Jakobson, method, and metaphor: a Wittgensteinian critique.
Having identified these essential components of language Jakobson begins looking for further evidence to confirm his claim. He graduates his arguments to facilitate comprehension of the increasingly complex uses to which language is put. He considers, for instance, a normal but controlled use of language in a psychology experiment performed with children. When a group of children were presented with the noun "hut," their verbal responses invariably exhibited metaphoric or metonymic preferences. There was, in other words, no response other than in terms of metaphor or metonymy. This consequence bolsters the findings based on aphasia: the defining poles of language are metaphor and metonymy, and the individual preference (or predilection) determines the priority of one over the other.
Jakobson's theory is brief and concise, and is very much in accord with the concept of a definition prevalent in the natural sciences. But he is not content with these two pieces of evidence; he goes on to consider an uncontrolled use of language. One can gauge the depth of Jakobson's conviction when one notices that in focusing on an uncontrolled use of language he selects an exceptionally creative form--literature. If the theory can handle literature, then unquestionably it must be acceptable. Jakobson considers the literary schools of Romanticism, Realism, and Symbolism and shows how they relate to the metaphoric and metonymic poles. Citing examples of literary products of these schools he shows that they exhibit the same pattern as the aphasia patients and the children in the psychology experiment. From literature it is an easy step to other artistic forms like the cinema; and again one finds the same evidence of metaphoric and metonymic processes at work. Jakobson then makes the larger claim that the same is true of other semiotic systems and suggests that collaborative, interdisciplinary research be undertaken to explore the implications of his discovery fully. (1) Such research could be carried out, he suggests, by experts from the disciplines of linguistics, psychology, psychopathology, poetics, and semiotics (93). The depth of Jakobson's conviction is perhaps best stated in a remark he made in 1980: "A linguistic study of aphasia closely linked to the study of language in general and to poetic language in particular not only contributes to the classification of aphasic disorders, but also to the comprehension of the structure of language and even to the improvement of the methods of poetics" (Jakobson and Pomorska 134).
As an example of what might result from such research and as further evidence for the metaphoric and metonymic poles of language, Jakobson discusses the case of Gleb Ivanovic Uspenskij, a Russian novelist with a strong metonymic bent. Jakobson contends that Uspenskij's writings constitute a case in point of the predominance of metonymy in Realist literature. Uspenskij suffered in his later life from a mental illness that involved a verbal disorder. Since he displayed a metonymical bent in his literary works, one would expect to see this characteristic reflected in his verbal disorder--and that is what exactly happened. Uspenskij was unable to relate his first and patronymic names (Gleb and Ivanovic), a disorder bound up with the metonymical bent. This kind of psycholinguistic analysis of verbal behavior is only one area illuminated by Jakobson's discovery. Other dichotomies in human actions and expressions, too, can be interpreted in this light. For instance, Freud's theory of the structure of dreams, the principles underlying magic and ritual, and the differences between prose and poetry can all be explicated on the model of Jakobson's bipolar division operating in human processes.
In what follows I attempt a Wittgensteinian critique of Jakobson's central claims. I shall begin with certain preliminary objections, especially those that present themselves on a first critical reading of his theory. These objections, however, can be refuted in turn by other preliminary and factual claims that Jakobson or other psycholinguists might make. This ability to offer further supplementary support for his theory is a product of certain conceptual assumptions that Jakobson makes in order to say what he does about language. I will therefore address those assumptions and argue that the conceptual errors Jakobson commits cannot be eradicated, that at best they can be compounded, contradicted, or recanted. The reason they cannot be eradicated is that they authorize his observations and analyses.
Jakobson's investigation of the structure of language begins with aphasia. Victims of aphasia, he says, are unable to use either metaphor or metonymy, depending on the type of aphasia involved. He offers this observation as evidence that metaphor and metonymy are polar opposites, and it is also part of his evidence that they are of fundamental significance to all verbal behavior. This use of aphasia raises some initial questions. How applicable are the conditions of verbal behavior in aphasia patients to the conditions of normal verbal behavior? There is a broad range of language applications in daily life. Is aphasia a deterrent to the use of language in all these aspects or is it limited to certain clinical condition? Although these questions are part of an initial response to his inquiry, it is reasonable to assume that Jakobson takes them into consideration and that they are irrelevant to the matter of metaphor and metonymy.
