla and Forma y sustancia; E. Coseriu and W. Vásquez, Para la unificación. Compare also A. Avram, Despre fonologia normej; N.D. Arutiunova, Ocherki po slovoobrazovaniiu, pp. 31–40; N.N. Korotkov, Norma, sistema i struktura, and others. 12. Coseriu’s concept of norm is genetically tied to an analogous concept in Hjelmslev (see L. El’mslev [Hjelmslev], Iazyk i rech’). 13. Violation of the norm while preserving the system creates the effect of an accent. The pronunciation [zh’ir] and [sh’es’t’] is typical of speakers of Turkic languages, and other peoples of the Soviet Union (Azeris, Chechens). 14. In making such a change we have left intact the phonological structure of the word, which relates to the system. For instance, the vowel of the pretonic syllable cannot be pronounced with a greater degree of reduction than the vowel of the prepretonic syllable.

JULY–AUGUST 2006

103

15. See Coseriu, Sinkhroniia, diakhroniia i istoriia, p. 176, n. 68. 16. Coseriu, Forma y sustancia, p. 74. 17. See, for example, N.S. Trubetzkoy, La phonologie actuelle; G. Laziczius, Die Scheidung langue—parole; N. van Wijk, La délimitation des domaines de la phonologie et de la phonétique, and so forth. 18. H. Ulaszyn, Laut, Phonema, Morphonema, p. 60. 19. See, for example, E. and K. Zwirner, Grundfragen der Phonometrie, and many other articles by E. Zwirner in various periodical publications. 20. Unless we include L. Hjelmslev, in whose work this delineation is narrowly defined. 21. Coseriu, Sinkhroniia, diakhroniia i istoriia, p. 335. 22. Ibid., p. 334. 23. His comment is interesting: “the concept of language, which can include the typical pronunciation of a word . . ., is broader than the concept that phonologists use to operate” (A.W. de Groot, “Structural Linguistics and Phonetic Law,” p. 190). After all, a phoneme as a “mark of a word” (de Groot’s term) is an identity and phonologists deal with oppositions. Thus, language-system opposes language-norms. 24. See A.A. Leont’ev [Leontiev], “Boduen de Kurtene i peterburgskaia shkola,” p. 122. 25. Zeichen und System der Sprache, vols. 1–2; “Tezisy dokladov na diskussii o probleme sistemnosti v iazyke”; A.A. Leont’ev, “Diskussiia o probleme sistemnosti v iazyke”; “Sens et usage du terme ‘structure.’” 26. Zeichen und System der Sprache, vol. 2I, p. 90. 27. Ibid., p. 46 28. Ibid., p. 110. 29. E. Polivanov, La perception des sons. 30. R.O. Shor, Kratkii ocherk istorii lingvisticheskikh uchenii, p. 112. 31. We will not go into the views of Herder and certain other forerunners of Humboldt here. 32. See G.V. Ramishvili, Nekotorye voprosy lingvisticheskoi teorii V. Gumbold’ta.

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.

Journal of Russian and East European Psychology, vol. 44, no. 4, July–August 2006, pp. 7–88. © 2006 M.E. Sharpe, Inc. All rights reserved. ISSN 1061–0405/2006 $9.50 + 0.00. DOI 10.2753RPO106104054404001

A.A. LEONTIEV

Psycholinguistic Units and Speech Generation
§ 22. The problem of “inner speech” appears to have been raised as a syntactic problem by L.S. Vygotsky. We will have to return to this problem to address it in detail, but for now it will suffice to state what is most important. Vygotsky drew a very clear distinction between inner and “external” speech, in terms of both their linguistic features and their psychological nature, attributing three specific features to inner speech. First, there is the abbreviation and “agglutination” of components of inner speech. Inner speech is elliptical speech par excellence, forming itself as a unique stochastic, linear connection between semantic “meanings” (smysly) (see below for more about this idea) that have yet to be cast in actual verbal form. This becomes possible due to the fact that inner speech is speech for oneself. Naturally, in order for these “meanings” to be able to function in linguistic thinking they must have a material reference, but this reference can be reduced to a minimum, for instance, to representation of only the initial letters of words. “Inner speech is, in this sense, speech almost without words.”1 Second, there is the predicative nature of the components of inner speech: it “from the psychological perspective, it consists entirely of predicates. . . . The law for inner speech is: always skip the subject.”2 Third, there is the specific semantic nature of the components of inner speech, which we have just mentioned, and a high level of situational and contextual dependence.
English translation © 2006 M.E. Sharpe, Inc., from the Russian text excerpted from A.A. Leont’ev, Psikholingvisticheskie edinitsy i porozhdenie rechevogo vyskazyvaniia, 2d ed. (Moscow: URSS, 2003), chapter 2, pp. 111–97. Published with the permission of Dmitry A. Leontiev. Translated by Nora Favorov. Notes and figures renumbered for this edition.—Ed.
7

8

JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

According to Vygotsky, the transition from “internal” to “external” speech is a “restructuring of speech, a transformation of absolutely distinctive and unique syntax, of the semantic and sound structure of inner speech into other structural forms characteristic of external speech . . . a transformation of predicative and idiomatic speech into syntactically articulated speech that is comprehensible to others.”3 This understanding of inner speech later drew extensive criticism. Indeed, it is not beyond reproach from the perspective of experimental validity, but, unfortunately, the criticisms voiced were even less valid. B.G. Anan’ev and L.I. Podol’skii4 criticize the idea of “predicativity,” believing rather that “nominalization” is typical for inner speech. Consider such sentences as: Night. Stars. But here critics are obviously confusing predicativity (or, rather, verbality) as a linguistic concept and, so to speak, psychological predicativity. Of course, Night. Stars. are nominative sentences, but psychologically they express something like “[It is] night,” “[There are] stars,” or “Night [has set in]” (not day), “Stars [are visible]” (and the moon is not), and so forth. The Kiev psychologist A.N. Raevskii, without evidence, decisively announced that “Inner speech is speech distinct from external speech not in terms of its nature, but only in terms of certain external structural features. Attempts to see it as speech with its own syntactic rules, different from ordinary speech, should be completely discarded.”5 It should be certainly clarified that the description of inner speech provided by Vygotsky relates specifically to inner speech, and not to the generation of “external” speech, that is, it has a somewhat indirect relationship to our subject. We introduced that description only because on the basis of Vygotsky’s ideas, his follower, A.R. Luria, developed a doctrine of an inner dynamic scheme of utterances that constituted, from his perspective, a specific stage of speech generation, the stage of “intention” (zamysel). This dynamic scheme breaks down in cases of so-called dynamic aphasia, as a result of which “overall thought, not taking the form of a known inner speech scheme, does not extend beyond the bounds of generally unformed intention.”6 Elsewhere, Luria speaks of “encoding a thought into a verbal utterance, which goes through the stage of inner speech.”7 Patients with dynamic aphasia
manifest extreme impairment in independent utterance: they note that they are able to generate individual words (the elements of utterance) out of order, but the scheme of the complete utterance (“the linear scheme of the phrase”) does not emerge. If, however, this linear scheme of the phrase is brought out into the open by placing guiding markers before the patient (for instance, buttons, pieces of paper) that correspond to the number of elements involved in the utterance, he will be able to produce the complete

JULY–AUGUST 2006 9

utterance, pointing to the markers in sequence with his finger (for instance, “I—want—to walk”); removing this external “linear scheme of the phrase” will again make this utterance impossible.8

It is significant that analysis of aphasic impairment leads to a need to attribute to this hypothetical link of speech generation the same characteristics that Vygotsky identified in inner speech.9 In any event, the cycle of works by A.R. Luria, L.S. Tsvetkova, and T.V. Ryabova [Akhutina] provides a solid basis for talking about the reality of the existence of an “inner program” or “inner scheme” component in the verbal act that holds one of the first places (in terms of order) in the overall mechanism of speech impairment, and that is essential to correct speech generation. The idea of such a link is completely alien to American psycholinguistics, and this clearly shows its longstanding “linguisticality,” about which we spoke earlier: allowing for the existence of such a link requires from the start a denial of the principle of direct reflection of linguistic elements on psychological elements, since there may be no direct equivalent of the components and principles of the organization of inner speech (or “inner scheme”) in a linguistic model of utterance. Therefore, it is very characteristic that, for instance, D. McNeill, in attempting to interpret Vygotsky’s concept of inner speech, reduces this concept to a certain rule of expansion of a grammatical “tree” that is characteristic of a small child.10 § 23. If we presume the existence of a special link in the mechanism of speech generation that corresponds to the inner utterance scheme, it creates a great temptation to project onto this link certain experimental data obtained in a different context. First, such a presumption permits the localization of the “order of expansion” that P. Gough writes about. If that is the case, we can presume that on this level there is an order of appearance of the components of an utterance that is independent of the actual sentences, with the subject of the utterance always appearing before the object. If this is true, then it can be presumed that the “reversibility” factor identified by Slobin is the result of an assessment of the subject–object relationship in the inner program link. On the basis of existing experimental findings, an attempt can be made to determine more exactly the order of elements within the utterance scheme. H. Clark’s findings are particularly interesting in this regard. In experiments on sentence production and on sentence association, he derived the following (identical) order of elements, if they are arranged by degree of interdependence probability (in accordance with the relative correlations between the probabilities of their appearance): modifier–subject of utterance–object of utterance–predicate.11 This order is seen in certain other cases. Experiments by W. Gutjahr

10 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

measured a value he called “Ergänzungspotenz” (the ability of a given class of words to support the production of an entire utterance—its mnemonic force, so to speak). It turns out that this value diminishes in the following sequence: S > O > P > Ad.12 Other authors obtained analogous results.13 In the previous book, we noted certain general patterns in the organization of utterance elements that are characteristic of communicative systems that lack a linguistically conditioned sequence of words or classes of words, or where such a sequence plays a subordinate role. For example, there is a syntactic model characteristic of sign lanugage. This model can be represented as S–(At)–O–(At)–V–Ad. The same model is characteristic of autonomous speech; of the initial stage of verbal speech among the deaf (Boy bird feeds); and of a certain stage in the development of child speech, specifically the stage at which the utterance elements have already been identified, but have not taken the shape of a morphemic paradigm.14 This deviates from Clark’s findings only in the placement of the attribute, but it is in this connection that Clark’s findings are not indicative, since the attribute in his test sentences always stood in the first place and served as a type of reference point in calculating the correlation. An experiment by Compton provided somewhat different results. He asked subjects to augment sentences of the type, The ball is rolling, either with an adjective modifier, an adverbial modifier, or through a negation. A single model emerged: first the verb was modified by an adverb, then the noun, and finally, negation was introduced. But during an analogous experiment using erasures, such consistent findings were not obtained.15 Unfortunately, this experiment was conducted on a very limited basis. Flores d’Arcais tried to establish a correlation between perception and the choice of the type of sentence syntax. It turned out that larger objects and objects on the left portion of a picture (that appear to be moving from left to right) have a tendency to be made into subjects: if objects appear to be moving from left to right, the usual construction will be: The car overtakes the wagon. While given the opposite movement it would be: The wagon is being overtaken by the car.16 § 24. In the works of D. Worth, and several other followers of N. Chomsky, we encounter attempts to expand the theoretical basis of the transformational model to allow it to cover all sentences intuitively characterized by one and the same semantic invariant, that is, not only the type, the professor examines the patient → the patient is examined by the professor, but also → the professor examining the patient.17 The “application model” of S.K. Shaumyan and P.A. Soboleva, who view their model as completely independent of the evolution of the transformational model, may be seen as a further step in this direction.

JULY–AUGUST 2006

11

Inasmuch as the application model has repeatedly, and quite thoroughly, been described in Russian-language publications,18 and its structure is exceptionally complex, we will not provide a detailed description here. We will merely point to certain specific features of this model that are interesting from a psycholinguistic perspective. First, unlike other models, it is a two-stage model in the sense that it consecutively generates objects of two types: language (genotypic) and speech (phenotypic). The first are called complexes: “a complex is a set of ordered elements, for which the order of registry is immaterial.”19 Information about the complex does not contain details about actual grammatical features and certainly not the sequence of grammatical elements in actual language; all of these details appear only in chains, that is, during the transition to the phenotypic stage. The primary task of the application model is the construction of correct complexes and correct transformational complexes. Second, transformations in the application model are not given (as, for example, in Chomsky), but are calculated. Such a calculation is performed on the basis of phrase complexes already provided (word combinations) using two procedures: deriving the set of classes and applying certain limitation rules to this set. Thus, from the “operand” R3OR2O (interpreted as a phrase such as high mountain), we first get two columns of classes with twenty-five potential combinations (these classes are generated by a “word generator” of the sort depicted in Figure 1). Then, applying certain limitations, we extract from among them eight transformations (the first stage): (1) vysokogornyi [high in the mountains]; (2) a transformation such as vysokii rostom [of tall height] that could not be used in this phrase, but would be possible in other, analogous phrases; (3) identical transformation; (4) vysota gory, gornaia vysota [height of a mountain, mountainous height]; (5) vysota byla goroiu [high as a mountain]; (6) gora byla vysokaia, vysitsia gora [the mountain was high, mountain towers]; (7) vysitsia goroiu [towers like a mountain]; (8) transformations such as khoroshii obed → khorosho poobedal [good supper → supped well]. Analogous operations can be carried out with more elaborate complexes, so the applicative model permits the most varied transformations within the bounds of the semantic invariant. The aforementioned example makes it evident that the application model has more “power” than the transformational model, allowing phrases and strings of phrases that are not amendable to a unified interpretation in the transformational model to be connected within a single system.20 However, it must be pointed out immediately that the path of speech generation of transformations in the application model is so elaborate that its psychological “reality” cannot be presumed, even for a minute.

12 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Figure 1. [Diagram of Word Generation]

white head reads whitens

head (dim.)

reading room

bleachery

There are other factors that make the prospect of psycholinguistic interpretation quite improbable when using the application model. First, there is the principle according to which this model is structured, going from more elementary to more complex objects. It is doubtful whether an analogous structure is applicable to objects within which psycholinguistics operates. Second, the application model operates at all levels with the transformational history of objects, and from the viewpoint of psycholinguistics, this appears improbable. Third, it introduces linearity only at the very end, while the principle of linearity (in the form of Markov dependency) clearly plays a more significant role in the process of speech generation. Fourth, this model is exclusively synchronous-descriptive, and in its present form it is clearly not applicable, for example, to the formation of child speech. At the same time, the application model has much to recommend. Perhaps it is best to view it as one of the possible LC variants, relating model interpretations with LP; in any event, the model requires psycholinguistic experimental verification, which has yet to be conducted. § 25. In this section we will briefly explore certain propositions developed within the framework of MP theory that are important for us, leaving aside the extensive literature in that area that is not relevant to this work. We will first explore the semantic synthesis model developed by I.A. Mel’chuk. This model is based on the idea of a “tree of meanings” constructed using a basic semantic language consisting of three kinds of elements: predicates, nouns, and adjuncts (an analogue of adjectives, adverbs, etc.). The successive stages of synthesis are as follows: “(1) semantic—the selection for the semantic tree of all syntactic constructions that are actually possible and the corresponding Russian lexicon; (2) syntactic—the determination of word order and syntactically conditioned morphological categories

JULY–AUGUST 2006

13

for each syntactic tree; (3) morphological—the structure and placement of actual word forms.”21 The tools used in synthesis are: (a) a Russian semantic dictionary (the rules of transition from “meanings” to lexemes or combinations of lexemes in the Russian language; (b) a Russian dictionary (a lexicon in which every word is given a morphological, syntactic, and lexical-semantic characterization); and (c) the rules for forming combinations of lexical units. The Mel’ chuk-Zholkovskii model is exceptionally appealing from a psycholinguistic standpoint and we will refer to it again. We will note that the Osgood model stands at the “wellspring” of this model. We will also mention a lesser-known article by Sestier that proposes the idea of identifying a number of individual segments within a sentence, the hierarchy among which is not indicated, the determining function of these segments in relation to the entire utterance.22 § 26. The viewpoint according to which the subject serves as a sort of passive receiver of information during the perception of speech—implying that the psychological mechanism permitting perception is not the same as the mechanism permitting speech generation—has been increasingly subject to doubt by psychologists and physiologists, as well as linguists. In this regard, there are three main orientations that can conditionally be called the “acoustic,” the “motor,” and the “analysis via synthesis” orientation. The traditional “acoustic” interpretation of speech perception is based on the flow of speech being perceived in segments, “step by step.” For separate communication segments to be identified, they must correspond to certain memory images. Once we have identified a sequential string of segments of a lower (more elementary) level, we can “copy” this sequence as a single segment onto a higher level. R. Jakobson rests his theory of binary differential features of phonemes on this interpretation. However, the fundamental premises of Jakobson’s model have not been borne out. First, it has been discovered that information about each given phoneme is not concentrated in one speech sound, but is spread over several. Second, it turns out that the transitions from sound to sound carry not less but, in many cases, more important information for recognition than the so-called “stationary sectors” do. Third and finally, it has been brought to light that “the physical characteristics of phonemes are not invariant and they change in relation to their phonetic position.”23 Contrasted with the “acoustic” theory, the “motor” or “articulatory” theory presumes that speech perception is possible only where sensory features elicit a certain reciprocal motor activity, as a result of which topological parameters of the perceived flow of speech are modeled within the listener’s speechauditory functional system, and only in this way are they identified.24 American

14 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

psychologists led by A. Liberman, belonging to the so-called Haskins group, are particularly avid proponents of “motor theory,” along with the Leningrad group of physiologists headed by L.A. Chistovich. Adherents of each of these two theories consider the other to be absolutely false, although both theories are supported by facts. A particularly heated polemic developed during the Eighteenth (Moscow) International Psychological Congress in 1966. At a symposium entitled “Models of Speech Perception,” R. Jakobson delivered an extremely aggressive talk, which was followed by a number of responses from adherents of the “motor” theory. Jakobson stated that the articulatory component of perception is optional; in speech perception, with the primary aspect belonging to the sensory plane; and speech perception is possible and logical without the participation of the motor link.25 This point of view is based, in particular, on certain data from Lenneberg relating to the anarthric child, who perceives speech without any measurable peripheral muscle activity.26 We will refrain from introducing the arguments of proponents of both theories, since these questions are only marginally relevant to our topic. We will limit ourselves to examining two aspects of the problem: first, the question of which of the two schools of thought more closely corresponds to the overall findings of psychology; and second, we will explore the extent to which both schools of thought are irreconcilable and whether or not some middle ground might be found. By and large, there is no doubt that “motor” theory corresponds much more closely to our current knowledge about the process of perception overall than to the “articulation” theory. A number of works (particularly in Soviet psychology) convincingly demonstrate the role of the motor component in touch and sight.27 A general theory of perception developed by the Soviet psychologist, V.P. Zinchenko, includes—as we mentioned in the first chapter [not translated here]—the idea of the body’s reciprocal action to a perceived object; his talk at the Eighteenth International Psychological Congress was appropriately entitled “Perception as Action.” From the perspective of speech perception, particularly relevant findings concern the human sense of pitch, since both abilities—hearing of speech and hearing of pitch—are uniquely human and genetically closely related. Zinchenko’s experimental study “demonstrated that the motor component of the process plays the decisive role in pitch perception.”28 Such a decisive conclusion does not, however, necessarily imply the participation of the motor component in every conceivable case of speech perception. The problem is that first of all, the arguments of adherents of the two competing theories do not take sufficient account—or actually take no account—of the physiological differences among various types of speech (such

JULY–AUGUST 2006

15

as spontaneous, or, on the other hand, imitative or “stochastic” speech, if we use the terminology of F. Kainz). At the same time, physiologically, speech behavior has to be quite multifaceted; suffice it to say that such common components of speech as vocatives correspond to first-signal speech mechanisms. Second, in perception, the fundamental possibility of reliance on an inadequate motor component, demonstrated in experiments by A.I. Ioshpe, and carried out under the direction of O.V. Ovchinnikova, have not been sufficiently taken into account. In these experiments the motor component of pitch perception was modified: instead of forming sound hearing based on the activity of the vocal cords, as would usually happen, for this purpose a device was used that related different pitches to the degree of force applied to a keyboard. It turned out that the development of the sense of pitch was not adversely affected by this. Third (and by no means least in importance), the perception of speech, in most cases, is not an elementary familiarization with its properties. When such a familiarization has been achieved, “it is possible to carry out an identifying (and reproductive) action. In this case, however, the act of identification relies on another system of reference points and features. . . . To the extent that the object has been identified, the observer distinguishes new features in it, groups them, and screens out a portion of the features that were distinguished initially.”29 He then combines the individual features into a sort of structure, into integrated images that become operative units of perception. If speech perception follows this course or a close approximation (and we have no reason to doubt that it does),30 then, evidently, it will turn out that both theories oversimplify this process. Finally, it should be kept in mind that one and the same process may be supported by both a constructive and a statistical mechanism.31 To the extent that it is voiced, the thinking surrounding all of these questions is often a priori in nature. Nonetheless, an experimentally based answer can eventually lead to an easing of the contradictions that have arisen. Of course, there is no reason to expect that this will undermine the fundamental correctness of the “articulatory” theory. We have yet to say anything about the third theory, the theory of “analysis by synthesis.” It is primarily represented in contemporary psycholinguistic literature by the works of M. Halle and K. Stevens. We will give an account of the model they describe in the article, “Speech Recognition: A Model and Program for Research.”32 This model, as seen in Figure 2, which is based on Figures 1 and 2 in the article cited, is also a model of speech generation, so the “analysis-by-synthesis” theory is, strictly speaking, a “theory for the ‘bearer’ of a language and not for the speaker and listener, taken individually.”33

Figure 2. [A Model of Speech Recognition]

Speech → signal Preliminary phonetic analysis Data from previous analysis

Spectrum analyzer

Storage mechanism

Data from previous analysis

Phonetic Control sequence Storage mechanism Preliminary phonetic analysis

Comparer

Generation rule

Comparer

Control

Phonological sequence

16 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Generation rule

Phonological sequence

Speech generation structure

Speech signal

JULY–AUGUST 2006

17

The substance of the “analysis-by-synthesis” model can be expressed as the idea that speech perception incorporates a number of blocks, each of which contains the rules of speech generation plus the rules for correlating the results of this speech generation with input signals. The figure shows phonetic and phonological blocks (or, using the terminology of transformational phonology, phonemic and morphophonemic blocks); however “the same operations take place . . . at the highest levels of analysis possible.”34 The principle on which the model works could be called “televisual”; it comes close to a hypothesis proposed by D. Worth.
The listener starts with the presumption of an input signal. On the basis of that presumption he generates an inner signal, which is compared to the perceived signal. The first attempt will likely be mistaken; if this is the case, then a correction is made and used as the basis for subsequent presumptions that may be closer. This cycle is repeated (almost certainly unconsciously) until the listener makes a choice that meets the corresponding requirements. . . . The output is not a transformed variation of the input; it is a program that must be followed in order to generate inner representations used for comparison.35

