Thinking and Speech. Vygotsky 1934

Chapter 1
The Problem and the Method of Investigation

The first issue that must be faced in the analysis of thinking and speech concerns the relationship among the various mental functions, the relationship among the various forms of the activity of consciousness. This issue is fundamental to many problems in psychology. In the analysis of thinking and speech, the central problem is that of the relationship of thought to word. All other issues are secondary and logically subordinate; they cannot even be stated properly until this more basic issue has been resolved. Remarkably, the issue of the relationships among the various mental functions has remained almost entirely unexplored. In effect, it is a new problem for contemporary psychology.

In contrast, the problem of thinking and speech is as old as psychology itself. However, the issue of the relationship of thought to word remains the most confused and least developed aspect of the problem. The atomistic and functional forms of analysis that dominated psychology during past decade led to the analysis of the mental functions in isolation from one another. Psychological methods and research strategies have developed and matured in accordance with this tendency to study separate, isolated, abstracted processes. The problem of the connections among the various mental functions – the problem of their organization in the integrated structure of consciousness – has not been included within the scope of the research.

There is, of course, nothing novel in the notion that consciousness is a unified whole, that the separate functions are linked with one another in activity. Traditionally, however, the unified nature of consciousness – the connections among the mental functions – have simply been accepted as given. They have not been the object of empirical research. The reason for this becomes apparent only when we become aware of an important tacit assumption, an assumption that has become part of the foundation of psychological research. This assumption (one that was never clearly formulated and is entirely false) is that the links or connections among the mental functions are constant and unvarying, that the relationships between perception and attention, memory and perception, and thought and memory are unchanging. This assumption implies that the relationships among functions can be treated as constants and that these constants do not have to be considered in studies that focus on the functions themselves. As we mentioned earlier, the result has been that the problem of interfunctional relationships has remained largely unexplored in modern psychology.

Inevitably, this had a serious impact on the approach to the problem of thinking and speech. Any review of the history of this problem in psychology makes it immediately apparent that the central issue, the issue of the relationship of thought to word, has been consistently overlooked.

Attempts to resolve the problem of thinking and speech have always oscillated between two extreme poles, between an identification or complete fusion of thought and word and an equally metaphysical, absolute, and complete separation of the two, a severing of their relationship. Theories of thinking and speech have always remained trapped in one and the same enchanted circle. These theories have either expressed a pure form of one of these extreme views or attempted to unify them by occupying some intermediate point, by moving constantly back and forth between them.

If we begin with the claim made in antiquity that thought is “speech minus sound,” we can trace the development of the first tendency – the tendency to identify thinking and speech – through to the contemporary American psychologist or the reflexologist. These psychologists view thought as a reflex in which the motor component has been inhibited. Not only the resolution of the problem of the relationship of thought to word but the very statement of the issue itself is impossible within these perspectives. If thought and word coincide, if they are one and the same thing, there can be no investigation of the relationship between them. One cannot study the relationship of a thing to itself. From the outset, then, the problem is irresolvable. The basic issue is simply avoided.

Perspectives that represent the other extreme, perspectives that begin with the concept that thinking and speech are independent of one another, are obviously in a better position to resolve the problem. Representatives of the Wurzburg school, for example, attempt to free thought from all sensory factors, including the word. The link between thought and word is seen as a purely external relationship. Speech is represented as the external expression of thought, as its vestment. Within this framework, it is indeed possible to pose the question of the relationship between thought and word and to attempt a resolution. However, this approach, an approach that is shared by several disparate traditions in psychology, consistently results in a failure to resolve the problem. Indeed, it ultimately fails to produce a proper statement of the problem. While these traditions do not ignore the problem, they do attempt to cut the knot rather than unravel it. Verbal thinking is partitioned into its elements; it is partitioned into the elements of thought and word and these are then represented as entities that are foreign to one another. Having studied the characteristics of thinking as such (i.e., thinking independent of speech) and then of speech isolated from thinking, an attempt is made to reconstruct a connection between the two, to reconstruct an external, mechanical interaction between two different processes.