If we allow for the validity of the application of the concept of aphasia to normal, everyday language, then we must consider the possibility of other diseases having similar implications for language. Are there other diseases affecting verbal behavior, and if so, do they suggest the same linguistic dichotomy of metaphor and metonymy? If abnormal verbal behavior is valid for an investigation of language, then it seems plausible to suppose that all conditions affecting speech should be studied and not just aphasia. This, again, is moot since one can only assume that Jakobson is familiar with various forms of verbal disorders and finds nothing contrary to the suggested dichotomy. Given, however, that the concept of aphasia is applicable to normal use and that there are no other diseases that might refute the findings based on aphasia, then the question arises whether verbal disorder is a symptom or criterion of aphasia. This question may be unanswerable but the point is nevertheless crucial, for we must know if the disturbance of the mechanisms of metaphor and metonymy is synonymous with aphasia or if there are other problems involved that are more fundamental. If there is something more fundamental, then the status of metaphor and metonymy as one of "primal significance and consequence" becomes suspect (Jakobson 93). Here, too, we may conclude that Jakobson either treats this point elsewhere or has taken it into consideration before using aphasia as providing definitive evidence about the nature of language.
The metaphoric and metonymic poles first suggested by the study of aphasia appear in a "well-known psychological test" (90). In this test children are told to give a verbal response to some stimulus. In all cases cited by Jakobson, the stimulus is the noun "hut." In every case the response can be characterized as either metaphoric or metonymic. The results of this test are used to substantiate the findings first discovered with aphasia--that metaphor and metonymy are opposite polar determinants of language. Jakobson's interpretation of the data from the experiment seems to me problematic. There is an implicit assumption here that some dichotomy is present in the responses of the subjects. Once this assumption is made, it seems plausible that the dichotomy revealed is that between metaphor and metonymy, but not necessarily to the exclusion of other possibilities. Even if we agree that the responses can or should be dichotomized, why not classify them according to some other scheme? If one were interested in the emotional foundation of linguistic expression, one could postulate the poles of positive and negative emotional response. In this case the child who responded that "a hut is a poor little house" might be considered pessimistic, while the child that responded that "a hut is a palace" would be optimistic. The same sort of categorizing could be done in terms of perception or the logical capacity of the subject, and we would say that the common factor in all of these interpretations is the interest of the interpreter and the assumption that there is a bipartite dichotomy involved.
Other evaluations of the data could postulate multipartite divisions in the language-structure, or for that matter none at all. The entire experiment as it is set up could be used, for instance, as an illustration of underdeveloped language concepts in children (perhaps in comparison to a group of adults presented with the same stimulus). Although the range of possibilities is rather broad and depends on the context in which the data is viewed, Jakobson infers the particular one he is looking for. We might also question the use of the word "test" in connection with the event. A test is generally an examination of an agreed-upon quality or condition. In this case the agreement would be upon the verbal predilections of children in a bipolar language scheme, but this requires that we grant the validity of the language scheme in the first place. An experiment, on the other hand, is an effort to collect data that will support (or refute) a hypothesis. This latter objection is no doubt quibbling and the former is probably irrelevant, since Jakobson must have had sufficient reasons for deciding on the metaphor/metonymy scheme, and he surely considered all the interpretive possibilities before favoring the one he did.
In any event, there is a more fundamental problem here than vague suspicions of a biased interpreter--the suspicion of a biased experiment. The structure and implementation of the "test" is both rigid and restricted. In terms of structure, the model presented by the experiment may not be applicable to language as it is normally used. Speech is seldom restricted to two people sitting in a room with one of them responding verbally to some stimulus. Language in a formal situation frequently differs from language in casual circumstances. Jakobson does not take into account the fact that there is an artificial context of authority in his experiment and that the impact of authority may be more pronounced on children than on adults.