Words serve as the “standard” unit during input (in intelligent speech). It can be seen that the organizational principle of this model corresponds to the overall conception of speech behavior set down in a book by Miller, Pribram, and Galanter; in particular, elementary action is seen as the unit of speech behavior. In addition to the version of “analysis-by-synthesis” theory described here, there is a later version primarily belonging to N. Chomsky. While the HalleStevens model is built on a “bottom up” principle, with the transition to highlevel units taking place after the operations with lower units is completed, Chomsky starts with the proposition that “phonetic representation . . . can be generated with a very small number of simple transformational rules used cyclically, where the order according to which they are used (as we know based on ideas arrived at independently) is determined by the syntactic structure of the utterance.”36 In contrasting his conception to the understanding of perception as a process of sequential segmentation and classification, Chomsky believes that
the process of understanding the utterance presented can in part be reduced to the construction of an inner representation of its full structural nature. . . . The listener perceives a certain ideal formation that corresponds to the signal that was actually heard and that was generated by the phonological portion of the listener’s grammar on the basis of the syntactic structural nature that he attributes to the signal concerned.37

18 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Such a “top-to-bottom” model, as can be seen, does not require “motor” theory as an essential component, and many authors who uphold Chomsky’s point of view at the same time regard “motor theory” with a great deal of skepticism.38 The “analysis-by-synthesis” model in its second version was used in a number of experimental studies. We will cite two of them. The first belongs to P. Lieberman. Investigating subjects’ perception of sentences with different degrees of redundancy in a variety of situations, Lieberman writes that,
The results of this experiment regarding perception can be interpreted in light of a model of perception in which the listener hears an entire sentence and produces a grammatical analysis of its meaning at the same time that it is preserved in some form of short-term dynamic memory. The listener in this model can use the results of operation on the higher levels to remove the uncertainty on the lower levels.39

Another significant experimental work written by Miller was the article he produced in collaboration with Stephen Isard, “Some Perceptual Consequences of Linguistic Rules,” which is particularly interesting for its explanation of experimental data, as well as for its theoretical reasoning (chronologically, it precedes Lieberman’s article, which, in part, is based on it). The authors start with the supposition that the speaker or listener uses three types of rules in selecting (or recognizing) words: grammatical, semantic, and pragmatic. The study is “based on the assumption that in order to understand a sentence it is necessary to process the acoustic signal in accordance with these linguistic rules. Linguistic rules usually serve to limit the number of alternatives from among which the listener can choose.”40 The units that the listener uses to operate will be syntactically completed, and, in order to distinguish them, he must conduct a syntactic analysis, that is, apply grammatical rules. He also uses semantic rules (to further reduce the number of alternatives). Finally, pragmatic rules primarily relate to nonlinguistic information about the context (or, rather, the situation) in which the sentence is used. One experiment studied the repetition of three types of sentences: grammatical (semantically and syntactically correct), anomalous (syntactically correct, but semantically incorrect), and ungrammatical chains (incorrect in both ways). It was anticipated that it would be hardest of all to repeat the last of these and the easiest to repeat the first type of sentence. And that proved to be the case—the proportions of correctly repeated sentences were respectively 88.6 percent, 79.3 percent, and 56.1 percent. Their primary focus, however, dealt with perception under noisy conditions. The difference here was very clearly demonstrated: given no differ-

JULY–AUGUST 2006

19

ence in the level of noise, for example, 63 percent of grammatical sentences, but only 3 percent of ungrammatical chains, were repeated correctly. Particularly important was the comparison of experimental results where sentences of differing types were either presented together or separately: when sentences of differing types were presented together the subjects were forced to constantly change the system of rules used to interpret the sentences. This was reflected, in particular, in the fact that anomalous sentences were understood much better when they were presented separately—there was no interference from grammatical sentences. Miller and Isard’s experiments rather convincingly demonstrated, first of all, the active nature of speech perception, that is, the fundamental correctness of the “analysis-by-synthesis” model, and second, the independent nature of semantic and grammatical rules used in the perception of utterances. Unfortunately, as far as we know, the authors did not continue this experiment, and, in particular, there was no attempt made to analyze speech perception in dealing with a nonnative language.41 Summarizing what has been said above, it can be asserted with a degree of confidence that: (a) The “analysis-by-synthesis” theory is the most satisfactory from the theoretical and experimental perspective; (b) In principle, it can be used with “motor” theory, that is, the idea of the body’s reciprocal action is generally true; and (c) A number of less-studied factors may create an impression that the “motor” principle is optional. We would like to stress that the “analysis-by-synthesis” theory absolutely does not equate semantic and grammatical rules in the generation and perception of speech, even at a given stage. This is particularly relevant in terms of various heuristics used in one case or another. But the “mechanism of understanding does not fundamentally differ from the mechanism of utterance planning while it is being produced.”42 This is what makes it possible— to a limited degree, of course—to apply speech perception data in analyzing speech generation—something we took advantage of in this work. We will base our subsequent discussion on these tenets. § 27. The second problem—to which we will devote a special section in this chapter, in light of its particular importance for modeling a speaker’s language ability—is the question of “innate knowledge.” Chomsky first raised this question in 1959 in his sweeping review, published in the journal Language, of B.F. Skinner’s Verbal Behavior. In the review he sharply attacked the theory explaining the development of speech solely in terms of inner reinforcement. In particular, the review raised the question of the “remarkable capacity of the child to generalize, hypothesize,

20 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

and ‘process information’ in a variety of very special and apparently highly complex ways which we cannot yet describe or begin to understand, and which may be largely innate.”43 Chomsky calls for a particular kind of research, which has been undertaken. What is the nature of this innate ability? To put it in layman’s terms, what must it “be able to do?” “The child who learns a language has in some sense constructed the grammar for himself on the basis of his observation of sentences and non-sentences (i.e., using corrections by the verbal community).”44 This question is examined in more detail in Chomsky’s book (Aspects of the Theory of Syntax) and in the chapter, “Innate Ideas” (in J. Katz’s Philosophy of Language). The basic hypothesis on which proponents of the idea of “innate knowledge” base their thinking is that:
the device that makes language acquisition possible contains, as an innate structure, all the principles established within the theory of language. This means that such a device presupposes: (1) linguistic universals that determine the form of linguistic description; (2) the form of phonologic, syntactic, and semantic components of linguistic description; (3) the formal nature of rules in each of these components; (4) a number of universal phonologic, syntactic, and semantic constructs on the basis of which specific rules and specific descriptions are formulated; (5) a methodology for selecting the optimal linguistic descriptions.45

In essence, this list is redundant, since the first four points are covered by the concept of linguistic universals, if “linguistics” is understood in the generative-grammar sense. “The study of linguistic universals is the study of the properties of any generative grammar in natural language.”46 Universals are either formal or substantive; the first are essentially universal relations, and the second, to put it briefly, are universal elements. According to the “innate knowledge” approach, a child formulates hypotheses about the rules of linguistic description of the language to which the sentences they hear belong (“primary linguistic data”). On the basis of these hypotheses, he then predicts the linguistic structure of future sentences and compares these predictions to the sentences that arise, rejecting those hypotheses that prove unjustified, and develops those that prove satisfactory. For him to be able to do all this, he must have, at a minimum, a capability providing an invariant of a linguistic system (on the basis of which all possible specific systems are built), or an “innate predisposition of the child to learn a language of a particular type” (in the sense of human language overall,

JULY–AUGUST 2006

21

and not its typological variations), and the ability to compare a specific system with “primary linguistic data,” or a “strategy for selecting an acceptable grammar comparable to primary linguistic data.”47 In formulating a theory of acquisition (we are far from having exhausted here even the most important tenets of this theory)—Chomsky and Katz maintained, for the most part, an a priori reasoning—in the work of two other authors the idea of “innate knowledge” was provided a specific basis. We are referring to the work of McNeill, who traced the mechanism of children’s acquisition of English, and Lenneberg, who studied the biological (i.e., in our usage—physiological) preconditions for such acquisition. McNeill’s main work that we have access to is his article in the collection, The Genesis of Language. There, following Braine,48 he introduces a distinction between “pivot-class words” (P) and “open-class words” (O). The former are words like “more,” “big,” “bye-bye,” in the phrases “more milk,” “big boat,” “bye-bye Daddy,” while the latter are, correspondingly, the second parts of these utterances. In other words, P corresponds to various predicating words, while O corresponds to objects of this predication. In and of itself, this distinction is entirely sensible and logical; incidentally, it corresponds to the idea of the “vector” in utterance generation, which is discussed in the following chapter. But McNeill presumes that the distinction between P and O is innate—or at least the ability to distinguish them is innate. “The child classifies randomly perceived elements of adult speech according to universal categories exemplifying speech.”49 McNeill later asserts that “baseline grammatical relations are also part of innate language ability.”50 Among such baseline relations, he includes “subject-predicate,” “predicate-object,” “modifier in a nominal phrase-noun,” and some others. McNeill also places their interrelations (hierarchy) within innate language knowledge. Further components of McNeill’s model are tied to this initial assumption. In the same collection with McNeill’s main work is a long article by Lenneberg entitled, “The Natural History of Language.” This and other articles by this author (“The Capacity for Language Acquisition” and “The Biological Perspective on Language”) summarize separate parts of his major book published in 1967.51 Lenneberg’s main ideas can be summarized as follows. The child’s capacity to acquire a language is the consequence of an innate capacity (maturation) for the following reasons. First, the physiological substrate of this acquisition is related to the physiological substrate of other child abilities that are clearly innate (or at least abilities founded in innate features). For instance, there is a close interconnection between speech and motor coordination. Second, if other abilities degrade (as happens with certain forms of

22 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

mental retardation), then the speech function degrades as well. Third, speech cannot be developed through exercise: if a child is at the babbling age, nothing will force him to truly speak before he “matures.”52 There are periods of child physical development that are optimal for language acquisition, referred to by Lenneberg as “critical periods.” With the exception of this last thesis and several others (that do not, however, belong exclusively to Lenneberg), his approach invites criticism, as J. Bar-Hillel has noted.53 One can point, in particular, to the fact that his viewpoint rests on various deductive conclusions, the validity of which are yet to be proved; the entire manner of proofs is too artificial—using a comparison of walking and writing and deriving from this comparison four criteria that are then used in evaluating language.54 Actually, this reliance on deduction (a “rationalistic viewpoint” as opposed to an “empirical” one) is also quite characteristic of Chomsky’s followers. While Lenneberg is readily and extensively published in editions produced by the followers and adherents of Chomsky and Miller—and while he is frequently cited in the works of authors belonging to this school—by no means do his own views coincide with our views. Returning to the “innate knowledge” approach, we will make certain points on this subject. We will start by saying that this approach very clearly reflects Chomsky’s general thinking, which we discussed above. In other words, if a child has certain abilities that cannot be explained based on what we know about the genesis and mechanism of speech activity, this does not automatically imply recognition of this as an a priori ability. What is a priori in regard to child speech activity is not necessarily a prior in regard to his overall (i.e., innate) mental activity; it is entirely possible to presume that the given ability is associated with a special functional specialization of mechanisms that were formed in some other context. Before a definitive statement on this matter can be made, it would be necessary to trace the genesis and the early stages not only of the formation of speech ability but also other child abilities, to examine how the child’s relationship with the world as an integral system is formed. This is not being done by adherents of the theory of “innate knowledge.” One of the rather paradoxical reasons for this self-limitation is that Chomsky and many other adherents of the viewpoint are not consistent “constructivists” in the sense in which this word is often used in our [Soviet] psychology. For them, it is as if human behavior is made up of two components, which we will call biological and social. The social component is a type of superstructure above the biological component. The biological component, which is common to humans and animals, does not interest them; true—it is always emphasized that all human social behavior is subject to the

JULY–AUGUST 2006

23

same laws as speech activity (it is guided by “rules”), but this applies specifically to social behavior. Behaviorism is criticized primarily for the fact that its adherents apply simple models based on animals to more complex forms of human behavior (such, for instance, is the spirit of Chomsky’s review of Skinner’s book); they raise no questions about the possible existence of a fundamental difference in the organization of processes that are not, strictly speaking, social (such as perception) in man as compared with animals, and even less so about whether indisputably biological behavior, including animals, may also rely on the principle of “rules.” Such a differentiation— between the biological and the social—is very characteristic of psychological thinking in the contemporary West—with the exception, of course, of Marxist works and the French “sociological” school. In fact, the answer to a question often does not take an “either–or” form; in the case of higher forms of behavior we most often have “and–and.”55 As it applies to the question at hand, the correctness or falseness of the “innate knowledge” approach to a large degree (if not entirely) depends on whether or not neurophysiological levels of organization of human speech (and nonspeech) behavior are innate. Here, we cannot address this question in detail, but one thing is completely obvious: that the a priori views we expressed above about the possible “open nature” of the speech mechanism, are fully supported by the physiological data regarding the nomenclature and interrelations of neurophysiological levels; if it is possible to talk about universals in speech behavior, then these universals are to a certain degree universals of human behavior overall. From the perspective of contemporary Soviet psychology, the problem of “innateness” must, evidently, be decided on the basis of the concept of a “functional system.”56 Applying that approach, however, the very framing of the problem turns out to be incorrect and demands reexamination. Regardless, the fundamental tenets of the views held by Chomsky and his followers about the problem of “innate knowledge” are not acceptable. In future, we will not rely on the arguments of these authors and, where necessary, will attempt to limit ourselves to introducing only specific observational and experimental data. § 28. As can be seen from the content of this chapter, it is extraordinarily difficult to put together, on the basis of numerous experimental studies, that minimum of data that has actual heuristic significance for our problem, and is not essentially tied to an a priori recognition of some model of speech generation. We will, however, attempt to do this. 1. There is no doubt that in the generation of speech a statistical-probability principle is used, but to a rather limited degree: first, where we do not make decisions about grammatical structures; second, only as applied to motor

24 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

integral and semantic units, but not to grammatical units; third, in the process of speech there can be a sort of “probability correction” associated with the constant influence of context. 2. A distinction should be made in a model of speech generation between the problem of choosing one of many possible solutions and the problem of making a particular decision (the carrying out of a chosen version of behavior). 3. A model of speech generation undoubtedly includes in some form a block (or blocks) of long-term memory and a block (or blocks) of short-term memory. 4. A model of speech generation undoubtedly uses some form of a “constructive” principle such as the PSG principle, so that sentences that are more complex in terms of their construction require more complex operational activity. 5. The transformational complexity of a sentence is relevant to its psychological complexity, although it is not clear in what form. In other words, some constructions are psycholinguistically “ordinary” and some are psycholinguistically “complex,” requiring additional operations for speech generation or, alternatively, for interpretation. 6. The psycholinguistic structure of an utterance depends to a significant degree on “prelinguistic” factors. 7. Within this structure one can tentatively identify the link of the “inner scheme” or the utterance program. 8. The utterance processing model must, evidently (at least in its grammatical component), be the same for the generation and perception of speech). Chapter 3 Plans: Their structure and realization § 1. Naturally, before attempting a detailed analysis of the separate components of a model for the generation of speech utterances, or even before listing such components, we should outline the most basic knowledge about the fundamental structure of such a model and point out any features that are not specific to our interpretation from one perspective or another and that will be common to any model of speech generation. It is obvious, first of all, that any such model must have (a) an utterance motivational stage. Actually, this stage lies beyond the bounds of a psycholinguistic model as such, but if we do not introduce it, we will simply not be able to understand the causality of speech behavior and, to a certain extent, the features of its structure.

JULY–AUGUST 2006

25

The second stage that undoubtedly holds a place in any model of speech generation is (b) the stage of intention (program, Plan) of the utterance. The third stage is (c) the stage of carrying out the intention (realization of the plan). It is possible that this stage takes place—partially, of course— concurrently with the second stage (i.e., while one part of the utterance is being realized, another is being planned). The fourth stage, which may take place overlapping with the third is (d) the stage of comparing the realization of the intention with the intention itself. It is easy to see that these stages correspond to the fundamental structure of any intellectual act. We will review this structure. First, we must bear in mind the essential difference between two forms of human behavior. We are referring to reflex behavior and intellectual behavior. The difference between them is that in the case of reflex behavior there is a rigid connection between the irritant or stimulus and the reflex reaction. We do not extrapolate our “model of becoming” onto the future, but react immediately to it in accordance with the scheme of behavior that has taken shape in us through past experience, or has been endowed genetically (this relates to an unconditional reflex activity). In the case of intellectual behavior there is no such rigid connection: we construct several potential “models of the future” based on the situation as it stands and make a more or less conscious choice of one of the possible reactions to the situation-irritant. In other words, at the foundation of intellectual behavior lies man’s ability to plan his actions based on the mediated nature of his specific activity. Human intellectual behavior is comprised of separate intellectual acts, each of which generally consists of three phases. The first is orientation within the situation, within the conditions of the problem, and the choice of a plan of action. The second phase is the execution of the plan that has been selected. The third phase is the result attained with the intended objective. A pupil who has to solve a math problem first weighs which way to solve it, then attempts to solve it, and, finally, looks at the answer key and is satisfied that the answer is correct. A scholar formulating a new theory first studies the data available to him and proposes a certain hypothesis; he then carries out an experiment to prove the correctness of this hypothesis, and, finally, is satisfied with the correctness. A taxi driver heading for a particular destination in the city first thinks about what route would be the best, and then drives until he reaches the desired destination. Thus, the concept of an intellectual act is interconnected with the concept of activity.57 An intellectual act is usually an act of activity or at least forms a certain aggregate of actions that together comprise a more complex activity-act. Speech can enter into an intellectual act at different stages, during differ-

26 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ent phases. First, the planning of actions may involve speech, with the planned actions being either verbal or nonverbal. In these two cases, the nature of planning is completely different. In the first case, there is a programming of a speech utterance without a preliminary explicit formulation of a plan using language; the second is specifically the formulation of a plan of action in the form of speech. We are interested only in the first case. These two functions of speech in the planning of activity should not be confused, as they occasionally are.58 The fact that both types of planning are often called “inner speech” seems to play a significant role in this confusion. Second, the actions themselves can be verbal. At the same time, the interrelations between verbal and nonverbal actions within an intellectual act can be very different. Again, this difference can be twofold. This interrelation can change first, due to a change in the length of a speech utterance given the exact sameness of the remaining components of the activity; and second, due to the proportion of verbal actions in an activity overall, that is, as a result of a change in the structure of that act. Third, the comparison of the result achieved and the intended objective can be verbal. This occurs in cases where the activity-act is rather complex, usually when the intellectual act is entirely or almost entirely theoretical (as is often the case, for instance, in scholarly activities). The first function—specifically, the use of speech in planning nonverbal acts—and the second function are the most typical speech functions in activity. It should be pointed out here that the very term “speech activity” contains a certain inner contradiction. Speech activity as an independent, complete activity-act is extremely rare; it is usually just one component of a higher-order activity. A typical speech utterance is an utterance that in one way or another regulates the behavior of another person.59 But this means that the activity can be considered complete only if this regulation has been successfully achieved. For instance, I ask the person sitting next to me at the table to pass me a piece of bread. The act of speech activity (an intellectual act), if taken as a whole, is not complete: the goal will be achieved only if and when the person sitting next to me actually passes me the bread. Therefore, in talking about speech activity we are not being precise: often this term does not indicate an isolated activity-act, but a complex of speech acts, each of which has its own intermediate goal that is subordinate to the overall goal of the activity as such. However, this complex is also organized in a certain way. It is not a linear chain of acts sequentially executed on the basis of a certain a priori program or on heuristic information. And the organization of this complex (in this case, simply the speech act), like the organization of any act that is a component of an activity-act, shares certain essential features with the organization

JULY–AUGUST 2006

27

of the activity-act overall, insofar as we understand acts as “relatively independent processes subordinate to the conscious goal.”60 In any event, speech action presumes the setting of a goal, the planning, and the execution of the plan (or rather program) and finally, the comparison of the goal and the result, that is, it is a variety of an intellectual act. From the perspective of psychology, a variety of types of speech utterances can be identified. One of the most complete classifications of such types was proposed by F. Kainz, who differentiated the following: (a) Initiative speech or spontaneous speech. In this case, “On his own initiative and with his own choice of linguistic material, a person formulates a subject and semantic content that he himself has acquired and developed using the expressive capabilities of language”;61 (b) Reactive speech, which Kainz understands as an answer to a question posed by an interlocutor, where speech is not strictly speaking spontaneous and—although Kainz does not explicitly formulate this—in part goes beyond the question of the nature of the capabilities used. In both regards, reactive speech, depending on the type of answer, can come significantly close to initiative speech (if the question demands something more than a “yes” or “no” answer); (c) Imitative speech, arising most often in pathological cases and allowing for both an understanding of its meaning and—in extreme cases—a lack of such understanding; (d) Automatic speech, usually associated with a dreamlike state of consciousness (sleep, trance) or manifesting itself as a symptom of a specific mental illness; (e) Stochastic speech (Reihensprechen), realized during the reproduction of excerpts of texts that have been memorized. “Common phenomena that fall into this category are a series of whole numbers, the alphabet, the names of the days of the week, and the months, prayers, poetry, and other works that have been memorized, articles of the legal code (for lawyers), famous quotations, well-learned and often-played roles (for actors), and so on.”62 Even such a cursory description shows that the types of speech identified by Kainz have much in common and the distinctions he draws obviously do not rest on unambiguous criteria. Further analysis reveals several aspects of these criteria: (1) the utterance motivation (a–b–c, d, e); (2) a constructive or stochastic principle of organization (a, b–c, e) (in the case of d it is not clear); (3) the extent to which consciousness is involved (c1, d–c2, e). From American psychological work, we will introduce here the classification of types of speech behavior found in B.F. Skinner, who relies exclusively on motivational criteria or rather on the nature of “reinforcement.”63

28 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

According to Skinner, utterances can be divided into four main types: (a) controlled behavior (see below); (b) “mands”; (c) “tacts”; (d) autoclitics. The term “mand” comes from the words command, demand, and countermand. Into this category, Skinner puts utterances motivated, so to speak, “from within.” “Mands” can be of the following types: (a) a request, such as The bread, please; (b) a command involving an additional, nonverbal reinforcement: Hands up (implying, Or I’ll shoot); (c) a question, such as What is your name? So “mand” is the equivalent of spontaneous speech. Controlled behavior can be divided into: (a) echo-answers (imitative speech); (b) textual behavior (behavior that follows a written text); (c) transcription, in which Skinner generally includes translation of a communication from one form to another, for instance putting dictation into writing; (d) intraverbal behavior, basically analogous to Kainz’s “stochastic speech,” but also including association and translation. “Tact” (short for “contact”) is a sort of speech behavior where the motivation comes from “without,” that is, it is the equivalent of reactive speech. Autoclitics are “answers to answers that have already been given” like assertions, negations, qualifications, or quantifications of speech utterances. C. Fries bases his classification of forms of utterances on two criteria: the ability of the utterance to begin a conversation where the speaker pronounces it. As a result, he distinguishes three types of utterances: “situational” (capable of starting a conversation), “continuing” (not beginning, but belonging to the same speaker), and “responding” (not beginning, associated with a switch to another speaker).64 J. Carroll65 also bases his classification of speech utterances on two criteria. The first is a functional criterion; the second is the “sentential” or “nonsentential” nature of the utterance, that is, whether or not it takes the form of a sentence. In Carroll’s works, we have the following (examples taken by us from Russian). I. Nonsentential forms of expression. A. Greetings, and so on: Hi [privet]; Greetings [zdorovo]; See you soon [poka]; It’s been ages [Skol’ko let, skol’ko zim]. B. Appeals and other means of attracting attention: Hey! [Ei!] Ivan! [Ivan!] C. Nonsentential exclamations: Oh! [Okh!] What the hell! [Chert poberi!] D. Nonsentential responses: Yes [Da], No [Net], Hmm [Aga], Agreed [Idet], Okay [Ladno]. II. Types of sentences. A. Existence assertions, usually expressed in English through the construction There is . . . B. Predication, expressed through the construction “subject + predicate.”