For example, a recent study of the relationship between these functions resulted in the conclusion that the motor processes associated with speech play an important role in facilitating the thinking process, in particular, in improving the subject’s understanding of difficult verbal material. The conclusion of this study was that inner speech facilitates the consolidation of the material and creates an impression of what must be understood. When inner speech was included in the processes involved in understanding, it helped the subject to sense, grasp, and isolate the important from the unimportant in the movement of thought. It was also found that inner speech plays a role as a facilitating factor in the transition from thought to overt speech.

As this example illustrates, once the researcher has decomposed the unified psychological formation of verbal thinking into its component elements, he is forced to establish a purely external form of interaction between these elements. It is its if he were dealing with two entirely heterogeneous forms of activity, with forms of activity that have no internal connections. Those who represent this second tradition have an advantage over those who represent the first in that they are at least able to pose the question of the relationship of thinking to speech. The weakness of this approach is that its statement of the problem is false and precludes any potential for its correct resolution. This failure to state the problem correctly is a direct function of the method of decomposing the whole into its elements, a method that precludes studying the internal relationship of thought to word. The critical issue, then, is method, if we are to deal with the problem successfully, we must begin by clarifying the issue of what methods are to be used in studying it.

The investigation of any mental formation presupposes analysis, but this analysis can take either of two fundamentally different forms. All the failures that researchers have experienced in their attempts to resolve the problem of thinking and speech can be attributed to their reliance on the first of these two forms of analysis. In our view, the second represents the only means available for moving toward a true resolution of this problem.

The first of these forms of analysis begins with the decomposition of the complex mental whole into its elements. This mode of analysis can be compared with a chemical analysis of water in which water is decomposed into hydrogen and oxygen. The essential feature of this form of analysis is that its products are of a different nature than the whole from which they were derived. The elements lack the characteristics inherent in the whole and they possess properties that it did not possess. When one approaches the problem of thinking and speech by decomposing it into its elements, one adopts the strategy of the man who resorts to the decomposition of water into hydrogen and oxygen in his search for a scientific explanation of the characteristics of water, its capacity to extinguish fire or its conformity to Archimedes law for example. This man will discover, to his chagrin, that hydrogen burns and oxygen sustains combustion. He will never succeed in explaining the characteristics of the whole by analysing the characteristics of its elements. Similarly, a psychology that decomposes verbal thinking into its elements in an attempt to explain its characteristics will search in vain for the unity that is characteristic of the whole. These characteristics are inherent in the phenomenon only as a unified whole. When the whole is analysed into its elements, these characteristics evaporate. In his attempt to reconstruct these characteristics, the investigator is left with no alternative but to search for external, mechanical forms of interaction between the elements.

Since it results in products that have lost the characteristics of the whole, this process is not a form of analysis in the true sense of that word. At any rate, it is not “analysis” vis vis the problem to which it was meant to he applied. In fact, with some justification, it can be considered the antithesis of true analysis. The chemical formula for water has a consistent relationship to all the characteristics of water. It applies to water in all its forms. It helps us to understand the characteristics of water as manifested in the great oceans or as manifested in a drop of rain. The decomposition of water into its elements cannot lead to an explanation of these characteristics. This approach is better understood as a means of moving to a more general level than as a means of analysis us such, as a means of partitioning in the true sense of the word. This approach is incapable of shedding light on the details and concrete diversity of the relationship between thought and word that we encounter in our daily lives; it is incapable of following the phenomenon from its initial development in childhood through its subsequent diversification.

The contradictory nature of this form of analysis is clearly manifested in its applications in psychological research. Rather than providing an explanation of the concrete characteristics of the whole that we are interested in, it subordinates this whole to the dictates of more general phenomena. That is, the integral whole is subordinated to the dictates of laws which would allow us to explain that which is common to all speech phenomena or all manifestations of thinking, to speech and thinking as abstract generalities. Because it causes the researcher to ignore the unified and integral nature of the process being studied, this form of analysis leads to profound delusion. The internal relationships of the unified whole are replaced with external mechanical relationships between two heterogeneous processes.