Within the restrictive structure of the experiment there is a further limitation due to the uniformity of stimulus. In every case the subjects were asked to respond to the noun "hut." This seems to ignore the possibility that different types of words may elicit different types of responses. Just within the category of nouns one could probably find stimuli that allow a greater range of response, let alone other categories like verbs or adjectives. An extended range of stimulus-response might also be achieved if the stimulus included more than one word, or even a full sentence or for that matter an entire paragraph. The point of such an experiment would be that there are many facets to language which the "test" Jakobson devises does not consider. We also use language in response to nonverbal stimulus from others ("You look sick, are you feeling well?") or to sounds made by others that are not considered language ("ugh!" implying "I know just how you feel!"). It seems that stimulus, interpretation, and context (if nothing else) influence our responses to other people. There are even uses of language in which we are, in the absence of others, like a child playing with dolls. Are all these cases to be construed as indicators or confirmations of metaphoric/metonymic predilection in the speaker? These objections--the restrictive structure of the experiment and the limited applications of it--are probably not fatal to Jakobson's theory, and we may grant that he would consider these objections rather primitive, since it is possible to argue that he is citing the experiment as incidental evidence for his theory rather than hard proof of it.
Jakobson's treatment of metaphor and metonymy is presented, however, as a discovery of the processes in the mind that prefigure language. For him, then, these processes constitute the essence of speech, and his theory seeks to explicate these processes because their significance extends not just to language but to "human behavior in general" (93). One is tempted here to counter such a claim by asking about other pre-existing processes. If metaphoric and metonymic processes occur in the absence of a language, why not have positive and negative emotional processes as well? The evidence from the psychology experiment could be interpreted in this way, and there is ample psychological or physiological evidence that could be used to show positive and negative poles of emotional behavior. These poles could be the determinants of language as easily as metaphor and metonymy could be. By postulating fundamental processes one is no closer to proving that metaphor and metonymy are the defining poles of language.
If one argues for internal processes in this fashion, there is a hidden premise, and the premise is exactly what is being concluded. When one trains attention on pre-verbal processes how does one talk about them to determine which is the more fundamental? It cannot be done--the only way to infer metaphoric and metonymic processes as the basis of language over some other pre-verbal process is to presuppose them, to take them as an article of faith. The logic of the argument depends on taking metaphor and metonymy as a given, a hidden premise, and thus Jakobson's argument, whether right or wrong, does not offer support for anything.
Another minor problem that suggests itself in connection with the logical structure of Jakobson's claims based on the psychology experiment involves the treatment of language as a tool. What we are examining, he avers, is something of "primal significance and consequence for all verbal behavior and for human behavior in general" (93). One is left wondering how it is that language can be used to explain something more fundamental than itself. As Wittgenstein might put it, this would be like having a rule that tells you where the rule comes from.
Having considered these initial objections to the use of aphasia and the psychology experiment, we can now turn to Jakobson's analysis of Uspenskij's speech disorder. This disorder affected Uspenskij's use of his first and patronymic names. The name Gleb was invested by Uspenskij with all his virtues, while Ivanovic was ascribed all his vices. According to Jakobson the "linguistic aspect of this split personality is the patient's inability" to use two symbols for the same thing, and it is thus a "similarity disorder" which "is bound up with the metonymical bent" (94).
With Uspenskij's ability to use metaphor impaired, there are two possibilities to consider: (1.) that he could not use metaphor at all, and (2.) that he could not use metaphor only when using his names. It is important to note that Uspenskij is using his name in some way and apparently is able to use language intelligibly. If, in the first case, Uspenskij could use no metaphor at all, then how profound is the role of metaphor in language when he could still be understood? In the second case (where Uspenskij is affected only when using his names) we have to wonder at the metaphoric process being superseded by something else--apparently something more fundamental must be determining his behavior. If Jakobson is right, how can a disturbance in verbal behavior be more fundamental than metaphor that is of "primal significance" for human behavior in general?
These objections lead us to the conclusion that there is something more fundamental at work in verbal behavior than Jakobson's metaphor-metonymy dichotomy. Whether it is pre-existing language-context from which metaphor and metonymy derive their meaning and application or whether it is something only indirectly related to language, Jakobson still presents an unsatisfactory argument. At this point we can return to the study of aphasia with some final objections. If aphasia patients can be understood (in any case) then how important is the role of metaphor and metonymy? Surely their role could not be that of the defining poles of language. If, on the other hand, aphasia patients cannot be understood, then how do we know that they are really using language and how can we apply it to language that we understand? In view of the numerous objections that can be raised against Jakobson's theory, one is compelled to question the integrity of Jakobson's assessment of the evidence from aphasia and literature as well as from the psychology experiment. Do "the varieties of aphasia" really lie "between the two polar types" (90) of metaphor and metonymy, and do literary schools really depend on metaphoric and metonymic processes?