JULY–AUGUST 2006

29

In Soviet psychological and linguistic literature, a similar classification of speech utterance forms, as far as we know, has been provided by two authors—one from the perspective of psychology and the other from a strictly linguistic perspective. We are referring to A.R. Luria and G.A. Zolotov. A.R. Luria classifies utterances on different levels. Primarily, he distinguishes four main forms of utterances. Affective speech is “exclamation, interjection or customary expressions: yea, yea; well, of course; oh, hell! These forms of speech do not generally originate from some idea, they do not form a thought, they mostly express certain inner affective states and express an attitude toward a situation.”66 Oral dialogic speech is the equivalent of Kainz’s “reactive speech” and Skinner’s “tact.”
The question of one conversant is the initial, beginning stage or the stimulus to speech; from it (and not from an inner idea) comes the answer of the second conversant. The function of formulating the thought is taken on by the conversant who is asking me a question. . . . A person who formulates a verbal response to this question always knows the overall subject being discussed.67

Oral monologic speech or a speech communication is different in that “the subject of communication originates not from the stimulus of the conversation partner, and not from the situation under which association is taking place, but from a person’s inner intention, which formulates this communication, from the thoughts of the subject, from the content that this subject wants to convey in an expanded utterance.”68 This is the equivalent of Kainz’s “initiative speech.” Finally, written monologic speech is a type of conversation without an interlocutor. It brings all features characteristic of oral monologic speech to their logical conclusion. Although Luria himself, like most other authors, does not explicitly formulate the criteria he uses, it is fairly easy to infer them from his classification. There are three: (a) motivation, (b) the constructive vs. the stochastic principle, and (c) the degree of consciousness, which distinguishes this last type from all previous ones. Furthermore, based on the work of G. Svedelius,69 Luria provides a classification of utterances for the “communication of events” and the “communication of relationships” (additionally, he compares utterances and “pointing to an object,” a variation on existence assertions). In the communication of events, the specific interaction of things is formulated and conveyed: the house is on fire, the girl is crying, the dog is barking. The communication of

30 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

relationships establishes the abstract-logical interrelations between a subject and a feature: Socrates is a man. In the case of “communications of relations,” “media of language play a role that is completely specific to language: they form certain relations, abstracting certain features and, with them as a basis, identifying abstract relations that cannot be expressed using visual, nonlinguistic means.”70 Such a distinction is by no means external, phenomenal; it is deeply rooted in the mechanism of linguistic thought.
Observation has shown that while “the communication of events” involves visual thoughts, merely expressed in verbal form, the “communication of relationships” demands the participation of operations using spatial relationships and the combining of correlated elements of the entire utterance in a single, simultaneous (quasi-spatial) structure. . . . The difference between these two constructions, which is clear for psychology, is supported by neurolinguistic research. . . . A characteristic feature of these patients [with damage to the parietooccipital region of the left hemisphere—A.L.] is the fact that logical-grammatical constructions that are different in terms of their inner structure do not experience the same fate. While “communication of events” as a rule remains intact, “communication of relationships” falls apart, and the patient loses the ability to use them.71

G.A. Zolotova followed a fundamentally different path, being unfamiliar with both Svedelius’s book and Luria’s work, and relied exclusively on Russian linguistic traditions. Without getting into all the ideas on this matter expressed in the history of Russian linguistics, we will limit ourselves to introducing the classification offered by V.A. Bogoroditskii. He distinguished between “sentences depicting a fact,” “sentences that determine the subject relative to its generic concept” (“this type of sentence is used when it is necessary to define what the given subject is [for instance, the horse is an animal]. In this case, the subject is a specific concept, the predicate indicates the generic concept to which the speaker relates the specific concept”),72 and he distinguished between “sentences that define the subject in terms of quality.” “Sentences of this type are used when it is necessary to indicate one or another quality of a subject, for instance, the wolf is gray.”73 Zolotova’s classification is somewhat more detailed.74 In particular, she identifies: sentences that communicate: a quantity of items; a subject and object of possession; a subject possessing a conditional feature; the state of a subject; a phenomenon and its evaluation, revealing the content of one abstract concept through another; the feasibility or the completion of an act; the correlation between an item and a place, an event and a time, phenomena and causes, and so on. Furthermore, Zolotova emphasizes that her classification

JULY–AUGUST 2006

31

is preliminary, and by no means covers the diversity of forms of utterances possible in the Russian language. Here we will bring our overview of the literature about classifications of speech utterances to an end, although it is by no means exhaustive, and we will try to explicitly formulate those criteria that (from our perspective) can and should be used in the psychological analysis of speech. These criteria can be of three main types. We can conditionally designate them as physiological, psychological, and linguistic. The reader, however, should keep in mind that such a naming is extremely conditional. We will include within the first group the criteria that are associated with the inner organization of speech mechanisms. The second group contains the criteria associated with the fundamental structure of an activity-act and the sociopsychological functions of speech (language) (“forms and types of speech interaction in connection with its specific conditions”).75 The third and final group includes the criteria associated with the particular features of linguistic realization of utterances (“forms of individual utterances . . . in close association with the interaction of which they are elements”).76 I. “Physiological” criteria (1) Orientation of the system of neurophysiological levels of speech behavior (what level is principal). This perspective gives us the following: (a) Communicative speech, (b) Nominative speech, (c) Echoing speech, (d) Choral speech. (2) The constructive, or stochastic, principle of speech generation, in particular the presence or absence of speech programming. This perspective gives us the following: (a) Active speech (“mand”), (b) Reactive speech. Its status in this regard is unclear (compare the viewpoints of Kainz, Luria, and Skinner), (c) Different types of “nonstandard speech”: their nomenclature is not clear. Compare Kainz’s “imitative,” “automatic,” and “stochastic” speech; Skinner’s “echo-answers,” “textual behavior,” “transcription,” and “intraverbal behavior.” This question can be left open, since in the present work we will be dealing exclusively with active and—to a lesser degree—reactive speech. (3) The degree to which consciousness takes part. From this perspective, at least the following can be identified: (a) Unconscious speech, (b) Controlled speech, (c) Conscious speech.77

32 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

II. “Psychological” criteria. (1) The place of the utterance within the structure of the intellectual act or the overall activity-act. From this perspective the following can be identified: (a) Planning speech, (b) Speech act, (c) Comparing speech. (2) Utterance motivation. From this perspective we can identify at least the following: (a) Spontaneous speech, (b) Situationally conditioned speech, (c) Contextually conditioned speech, (d) Unmotivated speech. (3) The functional orientation of the utterance. From this perspective we can identify quite a few types of utterances, of which we will note: (a) A command or request, (b) A question, (c) A greeting, (d) An exclamation (Luria’s “affective speech”), (e) Autoclitics, (f) A statement utterance. III. “Linguistic” criteria. (1) Sententiality, that is, the expansiveness or nonexpansiveness of an utterance in a sentence. Here, we have the following: (a) Nonsentential forms of utterance, (b) Sentential forms of utterance, (c) Supersentential forms of utterance, corresponding to the order of sentences, (d) Suprasentential forms of utterance, corresponding to a part of the sentence. (2) Logical-psychological types of utterances (“communication of events” or “communication of relationships,” classifications of Bogoroditsky and Zolotova). (3) The correlation of utterances with a speaker. Here, Fries provides us with the following: (a) Beginning utterances, (b) Continuing utterances, (c) Responding utterances. From this outline it can be seen that the types of utterances identified using various criteria often overlap (for instance, compare criteria I.2 and II.2; I.1 and I.3, etc.). This is to be expected, since it seems possible that such

JULY–AUGUST 2006

33

forms of communication could exist and be socially determined, differing from other forms in only one single feature, and not always an essential feature at that; and it is no accident that the most widespread and most typical types of speech communication have the greatest number of features in this outline. For instance, ordinary spontaneous speech is simultaneously defined (and is distinguished from any other type of speech) by features I.1.a; I.2.a; 1.3.b; II.1.b; II.2.a or b; III.3.a, and so on. Departing somewhat from the main topic of this section, we will point out that there are other criteria that can be used to classify speech utterances. Along with forms of speech, it is also possible to distinguish types of speech or types of speech activity by whether or not they can both exist in the same speaker (sound speech, mimic speech, etc.); by language function and speech function, attributing them to one or another utterance; proper speech and improper speech forms of utterance realization (i.e., the presence of functional equivalents of speech on different levels and the level at which this equivalence is established), and so forth.78 Not so long ago, A.A. Kholodovich79 provided an entire system of features that he used to characterize speech typologically. He distinguishes: (1) the means of expression (sound, writing, gesture); (2) the presence or absence of a conversation partner; (3) the directionality (uni- and bidirectionality) of the speech act; (4) the presence of one of many perceivers (individual or mass communication);80 (5) the close-contact or distance of the speech act. Other similar systems, which we will not now discuss, are also possible. We will now return to our main task and will set the boundaries—relying on a specific classification—of those types of speech that we will be dealing with. They will be: I.1.a—communicative speech, I.2.a and b—active and reactive speech, I.3.b and c—controlled and conscious speech, II.1.b—speech act, II.2.a, b, and c—spontaneous, and situationally and contextually conditioned speech, II.3.f—utterances that are primarily statements, III.1.b—sentential forms of utterances, II.2—this feature is not relevant for us now, III.3—this feature is not relevant for us now. While the types of speech that we will be further analyzing have had at least some psychological and psycholinguistic discussion, making it possible to use existing data, the types of speech that are not included here have not, as a rule, been the subject of serious scientific research, and,

34 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

therefore, their study has to start from the very beginning. It is natural that this—along with the limited scope of this book—has led to the necessity of completely abstaining from examining them, although making use of even those materials relating to the rather narrow area of our interest would be extremely worthwhile. § 2. We have already written quite a bit about the fundamental structure of activity in our other works.81 For this reason we will address this question only briefly, for purely heuristic purposes, so that the reader will not be forced to search. The specific nature of human activity is characterized by two main features. The first of them is goal-orientation, that is, the presence from the very start of certain goals, the attainment of which marks the cessation of an activity-act. Activity is organized (consciously or unconsciously) in just this way, so that this goal can be achieved with optimal means and the minimal expenditure of time and energy. In addition to a goal, an activity-act is characterized by a particular motive; one and the same activity may be carried out as the result of a variety of motives. One might save a drowning person under the influence of an inner moral imperative, but one can also throw oneself into the water to save him in just the same way from a vanity-driven desire to be a hero in the eyes of onlookers or to impress a specific person. The second feature is the structure of activity. It is made up of a sequence of acts, of the components of activity, which are characterized by an independent interim goal. These acts can be of two kinds: external, practical and inner, cognitive. Between the two (and, correspondingly, between practical and theoretical activity overall), there is no fundamental boundary: almost any activity-act incorporates both external and inner acts. Depending on the specific circumstances under which the act is carried out, it may comprise individual operations of either type. We will introduce an example of an extremely simple activity-act: a professor, interrupting his lecture, walks up to the window and closes it. There could be a variety of motives for this: there was a cold draft; or the professor noticed that his audience was getting cold; or there was a loud noise outside, making it difficult to be heard; and so on. The goal: close the window. The action: to descend from the podium, walk up to the window, and close the window. The operations associated with these actions are determined by the height of the podium, the distance to the window, the construction of the window frame, and so forth. If we attempt to find in this system a place for speech activity, or rather for the aggregate of speech acts,82 then, remembering that in the preceding paragraph we narrowed the subject of our discussion, it could be said that speech

JULY–AUGUST 2006

35

acts or a speech act (to simplify our discussion, from now on we will assume that a speech utterance does not extend beyond the bounds of a single speech act) constitutes a particular instance of an act within an activity-act. In other words, a speech act: (a) is characterized by its own goal or task (intermediate in relation to the activity overall and subordinate to the activity goal); (b) is generally determined by the structure of the activity overall and in particular by those acts that preceded it within the activity-act (in the same way that the act of closing a window requires a series of preceding acts and cannot be carried out or even programmed until this series has been completed); and (c) has a specific inner structure that is conditioned by the interaction of its features that are tied to the structure of the activity-act and are common to many activity-acts of the same type and to the specific conditions or circumstances under which this act is being carried out in this particular instance and at this particular time. Let us attempt to address all of these aspects of the speech act in greater detail. We generate a variety of speech acts, that is, we speak. How—if we address the question in all its possible aspects—is it determined what speech means we use in a given instance? To make it easier for the reader to follow the course of the discussion, we will take a hypothetical situation, one already introduced above: someone asks the person sitting next to him to pass the bread—and we will turn to this example from time to time. 1. The first factor influencing the speech act is the overall predominant motivation or motive of activity.83 In this case it will be a feeling of hunger, a need for food. So the range of speech utterances possible is immediately reduced. We now have only such speech acts that can lead to the eventual result, to the satisfaction of a need—to feeling full. 2. The second factor is what P.K. Anokhin refers to as situational afferentation: “What we mean by situational afferentation is the aggregate of all those external effects on the organism from a particular circumstance that, together with the initial motivation, most fully informs the being of the choice of action that best corresponds to the motivation at that moment. The physiological role and behavioral implication of the situational afferentation derives from the fact that it—primarily due to the relative constancy of action that characterizes it—creates in the central nervous system a rather widespread, integrated system of stimuli, a sort of neuronal model of the situation.”84 In other words, it is the limitations on the choice of action (and speech action, in particular) that are imposed by the situation at the start of the action. And this situation, in turn, is composed of two factors, which we will call A and B.

36 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Factor A is the element within the situation that does not depend on the given activity or on the actions that came before our action, but only passively participates in the choice of method for carrying out the speech act. Of course this “independence,” “absoluteness” is essentially relative; the very fact that I wound up sitting at the dinner table is the result of my prior purposeful activity. However, in addition to this component, Factor A also encompasses a component not dependent on my previous activity, for instance, the fact that the bread basket turned out to be out of my reach. It is essential to underscore that “situational afferentation” encompasses only what in the situation influences the choice of action and not what influences the variability in the way that the chosen action is carried out. Factor B is the element within the situation that is tied to previous actions within the framework of the activity-act, which is created by these actions. During the creation of the “neuronal model of the situation” (“a model of the past and of the present,” in the words of N.A. Bernstein), Factors A and B are equivalent and undifferentiated; the difference between them is purely genetic. 3. Thus, a “model of the past and of the present” has been created in us, in accordance with which we must carry out our speech action. But it is obvious that the choice of actions possible in the given situation—even taking into account the predominant motivation—is still exceptionally large. And the next factor to influence the choice of action is what Miller, Pribram, and Galanter call “the image of the result,” and what Bernstein calls “the model of the future.” “Modeling of the future” is possible “only through extrapolation of what the brain takes from information in the current situation, from ‘fresh traces’ of immediately preceding perceptions, from all of an individual’s preceding experience, and, finally, from those active trials and probings that relate to the class of actions that are still very summarily designated as ‘orientation responses. . . .’85 Unlike the model of the ‘past and present,’ the model of the future is probabilistic in nature. During any phase of extrapolation, ‘the brain is capable only of outlining for the impending moment a sort of table of probabilities of possible outcomes.’”86 In contemporary Soviet physiology, an organism’s appraisal of probabilistic experience accumulated in the past and guiding assessments of probabilities in a “model of the future” is referred to as “probability prognosis” of activity. What is the mechanism of such a probability prognosis?
The emergence of Situation A is a signal for the preparation of the organism’s response system appropriate to a Situation B, which is conditionally most probable following A. . . . The larger the range of events that have in the

JULY–AUGUST 2006

37

past followed A with equal frequency (i.e., the less clear the prognosis is), the broader the range of physiological systems that are mobilized in response to the A signal. Such a preadjustment to actions in the impending situation based on the probabilistic structure of past experience can be called “probability prognosis.”87

From the above it is clear that the probabilistic structure of past experience in such an understanding is associated not with the frequency of the appearance of particular stimuli in the past, but with the frequency of particular responses to them by the living being. This understanding correlates with data on the statistical organization of perception overall, and the perception of speech in particular, explained in part in §3 of chapter 2 of this book. I.M. Feigenberg has proposed the hypothesis that certain mental disorders seen in patients with schizophrenia are rooted in a breakdown in the mechanism of probability prognosis. The schizophrenic is unable to correctly assess the conditional probability of a particular outcome; his choice is random, and when it takes place, the phenomenon of “split personality” may occur, since the patient perceives his behavior as inadequate. To jump ahead, we could also point out that a breakdown in probability prognosis on the level of the fundamental organization of speech acts is associated with a breakdown of probabilistic structure on other levels as well. In the words of B.V. Zeigarnik, it can be observed that among schizophrenic patients, there is “a rich and multidimensional nature of associations,” the phenomenon of “disjointed speech,” constructed and based on random associative coupling of speech units. I.M. Feigenberg notes the
frequent use of words that are rarely encountered in the speech of healthy people. . . . While in a given situation a healthy person will choose from a large selection of possible associations or words that are frequently used by others (or that have led to success) under similar circumstances in the past; the mentally ill do not seem to take this frequency into account (there is a leveling of the probability of selection). From this viewpoint, the schizophrenic defect can be seen as a probabilistic disorganization of the use of information (past experience) stored in the brain, as an increased entropy in the brain as an informational system.88

A similar picture emerges among patients with damage to the brain’s frontal lobe, a condition referred to as “frontal lobe syndrome.” Using a method for probabilistic analysis of perception proposed by E.N. Sokolov, Luria arrived at the conclusion that “among this group of patients, the process of collating incoming information with an image or models of possible representation breaks down, as a result of which their cognitive activity loses the selectivity seen in a normal subject.”89 An analogous phenomenon occurs during the

38 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

reading of a text—“the impossibility of keeping oneself within the confines of the systems for selecting connections that are provided in the text and the emergence of incidental connections, and the inability to inhibit them.”90 Especially characteristic are intellectual activity disorders such as the solving of math problems: “In them, as a rule, the phase of preliminary orientation within the conditions of the problem is completely, or, to a significantly degree, absent, and the overall scheme (plan) needed to solve them does not emerge. This strategy-lacking process for solving problems turns into combinations of separate numbers, none of which usually have any relation to the ultimate goal.”91 However, as far as can be determined, this is not so much a matter of the breakdown of the probability prognosis per se, as a breakdown in the fundamental organization of the act, of its strategy. Up to this point we have been talking about the “probabilistic experience” factor, based on the premise that our subject has in the past encountered a situation analogous to the present one sufficiently frequent to be able to acquire this experience. However, one can imagine another possibility, one where the choice of action is not based on experience so much as on an ongoing assessment of the significance of a particular action for behavior. Here we run into a range of questions that are associated with mathematical problems about reaching an objective in a changing environment, questions that have been developed in recent years by I.M. Gel’fand, V.S. Gurfinkel, and the recently deceased M.L. Tsetlin. We are not here to look in depth at the results they achieved; their application to speech activity demands its own analytical monograph.92 We will merely note that the situations in which it is necessary to turn to an ongoing assessment of implications—“payoff functions”—in no way amount to an “extra-experiential” original search; the same mechanism must “come into play” in several other circumstances as well, for instance, when there is difficulty in carrying out an action, and during the search for an optimal solution to a situation that is taking shape. As has been noted more than once in the literature, these problems are associated with mathematical game theory. This theory, like the mathematical theory of optimal processes overall, of which it is essentially a part, is being called upon to lay the groundwork for a mathematical apparatus that will in the future (one can hope, the not too distant future) permit a quantitative analysis of human activity overall, and speech activity in particular. At the same time, many areas currently at the center of “mathematical linguistics,” for instance, speech statistics, will naturally be moved to the periphery. So, the current situation could have several possible outcomes, however only one outcome can be realized, that is, a person must bring the probability of one outcome to 100 percent, nullifying the probability of all other out-

JULY–AUGUST 2006

39

comes. Using this logic, so to speak, does he select this single realizable outcome? According to Bernstein, this is “an activity guided by juxtaposing a probability model of the future and the set objective.” In other words, a person, evaluating the probability of outcomes, at the same time correlates them with what can conditionally be called the action objective, that is, the closest expected result of this action. It may turn out that the most probable outcome—from what we already know about the situation, and based on our previous actions—is the version of behavior, that does not correspond to the objective of the action. Let us say that the person sitting between us and the bread basket has become involved in a conversation with someone else. We can predict that he will not hear our request and will not react to it. We can then, instead of addressing him, call on a third party also sitting near the bread. In so doing we have selected some other outcome (not the most probable one given the situation at hand) of this situation and we bring its probability to 100 percent by our active and conscious intervention in our behavior and its fundamental restructuring. But we could also take another path. We could reject this action objective or fundamentally restructure it. For instance, it would be possible, without any speech act, to stand and reach for a piece of bread ourselves. In this regard, the phenomena of so-called posthypnotic suggestion behavior is very interesting. During hypnosis, we can suggest to a patient that when he awakens he should take a particular course of action, for instance, he should go to the next room and tear a newspaper in half. In other words, we impose an action objective on him. The most interesting thing here is that the person begins to unconsciously construct a complete logical basis for this imposed objective; he constructs it in such a way that the imposed outcome becomes the most likely. Here we have, so to speak, a stark case of a person reconstructing his behavior in the most fundamental way with the sole objective—unconscious, of course—of making a specific, already-assigned outcome of a given situation the most likely. But how is an “action objective” determined? It is tied to our conception of the structure and objective of a complex activity-act. In other words, it is a limitation imposed on the choice of action by the activity-act as a whole. It should not be forgotten that in carrying out an activity, in addition to a conception of its goal, people always have a certain program (however vague it might be) for the attainment of that goal. The very emergence of a speech component in activity is the result of such complex programming. And it is this programming that serves as a factor determining the objective of individual actions (it is natural that along with “preprogramming” of the activityact as a whole, one can have a correction of its structure over the course of its realization that also influences this objective).