Nowhere are the negative results of this form of analysis more apparent than in the investigation of thinking and speech. The word is comparable to the living cell in that it is a unit of sound and meaning that contains – in simple form – all the basic characteristics of the integral phenomenon of verbal thinking. The form of analysis that breaks the whole into its elements effectively smashes the word into two parts. The researcher concerned with the phenomenon of verbal thinking is then faced with the task of establishing some external mechanical associative connection between these two parts of the integral whole.

According to one of the most important spokesmen of contemporary linguistics, sound and meaning lie unconnected in the word. They are united in the sign, but coexist in complete isolation from one another. It is no surprise that this perspective has produced only the most pathetic results in the investigation of sound and meaning in language. Divorced from thought, sound loses all the unique features that are characteristic of it as the sound of human speech, the characteristics that distinguish it from the other types of sound that exist in nature. As a result of the application of this form of analysis to the domain of verbal thinking, only the physical and mental characteristics of this meaningless sound have been studied, only that which is common to all sounds in nature. That which is specific to this particular form of sound has remained unexplored. As a consequence, this research has not been able to explain why sound possessing certain physical and mental characteristics is present in human speech or how it functions as a component of speech. In a similar manner, the study of meaning has been defined as the study of the concept, of the concept existing and developing in complete isolation from its material carrier. To a large extent, the failure of classic semantics and phonetics has been a direct result of this tendency to divorce meaning from sound, of this decomposition of the word into its separate elements.

This decomposition of speech into sound and meaning has also provided the basis for the study of the development of the child’s speech. However, even the most complete analysis of the history of phonetics in childhood is powerless to unite these phenomena. Similarly, the study of the development of word meaning in childhood led researchers to an autonomous and independent history of the child’s thought, a history of the child’s thought that had no connection with the phonetic development of the child’s language.

In our view, an entirely different form of analysis is fundamental to further development of theories of thinking and speech. This form of analysis relies on the partitioning of the complex whole into units. In contrast to the term “element,” the term “unit” designates a product of analysis that possesses all the basic characteristics of the whole. The unit is a vital and irreducible part of the whole. The key to the explanation of the characteristics of water lies not in the investigation of its chemical formula but in the investigation of its molecule and its molecular movements. In precisely the same sense, the living cell is the real unit of biological analysis because it preserves the basic characteristics of life that are inherent in the living organism,

A psychology concerned with the study of the complex whole must comprehend this. It must replace the method of decomposing the whole into its elements with that of partitioning the whole into its units. Psychology must identify those units in which the characteristics of the whole are present, even though they may be manifested in altered form. Using this mode of analysis, it must attempt to resolve the concrete problems that face us.

What then is a unit that possesses the characteristics inherent to the integral phenomenon of verbal thinking and that cannot be further decomposed? In our view, such a unit can be found in the inner aspect of the word, in its meaning.

There has been very little research on this aspect of the word. In most research, word meaning has been merged with a set of phenomena that includes all conscious representations or acts of thought. There is a very close parallel between this process and the process through which sound, divorced from meaning, was merged with the set of phenomena containing all sounds existing in nature. Therefore, just as contemporary psychology has nothing to say about the characteristics of sound that are unique to the sounds of human speech, it has nothing to say about verbal meaning except that which is applicable to all forms of thought and representation.

This is as true of modern structural psychology as it was of associative psychology. We have known only the external aspect of the word, the aspect of the word that immediately faces us. Its inner aspect, its meaning, remains as unexplored and unknown as the other side of the moon. However, it is in this inner aspect of the word that we find the potential for resolving the problem of the relationship of thinking to speech. The knot that represents the phenomenon that we call verbal thinking is tied in word meaning.