Though the objections raised above might be sufficient on their own to deny the claims made by Jakobson, there is a greater problem involved here than that of Jakobson's factual imprecision. In order to guard against the problems entailed in Jakobson's treatment of language, we heed to examine the assumptions underlying his theory and assess their implications for language when taken to their logical and necessary ends.
One of the assumptions implicit in Jakobson concerns the nature of language--that it is uniform in application. The evidence about language found in the study of an abnormal user, the evidence from the psychology experiment, and the evidence from literature can all be applied to verbal behavior; and indeed Jakobson would have us believe that it all applies in the same way to human behavior in general. This assumption about the nature of language is central to Jakobson's investigation of language. The uniformity of language in its use and as an object of study suggests that language is not function-oriented. If language can be separated from a functional context, then verbal behavior is not necessarily related to the application of the language at any given time. This would be true of behavior in general: whatever governs behavior is independent of a specific context.
There are two initial difficulties here. First, when Jakobson draws on the example of aphasia patients and their use of language, he is using the abnormal case as a standard for the normal use. A more appropriate way to solve the problem presented by aphasia is to look at it through the context provided by the ordinary use of language. How do we know, after all, that these patients are using the language incorrectly? Only because we are already acquainted with the correct use. If we only understood language on the basis of information as provided by this kind of phenomenon, then we would never discover that there was a problem with the patients' usage in the first place. There would also be a problem in justifying the use of evidence from something like the psychology experiment because the use of language in that instance would be independent of the stimulus. Whatever is being measured by the experiment would not be the reaction to the single noun "hut."
If language and behavior are not ruled by the situation then what are they oriented to? Language must be dependent on some nonverbal process that determines the mode and range of response. In Jakobson's experiment, language itself is not the object of study; rather our linguistic predilections are. But having arrived at our natural responsive mode as the determinant of language, one is tempted to ask again just what it is responsive to. Labeling some nonverbal process as the source of linguistic predilection gets us no nearer to understanding the governing process at work. Placing the emphasis on internal processes like this leads one to a model of language as essentially private. Once language is separated from an external and situational context it gravitates toward an internal and private context. Then if language tells us anything at all, it tells us not about the world we inhabit but about the world within us. Jakobson's initial assumptions about language--those he must rely on to accept the evidence he does--can lead to solipsism, on the one hand, and to behaviorism on the other.
Once we have created this essentially private environment for language, our problems are legion. We are forced to consider all articulations as language, and this is borne out by Jakobson's use of aphasia patients. Though he admits that patients suffer from a fundamental disorder of speech, he never doubts that they are using language and that their use of it is applicable to the study of (ordinary) language as such. Since, in his account, language is governed by something private, then anything, even nonsense, can be considered as constituting a species of language. This "anything" could range from completely unfamiliar sounds to words put together in a numerical arrangement. However, as Wittgenstein remarks, commenting on the inability of a private language, "a word stands in need of a justification which everybody understands" (Investigations 1.261). By taking language out of the public realm we lose that justification for what we say--in fact we lose the ability to say anything at all. Wittgenstein says about an essentially private phenomenon that it "is not a something, but not a nothing either! The conclusion was that only a nothing would serve just as well as a something about which nothing could be said" (1.304). And he adds that the "paradox disappears only if we make a radical break with the idea that language always functions in one way, always serves the same purpose: to convey thoughts--which may be about houses, pains, good and evil, or anything else you please." Jakobson seems to treat language in this way: it has one uniform application in any situation, concerning any subject, and is governed by something private.