40 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

What form does the action objective take for the person performing the action? In outlining the basis of set theory, I.T. Bzhalava used the fortuitous term “imagined situation,”93 which is ideally suited to our case. The action objective is presented to the acting person in the form of a model of subsequent action, or rather, certain parameters of it. When I am preparing to say “Please pass the bread,” I am not thinking of bread as much as I am thinking about the future action of my dinner companion, and only as a result of this action, getting my bread. When I am looking for chalk, I am not thinking about chalk, but about the future process of writing on the board with this chalk, and so forth. But it is easy to imagine a case where we would not have a clear imagined model of subsequent action, for example, when we do not know the potential reaction of our companion to what we say. Under these conditions, the very nature of the action changes most fundamentally. First, instead of a single outcome we wind up with several equally possible outcomes, each of which requires a newly programmed action; but we can program only one version of action at a time. For this reason, there may be something akin to “feeling one’s way” toward the correct path while a person gets his bearings, where “trial balloons” are sent up to clarify the reaction of the person he is dealing with. As soon as this reaction becomes clear to us, that is, as soon as the choice of a single outcome has been made, our vacillation comes to an end and we program our speech action without ambiguity. But “feeling one’s way” is possible only when there is time for it: “if events unfold quickly and an individual finds that time is running out, he is forced to limit himself to the first, crudely assessed evaluations.”94 Then, typically, there is a restructuring of the speech act “off the top of our head,” sometimes with a simple rejection of one of the versions that has already been put into play and the choice and realization of a different version from the very start. Second, the absence of a clear model of subsequent action triggers a process of awareness of action. We will not address that here. But let us return to our action. For the time being we have a “model of the past and of the present” and the action objective. The individual carrying out the act correlates the possible outcomes of a “model of the past and the present” with the future and the possible actions, and as a result he chooses one of the possible actions based on its content. In other words, he has a certain “speech intention,” he knows what he will have to say and what effect that, evidently, will have. What remains is a question that is not of interest to cybernetics, but that is of fundamental important to us: how will he say it? However, that is a question outside the bounds of a purely psychological examination. Let us recall where in activity the boundary lies between the planned and

JULY–AUGUST 2006

41

the unplanned, between the components of the activity-act that in advance, consciously or unconsciously, are separated out by the individual during the planning of this act, and its components that depend on specific conditions of activity and take shape as a result of an automatic “harmonizing” of the activity with these conditions. Such a boundary lies between action and operation. If we turn now to the speech act, we can see that programmed and nonprogrammed elements are rather clearly distinguished within it. This distinction is easiest to see if we examine several languages, instead of just one, and compare how the speech act is realized in them. Since this act is a psychological concept and its status is wholly dependent on extralinguistic factors, we must recognize that when we have people expressing the same content but speaking different languages, we are dealing with one speech act. In different languages, however, this act is carried out on the basis of different operational structures, using a different nomenclature of operations. As far as can be judged based on the extremely meager findings of psychological and psycholinguistic research of speech up to this point, the variation of operations is tied not only to the differences between languages. Another essential factor is the specific circumstance of the act (to the extent that it does not enter into the model of the past and the present). Other factors include the speech context of the act; individual differences in speech experience, which is particularly evident in the specific way the motor program is realized (see below); the sociolinguistic or functional-stylistic factor, which determines the choice of a particular linguistic tool from among a number of potentially possible ones based on the nature of the relationship of the conversants and on certain other features of communication; the affective factor, and so on. We will address many of these factors below. It is therefore helpful to talk about the speech act only to the extent to which we are dealing with the particular features of activity that are irrelevant to the conditions of the act and that are determined solely by the structure of the activity-act. When we turn to the specific features of activity that are determined by the conditions of the act and that are irrelevant to the structure of activity (leaving aside, of course, the fact that the very boundaries of variability in operations are conditioned specifically by the structure of activity) we will address speech operations. In that case, what are the actual parameters of the speech act, of what does it consist? Evidently, its structure reflects, first of all, certain features common to any act or activity overall, in particular, they comprise programming, the realization of programming, and a comparison of one with the other.95 Second, we have those features of activity that are conditioned by a “model of the past and the present.” Such a feature could be, for instance, the pres-

42 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ence or absence of the person or object that is being discussed within the view of the conversants, his/its place in the structure of the preceding action, and so on. Third, we have those features of activity that are connected with the place of the given action within the structure of activity overall; we can thus build a speech act presuming one or another reaction by our conversation partner, focusing in on or skipping over certain elements of this reaction. All of these features of the speech act as such can be concentrated only in those of its links that, while directing its specific realization, at the same time do not depend on its realization. The program of the speech act or of the speech utterance is such a link. From the above it is obvious, by the way, that from the perspective of the theory of speech activity, the problem of “functional grammar” or grammar “from content to expression” is the problem of the transition from the level of speech acts to the level of speech operations, the problem of the realization of the speech act program. For this reason, the following paragraphs are most directly relevant to the psychological underpinnings of the methodology of foreign language instruction.96 The hypothesis stated above, based on general psychology and indirect experimental data, is very promising. But it remains a hypothesis. For it to become fact, it must be confirmed experimentally. § 3. In the preceding section, we explored the way a subject chooses one of the possible speech act programs. Before turning to an analysis of the structure of such a program, let us examine the very possibility of such a link in speech. In contemporary science there have been numerous studies and even whole schools of science that to some degree are based on the idea of a certain organization of utterances that precedes speech, language organization itself. Let us begin with the fact that most works associated with this subject, usually called “logic and grammar” or “judgment and sentence” are usually of this sort. Such works most often belong to the sphere of logic, but sometimes—when their authors are linguists—they pass entirely into the realm of linguistics. Typically, the researcher’s thinking is as follows. There is thought, there is language (or speech). They go “hand in hand.” If in language (speech) we distinguish a unit such as the sentence, for instance, then there must be an analogous unit in thought. Upon close examination, it turns out that the judgment found in Aristotelian logic does not fit the role of such a unit, as it is too narrow and does not encompass all types of utterances. This is “the problem of judgment and sentence,” for which the majority of authors find a simple solution—the creation of a broader concept in which the concept of

JULY–AUGUST 2006

43

“judgment” constitutes a special case (a “proposition,”* “logical phrase” or “logema” in P.V. Chesnokov).97 The boundaries of this broader understanding are set in such a way that it covers various types of sentences. For linguists who adhere to this thinking, “between language and logical operations there is no room for ‘psychic reality.’”98 From here it is already clear that this approach is incapable of satisfying us. It is completely abstracted from true regularities of linguistic thinking; A.A. Potebnia was completely correct when he wrote eighty years ago that “the reasoning of the logician does not examine the process of utterance, but with its one-sided point of view, it evaluates the result of the completed process.”99 Furthermore he bases his thinking on an a priori recognition of the structural parallelism of language and thought, which is clearly not justified and in the final analysis is untrue. A.A. Shakhmatov already understood this, and before him, the same A.A. Potebnia, who very reasonably asserted that “the grammatical sentence is in no way equivalent or parallel to logical judgment. . . . For logic, what is essential in judgment is only the compatibility or incompatibility of two concepts.”100 Recently, G.P. Shchedrovitskii provided a serious analysis of the falsehood of the idea of such a structural parallelism.101 It goes without saying that the idea of language and thought “going hand in hand” is baseless. For some reason it is believed to be an essential component of the Marxist understanding of consciousness and thought, and that in expressing it, we are defining our philosophical position. This is not the case. First, it is not only Marxists who will eagerly endorse this thesis, but scholars far from Marxism—the Neohumboldtian of the Weisgerber sort, the English “linguistic philosopher,” the American “general semanticist.” Second, from the Marxist perspective, it is not enough simply to talk about the connection between language and thought, at least about their going “hand in hand” or about a “direct” connection: Marxist psychology and, following in its footsteps, gnoseology point to the linguistic nature specifically of human thought and to the fact that human thought is mediated by socially developed sign systems, including language. When Marx says that “language is practical real consciousness,”102 this is by no means just a simple metaphor (and it is even less just to implicitly substitute only thought in the place of consciousness in Marx’s formula, as is often done, relying on his other famous statement). 103 It must not be forgotten that Marx endowed a quite specific philosophical content in the terms “practical,” and “real.” For him, conscious-

*In Russian, the word predlozhenie can mean either sentence or proposition. Here the author uses the Russian cognate for “proposition,” not the word predlozhenie used up to this point in the paragraph.—Trans.

44 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ness is a virtual correlate of real language; it becomes real in language, takes on its being in language, its “body.” Finally, the very word “language” for Marx is by no means terminological. It does not stand in opposition to speech or speech activity and covers all these concepts; if we allow ourselves to interpret Marx’s cited thought in the terms that are now used, then it would have to be said that consciousness is realized in speech activity or that consciousness is realized with the help of language. But allow us to return to our main topic. We have been convinced that a purely logical approach to the problem of the “prelinguistic stage” of speech for us would be absolutely fruitless. But there are other approaches to this problem, one of which is the logical-psychological approach. We already touched on this aspect of the question to some degree when we spoke in the first section of this chapter about the concept of “communication” in connection with the works of Svedelius and Luria. Here we will examine the views on this question of only one researcher, Shakhmatov. For him, “communication” forms the psychological basis of the sentence, that is, it is the act of thinking, as it is for many other authors, with G. Paul at their forefront. But unlike these other authors, Shakhmatov believes that while “the beginning of communication lies beyond the bounds of inner speech, its completion is reached in the process of inner speech.”104 Thus, for him, communication is not an external category as far as speech is concerned, it is not a logical-psychological category; it is part of speech thought as a process, it enters the model of speech generation as one of its levels. According to Shakhmatov, any communication is binary/dual: “The sentence ‘The crow that was frightened by us flew up onto the tall linden tree,’ is communication, the subject of which is ‘the crow that was frightened by us’ and the predicate of which is ‘flew up onto the tall linden tree.’”105 Shakhmatov’s idea about binary/ dual nature of speech as a characteristic of speech generation at one of its earlier stages is supported by several facts. In experiments by the Leningrad psychologist V.V. Oppel, first-graders who were asked to break an utterance into “words” (without being told in advance just what a word is) divided it primarily into the subject and predicate of the communication: Theapples— areinthebowl; Onthestove—isthekettle; Thedog—bristledandbarked. However, in those cases where the image suggested by the subject was unchanged by predication, then the subject and predicate were seen as one “word”: Itsraining; Thesunshines.106 Analogous data can be obtained through the analysis of early childhood speech, from the study of aphasia,107 and so on. In essence, the theory of “actual sentence articulation” reflects the development of similar ideas.108 According to this theory, it is possible to approach an analysis of a speech utterance from at least two angles: from the perspec-

JULY–AUGUST 2006

45

tive of its formal structure and from the perspective of how the given utterance conveys new information, and which of its parts conveys facts already known to us, and which part conveys new facts and information. Unfortunately, the development of this theory up to now has proceeded almost exclusively within the framework of pure linguistics and for this reason alone may not be particularly productive; we know of only one work that attempts to apply the theory of “actual articulation” to the psycholinguistic aspect of speech generation. On the basis of experiments performed, the author of this work, K. Pala, reaches the following idea regarding the fundamental structure of speech generation:
First, the speaker has a structure of representations, that is, a semantic structure of a given communication that at that moment has nothing to do with the specific syntactical realization of the given communication. But when the speaker begins to generate this communication, he begins to use the syntactic realization of the communication’s semantic structure, and during this process he can select different syntactic realizations for one semantic content of the given communication.109

Elsewhere in this book [not translated here], we described the views of certain American psycholinguists regarding the “prelinguistic” stage in speech generation, as well as L.S. Vygotsky’s conception of inner speech, developed and substantiated by A.R. Luria through studies of aphasic disorders. We will not repeat that description here, concentrating our attention instead on individual questions associated with the idea of “the inner program” or “the inner scheme” of the speech act. First, we will try to clearly contrast the often-confused110 concepts of “inner speech,” “inner talking,” and “inner programming of a speech utterance.” Inner talking (“external speech to oneself,” as P.Ia. Galperin fortuitously defines it) is the hidden physiological activity of the organs of articulation that emerges under certain circumstances and that imitates to a greater or lesser degree the processes that occur during real speaking. When might this happen? Inner talking emerges “during the solving of difficult (i.e., atypical) problems, such as multistep arithmetic problems; during the reading and translation of foreign texts by those with a poor understanding of the language; while rephrasing texts (explaining something ‘in one’s own words’); during the memorization and recollection of verbal material; during the written recording of thoughts, etc.”111 It is associated with two main types of situations: (a) the perception of speech—according to the “motor” theory of perception; however, as we noted above, there is reason to believe that not all speech perception occurs according to the scheme proposed by “motor” theory and

46 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

(b) mental acts occurring in expanded (not automatized) form. In the first of these situations, inner talking occurs outside of inner speech as such and completely outside of “deep” psychic processes. As experiments by A.N. Sokolov demonstrate, we can engage the speech-motor apparatus with another activity and at the same time manage a sufficiently adequate understanding and reproduction of the text being heard.112 During the second of these situations, inner talking accompanies inner speech. Inner speech is a speech action moved “inside,” that is, produced in a compact, reduced form. In a typical situation, it emerges during the solving of a problem. Inner speech may be accompanied by inner articulation (with rather complex types of problems), but this is absolutely not an essential condition; inner speech may also be “represented” in talking by separate “hints,” by the supporting speech-motor features of words and word combinations. Inner speech is most often accompanied by inner articulation in cases where it is closest to conversational, discursive speech. This is one of its “poles.” The second “pole” is maximally compacted inner speech, least of all tied to inner talking and hovering on the edge of not being an intellectual act and of turning into a simple reflex act. Finally, as may be understood from the aforementioned, inner programming is the unconscious construction of a certain scheme on the basis of which a speech utterance will be generated in the future. The second essential question is tied to the relationship between inner speech and inner programming. The difference between them is the difference between the intermediate link in the process of speech generation and the final link. In other words, inner programming can unfold either in external speech (bypassing inner speech), or in inner speech, depending on the functional specialization of certain other factors. Speech “for oneself” or speech as a means of programming a nonverbal act is usually inner speech; speech “for others” is external speech. But speech “for oneself” can also, under certain circumstances, acquire structural features characteristic of external speech.113 The transition from inner programming to external speech takes place using the rules of the grammatical and semantic expansion of the program. It appears that the transition from programming to inner speech is also associated with the use of some rules—a sort of “minimal grammar.” In this connection Vygotsky’s predicativity as a feature of inner speech should be mentioned. One would think that this feature must belong to inner speech and not to inner programming; it is this feature that is associated with the “minimal grammar” of inner speech. But compare below.

JULY–AUGUST 2006

47

Figure 3. [Inner Programming and Inner Speech]
inner programming (rules of grammatical and semantic expansion) (special rules for semantic expansion)

external speech

inner speech

Thus, the relation between inner programming and inner speech, as we see it, is depicted in Figure 3. The third problem is to determine what features of the sentence are included in the scheme (program). Let us try to approach this problem logically. It would seem that the scheme of the utterance should not reflect features that belong to all speech utterances and can appear during the course of the program’s realization. Further, the utterance scheme should not reflect (if we presume it to be “supralinguistic”) features specific to the language, those that are not universal: its “grammar” is independent of the grammar supporting its realization.114 Finally, if we allow for the existence of a special motor program (which we will address separately), then the program should not reflect features associated with this aspect of programming. Based on this reasoning, it should be presumed that the program is put together from something akin to semantic “signposts,” that is, it incorporates the individual correlates that are particularly important for the utterance and its components, such as subject, predicate or object, and all this to the extent that their interrelations are essential to the future utterance. During the transition to external speech there is an expansion of each of these components into a syntactic group or a separate word plus the forms of connection of these groups in a sentence, if they exist in the given language. We will examine certain aspects of this process more closely. The fourth problem is associated with the form in which these features are recorded in the scheme. Here, we should look at N.I. Zhinkin’s experiments on inner speech that use a very simple method (the subjects were required to knock on the table in a certain rhythm during the process of solving a problem). It turned out that in most cases (at approximately the same point at which there was a reduction in the muscle tension of the speech apparatus in Sokolov’s115 experiments) the knocking does not interfere with inner speech, that is, inner speech makes a transition from a speech-motor code to some

48 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

other code, a code that is, by nature, one of images and schemes or an objective-representational code. On this subject N.I. Zhinkin wrote that “the language of inner speech is free from the redundancy of all natural languages. The forms of natural languages are determined by strict rules, as a result of which interrelating elements are specific, that is, the presence of certain elements presume the appearance of other elements, which is the source of redundancy. In inner speech the connections are objective, that is, they have to do with content and not form, and the conventional rule comes into play only when it is essential for a given thought operation.116 This idea is not new: a particular work of M.S. Shekhter is devoted to the nature of secondary images of a similar sort,117 and Vygotsky has also written about this. Numerous observations supporting Zhinkin’s findings have been revealed in studies of mechanisms of creative thought. Consider, for instance, the following statement of Albert Einstein:
The words or language, as they are written or spoken, do not play any role in my thought mechanism. The psychic/mental entities which seem to serve as elements in thought are certain signs and more or less clear images which can be voluntarily reproduced and combined. There is of course a certain connection between these elements and relevant logical concepts. . . . Conventional words or other signs have to be sought for laboriously only in a second stage, when the mentioned associative play is sufficiently established and can be reproduced at will.118

But it was N.I. Zhinkin who put this problem into explicit form. Nevertheless, it is easy to see that what Zhinkin is analyzing in the work cited and in others presenting results of the same experimental study is not inner programming, but either “pure” inner speech or something on the border between pure inner speech and inner programming: the subject is given prepared words, which he has to use to create a meaningful utterance. What this and inner speech have in common is that the subject is faced with the task of operating with already prepared speech elements. Programming and this experimental situation share the necessity of “guessing the grammatical construction of the phrase” at a particular stage of solving the problem.119 (But Zhinkin is not precisely correct here. What remains unknown is whether his subjects are “guessing” specifically the grammatical construction as such or the program of the utterance.) Overall, however, this case is closer to inner speech; strictly speaking we typically have an experimental situation that is almost never encountered in real activity. To what extent is it possible to apply data about inner speech to speech programming?

JULY–AUGUST 2006

49

We will start by pointing to one obvious difference. The program is by no means formed ad hoc; it was not by chance that Vygotsky spoke of “the grammar of thought.” The above quotation of Einstein provides an interesting illustration of this proposition. In speaking about a “second stage,” he was referring to inner programming and a “game of associations” that could be, in his words, “reproduced” once it had been “sufficiently established.” There is also at least one obvious similarity. Whatever “meanings” might be attributed to program components, to the components themselves—in their phasic, external aspects—may differ in terms of their origin, belonging to different code systems. They may relate to the speech-motor code. They may relate to the auditory or visual image codes. They may relate to the code of secondary images and schemes identified by Zhinkin for inner speech, and so forth; the nature of the code used to reinforce the speech program can vary within a broad range. But any unit of such a code is two-sided. In addition to the “signifier”— how a certain content is fixed to it, it also has a “signified”—what is fixed to it. The latter does not depend on the specific nature of the code and can be equally realized in different codes. What is the nature of this “signified,” and what attaches the subject to the code unit? The first possible answer to this question says—signification [znachenie]. However signification clearly has a specifically linguistic nature and depends on the structural-grammatical and lexical-semantic features of the specific language. On the other hand, signification is too “objective” to be used in programming; nothing ties it to the structure of activity and it is simply incorporated in it, not undergoing any essential transformation in this process. But a program by its psychological status—as the most important component of the speech act, concentrating itself on its main specific features— must enter into the system of activity and into a relationship of interaction and interdependence with other components of this system. For this reason, a more likely answer would be sense [smysl], as this word is understood by psychologists of the Vygotsky school, A.N. Leontiev in particular.120 Briefly, this concept can be defined in the following way: sense is the reflection of fragments of reality in the consciousness through the prism of the place that the fragment of reality occupies in the activity of a given subject. The concept of sense is broader than the concept of signification, since in addition to the “objective” features of the structure of activity that depend on the objective-situational determination of that activity, its structure also includes “subjective” features that determine the specific psychological nature of that activity. These meanings, it appears, are “tied” to units of the inner programming

50 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

code. It is they, or rather, the psychological features of the structure of activity, that determine the “grammar of programming.” Before turning to this “grammar,” we will emphasize one feature of the “programming code” that is very important in our view. It can be assigned more or less arbitrarily and interiorized (actually, this is the usual path of formation of this code—after all, secondary images are essentially compacted mental acts on an object). This opens wide possibilities for the experimental study of programming. If we now try to discover the inner structure of the program, the first thing that may determine such a structure is the features of “the model of the past and of the present,” the situation in one of its aspects. We will explore this question. The situation as a factor in the structure of a speech program, as has already been noted, must be distinguished from the situation as a factor influencing the specific realization of the program. We will not be examining the situation from this second perspective. As concerns the first aspect, in light of the clear insufficiency of materials, we will examine only the difference between the two most characteristic types of situations, specifically, the sensory and extrasensory situations. What do we mean by these terms? The formation of a speech utterance can follow different paths and can be motivated in different ways. On the one hand, it may be a description of the speaker’s direct perceptions (or what he is imagining, which is fundamentally the same thing), for example, pictures. We will call this a sensory situation. It would generally correspond to Svedelius and Luria’s “communication of events.” On the other hand, we might be operating with abstract concepts that are not immediately before us, or with certain generalized sensory data: not the barking of a particular dog, but the barking of dogs in general. This is analogous to the “communication of relations.” This distinction relates to the distinction between two types of “semantic states” indicated by A.A. Brudnyi: “There is a reciprocal influence between two types of semantic states of the word: (1) an extrasituational (or systematic) state in which the word possesses a certain semantic potential, and (2) a situational state in which the semantic potential is realized in the form of a number of “contextual signifiers.”121 It appears that with the first of these types we are dealing with an extrasensory situation and with the second, a sensory situation. The problem of speech generation in the context of a sensory situation is particularly complex because here this speech generation is preceded by a specific type of perception activity. In essence, almost no research has been devoted to the laws that govern how a subject selects and processes perceived