A brief theoretical discussion of the psychological nature of word meaning is necessary for clarifying this point. Neither associative nor structural psychology provides a satisfactory perspective on the nature of word meaning. As our own experimental studies and theoretical analyses will show, the essence of word meaning – the inner nature that defines it – does not lie where it has traditionally been sought.

The word does not relate to a single object, but to an entire group or class of objects. Therefore, every word is a concealed generalization. From a psychological perspective, word meaning is first and foremost a generalization. It is not difficult to bee that generalization is a verbal act of thought; its reflection of reality differs radically from that of immediate sensation or perception.

It has been said that the dialectical leap is not only a transition from matter that is incapable of sensation to matter that is capable of sensation, but a transition from sensation to thought. This implies that reality is reflected in consciousness in a qualitatively different way in thinking than it is in immediate sensation. This qualitative difference is primarily a function of a generalized reflection of reality. Therefore, generalization in word meaning is an act of thinking in the true sense of the word. At the same time, however, meaning is an inseparable part of the word; it belongs not only to the domain of thought but to the domain of speech. A word without meaning is not a word, but an empty sound. A word without meaning no longer belongs to the domain of speech. One cannot say of word meaning what we said earlier of the elements of the word taken separately. Is word meaning speech or is it thought? It is both at one and the same time; it is a unit of verbal thinking. It is obvious, then, that our method must be that of semantic analysis. Our method must rely on the analysts of the meaningful aspect of speech; it must be a method for studying verbal meaning.

We can reasonably anticipate that this method will produce answers to our questions concerning the relationship between thinking and speech because this relationship is already contained in the unit of analysis. In studying the function, structure, and development of this unit, we will come to understand a great deal that is of direct relevance to the problem of the relationship of thinking to speech and to the nature of verbal thinking.

The methods we intend to apply in our investigation of the relationship between thinking and speech permit a synthetic analysis of the complex whole. The significance of this approach is illustrated by yet another aspect of the problem, one that has remained in the background in previous research. Specifically, the initial and the primary function of speech is communicative. Speech is a means of social interaction, a means of expression and understanding. The mode of analysis that decomposes the whole into its elements divorces the communicative function of speech from its intellectual function. Of course, it is generally accepted that speech combines the function of social interaction and the function of thinking, but these functions have been conceptualized as existing in isolation from one another, they have been conceptualized as operating in parallel with no mutual interdependence. It has always been understood that both functions are somehow combined in speech. But traditional psychology left entirely unexplored issues such as the relationship between these functions, the reason that both are present in speech, the nature of their development, and the nature of their structural relationship. This is largely true of contemporary psychology as well.

However, in the same sense that word meaning is a unit of thinking, it is also a unit of both these speech functions. The idea that some form of mediation is necessary for social interaction can be considered an axiom of modern psychology. Moreover, social interaction mediated by anything other than speech or another sign system – social interaction of the kind that occurs frequently in non-human animals for example – is extremely primitive and limited. Indeed, strictly speaking, social interaction through the kinds of expressive movements utilized by non – human animals should not be called social interaction. It would be more accurate to refer to it as contamination. The frightened goose, sighting danger and rousing the flock with its cry, does not so much communicate to the flock what it has seen as contaminate the flock with its fear.

Social interaction based on rational understanding, on the intentional transmission of experience and thought, requires some system of means. Human speech, a system that emerged with the need to interact socially in the labour process, has always been and will always be the prototype of this kind of means. Until very recently, however, this issue has been seriously oversimplified. In particular, it has been assumed that sign, word, and sound are the means of social interaction. As one might expect, this mistaken conception is a direct result of the inappropriate application of the mode of analysis that begins with the decomposition of the whole into its elements. It is the product of the application of this mode of analysis to the entire range of problems related to the nature of speech.

It has been assumed that the word, as it is manifested in social interaction, is only the external aspect of speech. This implied that sound itself can become associated with any experience, with any content of mental life, and consequently, that it can be used to transmit or impart this experience or content to another human being.