Aside from his assumption that language is uniform and private, Jakobson also assumes that language is governed by a set of extrinsic rules. This is partly connected to his first assumption and follows from his treatment of language. Usually, extrinsic rules for language are attributed to their having a correspondence with certain "facts" about external reality. For Jakobson, the rules of language (independent of its use) correspond to certain "internal" facts about human processes. Once again the criteria for determining the intelligible and unintelligible use of language become meaningless. This is a refinement of the previous charge that language can be any arrangement of sounds or words. That previous charge was intended to cover cases like "colorless green ideas" or "aargh fump." The refinement here covers the logic of the language, not just the articulated form. One can make statements that seem properly constructed grammatically but which deny the sense and logic of language. "If," as Wittgenstein says, "people always said things only to themselves, then they would merely be doing always what as it is they do sometimes" (1.344). Jakobson would necessarily take this statement as sensible, since it tells us about the internal process or preference of the speaker. Wittgenstein explains such a misapplication of language in the statement that follows:
"What sometimes happens might always happen."--What kind of proposition is that? It is like the following: if "F(a)" makes sense "(x).F(x)" makes sense. "If it is possible for someone to make a false move in some game, then it might be possible for everybody to make nothing but false moves in every game."--Thus we are under a temptation to misunderstand the logic of our expressions here, to give an incorrect account of the use of our words. Orders are sometimes not obeyed. But what would it be like if no orders were ever obeyed? The concept 'order' would have lost its purpose. (1.345)
Jakobson's treatment of language as dependent on extrinsic rules is doubly in error. On the one hand, reliance on such rules violates any application of the language, which can then be treated as separate from pertinent "facts." As a consequence, language can be misapplied in light of those facts, creating factual and conceptual nonsense or leading us to "discoveries" about language and reality that only compound the error. In effect, anything can be said. On the other hand, Jakobson's rules correspond to private processes and thus language has no relation to the "facts" externally. It is only a matter of the self that allows us to make "discoveries" about other "facts." This, too, is a misconstrual of the logic of our expressions, but because of the privacy of language, now in effect, nothing can be said to dispute or question it. Language is thus reduced to meaninglessness in at least two ways: first by assuming extrinsic rules and then by gearing those rules to private phenomena. The result is a ludicrous conception which is highly entertaining but in no way enlightening.
Aside from a trend toward solipsism, there is another trend here that radically opposes the first--behaviorism. This stems from Jakobson's insistent use of a uniform stimulus-response model as a viable means of gathering information about the self. Confronted with a stimulus, the human machine will respond in a given number of ways, and those responses indicate the relative roles of certain private processes fundamental to language. These two assumptions, as pointed out earlier, are mutually exclusive: if language is taken as responsive to external stimuli then the assumption that it is governed by essentially private processes is untenable.
Behaviorism reduces human action to outward measurements of stimulus and response and amounts to a form of skepticism just as much as solipsism does. In a given situation, the behaviorist argues, human beings will respond according to the mechanism of behavioral determinants. Wittgenstein's critique of behaviorism's seemingly innocuous first step here is pertinent: "We talk of processes and states and leave their nature undecided. Sometime perhaps we shall know more about them--we think. But that is just what commits us to a particular way of looking at the matter" (1.308); the problem is that "we have a definite concept of what it means to learn to know a process better." Once we assign an unknown verbal or behavioral determinant of an inner process we are free to interpret the sense of "process" in a misleading way. Here we sublime the logic of the language on the basis of the picture that we acquire from this vague idea of process. Yet the very fact that this "process" is nonverbal indicates that it is not contained in the language we seek to understand (1.115). It is not permissible to speak of it as a process. This is something else about which nothing can be said.
Another conceptual difficulty Jakobson faces is that of reductionism. Reductionism is a way of looking for the essential characteristics of some object or process that one fears is otherwise ill-defined. What Jakobson attempts to do is to define language more fully by finding its essential characteristics. The characteristics he has discovered (by examining a large number of instances of language) amount to metaphor and metonymy, or rather metaphorical and metonymical modes of human behavior. However, if the reduced sense of language does not coincide with its ordinary sense, the argument will not hold. But how do we decide from the enormous variety and complexity of language-use just what the common and essential elements are? Wittgenstein's well-known remark that some concepts defy clearer definition than what they already have is pertinent here: "the degree to which the sharp picture can resemble the blurred one depends on the latter's degree of vagueness" (1.77). Although this remark was aimed at concepts in aesthetics and ethics, it works here too. The sharp picture in our case is Jakobson's proposed concept of language, while the instances of language he observes in his experiment constitute the blurred one. His analysis proceeds in the way it does because he thinks there is something concealed or buried in the language that should be brought out for a clearer view, but, as Wittgenstein would suggest, there is nothing hidden there (1.92).