JULY–AUGUST 2006

51

information; there are only a few interesting observations of aphasics. These observations lead to a clear delineation of the mechanisms governing, so to speak, nomination and governing predication. Someone with “efferent motor aphasia” is not only incapable of constructing a sentence, but at the same time “substitutes an entire predicative utterance with separate words, usually nouns in the nominative case”: “Here’s . . . the front . . . and here’s the invasion . . . and here’s the explosion, and here’s . . . nothing . . . here’s . . . a fragment . . . speech, speech, . . . speech”122 (so-called telegraphic style). On the other hand, the “activation” of the mechanism that governs predication simultaneously prompts the appearance of verbs and other “predicative” words. This leads to the idea that there is a link in the mechanism of speech generation that is constructed according to the “vector” principle: we have some sort of initial components that correspond to the sensory images and a certain “predicative activity” that is attached to them and that forms the utterance (personal communication from A.R. Luria). One can presume that this link corresponds to the inner programming link. Let us note that the most significant data (the experiment by Slobin et al.) related specifically to the nominative part of the sentence. Further, it is entirely possible to interpret the utterance model in sign language not as a sequence of homogeneous elements, but in the following form (predication is signified by an arrow ↑ after the predicated part): [(S ↑) (O ↑)] ↑ From this point of view, Compton’s experimental results reflect consistent movement “inside the brackets” (negation is associated with an additional processing of the sentence at the stage of purely grammatical formation). We will cite one more work where a distinction between the image and “dynamic” components of the utterance corresponds to the distinction between “lexemes” and “taxemes.”123 If we accept this scheme, then it appears that Osgood’s idea about a “compacting” mechanism—where a combination such as S–A will be interpreted as (S ↑) with the subsequent dynamic unification of all components into the features of one “main” word—turns out to be close to the truth, the only difference being that the verb, it appears, will wind up on the periphery of the utterance. It is extremely interesting that similar ideas regarding the sentence as a kind of propositional function and generation as a process of sequential lexical-grammatical, and then phonetic derivation of a certain primary relation (relational proposition) that stem from completely different premises than ours, have been expressed recently by S.D. Katsnel’son and certain other authors. Katsnel’son proposes the following phases of speech generation:

52 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

(1) the identification of the topic and plan for the expansion of speech; (2) the quantification of elements of consciousness into separate “propositions” and the refinement of elements of “proposition” with the help of predicative significations; (3) the actualization of “propositions” by determining the volume and content of separate significations, as well as the intention and modality of the utterance, and so on; (4) the morphological formatting of the utterance; and (5) its phonological execution.124

So in cases where we are dealing with something clearly sensory, that is, when the subject and object of our utterance are provided to us as perceptible images (and perhaps in all other cases as well), the abstract-logical sequence of the program formation can be hypothetically imagined as follows: (a) the transference of perceived data into a sequence of sensory images or other elements of the object-scheme code (I → SiOi); (b) the optional stage—the attribution of certain features to elements of the object-scheme code, a sort of preliminary predication (Si + ↑ → S ↑, Oi + ↑ → O ↑); c) predication, possibly on the basis of the addition of a “verbal” component to the program, but not necessarily [(S ↑ O ↑) + (V)] ↑; d) the optional stage—the attribution of certain features to the utterance overall. In fact, the program elements appear to be elements such as S ↑, O ↑, and it is with them (as givens) that the real programming sequence begins. In the typical case, by the beginning of speech we always have sensory images of the objective components of the utterance, already “loaded” with a particular “meaning,” alongside subjective associations that are inseparable from this image.125 Up until now we have been talking about the situation factor. Now the time has come to talk about context as well. Worth is evidently correct when he points to its significance for the scheme of speech generation; of course, what we have in mind now is not context as a whole, that is, the influence of the preceding speech element on the one that follows it, but about the element within it that is associated with the nature of program components. This brings us back to the theory of “actual sentence articulation.” What interests us now is the fact that in languages with different structures, the actual articulation corresponds to a different linguistic realization. Furthermore, even within a single language this realization can be different due to a special intonation (“logical emphasis”), inversion, structural-grammatical features, an so on. It appears that there must be some code in which the “actual articulation” is expressed without variation, and only later is there a transition from this code to the actual language code. It is natural to presume that the inner programming code is such a code.

JULY–AUGUST 2006

53

There is, however, another possible way in which context (and situation) influence programming, which can be generally called ellipsis. Of course, it should be kept in mind that ellipses can have very varied origins. Previously,126 we analyzed certain types of ellipses; but all these types (phonetic reduction; the Apfel und Pflaumenbäume type; partial degrammaticalization, such as the majority of the kids are staying here; semantic neutralization, such as that one, you already did that thing to him? etc.) are connected not with the particular features of the program, but with the particular features of its realization—grammatical, semantic, or phonetic. However, since, for example, a vast number of utterances have the same subject, this subject may not, most likely, serve in the program of a particular utterance as a special component and may be replaced by some even more “conditional” “sign.” From this perspective, research into how utterances are programmed where the subject is the speaker himself is of particular interest. Consider, in this connection, the well-known fact that small children always refer to themselves in the third person and only later, and with some difficulty, switch to using the pronoun “I.” Until the age of ten or eleven, the author of this book was in the habit of putting himself to sleep by telling himself made-up stories in which he was one of the characters; it is characteristic that not only did he always refer to himself by name and in the third person in these stories, but in imagining himself visually in the situation described, he separated himself as the hero of the tale from himself as the storyteller. Finally, the third aspect concerning the influence of context is associated with emphatic inversion and the simultaneous grammatical deformation of the sentence (in ways that may be canonical in the given language), such as the French, “Cette lettre, tu l’as lue?” or the Russian Eto pis’mo, ty ego chital? [This letter, did you read it?]. We will encounter more such cases. Before turning to certain peripheral cases of the use of a speech program, we will point out that a speech program can be (and usually is) subordinate to a program that is hierarchically more complex and that supports not the generation of the individual utterance, but the construction of the speech whole; this is long-range programming and not just programming one utterance in advance. The nature of this programming—which has been investigated, in particular, by L.S. Tsvetkova127—is unclear. This is of much less interest to us, however, than programming of the individual utterance. Turning to the peripheral cases mentioned above, we will examine two of them: first, the use of a program in memorization, and second, what is called “simultaneous interpretation.” In connection with the latter we will also have to address the question of translation in general. The use of a speech program in memory is tied to its rather complex and

54 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

specifically human forms. This use is encountered, evidently, in those cases where we need to commit the content of a specific utterance to memory; such cases can be contrasted with others, where we must reproduce the utterance exactly (with the motor program serving primarily—but not exclusively—the role of the Plan that mediates memory) or when we must reproduce a more complex speech whole, without fixing the individual utterance (with the “longterm program” just mentioned serving as such a Plan). As far as we can determine, the cases of interest to us correspond to what V. Gutjahr128 describes as reproduction of the “semantic skeleton” (Sinngerüst) of the utterance. It is this use of a program that allows us to turn, with heuristic objectives (for the study of its structure), to a value referred to by Gutjahr as “Ergänzungspotenz,” that is, the ability of a given class of words to serve as a mnemonic basis for the reproduction of an entire utterance. Evidently, “momentary amnesia”—observed in Sokolov’s experiments, which feature the combination of reciting poetry and listening to speech, where “the subjects have heard and understood everything and almost at the same instant forget most of what they heard”129—is associated with the fact that the program was “busy”: it is not surprising that Sokolov observed his subjects attempting to rely on some external elements as a replacement for the program (bending their fingers, etc.). On the other hand, in Sokolov’s experiments there was a case where the subjects attempted to capture the speech they were hearing without the help of a program, using sequential visual or auditory images:
Subject S registered rather rich visual images that he did not usually experience during ordinary listening. Subject B attempted to capture individual thoughts in the speech she was hearing with the help of auditory images, making them “carriers of the overall sense.” As regards auditory images, however, it should be noted that at first, before speech comprehension, they did not lead to remembering: then, although individual words and phrases of the speech being heard “resonated innerly” (evidently, there appeared what is called initial auditory memory images), they did not create any logical connection and were immediately forgotten.130

Evidently, one of the types of speech—specifically, dialogic speech—is constructed on the same general principle as that kind of “image-driven remembering” of speech. In other words, dialogic speech does not have a program. It “does not necessarily issue from an idea or a thought that has been innerly formed by the subject.”131 In most cases it is situational and cannot be comprehended without knowledge of the given real or imagined situation. Furthermore, dialogic speech is typically reactive. The response-answer of the conversant usually takes the form of a rephrasing or even repetition of

JULY–AUGUST 2006

55

the question or comment: “It’s cold.” “I’ll say—quite a frost!”; “You heading home?” “Heading home, sure”; “Well what?” “Nothing,” and so on. In essence, dialogic speech is constructed according to the “stimulus– response” scheme. What the first conversant says in most cases allows for a relatively small number of possible answers, at least in terms of content. The speech “function” of the second conversant boils down to the selection of the most fitting of these possible answers—for the given situation and the given subject132 (one answers “I’ll say—quite a frost!” while another in the same situation might say “Brrr.”). Thus the connection between what the first and second conversants say is most easily interpreted with the help of the usual conditioned reflex connection. This makes it possible to “only half listen” to the other conversant when he is saying something for which there is an obvious response and allows for the possibility of overlaying lines of speech: one conversant has not yet finished saying something and the other has already started talking, interrupting the first.133 In some ways, dialogic speech and simultaneous interpretation are psychologically similar. Unfortunately, simultaneous interpretation has been little studied. We know of only four works on the subject (in Russian), of which only two touch on issues of psychology. Z.A. Kochkina believes that the “simultaneity,” “the synchrony” of such interpretation is a fiction and that it is “achieved by means of: (1) shortening the communication, (2) pronouncing the text more rapidly than the speaker. . . . The comprehension of another’s thoughts and the formation of one’s own cannot occur simultaneously, and these two types of speech activities are associated with the shifting of attention.”134 M. Tsvilling merely states certain hypotheses, outlining the course for further experimental confirmation. From among these hypotheses, we will recall the concept of the “minimal translation unit” (“the size of this unit will, in particular, also depend on the opportunities the target language presents for constructing the start of the phrase on the basis of minimal information. So, in a language that has only a limited number of syntactic sentence models, the amount of information in any fragment from the start of a sentence will be less than in a language with a richly complex system of syntactic structures. Consequently, the time lag in simultaneous interpretation into this language, all other things being equal, will be shorter”),135 as well as something that is particularly important to us: the idea that within the process of simultaneous interpretation there can exist “translation into a certain inner intermediate language.”136 This is, evidently, the “language” of the program. We will attempt to provide a more or less general psychological characterization of simultaneous interpretation. Under optimal circumstances, that is, when a highly qualified interpreter with a high level of automatic verbal

56 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

ability is in the booth, we see a type of monologic speech where the program is externally imposed. This feature of simultaneous interpretation, incidentally, is shared by any other type of translation. Another feature common to several types of translation (although not to all) is what might be described as reliance on identifying linguistic units; while in ordinary speech, programs may be enough (a person understands himself perfectly well and does not need any additional means for clarifying and reinforcing the content of the program), in “simultaneous speech,” the interpreter, as a rule, in addition to the work associated with programming the utterance overall, performs work to expressly “recode” certain linguistic components of the utterance (usually writing them down on a sheet of paper “to help remember”) (Figure 4). In simultaneous interpretation there is, however, a specific feature that distinguishes it from all other types of translation. This is the “discontinuity” of the program. In other words, the simultaneous interpreter does not “gather up” the entire sentence into the program in order to then “unfold” this program into a sentence in the other language; he performs this operation on separate components of the sentence, specifically those that are universal, in a certain sense, for any language, whatever its structure. For example, in translating the sentence, “The skill of simultaneously listening and speaking has to be developed in parallel with learning a foreign language, and not after it has been mastered,” the interpreter will wait for the word “speaking” to render a translation of that component. Then there is a break. “Has to be developed.” Translation, break. “In parallel with learning a foreign language.” Translation, break, and so on. Even from this example (based on the author’s experience and intuition, but not confirmed experimentally), it is evident how interpretation depends on the actual articulation of the utterance and its overall logical structure. If the phrase “in parallel” is clearly opposed speech with “after,” then we would expect a pause and a “break” to come before it, but if there is no such opposition (and the words “and not after” come as a surprise to the interpreter), it is possible that the fragment “has to be developed in parallel with learning a foreign language” will be the “minimal translation unit.” Finally, as noted by Tsvilling, it is evident that the order of components in the source language sentence and in the “target” language sentence exerts a certain “unsettling” effect during simultaneous interpretation. Everyone has heard how simultaneous interpreters who are not very good generate Russian phrases with syntactical organization that is far from Russian, preferring this to the risk of forgetting some essential utterance component if they extend the “minimal translation unit.” Using our terminology, that problem

JULY–AUGUST 2006

57

Figure 4. Linguistic Components of the Utterance

d c b a

d1 c1

b1 a1 abcd = a1b1c1d1 a = a1 c = b1 d = c1d1

can be characterized as the problem of grouping program components: the translation is not of the a–an–a1, b–bn–b1 type, but the ab–anbn–b1a1 type. We just mentioned that the external assignment of the program is a feature typical of all types of translation. This same thought can be expressed in yet another way: during translation, as is well-known, a certain invariant is constant in the transformation of an utterance in language I into an utterance in language II. (Here, we are not talking about literary translation; in this case the invariant is the dominant literary structure of the work.) What is this invariant? It is defined as “the commonality of the expressed . . . semantic content,”137 as the sameness “of basic semantic units of the intermediate language placed in correspondence with the given expression,”138 as “intuitivesensory experience” (and even—on the neighboring page—“immediate, intuitive reflection”) “of connections and relationships,”139 and finally, “the very same thoughts, feelings, desires.”140 These definitions fall rather clearly into two groups, depending on whether they are expressed by linguists or psychologists. (Here is yet another typical example of the difference between “linguistic” and “psychological” modes of scientific thought!) The first two definitions are based on a belief in the objective, logical sameness of the translation units. Here we find a variation on a well-known theme, which we addressed quite thoroughly in a previous book. We will cite a few lines from it:
Language—or, rather, speech activity—models a system of relations of the social person toward the world. Signification is not a reflection of denotation

58 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

in the mirror of language. It is a system of constant speech activity that creates conditions for relative consistency between its structure and one or another class of objects or phenomena of reality. . . In most cases . . . generalization and communication . . . are viewed very simplistically, as rigid systems of relations between words and the subjects and objects of activity that have been set for eternity. In actuality, both of these processes remain processes at any moment of activity, preserving their dynamic nature.141

The second two definitions are much closer to the true state of things. Nonetheless, neither of them is entirely correct. It is clear that “the very same thoughts, feelings, desires” is not a terribly strict formulation, even if we rephrase it and speak of one and the same objective logical and emotional content and the same motivation. As concerns B.V. Beliaev’s definition, in analyzing his book we have already been in the position of pointing to the fundamental (from our perspective) mistakenness of the very idea of an “immediate sensory” reflection of reality using intuition, with all the attending consequences.142 Therefore, it will not be possible to support this definition without confusion. What then, in this case, is an invariant in translation? We would argue the inner program of the speech utterance is such an invariant, the system of elements that are functionally “loaded” with meanings of the representational code or of actions on such elements. And inasmuch as meaning is a function of the interrelation of motivation and goal orientation of activity, the choice of program is conditioned by the living being’s previous experience (probability prognosis) and the structure of the program, in particular, by factors of situation and context, since all these factors are relevant in translation and must be considered during its psychological analysis (although in the typical case of translation, some of these factors are neutralized, for instance the experience factor and the situation factor). Another thing to note in regard to programming relates to the problem of biand multilingualism. Without going into these questions in depth (a special section will be devoted to them in another work), we will only point to the fact that various levels of mastery of a second language are associated with the correlation of speech abilities at various stages of speech generation. While “partial” mastery of a language is associated with the recoding (or a slowing of an operation) on the level of grammatical realization of the program, “full,” “fluent” mastery is a volitional and easy transition from one means of such realization to another means, conditioned primarily by factors of the speech act, of situation and of context (I speak English rather than Russian when I want an Englishman to understand me or because I am in London or at a reception at the British embassy or because I was asked a question in English).

JULY–AUGUST 2006

59

§ 4. In talking about the inner program of a speech utterance as the equivalent of what Miller and his coauthors call the “grammatical Plan,” we naturally must also ask ourselves what role the “motor Plan” plays in speech generation and what specific place it occupies in the overall structure of this generation. In other words, are stages (b) and (c), which were indicated at the beginning of this chapter, sequential in the absolute sense (the realization begins only after the completion of any planning) or in the relative sense (i.e., planning of one act can occur after the realization of another act)? In this regard we are in full agreement with T.V. Ryabova’s [Akhutina’s] reasoning based on the analysis of aphasic disorders.143 The scheme she provides appears as in Figure 5. The left portion of this figure shows combinatorial operations (successive) and the right shows selection processes (simultaneous). Our divergence with Ryabova in interpreting this diagram is not fundamental. It involves two main propositions. First, we doubt that the stage “choice of words by meaning” fully captures this aspect of selection, and, accordingly, that the stage “finding the full form of the word” amounts to finding its sound form. Later, we will explore another possibility, specifically the possibility of finding not all features, but only the supporting semantic features of words during the second stage of speech generation. Then “finding the full form” includes not only the sound but also the semantic search at the level of the grammaticized utterance. Second, Ryabova is far from correct when talking about the kinetic scheme of the utterance. In principle, the idea of this link is indisputable. Consider Chistovich: “construction in the human brain of a program of articulatory movements,” and “realization of this program, that is, its transformation into a certain sequence of sets of articulatory movements.”144 But the corresponding operation should not take shape as kinetic or motor utterance programming, but as motor programming of the syntagma, as L.V. Shcherba understands it. In other words, Osgood’s idea of functional encoding units finds a certain correlate here. This assumption is supported by a series of studies conducted in Chistovich’s laboratory. Her proposal that the syntagma is the unit of motor speech programming is also indirectly supported by something established by L. Kaiser:145 that the average length of the syntagma is seven syllables, which corresponds to the capacity of the human operative memory (if we count the elementary units of syllables). A special experiment was conducted to prove this hypothesis. In planning it, Chistovich presumed two possible mechanisms for the pause.
If the pause is assigned in the program, then there is no reason to expect

60 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Figure 5. [Diagram of T.V. Ryabova [Akhutina] on the Analysis of Aphasic Disorders]
Combination Inner speech utterance scheme Choice Choice of words by meaning

Gramatical structuring

Choice of words by sound

Utterance motor program

Choice of articulation

that a fluctuation in the activity of the pause between syntagmas during repetitions of a sentence will be greater than the fluctuation in the duration of syntagmas. The second mechanism may be that the pause reflects a stop in the working of automatic processes that is associated with a switching of articulatory programs. In this case the length of the pause . . . will fluctuate more than the length of the actual syntagmas.146

The results of the experiment fully supported the initial hypothesis. The second question confronting the researchers concerned the connection between the articulatory program and the pronunciation rate. It was proposed that the rate was not related to the program itself, but to its realization. In this case the temporal organization of the syntagma “must be characterized specifically by such relations within it that do not depend on the rate and are invariant given a change in the latter. . . . Only relative durations can be invariant.”147 The experiment supported this hypothesis, demonstrating the dependence of relative syllable duration within a word on the properties of the syntagma (the place of the word within the syntagma, the position of the logical stress). The researchers therefore concluded that the articulatory program of the syntagma is not a simple sequence of verbal subprograms, but that the word (as well as the phoneme) is related to a set of instructions that must be used in constructing the program of articulatory movements of syntagmas. The syntagma’s rhythmic pattern turns out to be invariant only in relation to syllables. “In the syntagma program, the only thing rhythmically organized is the syllable commands. . . . The expansion of the syllable in the sequence of speech sounds occurs according to some law of its own.”148 Then the sense was clarified in which one can talk about a preceding

JULY–AUGUST 2006

61

syllable as a signal to the syllable that follows it. Data indicate that each consecutive syllable command is automatically assigned, that is, not using information from afferent or proprioceptive impulsation arising during the pronunciation of the previous syllable. In studying the pronunciation of syllables in different structures, Chistovich and her colleagues demonstrated, using the Fairbanks-Guttman method (speech under conditions of delayed auditory feedback),149 that “the basic elements of speech are the most simple articulatory combinations such as CV, and more complex combinations such as CCV, CCCV are no more than groups of these simple combinations organized in such a way that the following combination begins before the preceding one has had time to finish.”150 In other words, the mechanism for syllable formation would “count” CCV as two syllables. All of this reveals the nature of the motor program in the most convincing way possible. And, returning to Ryabova’s diagram [Figure 5], the corresponding operation in it should be denoted as “kinetic programming of the syntagma.” But this means that somewhere during previous stages of speech generation the semantic benchmarks, so to speak, of a future motor program are established, that is, some sequence of utterance elements, within the bounds of which the motor programming will then take place, is selected on the basis of some contextual criteria (setting aside, for the moment, automatic speech behaviors—various clichés, set phrases, etc.). This selection can take place either in link 1 (programs), or in link 2 (the initial encoding), or in links 3 or 4. The first two can be eliminated because in them elements of the utterance are not yet articulated. It is therefore natural to propose that the syntagmatic structuring of speech occurs during the translation of the program into the actual syntactic structure or during the “filling in” of this structure with a lexicon.151 Let us take a closer look now at the data we have regarding this stage (or, as it were, these stages) of speech generation. In other words, let us move on to the problem of realization (the plan, the program). § 5. Let us presume that we have already “generated” a speech utterance program. What happens with it now? It is utterly clear that the next stage must be—as we stated at the beginning of this chapter—the realization of the program, that is, its recoding into some other structures that correspond to the specific lexical, grammatical, and phonetic (in the broad sense) features of a specific language. Formally speaking, this stage (or aggregate of stages) constitutes two different processes that take place with the utterance: (a) the selection of language units from the “memory” of the language speaker, and (b) the construction of the