A more sophisticated analysis of this problem and of related issues concerning the processes of understanding and their development in childhood has led to an entirely different understanding of the situation. It turns out that just as social interaction is impossible without signs, it is also impossible without meaning. To communicate an experience or some other content of consciousness to another person, it must be related to a class or group of phenomena. As we have pointed out, this requires generalisation. Social interaction presupposes generalization and the development of verbal meaning; generalization becomes possible only with the development of social interaction. The higher forms of mental social interaction that are such an important characteristic of man are possible only because – by thinking – man reflects reality in a generalized way.

Virtually any example would demonstrate this link between these two basic functions of speech, between social interaction and generalization. For example, I want to communicate to someone the fact that I am cold. I can, of course, communicate this through expressive movements. However, true understanding and communication occur only when I am able to generalize and name what I am experiencing, only when I am able to relate my experience to a specific class of experiences that are known to my partner.

Children who do not possess the appropriate generalization are often unable to communicate their experience. The problem is not the lack of the appropriate words or sounds, but the absence of the appropriate concept or generalization. Without the latter, understanding is impossible. As Tolstoy points out, it is generally not the word itself that the child fails to understand but the concept that the word expresses (1903, p. 143). The word is almost always ready when the concept is. Therefore, it may be appropriate to view word meaning not only as a unity of thinking and speech but as a unity of generalization and social interaction, a unity of thinking and communication.

This statement of the problem has tremendous significance for all issues related to the genesis of thinking and speech. First, it reveals the true potential for a causal-genetic analysis of thinking and speech. Only when we learn to see the unity of generalization and social interaction do we begin to understand the actual connection that exists between the child’s cognitive and social development. Our research is concerned with resolving both these fundamental problems, the problem of the relationship of thought to word and the problem of the relationship of generalization to social interaction.

However, in order to broaden our perspective on these problems, we would like to mention several issues that we were not able to address directly in our research, issues that became apparent to us only as we were carrying it out. In a very real sense, our recognition of the significance of these issues is the most important result of our work.

First, we would like to raise the issue of the relationship between sound and meaning in the word. We have not dealt with this issue extensively in our own research. Nonetheless, recent progress on this issue in linguistics seems to relate directly to the problem of analytic methods that we discussed earlier.

As we have suggested, traditional linguistics conceptualized sound as independent of meaning in speech; it conceptualized speech as a combination of these two isolated elements. The result was that the individual sound was considered to be the basic unit of analysis in the study of sound in speech. We have seen, however, that when sound to divorced from human thought it loses the characteristics that makes it unique as a sound of human speech; it is placed within the ranks of all other sounds existing in nature. This is why traditional phonetics has been primarily concerned not with the psychology of language but with the acoustics and physiology of language. This, in turn, is why the psychology of language has been so helpless in its attempts to understand the relationship between sound and meaning in the word.

What is it, then, that is the most essential characteristic of the sounds of human speech? The work of the contemporary phonological tradition in linguistics – a tradition that has been well received in psychology – makes it apparent that the basic characteristic of sound in human speech is that it functions as a sign that is linked with meaning. Sound as such, sound without meaning, is not the unit in which the various aspects of speech are connected. It is not the individual sound but the phoneme that is the basic unit of speech. Phonemes are units that cannot be further decomposed and that preserve the characteristics of the whole, the characteristics of the signifying function of sound in speech. When sound is not meaningful sound, when it is divorced from the meaningful aspect of speech, it loses these characteristics of human speech. In linguistics, as in psychology, the only productive approach to the study of sound in speech is one that relies on the partitioning of the whole into its units, units that preserve the characteristics of both sound and meaning in speech.

This is not an appropriate place for a detailed discussion of the achievements that have been attained through the application of this mode of analysis in linguistics and psychology. In our view, however, these achievements are the most effective demonstration of its value. We have used this method in our own work.

The value of this method could be illustrated by applying it to a wide variety of issues related to the problem of thinking and speech. At this point, however, we can only mention a few of these issues. This will allow us to indicate the potential for future research utilizing this method and to clarify the significance of the method for this whole system of problems.