62 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

speech whole from these elements. We will attempt, however, to demonstrate below that the binary nature of this process is a fiction. One way or another, we cannot get by without the concept of memory. We will point out that in Yngve’s model, and in Chomsky and Miller’s model, this concept occupies an essentially central position: the entire grammatical model of speech generation is localized specifically in (long-term or shortterm) memory, and short-term memory is viewed as a factor limiting the capability for realizing this model. In short, it can be said that neither Yngve nor Chomsky-Miller are dealing with anything but memory. For this reason, we must pause to examine the concept of memory. We will attempt to grasp first and foremost the ideas that are encompassed in this concept in contemporary research. We will begin by indicating the rather serious divergence between general psychology’s interpretation of memory as one of the highest human psychological functions (this concept usually includes the more elementary ways in which external influences are registered) and the interpretation current among most of the contemporary schools of psychology abroad and in almost all psycholinguistic works. Memory in its primary sense is first and foremost a process. “Memory should be understood as a process that allows us to preserve and reproduce the traces of past experience and to react to the signals or situations that have ceased to immediately impact a person.”152 Memory, in the second sense of the word, is a sort of “warehouse,” a passive repository of impressions; the processual principle emerges only at the “entry” and “exit” from this repository—in the form of storage and retrieval; but first there is specifically a “storing” of the remembered material in the warehouse, and second, the “withdrawal” of this material from the repository for a particular purpose. The typical thinking of adherents of such an understanding is that:
One of fundamentally essential operations for statistical encoding is the recoding of large or small fragments of communication into memory. In the process of learning, the more probable, often-encountered combinations, evidently, are recorded in abbreviated form in long-term memory. Some of these combinations are genetically fixed in memory.153

Second, the statistical concept of memory becomes particularly prominent in the psycholinguistic research with a “cybernetic” slant (e.g., the studies based on the Miller-Chomsky model). Works by authors of this school are filled with expressions like “to place in long-term memory,” “to be stored in short-term memory until . . .,” and so on. Here we are dealing with the phenomenon mentioned in chapter 2 in connection with an analysis of Chomsky’s psycholinguistic views: a material equivalent is postulated for each function. Furthermore, when we encounter the obvious multitude of

JULY–AUGUST 2006

63

memory processes in experiments (something previously not subject to doubt), this brings with it assertions of a multitude of memory “blocks”—here, we mention the article by Savin and Perchonock, which will be cited later. It is true, on the other hand, that there is also the opposite tendency—to view memory as a continuum and even to limit the distinction between types of memory to the distinction between “exits” (retrieval). It is absolutely impossible to summarize everything that we now know about memory for two reasons: the enormous quantity of material and the amazing breadth of its theoretical interpretation. We will merely indicate the primary distinctions. As noted above, three components or three processes of memory are usually identified independent of any type of memory (but most often in reference to long-term memory). They are: storage, memory, and retrieval. Sometimes storage and retrieval are defined as engramming and ecphoria, respectively. There are two main types of memory—involuntary memory and voluntary memory. The latter, in turn, can be immediate or is mediated. Mediated voluntary memory is the epitome of “human” memory and forms relatively late in the child. Its fundamental mechanisms were studied in the 1920s and 1930s by L.S. Vygotsky and A.N. Leontiev.154 Next, the different types of retrieval should be distinguished. First, we have something that is not actually retrieval, but rather is referred to as recognition. Second, we have immediate retrieval. Third, we have mediated retrieval, that is, retrieval by association. There is extensive literature, including works in Russian, regarding all of these types of memory and retrieval in particular.155 We must now limit the focus of our further analysis to the forms and aspects of memory that are relevant to us. It is obvious that within the process of speech formation we can (in principle) encounter the following instances that require us to operate with traces of previous influences: (a) “situational memory,” allowing for consistent reactions to the repetition of one and the same set (in whole or in part) of external factors conditioning a speech utterance; (b) preservation of certain components of a given utterance in the memory during the realization of the utterance (as in V. Yngve’s types of “obligations”); (c) storage and retrieval of the utterance plan or its program; (d) storage and retrieval of the utterance content; (e) storage and retrieval of the forms of the utterance, or, more precisely, its learning and retrieval “by heart”; (f) storage and retrieval of grammatical structures; (g) storage and retrieval of words; (h) storage and retrieval of set phrases; and finally (i) storage and retrieval of sound sequences.

64 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

Below we will now examine in detail most of these cases. For now, we will direct the reader’s attention to certain aspects that may cause confusion. First, there is the fact that we are dealing with fundamentally different processes. They can be classified according to the following features: I. What is being stored (retrieved). This can be: (1) the external circumstance of the utterance (a); (2) the content of the utterance (c, d); (3) the form of the utterance (e, h); (4) the linguistic components of the utterance (f, g, i); and (5) the linear components of the utterance (b). II. How it is being stored. This can be (1) “automatic” immediate memory (a, b); (2) voluntary memory (c, d, e);156 (3) obligatory, so to speak, memory, that is, a person essentially cannot fail to commit certain material to memory if he is living in human society (f, g, h, i). III. How it is retrieved. Here, the following versions are possible: (1) retrieval that is completely automatic, not leaning, in most cases, toward conscious retrieval (a, b); (2) retrieval that is automatic, but generally subject to consciousness (f, g, h, i); (3) retrieval that is semi-conscious or fully conscious (c, d, e). IV. Whether it is held for a long time. Here, the following types can be identified: (1) memory within the confines of the utterance (b); (2) memory within the confines of certain utterance sequences or within the confines of a relatively small time interval (c, d, e); (3) “permanent” memory (a, f, g, h, i). Second, despite any superficial similarities between some of the cases listed and well-researched experimental situations, there are fundamental differences between them. Primarily, what we have in mind is the fact that “word memory” is by no means the same as what is studied under the name of “verbal memory,” although researchers exhibit a tendency to apply the results of studies of “verbal memory” to “word memory.” “Verbal memory” is memory that is generally voluntary and mediated; “word memory” is immediate and involuntary, obligatory in the sense indicated above. “Verbal memory” is characterized by a voluntary, often conscious and generally mediated retrieval; “word memory” is the opposite in this sense. Finally, if “verbal memory” is temporary memory, then “word memory” is “permanent.” We will attempt to elucidate the individual types of memory that are relevant to speech generation in greater detail. A. Situational memory. We already encountered this phenomenon above when discussing P.K. Anokhin’s situational afferentation and N.A. Bernstein’s “model of the past and of the present.” In essence, it should not be called memory at all; it is simply the establishment of classical conditioned-reflex associations.
If in the past event A regularly preceded event B, triggering reaction b; then event A becomes a signal leading the organism to develop reaction b.

JULY–AUGUST 2006

65

. . . If event A always comes before event B and in the life of a given organism no other combinations are encountered (in other words, the conditional probability of event B given the onset of A is 100 percent), then a rather small number of pairings of A and B are needed before event A to begin to trigger reaction b, that is, to develop a conditioned reflex. But if event A is encountered preceding other events (B, C . . . N), but the conditional probability of event B given event A is greater than the conditional probability of events C . . . N, then a greater number of pairings of A and B is needed for the development of a conditioned reflex. The closer the conditional probability of events B, C, N, given the occurrence of event A, the harder it is to develop a conditioned reflex, the greater the number of pairings needed for that.157

Engramming, the consolidation of past experience, occurs in this case in the form of electrophysiological activity in individual cells or groups of cells. If, following the example of L.P. Kraizmer, we talk about four levels of biological memory—structural, network, cellular, and molecular158—then “situational memory” is localized, most likely, at the “network” level. Research in this area, however, is still in the preliminary stage.159 We will not examine this type of memory in the present work. B. Commitments [ obiazatel ’stva]. This is the type of memory that psycholinguists most often deal with, referring to it as “short-term” memory. As has already been stated, they study this type of memory almost exclusively in the context of the limitations it imposes on the generation of utterances. A separate section will be devoted to this type of memory in the present chapter. C. Program memory. We encounter this type of memory on those quite numerous occasions when a person has planned his future speech in advance, at least several sentences (or more) ahead. It is completely obvious that there is no planning of the external linguistic form of the utterance, and if there is any, it is only among those speakers of a language who lack the established skills of spontaneous monologic speech.160 This question is extremely intriguing in and of itself. While in our theoretical research we—linguists, psychologists, philosophers—operate as if we are dealing with a “typical” form of speech, specifically with active, spontaneous speech (or at least speech that is in part situationally conditioned), and so on, such speech becomes possible and somewhat usual only at a relatively high level of social development, and is a product of the conscious nurturing of certain abilities and skills. “Uneducated” speech is reactive speech, conditioned through contact and situation. And for such an “uneducated” speaker, the planning of a speech essentially amounts to its preliminary generation with subsequent precise retrieval—a person reads “word by word.” (In some

66 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

cases he is not even the one who wrote what he is reading.) But the more automatized a person’s spontaneous speech skills are, the less he will have to establish for himself his utterance. For a well-educated adult, especially a scientist, brief notes are completely sufficient. Professional lecturers and teachers manage, as a rule, without any notes at all, but with just an outline. Of course, another factor arises here, a factor that could be called the degree of responsibility. One and the same person will put more planning into one and the same talk if he will be delivering it in the assembly hall of Moscow University on the Lenin Hills or before the annual gathering of the Academy of Science in the House of Scholars, or in a more cursory manner, without filling in the details, if it is intended for a small gathering of colleagues. Although things like outlines and notes are not mirror images of the plans and programs formed and memorized by a person, but just the products of the exteriorization of these programs on the basis of a special subsystem of rules for their realization with special rules for their lexical, grammatical, and phonetic (or in this case, probably graphic) expansion (in all probability close in pattern to the expansion of the program into inner speech, but essentially not studied at all). The study of such outlines and notes can provide a great deal for the understanding of the specific nature of the programs that lie at their foundation. However, such a study has yet to be undertaken. There is, however, at least one source that provides material for such a study. We are referring to the collected works of Vladimir Ilyich Lenin, which in many cases includes the preliminary notes prepared by Lenin himself, side by side with the stenographic records of his speeches and appearances, allowing one to trace the form in which a given content was coded (or, alternatively, the form into which a given point was expanded). Returning to our main topic, we will point out that the recoding of a stored program in written form that we describe here and its consolidation through an “external” rather than “inner” path is of less interest to us in this case. A more important case, from our perspective, presumes the complete absence of “external” elements. What form does the program encoding take in such a case? Evidently, we are dealing with elements of an “object-representational” code. We either write in our heads and then, speaking, see the notes before us, or we bring together the real or imagined visual images into an orderly chain and “attach” to it components of the utterance, or do something of that sort. Among certain people, specifically those who do not have a developed sensory (primarily, visual) imagination and visual memory, according to our observations, the program can be coded in a complicated form, similar to discursive, logical thought. Instead of retaining separate reminders for the expansion of

JULY–AUGUST 2006

67

thoughts, we retain the inner logic of its content. This brings us to the next form of memory. D. Content memory. Vygotsky puts it very concisely: “We retain meaning independent of words. For instance, in today’s lecture I will have to convey the content of a large number of books and papers, and I remember their meaning, their content, but I would have a hard time reproducing their verbal form.”161 Experiments by Bühler have shown that the storage of thoughts follows a different sequence than the storage of words. This idea was then repeatedly used in experimental studies of memory, especially by Soviet psychologists. We can turn again to Vygotsky for a summary of the results of these studies: the retention of thoughts “follows the laws of the semantic reference of one thought to another.”162 “Content memory” epitomizes human memory, especially, higher-order memory. After all, in order to realize “semantic reference,” in order to organize stored content, a complex, analytic-synthetic activity is needed, mediated by language or other auxiliary means. It is the memory that we encounter in the book by Miller, Pribram, and Galanter about Plans; it is the memory that was studied by L.S. Vygotsky and A.N. Leontiev in their time. It could conditionally be called “structuring” memory. We will return to it in the next section. E. Form memory. This is different from “content memory” in that here the semantic organization of the text serves solely as a means of retention, as a kind of mnemonic device. A typical example of this sort of memory would be a pupil memorizing the poem, “Upon a chilly and wintry day. . . .” As anyone knows, understanding this or any analogous text makes it easier and faster to memorize, but is not an absolute requirement. A pupil with a trained memory can memorize a nonsense text without any great difficulty or a text in any foreign language. And the text memorized, whether or not it was comprehended during the process of memorization, can become nonsensical— something referred to in American psycholinguistics as “satiation.”163 What is being “placed in memory” in this case? Unlike “content memory,” where the utterance program lies at the basis of memory, the “grammatical Plan,” “form memory” deals with the kinetic program, with the “motor Plan.” Although in ordinary speech (and, correspondingly, in ordinary memory) both Plans appear simultaneously and in association with one another, this does not mean that they are unable to function in isolation. To jump somewhat ahead, we will point to one type of speech that relies exclusively on motor programming. This is what is known as “glossolaliac” speech, encountered, in particular, in certain Christian sects. V.B. Shklovskii studied it from a somewhat different angle in his lifetime, introducing a number of

68 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

interesting examples. Here is one pertaining to the Irvigian sect, which existed in Scotland in the 1830s:
Hippa gerosto hippo boorus senote Foorime corin haoro tauto noostin . . . and so on.164

It is easy to see that this zaum (something beyond sense or reason) is organized in a certain way in terms of sound relations and bears a resemblance to Greek. Everything depends on that initial motor program, which— most likely borrowed from “meaningful” Greek speech—achieved an independent “currency” and acquired a rich acoustic instrumentation, serving in a variety of specific realizations. The mechanism of such a transition from the motor program of meaningful speech to its “instrumentation” along a “trans-sense” line is wonderfully evident in the best-known example given by Shklovskii from A.M. Gorky’s My Childhood. The original text reads:
Bolshaia doroga, priamaia doroga, Prostora nemalo beresh’ ty u boga. Tebia ne rovniali topor i lopata, Miagka ty kopytu i pyl’iu bogata. (The road is broad, the road is straight, Quite an expanse you take from God. Ax and shovel have not leveled you. You are soft on the hoof and rich in dust.)

The text as reproduced by little Alesha was as follows:
Doroga, dvuroga, tvorog, nedotroga, Kopyta, popyto, koryto . . . (The road has two horns, the curd cheese is out of reach, Hoof, tooth, roof . . .)

If such an interpretation is correct, it would be interesting to use a computer to process “glossolaliac” texts (no other method would be possible due to the tremendous volume of information that would need to be processed) to attempt to reveal the motor programs at their foundation. One more comment in passing: if, when speaking in “prose” we store, consider, and, to a certain degree, retrieve only the grammatical Plan (program) of the previous utterance or utterances, in composing verse we rely on the storage and retrieval of the motor program as well. It is the motor program, according to all the evidence, that makes up the mechanism of the

JULY–AUGUST 2006

69

phenomenon that B.V. Tomashevskii in the 1920s called the “rhythmic impulse.”165 But this takes us beyond the bounds of the present work. Native-language memory. This is what is usually referred to in psychology and psycholinguistics as “long-term memory.” We will encounter it below. For now we will merely emphasize that an exceptionally important component of this memory is the H sort of memory, which is underappreciated in the works of transformationalists and which holds a place of honor only within the system of views adhered to by Osgood and his associates. The distinction between “situational memory,” “content memory,” and “native language memory” has a curious parallel in the distinction proposed by W. Penfield166 between the three principle types of memory: memory of individual experience, memory of concepts or generalizations, and memory of words. It appears that the distinction between short-term and long-term memory was first coherently expressed by D. Broadbent in 1958,167 but a systematic study of this distinction was carried out only in the sixties. Two opposite viewpoints were developed. Broadbent himself viewed both types of memory as freestanding and essentially independent of one another. E. Melton, on the other hand, considers storage a single process and looks for a distinction in memory output. Questions associated with this distinction were discussed in detail at the symposium “Short-Term and Long-Term Memory” during the Eighteenth International Psychological Congress in Moscow during the summer of 1966. However, if we look at the basis there is for the differentiation between these two types of memory, it turns out that this differentiation stands on rather shaky ground. “In an area where so much is unknown and where it is so difficult to obtain direct answers to many important questions, we would be ill-advised to completely ignore any source of information, however distorted it might be. What do introspection and self-analysis tell us about memory? . . . Upon consideration we will probably agree that there must be at least two different types of memory. . . . These subjective impressions about short-term and long-term memory are consistent with a number of physiological observations.”168 “Common sense, along with experimental data, supports the idea that our ability to retain the data of an event is associated with at least two processes.”169 And so on. It is natural that, due to the extremely vague criteria for what constitutes “long-term,” psychologists—especially Soviet psychologists—started using the functional orientation of memory types as a criterion.
Short-term memory is primarily associated with an initial orientation in the environment, and, therefore, is mainly aimed at determining the overall

70 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

number of new signals without regard to their informational content (which, in any event, cannot be determined, since the first appearance of a signal does not convey anything about its probabilistic nature). . . . The task of long-term memory consists in preserving what will be needed in the future. It is associated with the organization of behavior at the level of lengthy time intervals, which will inevitably require foresight. For this reason the probabilistic profile of signals takes on a high degree of significance. . . . And this means that the main determinant of long-term memory is the informational content of the signal.170

This assertion is founded on the results of a cycle of studies by P.B. Nevel’skii.171 He demonstrated that the number of stored units is relevant to short-term memory, while the informational load is relevant to long-term memory; meanwhile, information, according to Nevel’skii, is put together from the characteristics of uncertainty (more will be stored if there are fewer symbols or less information per symbol), probability (more symbols will be stored if different symbols have differing probabilities of appearing—some appear with greater frequency and others will lesser frequency), and variety (more information is retrieved if there is a more extensive source from which symbols are taken). Miller’s well-known experiements172 led to a belief that during storage there is a mandatory recoding of material, in particular its “consolidation.” However, Miller’s conclusions were one-sided. As the authors of the chapter on memory in the book Engineering Psychology point out, he does not go beyond the bounds of the “stimulus–response” schema and does not view “chunks” of information as products of a specific human activity.
It is necessary to study the particular features . . . of activity that lead to the formation of chunks of stored material that are different in terms of capacity and content. Only the patterns of this activity can serve as an objective basis for the psychological analysis of features of these chunks, including their length. The process of encoding and decoding can also be explained by the features of activity. Finally, the patterns of storage activity that led to the formation of these chunks must also serve as a basis for the mathematical analysis and evaluation of pieces of information.173

This general assertion, made on the basis of experimental findings, takes the form of the thesis that
it is not so much the reduction in the amount of information that affects the capacity of memory, as the activity that reduces the quantity of information. The dependence of storage on the amount of information is fully evident in cases where the amount of information in the stored material is

JULY–AUGUST 2006

71

reduced by the subject himself in the process of active mnemonic, cognitive activity. . . . Research has also shown that memory capacity also depends to a great extent on relevant information, that is, on information that relates to the goal of human activity.174

The above quote compels us to be guided by the activity principle in our analysis of the phenomena of memory, and to base our classification of types of memory on the particular features of the memory activity—its structure, goal-orientation, and so on. In general terms, this principle corresponds to feature II (how something is stored) in the classification of types of memory given above. We will remind the reader that from this perspective immediate speech memory, then operational memory, “structuring” memory, and, finally, obligatory or permanent memory have been delineated. From here on we will be using the following terms: (1) immediate memory (type b); (2) operational memory (types c, d, e); and (3) permanent memory (types f, g, h, i). It is easy to see that this use of terminology originates from that accepted within Soviet psychology, in particular in the works of psychologists of the Kharkov school. Psychologists of the Kharkov school distinguish operational and immediate memory as two different types within short-term memory.
In short-term memory, we are dealing with the psychophysiological capabilities of impressions and initial retention of what has just been perceived, which are general and abstracted from the objective of activity. Operational memory is subject to the goals of the specific activity. Its short-term quality is relative. It is defined by features of the particular activity. While relating short-term to long-term memory, operational memory itself can maintain a significant duration in comparison to immediate memory.175

Before moving on to a more detailed clarification of the types of memory that are pertinent to the main focus of our work, we will examine the currently available data concerning the psychophysiological mechanisms of different types of memory. These data are extremely fragmentary and inadequate. According to the findings of Lorente de No and other contemporary authors,176 at the basis of immediate memory (and short-term memory generally) lie the recirculating processes within the closed circuits of the neural net, within circular chains of neurons. By giving an impulse to a link in this chain, we elicit electrophysiological excitation in other cells of the circuit until it returns to the starting point and begins the next “circle.” According to a current hypothesis, these “reverberating circuits” [reverberatsionnye krugi] represent the physiological substrate of short-term memory. They are associated with certain biochemical changes, specifically

72 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

with an increase in ribonucleic acid (RNA) in the excited cells; the change from “electrophysiological” storage to “biochemical” storage corresponds to a change from short-term to long-term memory. And, of particular importance, from work by Hyden177 it appears that the mechanism for “reading” the biochemical memory code is the same mechanism of circuit excitation, since RNA fixes different excitations differently, and, encountering a stimulant that had in the past occasioned a given change in cell biochemistry, we achieve a unique “resonance.” “The cells of the nervous tissue begin to respond in different ways depending on whether old or new stimulants are acting on them. If new, unfamiliar stimulants are acting on them, they do not elicit the corresponding resonating structures in the protein acids; if familiar stimulants are acting on these cells, they elicit the corresponding resonance.”178 There is more, however, to the physiological mechanisms of memory. “The nerve cell reacts very quickly and very quickly discharges its charge: for the nerve cell it is typical that, having reacted quickly to a certain stimulant, after a certain period it is always again ready to receive new stimulation. It is unlikely that such an apparatus would turn out to be suited for preserving the traces of long-term memory.”179 Evidently, the brain’s gray matter, the socalled glia, takes an active part in such storage. It has been established that biochemical changes in the glia (primarily affecting RNA) have an inverse relation to corresponding changes in the nerve cell. The glia, in a certain sense, “discharge” nerve cells. At the same time, evidently, it directs the movement and growth of nerve cell dendrites, and in so doing forms a permanent structural basis for memory. The “potentials” in glial cells are much slower than in nerve cells, and there is reason to believe that the glia, together with nerve cells, form a unified system and, in particular, stabilize excitation arising in nerve cells. The data introduced about the physiological mechanisms of memory have tremendous fundamental significance for us. If the processes described actually do constitute the basis for memory, then it turns out that the problem of distinguishing types of memory is associated less with differing localization than with sequential links of one and the same process: immediate storage on the basis of “reverberation” excitation (and the “reading” of traces directly from this link)—storage on the basis of biochemical processes (and “reading” based on resonance, that is, where the excitation resonates)—storage on the basis of a neuroglial regulation. Such an understanding best corresponds to the spirit of the psychological conception on which we base our analysis of the speech process. § 6. We will begin a more detailed discussion of the types of speech memory starting with what is referred to above as operational or structuring memory.

JULY–AUGUST 2006

73

As has already been noted, operational memory relates to a specific activityact. This is storage “for as long as needed.” At the same time, operational memory, being immediate memory, is specifically human, as are man’s other higher psychic functions. As it pertains to speech processes, operational memory is program memory. This means that within the framework of the system of speech acts, we are, so to speak, constantly holding some program data, both from syntactic and, mutatis mutandis, motor programs. This is essential for us to be able to further generate the utterance, since during the course of that subsequent speech generation, there seems to be a systematic correlation between what output we have and the program structure. The picture that most closely matches our conception180 of such a correlation was drawn by D. Worth when he presented a model of speech generation in the form of a television screen linked to two mechanisms—an expansion mechanism and a mechanism for reading and correction, which conveyed the results of this correction back to the expansion mechanism until the correction ceased, after which the sentence moved to the next “bloc”—the phonetic bloc. What is stored using operational memory? Can it be said that we remember the program as such? This is a very complicated question, because at present we know very little about the program itself. Approaching this from the perspective of general psychology, its answer would be as follows. The program, as has already been noted, is not a given, not a readymade program, but a process: programming. And—as applicable to memory—this process mediates storage of the content aspects of the utterance.181 The “semantic ‘load’” of elements of the programming code and the program overall is what we “transfer” in space (if we are dealing with programming of speech in progress) or in time (if we are dealing with operational memory). In this connection, Vygotsky said something of interest. We can express one and the same thought, he said, in different meanings: for example, the thought I am not guilty can be realized as The clock fell down on its own. And so “semically, the clock fell relates to the corresponding thought, in the same way that a semantic connection during the mediated storing relates to what is being stored.”182 The second aspect of using the program in operational memory is associated with what happens outside the bounds of the utterance. It appears that Worth is correct as well in that previous utterances participate in generating successive utterances—in accordance with the Markov chain principle, that is, exclusively at the level of linear discharge; but in conveying his thoughts on this, we stated that this carries with it a presumption of the necessity of selectively encoding and storing data about the content of previous utterances.

74 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

This is exactly where it makes the most sense to presume that such a form of encoding and fixing is the program. If that is the case, then we eliminate the problem of an “unnecessary” stage of speech generation (in addition to everything else, it is still not known what content encoding is for): it is quite plausible that the utterance program is not “removed” immediately, but continues to remain fixed for some time—perhaps until the end of the generation of the next utterance. In this sense, V. Shevchuk’s observations on sentence comprehension and retention are extremely interesting.183 While we are not able to convey his data in detail, we will indicate that Shevchuk views the process of sentence comprehension as a multistep process in which a step corresponding to programming has its place. There are also works that support our point of view more directly. Mehler and Miller’s data point to the storage of the sentence’s semantic content as a first stage and the grammatical structure as a second stage; from the perspective of the Chomsky–Miller model, it would seem that the reverse sequence should be expected. Jumping somewhat ahead, we will say a few words about the interpretation of data from Mehler, Colemen, Prentice, and Savin and Perchonok. We will remind the reader that all of these authors stated that active (kernel or core) sentences are retained better than their transformations—passive constructions, nominalizations, and so on. Does this mean, as J. Mehler presumed, for example, that in the usual case, the core sentence plus a certain aggregate of actions are stored? This would seem far from certain, especially if we consider that there is a general tendency toward a better understanding of a sentence “the closer the grammatical structure of a sentence is to the semantic structure of its content.”184 Since, in the experiments of the authors mentioned, core sentences were given to subjects as the original sentences, core sentence” was and from the very start the direct relationship “program → established, and all other sentences turned out to be psychologically produced. But it is entirely possible that during storage, instead of core sentences, programs were used, the realization of which, in that case, occurred during retrieval and not during storage. There is nothing to contradict such an assumption in the results of the experiments mentioned above.185 We have advanced the hypothesis that when there is a delay in the storage of utterances, it is not the utterance as such that is stored, but its program. This perspective allows us a fresh interpretation of certain earlier data, specifically data about the “compression” of sentences under certain circumstances. We know of at least two such works (not counting Shevchuk’s book mentioned above; unfortunately this book does not adhere in every way to

JULY–AUGUST 2006

75

contemporary standards of psychological research), specifically, articles by E.L. Ginzburg and his coauthors and by G.V. Eiger and M.M. Gochlerner. Ginzburg and his coauthors studied mistakes in reproducing sentences that were tachistoscopically exposed (on a television screen) in the subject’s native language; it was found that out of sixty-nine sentences shown, forty-five underwent compression.186 Eiger and Gochlerner conducted an analogous study, for which, in addition to native speakers of Russian, they used native speakers of German. The most important conclusion they reached was that “transformations in speech perception are a constant factor. However their types, prevalence, coincidence, and frequency of use with the given structures must be determined by the style and type of speech.”187 that is, they are more likely to be conditioned by psychology than by structural and grammatical concerns. The authors assume that “transformation occurs in the mnemonic action during the first stage, during the stage of cognitive orientation within the subject matter.”188 However, under the conditions present in their experiment on operational memory, the workings of immediate memory undoubtedly also could not fail to appear. The delimitation of these two aspects was not drawn in this work. In the authors’ interpretation of experimental results, some points are distorted. Thus, program storage, or, more precisely, the use of programming during storage, is carried out in two stages: (a) storage of the program during the process of its realization; and (b) storage of the program as a basis for the generation of subsequent utterances. We will point to certain other features of these two types of usage of programming. As far as the first is concerned (if we again jump somewhat ahead), there arises the problem of how the results of program realization are combined with the program itself. In the end, what is being combined is obviously different from the thing it is being collated with! This is, however, a pseudoproblem, as we will attempt to demonstrate below. As concerns the second point, up to now we have been operating essentially with two utterances: the preceding and the following. But during the generation of an utterance an entire chain of preceding utterances is kept in mind, and there is a sort of accumulation of their content. How does this process take place? It can be presumed that its specific structure in a certain sense stands in an inverse position to the programming of large speech aggregates. If in this case we are dealing with some form of dynamic interaction between “smaller” and “larger” programs, associated, evidently, with the elaboration of “larger” programs in “smalerl” ones, in the mechanism of

76 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

operational memory that interests us here, “smaller” programs can provide, under certain conditions, a sort of qualitative leap, being bundled into a “larger” program. Unfortunately, systematic study of such compression has yet to be conducted. It is also unknown what role a subject’s creative activity plays in such compression, that is, to what degree it is automatic. Considering that operational memory is the type of memory that is most dependent on the specific features of the activity-act, one should expect that such creative activity serves as a determining factor and is related to the structure of activity. In any event, it can be presumed that operational “program memory” arises only in cases of delayed retrieval and that where it follows immediately, we use a mechanism of semantic accumulation of the type described by Osgood. § 7. We have in mind earlier developments such as “obligatory measures, or requirements,” that is, the temporary preservation during the delivery of an utterance, including characteristics that are essential to the maximal “correctness” of speech generation. The first question that arises, of course, is the question of what is exactly retained by immediate memory, and how much actually is retained. First of all, it is impossible that the grammatical structure of the sentence is not fixed, in whatever form, since otherwise it would be simply impossible to bring the sentence to its conclusion in the same grammatical “key.” Cases in which this key is lost for one reason or another (losing or changing focus, changing the program in mid-course, etc.) are by no means rare: we are approaching a particular word and suddenly begin to construct the grammatical structure of the sentence as if anew, giving the word a different structural characteristic than the one it had before. One would assume that precisely in immediate memory, grammatical “prognoses” are fixed that replace one another on the “television screen” (if, as we will see in the future, we rely on Worth’s hypothesis). In what form are they fixed? Perhaps a branch of the syntactic tree of phrase structures is such a form; in other words, it is not the linguistic construction overall that we predict, but—with whatever degree of “depth”—the nature of the hierarchy of grammatical relations linking the given prediction to part of the utterance (to the extent possible to predict it based on a given part). Second, certain semantic elements must also be fixed, since otherwise it would not be possible to keep track of the inner context of the sentence (we are not, after all, aware of the program during speech generation!). These elements will be examined below. As concerns what could be conditionally called the “capacity” of immediate memory, it is well known since the works of G. Miller.189 First of all, it is known that the volume of immediate memory is measured in storage units and—where necessary—the number of such

JULY–AUGUST 2006

77

units is actively reduced. Second, it is known that the number of such units, not only in speech memory, but in other types of immediate memory as well, does not exceed 7 ± 2. This has been confirmed in working with several languages (see information about V. Yngve’s hypothesis above in chapter 2). With respect to semantic units, data can be drawn primarily from G. Miller’s well-known body of works. Based on Brener’s results, Miller provides the following average figures: abstract words (visual)—5.24; abstract words (oral)—5.58; concrete words (visual)—5.76, concrete words (oral)—5.86. He also introduces Reed’s data, which differ slightly from Brener’s—6.55. We will introduce several other interesting figures: in immediate memory 1.75 simple sentences can be stored (with an average of 6 or 8 words), that is, approximately 10.5 organized words. We will point out, however that these numbers are not entirely correct, since the authors did not spell out how to distinguish between utterance phrases that follow a common template and those that do not.190 Thus, the capacity of immediate memory for meaningful words generally corresponds to the memory’s capacity for grammatical elements. Naturally, the question arises, at what expense does an increase in the volume of memory take place during the storage of an entire sentence rather than isolated words? We just mentioned one means of compensation—set phrases. But there is a much more effective means of compensation associated with the existence of a program; in other words, the storage of a phrase can be in a certain sense mediated by the storage of the program. We will remind the reader about the existence of experiments (primarily by Epshtein) that show that the level of storage increases overall with an increase in the level of the utterance’s grammaticalization. There is another especially attractive possibility in light of our understanding of the psychological nature of signification, described below—the possibility that not whole lexical units, but semantic components are stored. No one has yet attempted to measure the volume of memory from this perspective. Immediate speech memory can be characterized—if we make use of S. Sternberg’s191 fortuitous terminology—as memory that informs about the presence of an element (unlike operational memory, which evidently signals both the position and sequence of an element). In this connection, let us say a few words about the particular features of the process of retrieval. It is important, at the same time, to keep in mind that here, and in the future, we are dealing with processes that take place during the generation of speech, but not during perception, where things work differently. It was mentioned above that retrieval is based on a resonance mechanism (in the broad sense of the word); a repeat appearance of certain initial conditions prompts a repetition of the reaction process. As it pertains to immediate

78 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

memory, this means that during the generation of the second part of an utterance, certain features (or rather features of the processes of its generation) arise, which coincide or come close to the features of the speech-generation processes of its first part. In other words, if in the beginning of a sentence we use the masculine noun “rooster” [petukh], no matter how far the distance between subject and predicate may be, when we reach the predicate, the mechanisms of immediate memory will undoubtedly be activated and the predicate will be realized in the needed form of agreement: “crowed” [zakukararekal]. In this way, storage and retrieval at this level rests on a principle that can be characterized as the principle of relevant recognition criteria. A certain word appears; it is recognized as corresponding to certain criteria; it is given a certain specific realization on the basis of immediate memory data. This, however, is merely the crudest approximation of a description of the actual processes that take place. In fact, the very appearance of a given word is most likely not independent of the available criteria. These criteria to a certain degree determine the very search for this word. We will discuss this further below. As far as the nature of criteria used here, we should note the fact that there is a grammatical relationship between distant elements of an utterance (whatever the specific grammatical structure of the language) and only between: (a) the subject and predicate, (b) the modified and the attributive, (c) and, less frequently, between the object and the predicate. In other words, agreement (or form, it is all the same) is a certain feature of the process of predication (in the broad sense, as we used this expression previously in discussing the “vector” model. What we are probably talking about here is the actual execution (in one or another specific form of realization) of a process that is potentially tied to the appearance of some reference element; the criteria we are referring to in this case are essentially some features of reference, predicated elements related to the process of predication. This entire problem demands additional experimental and theoretical analysis that does not fall within the scope of this book. We will limit ourselves to spelling out the most fundamental aspects of the organization of this link in memory, as we have done above. It is of particular interest to us that in the process of retrieving data stored in immediate memory, we are dealing not so much with a process of memory itself as with a process of recognition or identification. Remembering is a process stimulated by identification. How is this identification carried out? In recent years its mechanisms have been studied rather extensively. There are two identification models: the probabilistic model and the two-state model. In the first model,192 a probabilistic quantitative characteristic is attributed to each stimulus, about which the subject must say whether or not he has encountered it previously. If it exceeds a specific threshold, the stimulus is identified as having been encountered

JULY–AUGUST 2006

79

previously. In the second model, each of the stimuli is assigned only a plus or minus.193 In this case any mistakes (10 percent during retrieval of words, 12 percent for sentences, and 2 percent for pictures, according to Shepard’s data) are explained not by movement of the threshold up or down, but by a simple statistically determined selection error. According to Shepard, the “twostate” model has greater experimental validity. As it applies to immediate memory, this is probably the case. The first type of model, however, appears to be better suited for operating on words using permanent memory. Staying for the moment with immediate memory, during the generation of utterances we have, on the one hand, a set of “old” stimuli, and on the other, a sequence of “new” ones. During speech generation there is a comparison of both of these, the separating out of the “old” stimuli from “new” and the assignment of specific linguistic characteristics to them. The process of such a comparison, as demonstrated by Sternberg’s data in the work cited above, is extremely fast—up to thirty units per second. Keeping in mind that the pronunciation rate is approximately one to four words per second,194 and each word carries no more than five or six syntactically relevant grammatical features, this should be more than enough. But it would be interesting to see whether or not “disagreement” of speech occurs given a sharp acceleration of its rate in relation to the grammatical “load.” Up to now we have been dealing primarily with grammatical “requirements,” and we have not touched on the essence of operating with semantic units. If we now turn to this aspect of the question, then the most important thing will be that in speech memory we are dealing not with actual semantic features, but with associative features.195 Here it is appropriate to emphasize that since we are not providing an actual linguistic analysis but are working with speech processes, we are dealing only with associations. Any kinds of “semantic fields,” “semantic groups,” and so on, determined on the basis of a conscious comparison of lexical meanings, are irrelevant to speech mechanisms. In speech activity, we encounter only what T. Slama-Cazacu fortuitously labeled “dynamic structurization” of meanings,196 that is, the labile, individual, dynamic interlinking of directed processes associated with the reproduction in speech of elements of the so-called lexicon197 that is subject to the influence of the most varied subjective and objective factors. We will return to this question later. Notes
[Bibliographic information is incomplete in the original.—Ed.] 1. L.S. Vygotskii [Vygotsky], Myshlenie i rech’, p. 368. 2. Ibid., p. 364. 3. Ibid., p. 375.

80 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

4. See. L.I. Podol’skii [Podolskii], “O vzaimovliianii vnutrennei i vnseshnei rechi”; B.G. Anan’ev, K teorii vnutrennei rechi v psikhologii. 5. A.N. Raevskii, Psikhologiia rechi v sovetskoi psikhologicheskoi nauke za 40 let, p. 45. 6. A.R. Luriia [Luria], Travmaticheskaia afaziia, p. 99. 7. A.R. Luriia, Problemy i fakty neirolingvistiki, pp. 210–11. In one recent publication by A.R. Luria’s colleague, L.S. Tsvetkova, it is explicitly stated that Vygotsky also viewed inner speech “as a convoluted program of coherent oral utterances” (L.S. Tsetkova, K neiropsikhologicheskomu analizu tak nazyvaemoi dinamicheskoi afazii. Soobshchenie 1, p. 162). This assertion is very plausible, but needs specific evidence. In general, Vygotsky should be read “analytically”—many of his ideas for one reason or another lie “beyond” the text and demand additional thought by the reader. This is especially important to underscore that it has now become fashionable to cite Vygotsky (as well as N.A. Bernstein), but these citations, as a rule, involve random catchy formulations, which have often been misinterpreted. 8. A.R. Luriia and L.S. Tsvetkova, O nekotorykh neiropsikhologicheskikh mekhanizmakh rechevogo vyskazyvaniia], p. 2. Consider as well A.R. Luriia and L.S. Tsvetkova, K voprosu o narushenii vyskazyvaniia pri lokal’nykh porazheniiakh mozga. 9. See T.V. Ryabova, “Mekhanizm porozhdeniia rechi po dannym afaziologii,” p. 87. 10. D. McNeill, Developmental Psycholinguistics, p. 52. Evidently, independent of Luria, N.I. Zhinkin arrived at the same idea of an inner program of speech action (which he called zamysel [conception]) (see, for example, Vnutrennie kody iazyka i vneshnie kody rechi, p. 2375). 11. H.H. Clark, “The Prediction of Recall Patterns in Simple Active Sentences,” p. 103. 12. W. Gutjahr, “Zur Psychologie des sprachlichen Gedächnisses,” p. 76. 13. See B. Anderson, “The Short-Term Retention of Active and Passive Sentences”; E.B. Coleman, “Learning of Prose Written in Four Grammatical Translations”; see also H.H. Clark, “Some Structural Properties of Simple Active and Passive Sentences.” 14. See A.A. Leont’ev [Leontiev], Slovo v rechevoi deiatel’nosti, pp. 203–4. 15. A.J. Compton, “Studies on the Psychological Correlates of Linguistic Theory.” 16. Harvard University, The Center for Cognitive Studies . . . , pp. 28–29. 17. See D. Wors [Worth], “Transformatsionnyi analiz konstruktsii s tvoritel’nym padezhom v russkom iazyke”; R. Ruzhichka, “O transformatsionnom opisanii tak nazyvaemykh bezlichnykh predlozhenii . . .” 18. Among the most important publications, see: S.K. Shaumian [Saumjan] and P.A. Soboleva, “Applikativnaia porozhdaiushchaia model’”; S.K. Shaumian [Saumjan], Strukturnaia lingvistika. Initial information about the features of the model is provided by Iu.D. Apresian (see Idei i metody sovremennoi strukturnoi lingvistiki, pp. 223–32). 19. Shaumian, Strukturnaia lingvistika, p. 184. 20. In a recent article, however, F. Jüttner demonstrates that the calculation of transformations in the applicative model is far from faultless (“Zum Transformationenkalkul bei S.K. Saumjan”). We will not, however, involve ourselves with this polemic. 21. A.K. Zholkovskii and I.A. Mel’chuk, “O vozmozhnom metode i instrumentakh semanticheskogo sinteza,” p. 24. 22. A. Sestier, “Sur la nécessité et la possibilité. . . .” In the book by I.A. Mel’chuk

JULY–AUGUST 2006

81

and R.D. Ravich, Avtomaticheskii perevod, where we found a reference to this article, this thought is marked with the letters NB. 23. L.V. Bondarko and L.R. Zinder, “Differentsial’nye priznaki fonem i ikh fizicheskie kharakteristiki,” p. 37. See also A.M. Liberman et al., “Some Observations on the Efficiency of Speech Sounds.” 24. Regarding this point of view, see also A.A. Leont’ev [Leontiev], Slovo v rechevoi deiatel’nosti, pp. 105–8, and Psikholingvistika i problema funktsional’nykh edinits rechi. 25. R. Jakobson, The Role of Phonic Elements in Speech Perception. 26. E.N. Lenneberg, “Understanding Language Without Ability to Speak: A Case Report.” 27. See A.N. Leont’ev [Leontiev], “O mekhanizme chuvstvennogo otrazheniia”; V.P. Zinchenko, “Teoreticheskie problemy psikhologii vospriiatiia.” 28. O.V. Ovchinnikova, “Opyt formirovaniia zvukovysotnogo slukha,” p. 18. 29. Zinchenko, “Teoreticheskie problemy,” pp. 252–53. 30. E.I. Isenina rather convincingly demonstrated that during speech perception there is no constant unit of decision (“Differentiation and Recognition as Mechanisms of Phonemic Hearing” [Razlichenie i uznavanie kak mekhanizmy fonematicheskogo slukha], pp. 8, 21). Also compare the conclusion reached by R.G. Piotrovskii on a completely different basis: “over the course of speech communication the very same linguistic units are constantly changing their signification both in the semanticpragmatic sense and in the syntactic (statistical-probability) sense” (Informatsionnye izmereniia iazyka, p. 100). N.I. Zhinkin has written extensively about the difference between differentiation and recognition as they pertain to speech. 31. Zinchenko, “Teoreticheskie problemy,” pp. 254–56. 32. M. Halle and K.N. Stevens, “Speech Recognition: A Model and Program for Research.” 33. Dzh. [G.] Miller, Psikholingvisty, p. 251. 34. Halle and Stevens, “Speech Recognition,” p. 610. 35. Miller, Psikholingvisty, p. 251. 36. N. Chomskii, “Ob”iasnitel’nye modeli v lingvistike,” p. 268. 37. N. Chomskii, “Logicheskie osnovy lingvisticheskoi teorii,” p. 569. 38. See S.M. Ervin-Tripp and D.J. Slobin, Psycholinguistics, pp. 448–49. 39. P. Lieberman, “Some Effects of Semantic and Grammatical Context . . . ,” p. 185. 40. G.A. Miller and S. Isard, “Some Perceptual Consequences of Linguistic Rules,” p. 217. 41. Similar work is being conducted in the Russian Language Research and Methodology Center [Nauchno-metodicheskii tsentr russkogo iazyka] (Moscow) under our direction. 42. E. Lenneberg, Biological Foundations of Language, p. 106 (emphasis added). 43. N. Chomsky, “A Review of B.F. Skinner’s Verbal Behavior,” p. 563. 44. Ibid., p. 577. Criticism of this work in a review by Y. Bar-Hillel (“The Structure of Language,” pp. 44–45), in our view, is not sufficiently well-founded, although we in no way share Chomsky’s understanding, as is well known. 45. J.J. Katz, The Philosophy of Language, p. 269. 46. N. Chomsky, Aspects of the Theory of Syntax, p. 28. 47. Ibid., p. 25. 48. M. Braine, “The Ontogeny of English Phrase Structure.” As McNeill indicates, an analogous distinction is introduced by S. Ervin and R. Brown (corresponding to “operators” and “modifiers”).