As we suggested earlier, the problem of the relationships and connections among the various mental functions was inaccessible to traditional psychology. It is our contention that it is accessible to an investigator who is willing to apply the method of units.

The first issue that emerges when we consider the relationship of thinking and speech to the other aspects of the life of consciousness concerns the connection between intellect and affect. Among the most basic defects of traditional approaches to the study of psychology has been the isolation of the intellectual from the volitional and affective aspects of consciousness. The inevitable consequence of the isolation of these functions has been the transformation of thinking into an autonomous stream. Thinking itself became the thinker of thoughts. Thinking was divorced from the full vitality of life, from the motives, interests, and inclinations of the thinking individual. Thinking was transformed either into a useless epiphenomenon, a process that can change nothing in the individual’s life and behaviour, or into an independent and autonomous primeval force that influences the life of consciousness and the life of the personality through its intervention.

By isolating thinking from affect at the outset, we effectively cut ourselves off from any potential for a causal explanation of thinking. A causal analysis of thinking presupposes that we identify its motive force, that we identify the needs, interests, incentives and tendencies that direct the movement of thought in one direction or another. In much the same way, when thinking is isolated from affect, investigating its influences on the affective or purposive aspects of mental life is effectively precluded. A causal analysis of mental life cannot begin by ascribing to thought a magical power to determine human behaviour, a power to determine behaviour through one of the individual’s own inner systems. Equally incompatible with a causal analysis, is the transformation of thought into a superfluous appendage of behaviour, into its feeble and useless shadow.

The direction we must move in our attempt to resolve this vital problem is indicated by the method that relies on the analysis of the complex whole into its units. There exists a dynamic meaningful system that constitutes a unity of affective and intellectual processes. Every idea contains some remnant of the individual’s affective relationship to that aspect of reality which it represents. In this way, analysis into units makes it possible to see the relationship between the individual’s needs or inclinations and his thinking. It also allows us to see the opposite relationship, the relationship that links his thought to the dynamics of behaviour, to the concrete activity of the personality.

We will postpone the discussion of several related problems. These problems have not been the direct object of our research in the present volume. We will discuss them briefly in the concluding chapter of this work as part of our discussion of the prospects that lie before us. At this point, we will simply restate the claim that the method that we are applying in this work not only permits us to see the internal unity of thinking and speech, but allows us to do more effective research on the relationship of verbal thinking to the whole of the life of consciousness.

As our final task in this first chapter, we will outline the book’s general organization. As we have said, our goal has been to develop an integrated approach to an extremely complex problem. The book itself is composed of several studies that focus on distinct though interrelated issues. Several experimental studies are included, as are others of a critical or theoretical nature. We begin with a critical analysis of a theory of speech and thinking that represents the best thought on the problem in contemporary psychology. It is, nonetheless, the polar opposite of our own perspective. In this analysis, we touch on all issues basic to the general question of the relationship between thinking and speech and attempt to analyse these issues in the context of our current empirical knowledge. In contemporary psychology, the study of a problem such as the relationship of thinking to speech demands that we engage in a conceptual struggle with general theoretical perspectives and specific ideas that conflict with our own.

The second portion of our study is a theoretical analysis of data related to the development (both phylogenetic and ontogenetic) of thinking and speech. From the outset, we attempt to identify the genetic roots of thinking and speech. Failure in this task has been the underlying cause of all false perspectives on the problem. An experimental study of the development of concepts in childhood, a study that is composed of two parts, provides the focus for this second part of the investigation. In the first part of this study, we consider the development of what we call “artificial concepts,” concepts that are formed under experimental conditions. In the second, we attempt to study the development of the child’s real concepts.

In the concluding portion of our work, we attempt to analyse the function and structure of the general process of verbal thinking. Theory and empirical data are both included in this discussion.

What unifies all these investigations is the idea of development, an idea that we attempt to apply in our analysis of word meaning as the unity of speech and thinking.