82 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

49. D. McNeill, Developmental Psycholinguistics, p. 36. 50. Ibid., p. 45. Another work by McNeill (“The Creation of Language”) is specifically devoted to the question of baseline relations. 51. E.H. Lenneberg, The Biological Foundation of Language. 52. E.H. Lenneberg, “The Natural History of Language,” p. 246. 53. Bar-Hillel, “The Structure of Lanugage,” p. 45. 54. E.H. Lenneberg, “The Capacity for Language Acquisition,” p. 583 ff. Overall, a serious “biological,” or rather, physiological analysis is provided by Lenneberg only in regard to articulation. When he discusses semantics and grammar, he essentially repeats truisms from transformational theory. It is especially noteworthy that Lenneberg absolutely ignores the new quality involved in the transition between animal and human worlds, specifically the social nature of the latter. See below. 55. See A.N. Leont’ev, “Biologicheskoe i sotsial’noe v psikhike cheloveka,” and “Ob istoricheskom podkhode k izucheniiu psikhiki cheloveka.” 56. See A.N. Leont’ev, “Ob istoricheskom podkhode.” 57. See A.A. Leont’ev, Slovo v rechevoi deiatel’nosti, ch. 1. 58. See, for instance, “Certain thoughts are expressed in external speech only because they are preliminarily expressed in a verbal way in inner speech” (B.V. Beliaev, Ocherki po psikhologii obucheniia inostrannym iazykam, p. 92). 59. See A.A. Leont’ev, “Obshchestvennye funktsii iazyka i ego funktional’nye ekvivalenty,” p. 102. 60. A.N. Leont’ev and D.Iu Panov, “Psikhologiia cheloveka i tekhnicheskii progress,” p. 415. 61. F. Kainz, Psychologie der Sprache, vol. 3, p. 112. 62. Ibid., p. 113. 63. B.F. Skinner, Verbal Behavior, ch. 3–5. 64. C.C. Fries, The Structure of English. 65. J.B. Carroll, Language and Thought, p. 24. 66. A.R. Luriia, Slovesnaia sistema vyrazheniia otnoshenii, p. 33. 67. Ibid., p. 34. 68. Ibid., p. 37. 69. G. Svedelius, L’analyse du langage. 70. A.R. Luriia, Problemy i fakty neirolingvistiki, p. 208. 71. Ibid., p. 209 (emphasis in original). 72. V.A. Bogoroditskii, Obshchii kurs russkoi grammatiki, p. 208. 73. Ibid., pp. 210–11. 74. G.A. Zolotova, “O structure prostogo predlozheniia v russkom iazyke.” 75. V.N. Voloshinov, Marksizm i filosofiia iazyka, p. 114. 76. Ibid. 77. Compare the analysis of this problem in the book A.A. Leont’ev, Slovo v rechevoi deiatel’nosti, p. 120 ff., as well as in ch. 4 of this work [not translated here]. 78. See [A.A.] Leont’ev, “Obshchestvennye funktsii iazyka.” 79. A.A. Kholodovich, “O tipologii rechi.” 80. In this connection, compare the distinction between “axial” and “retial” communication in A.A. Brudnyi, “O nekotorykh prilozheniiakh teorii informatsii.” 81. See, for instance, A.A. Leont’ev, “Teoriia rechevoi deiatel’nosti i problemy obucheniia russkomu iazyku,” Article I. 82. See A.A. Leont’ev, “Ob”ekt i predmet psikholingvistiki i ee otnoshenie k drugim naukam o rechevoi deiatel’nosti,” p. 31.

JULY–AUGUST 2006

83

83. On the physiological mechanisms of dominant motivation, see P.K. Anokhin, Kibernetika i integrativnaia deiatel’nost’ mozga, pp. 6–9. The concept of motive, generally speaking, is more complex than it appears to Anokhin. In particular, human needs are typically objective, mediated by objects; in the example above, the motive will be a vision of bread or the concept of it. “A motive is an object that meets one need or another and that, in one form or another, is reflected by the subject and guides his activity” (A.N. Leont’ev, “Potrebnosti, motivy i soznanie,” p. 5). 84. Anokhin, Kibertnetika, p. 9 (emphasis in original). As N.I. Zhinkin pointed out that it would be more correct here to talk not about “choice of action,” but about “conditions of a choice of action.” 85. N.A. Bernshtein [Bernstein], Ocherki po fiziologii dvizhenii i fiziologii aktivnosti, p. 290. 86. Ibid. 87. J.M. [I.M.] Feigenberg, “Veroiatnostnoe prognozirovanie i prednastroika k deistviiam . . . ,” pp. 127–28. For a detailed analysis of the possible physiological mechanisms of such prognosis, see also Feigenberg, “Veroiatnostnoe prognozirovanie v deiatel’nosti mozga.” 88. Feigenberg, “Veroiatnostnoe prognozirovanie,” pp. 130–31. See also Feigenberg and V.M. Levi, “Veroiatnostnoe prognozirovanie i eksperimental’noe issledovanie ego pri patologicheskikh sostoianiiakh”; B.V. Zeigarnik, Narusheniia myshleniia u psikhicheski bol’nykh. 89. A.R. Luriia, Vysshie korkovye funktsii cheloveka, p. 236. 90. Ibid., p. 244. 91. Ibid., pp. 248–49. See also A.R. Luriia, Mozg cheloveka i psikhicheskie protsessy, pp. 430–32 (where an association between the frontal lobe and a programming mechanism is proposed [not only at the stage of probability prognosis, but during subsequent stages as well]). 92. See, for instance, I.M. Gel’fand [Gelfand], V.S. Gurfinkel’ [Gurfinkel], and M.L. Tsetlin, “O taktikakh upravleniia slozhnymi sistemami v sviazi s fiziologiei”; Gel’fand and Tsetlin, “O matematicheskom modelirovanii mekhanizmov tsentral’noi nervnoi sistemy.” 93. I.T. Bzhalava, Psikhologiia ustanovki i kibernetika, p. 76. 94. Bernshtein, Ocherki po fiziologii dvizhenii, p. 291. 95. We will remind the reader in this connection of Miller, Pribram, and Galanter’s idea of the TOTE unit (see ch. 1, pp. 26–27) [not translated here]. 96. Compare A.A. Leont’ev, “Vneiazykovaia obuslovlennost’ rechevogo akta . . .”. 97. See P.V. Chesnokov, Logicheskaia fraza i predlozhenie and Osnovnye edinitsy iazyka i myshleniia. 98. Zh. Piazhe [J. Piaget], Psikhologiia, mezhdistsiplinarnye sviazi i sistema nauk [Psychology, Interdisciplinary Relations and the System of Sciences], p. 26. 99. A.A. Potebnia, Iz zapisok po russkoi grammatike, p. 70. The Soviet philosopher E.V. Il’enkov [Ilyenkov] points to the differences between forms of thought and forms of logical knowledge (see “K istorii voprosa o predmete logiki kak nauki”). 100. A.A. Potebnia, Iz zapisok, p. 68. 101. See, for example G.P. Shchedrovitskii and N.G. Alekseev. The principle of “the form and content of thought.” 102. K. Marks [Marx] and F. Engel’s [Engels], Nemetskaia ideologiia, p. 29. 103. L.S. Vygotsky stated in this connection that “speech is a correlate of consciousness, and not of thought” (L.S. Vygotskii, Problema soznaniia, p. 157).

84 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

104. A.A. Shakhmatov, Sintaksis russkogo iazyka, p. 20. 105. Ibid., p. 28. 106. See V. Oppel’, “Nekotorye osobennosti ovladeniia rebenkom nachatkami gramoty,” pp. 59–60. 107. Compare in this connection the idea of “the grammar of thinking,” which A. Gel’b contrasts with actual grammar. 108. On this subject, see V. Mathezius, “O tak nazyvaemom aktual’nom chleneniia predlozheniia”; K.G. Krushel ’nitskaia, “K voprosu o smyslovom chlenenii predlozheniia”; O.A. Lapteva, Chekhoslovatskie raboty poslednikh let; I.P. Raspopov, Aktual’noe chlenenie . . .], and so on. 109. K. Pala, “O nekotorykh problemakh aktual’nogo chleneniia,” p. 87. 110. See, for example, B.F. Baev, “Protsess obshcheniia i vnutrenniaia rech’,” pp. 313–14; B.V. Beliaev, Ocherki po psikhologii obucheniia inostrannym iazykam, p. 92, and others. Surprisingly, even in a fundamental work like P.K. Anokhin’s monograph (see Biologiia i neirofiziologiia uslovnogo refleksa), the example provided by the author on p. 249 reveals this confusion. 111. A.N. Sokolov, “Dinamika i funktsii vnutrennei rechi,” p. 178. 112. See Sokolov, “Vnutrenniaia rech’ i ponimanie,” p. 68. 113. Ibid., p. 135. 114. In this connection, compare L.S. Vygotskii (“Problemy soznaniia,” pp. 63 and 65): “There are two ways of structuring syntax: in terms of meaning and physically. . . . The grammar of speech does not correspond to the grammar of thought.” 115. And this occurred “as a result of: (1) the generalization of mental acts and the formation on this basis of speech and thought stereotypes characteristic of ‘compacted reasoning,’ (2) the substitution of other speech components for the speech-motor component (substituting auditory components while listening to speech and visual components during reading), (3) the appearance of visual components in thought” (A.N. Sokolov, “Dinamika i funktsii vnutrennei rechi,” p. 178). 116. N.I. Zhinkin, “O kodovykh perekhodakh vo vnutrennei rechi,” p. 36. 117. M.S. Shekhter, “Ob obraznykh komponentakh rechevogo myshleniia.” 118. A. Einstein, letter to Jacques Hadamard, p. 28. 119. N.I. Zhinkin, “Issledovanie vnutrennei rechi,” p. 121. It seems to us that N.I. Zhinkin is too decisive in generalizing the data of his experiment, applying them beyond the scope of the experimental situation they studied. The basis for such a generalization remains to be proved. 120. See A.N. Leont’ev, “Psikhologicheskie voprosy soznatel’nosti ucheniia,” pp. 25–34, Problemy razvitiia psikhiki, pp. 223–27, and “Potrebnosti, motivy i soznanie”; N.G. Morozova “O ponimanii teksta”; A.A. Leont’ev, Slovo v rechevoi deiatel’nosti, pp. 168–74. 121. A.A. Brudnyi, “K probleme semanticheskikh sostoiannii,” p. 6. We do not feel it is possible, however, to agree with some of the author’s subsequent ideas. 122. A.R. Luriia, Vysshie korkovye funktsii cheloveka, p. 180. Additionally, one encounters (associated with so-called frontal lobe syndrome) defects in the analysis of pictures. A common occurrence in such cases would be the selection of random elements in a picture and a resulting inappropriate understanding of the situation as a whole. But this case lies outside the scope of this work. 123. O. Fujimura, “Some Remarks on the Analysis-by-Synthesis Thesis as a Model of Speech Perception.” We would like to express our deepest gratitude to K. Stevens, who was kind enough to acquaint us with the text of this paper, which was prepared

JULY–AUGUST 2006

85

for a seminar on speech formation and perception (Leningrad, August 1966), but not read due to the absence of the author. 124. S.D. Katsnel’son, “Soznanie–myshlenie–iazyk,” p. 13; compare as well Katsnel’son’s “Porozhdaiushchaia grammatika i printsip derivatsii,” and I. Bellert, “On the Problem of a Semantic Interpretation.” Compare also certain ideas expressed (unfortunately, in very abbreviated form) by T. Moore, “The Topic-Comment Function.” 125. On the relationship between the word and the actual image in this context, compare A. Staats, “Emotions and Images in Language.” After the book was submitted for publication the collection Psikhologicheskie issledovaniia [Psychological Studies] was released, including an article by A.N. Leontiev and Iu.B. Gippenreiter that raises the question of a connection between postexpositional activity in the visual system and speech activity. The authors write, in particular, “The process of verbal signification in identification is rightly understood not as a special process, separate from perception and subsequent to the processing of its product through thought, but as a process incorporated in the activity of perception itself. In terms that we have adopted we can in this case talk about a linguistic level of working in the visual system. Identification, that is, the actual perception of the object, necessarily demands the correlation of pre-information received with a reference that is maintained in man in generalizing systems that have a linguistic basis” (“O deiatel’nosti zritel’noi sistemy cheloveka,” p. 19). If this is truly the way things are, then the hypothesis that we have just stated has been given a real basis. However, in our opinion, here we can speak only of a “linguistic basis,” but likely not of a “linguistic level”—this is more likely the level of the code of inner speech. 126. Slovo v rechevoi deiatel’nosti, pp. 201–2. 127. L.S. Tsvetkova, “Narushenie analiza literaturnogo tekstva.” 128. W. Gutjahr, “Zur Psychologie des sprachlichen Gedächtnisses,” vol. 2, p. 59. 129. A.N. Sokolov, “Vnutrenniaia rech’ i ponimanie,” p. 123. 130. Ibid., p. 124. 131. A.R. Luriia, “Slovesnaia sistema vyrazheniia otnoshenii,” p. 34. 132. See I.A. Zimniaia, “Nekotorye psikhologicheskie predposylki modelirovaniia rechevoi deiatel’nosti,” p. 167. 133. J. Jaffee, S. Feldstein, and L. Cassotta, “A Stochastic Model of Speaker Switching in Natural Dialogue.” 134. Z.A. Kochkina, “Nekotorye osobennosti deiatel ’nosti sinkhronnogo perevodchika,” p. 109. 135. M. Tsvilling, “Sinkhronnyi perevod kak ob”ekt eksperimental’nogo issledovaniia,” pp. 91–92. 136. Ibid., p. 93. 137. L.S. Barkhudarov, “Obshchelingvisticheskoe znachenie teorii perevoda,” p. 11. 138. I.I. Revzin and V.Iu. Rozentsveig, Osnovy obshchego i mashinnogo perevoda, p. 68. 139. B.V. Beliaev, “Psikhologicheskii analiz protsessa iazykovogo perevoda,” pp. 168, 169. 140. I.V. Karpov, “Psikhologicheskaia kharakteristika protsessa ponimaniia i perevoda uchashchimisiia inostrannykh tekstov,” p. 72. 141. A.A. Leont’ev, Slovo v rechevoi deiatel’nosti, p. 213. 142. See A.A. Leont’ev [review], “B.V. Beliaev. Ocherki po psikhologii,” pp. 88–89. 143. T.V. Ryabova, “Mekhanizm porozhdeniia rechi po dannym afaziologiia.” 144. Rech’. Artikuliatsiia i vospriiatie, p. 6.

86 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

145. L. Kaiser, “Biological and Statistical Research.” 146. Rech’. Artikuliatsiia i vospriiatie, p. 84. 147. Ibid., p. 86. 148. Ibid., p. 96. 149. G. Fairbanks and N. Guttman, “Effects of Delayed Auditory Feedback Upon Articulation.” 150. Rech’. Artikuliatsiia i vospriiatie, p. 153. 151. Here there is an association with the model proposed by A. Sestier (see note 22), and with the ideas of other authors. We will return to them. 152. A.R. Luriia, “Osnovnye iavleniia i vidy pamiati,” p. 7; compare “Memory is not a separate compartment, but a process” (W. Grey Walter, Zhivoi mozg [The Living Brain], p. 177). 153. I.M. Tonkonogii and I.I. Tsukerman, “O funktsional’noi strukture operativnoi pamiati,” p. 79. 154. See L.S. Vygotskii, “Razvitie vysshikh psikhicheskikh funktsii,” pp. 258–75; A.N. Leont’ev, Razvitie pamiati. 155. See, in particular, A.A. Smirnov, “Razvitie pamiati” and Problemy psikhologii pamiati; P.I. Zinchenko, Neproizvol’noe zapominanie; P.P. Blonskii, “Pamiat’ i myshlenie” and “Psikhologicheskii analiz pripominaniia,” and others. 156. Within an utterance sequence, this type of memory can also be involuntary. See below. 157. J.M. Feigenberg, “Veroiatnostnoe prognozirovanie v deiatel’nosti mozga,” pp. 61–62. 158. L.P. Kraizmer, “Khranenie informatsii v kiberneticheskikh sistemakh,” pp. 192–93. 159. Compare K. Pribram, “K teorii fiziologicheskoi psikhologii,” and other works. 160. Here and elsewhere in this chapter we rely to a certain degree on the thinking expressed by N.I. Zhinkin. He writes, in particular, “Let us imagine that we have pronounced several sentences and are continuing our story. The question arises: are these sentences preserved in our memory, and if so, in what form? It is obvious that they cannot be preserved in their entirety, word for word. . . . It must then be assumed that the words already pronounced have been transformed into some other signs. This is the code of inner speech. . . . This is the same code in which the planning of the continuation of our story occurs. The semantic mileposts of a story’s development are planned in inner speech, taking into account all that has already been said up to that point” (N.I. Zhinkin, Psikhologicheskie osnovy razvitiia rechi, p. 20. Compare Zhinkin, Vnutrennie kody iazyka i vneshnie kody rechi, p. 2375). 161. See L.S. Vygotskii, “Razvitie vysshikh psikhicheskikh funktsii,” pp. 266–67. 162. Ibid., p. 266. 163. See, in particular, H. Amster, “Semantic Satiation and Generation.” 164. V.B. Shklovskii, “O poezii i zaumnom iazyke,” p. 11. 165. B.V. Tomashevskii, Russkoe stikhoslozhenie. Metrika, pp. 65–66. 166. W. Penfield, “The Nature of Speech,” p. 64. 167. D.E. Broadbent, Perception and Communication. 168. D. Vulldridzh [Wooldridge], Mekhanizmy mozga [The Machinery of the Brain], pp. 257–59. 169. J.L. McGaugh, “Memory Storage Processes,” p. 9. 170. B.F. Lomov, Chelovek i tekhnika, pp. 213–15.

JULY–AUGUST 2006

87

171. See P.B. Nevel’skii, “Pamiat’, informatsiia i deiatel’nost’”; “Sravnitel’noe issledovanie ob”ema kratkovremennoi i dolgovremennoi pamiati”; and “Ob”em pamiati i kolichestvo informatsii.” 172. See, for example, G.A. Miller, “Human Memory and the Storage of Information.” 173. P.I. Zinchenko et al., “Voprosy psikhologii pamiati i teoriia informatsii,” pp. 275–76. 174. Nevel’skii, “Pamiat’, informatsiia i deiatel’nost’,” pp. 222–23. The author is grateful to P.B. Nevel’skii for his stated thinking about the types of memory (first and foremost about the concept of “permanent memory” that is used subsequently). 175. P.I. Zinchenko, “Nekotorye problemy psikhologii pamiati,” pp. 10–11. See also G.B. Repkina, “Ob ob”eme operativnoi pamiati,” p. 135 ff. 176. R. Lorente de No, “Analysis of the Activity of the Chains” and “Macromolecular Specificity and Biological Memory”; U.R. Eshby [W.R. Ashby], Konstruktsiia mozga [The Construction of the Brain]; O. Bureshova and Ia. Buresh, “Fiziologiia neposredstvennoi pamiati.” See also A.R. Luriia, “Fiziologicheskie osnovy pamiati.” 177. Kh. Khiden [H. Hyden], “Kletki-satellity v nervnoi sisteme” [Satellite Cells in the Nervous System]. 178. A.R. Luriia, “Fiziologicheskie osnovy pamiati,” p. 15. W. Grey Walter, who shares these views, writes very well in one of his last books “about the particular value of storing information using an oscillating mechanism. This is not a thing, but a process, not a coin lying on a table, but a candle burning on the altar. Being dynamic, such a means of storing information opens the way to action . . . and having frequency, it can be evaluated numerically. Furthermore, this method allows for the activating of recollection (‘recalling the forgotten’) through other recollections with a similar frequency, although the similarity of frequencies may turn out to be utterly random. Such a mechanism . . . is a possible explanation of the idiosyncrasy and caprice of verbal associations” (Zhivoi mozg, p. 176). 179. Luriia, “Fiziologicheskie osnovy pamiati,” p. 18. 180. Our view also corresponds to the conception of N.A. Bernstein. 181. Compare on this subject: A.S. Novomeiskii, “O vzaimootnoshenii obraza i slova pri zapominanii”; and A.A. Smirnov, Problemy psikhologii pamiati, pp. 391–97. 182. Vygotskii, “Problema soznaniia,” p. 95. 183. W. Szewczuk, Badania eksperymentalne nad rozumieniem zdán. 184. Ibid., p. 215. In a recent work by Clark and Clark (based on an experiment on the storage of the temporal relations between events encoded in different linguistic forms) a shortcoming of the Mehler–Miller model of memory was noted (H.H. Clark and E.V. Clark, “Semantic Distinctions and Memory for Complex Sentences”). 185. Generally in psycholinguistics and in particular in the works of American authors, increasing attention is being paid to the possibility that experience acquired over the course of the experiment could influence its results. Compare, for example, the results of an experiment by Mehler and Carey (J. Mehler and P. Carey, “Role of Surface and Base Structure in the Perception of Sentences”) and in particular those by Salzinger and Eckerman, in which the influence of presentation order on the relative number of mistakes in transforms of nuclear sentences is obvious (K. Salzinger and C. Eckerman, “Grammar and the Recall of Chains of Verbal Responses,” p. 236). The authors themselves associate this phenomenon with the frequency of grammatical structures—without sufficient justification, in our view.

88 JOURNAL OF RUSSIAN AND EAST EUROPEAN PSYCHOLOGY

186. E.L. Ginzburg, V.A. Pestova, and V.G. Stepanov, “O psikhologicheskoi real’nosti operatsii szhatiia,” p. 47. See also E.L. Ginzburg et al., “Eksperimental’nye dannye po vospriiatiiu i zapominaniiu szhatykh i polnykh tekstovykh soobshchenii.” 187. G.V. Eiger and M.M. Gokhlerner [Gochlerner], “O roli ovladeniia priemami.” 188. Ibid. 189. See, for example, “Magicheskoe chislo sem’, plius ili minus dva,” “Human Memory and the Storage of Information,” and others. 190. See G.A. Miller, Langage et communication, pp. 283–84 ; R. Brener, “An Experimental Investigation of Memory Span”; H.B. Reed, “Repetition and Association in Learning.” 191. S. Sternberg, “High-Speed Scanning in Memory.” It makes sense, following H. Buschke’s lead (“Types of Immediate Memory”), to also introduce a distinction between “marking” and “addressing” memory. Immediate memory is a typical example of the former, while permanent memory is typical of the latter. 192. J. Nachimas and S. Sternberg, “An Analysis of the Recognition Process”; I.K. Egan, “Recognition Memory and the Operating Characteristics.” Compare R.N. Shepard, “Recognition Memory for Words, Sentences and Pictures,” p. 161. 193. See R.D. Luce, “A Threshold Theory for Simple Detection Experiment”; see also Nachimas and Sternberg, “An analysis of the Recognition Process,” and R.N. Sherpard, “Recognition Memory for words, Sentences and Pictures.” 194. Compare data from the experiments of L.A. Chistovich (see “Rech’, artikuliatsiia i vospriiatie,” p. 87 and the following). 195. See G.A. Miller and J. Selfridge, “Verbal Context and the Recall of Meaningful Material.” 196. T. [Slama-] Cazacu, “La ‘structuration’ dynamique des significations.” 197. See in this regard A.A. Leont’ev, Slovo v rechevoi deiatel’nosti, pp. 184–86. The reorientation in this sense of certain linguists is interesting, in particular S. Vidlak, “Problema evfemizma na fone teorii iazykovogo polia.” E. Coseriu recently published a major work clearly contrasting the “semantic field” as an abstract-logical concept and the concept of “solidarity” relating to speech (Lexikalische Solidaritäten). See also G.S. Shchur, “Ob assotsiativnykh gruppakh slov.”

To order reprints, call 1-800-352-2210; outside the United States, call 717-632-3535.