For argument and analysis to function successfully, one must agree on terms being employed in a given inquiry, and if the sense of the terms changes, then one must acknowledge that the subject of inquiry has accordingly changed. In Jakobson's inquiry the language that is studied is different from the notion of language on which his conclusions rest. He is, in other words, talking about two different things and consequently brings us no closer to understanding either of them. This analytical twist is best illustrated by considering what happens to the senses of metaphor and metonymy. The terms are originally part of a pre-existing linguistic context, one that leads him to a"discovery" about the nature of metaphor and metonymy and their relationship to language. This discovery, however, springs from a contextual misplacement of the terms, for these terms derive their meaning from the language of which they form a part. Jakobson simply reverses this relation and defines language itself in terms of metaphor and metonymy. If metaphor and metonymy are to retain their original meaning, they must retain their original content. When they do not, their meaning is changed. The metaphor and metonymy that Jakobson concludes are the determinants of language are very different from the metaphor and metonymy that originally had their place in language.
In the absence of language it may be possible to make associations or draw analogies between disparate elements in some way, but without articulation, without the structure and meaning imposed on articulation by language, the associations or analogies made cannot be construed as metaphor. The same kind of argument will hold true in the case of metonymy. It is only through articulation that we know and use metaphor and metonymy, but Jakobson claims that these figures of speech actually prefigure speech. When the terms are abstracted from their functional environment we must determine if the same application holds. If this new understanding of metaphor and metonymy does not account for their application in the language of which they are a part, then the "discovery" is only a change in definition. As already stated, however, metaphor and metonymy in their ordinary use are figures of speech; they do not account for pre-verbal or nonverbal processes that determine language. If we insist that metaphor and metonymy are indeed the determinants of language, then again we have a hidden premise because language will not tell where language comes from; the primacy of those processes would have to be taken as an article of faith--but then why argue as though it were an empirical discovery?
As Wittgenstein says, the given is the form of life; he also equates the form of life and language: language must be taken as the given (1.19 and p. 226). If we take language as the given, any assessment of it must account for itself in the game, and this is what Jakobson's language fails to do. If we measure his modified definitions against the given, we find they fall short of the mark.
When Jakobson analyzes aphasia, he arrives at conclusions about the structure of language that are already at work in his analysis. His conclusions about the way language functions in aphasia patients are already given in the systematic presuppositions about language he already holds. Jakobson, however, argues as though he has made some factual discovery about aphasia that also reveals to him the fundamental structuring elements of language. Jakobson, of course, claims to be carrying out a scientific investigation of language, and his study of aphasia is meant only to confirm the view of language he has already reached. The Wittgensteinian view of language underlying my own effort insists on the importance of flexible, analogical, and tentative formulations of questions and answers in exploring language, and that view in turn values informal, intuitive procedures of reasoning which do not culminate in either easily transportable methods or absolutist characterizations of language. A linguistic study of aphasia can certainly be useful for classification of aphasic disorders, and it can be useful to explore peculiar manifestations of language in various conditions of linguistic impairment. It is useful, too, to study peculiarities in the literary works of a writer suffering from aphasia, and thus useful to articulate the poetics underlying the works of such a writer. My discussion of Jakobson here centers on the question of why it should reveal workings of language or literature as a whole. Even if it could reveal certain features of Realism, could it conclusively pin down all forms of Realism? (2)
In Wittgensteinian terms the claim "metaphor and metonymy are the defining poles of language" is not just a move in a language-game. This claim is purported to be empirical, but what would happen if it was? What this suggests is that we are familiar with the concepts of metaphor and metonymy and then we apply them independently to language and make a discovery about it as a result. To make such a discovery we must be familiar with this sense of these concepts beforehand since ordinary language does not supply them and the empirical evidence, while it may be interpreted to indicate processes, cannot reveal their nature. But then, where did we learn these concepts in the first place if not through language itself? Privately? However, as the later Wittgenstein has shown, a private language makes no sense. Thus if we do not learn the concepts in some private fashion and if we must be familiar with them to make the claim that Jakobson makes, we are caught in a logical bind. Apparently the claim cannot be empirical.
Well, then, can it be something other than empirical? The answer is yes--it could certainly be a metaphysical claim. A metaphysical question, as Wittgenstein states in The Blue Book, is an expression of "an unclarity about the grammar of a word in the form of a scientific question" (35). Jakobson's implicit question--are metaphor and metonymy the fundamental poles of language?--is an illegitimate construct in the grammar precisely because of the essential privacy of the inner processes and because language is the given. Although Jakobson does not present this as an empirical question, as shown by his use of empirical evidence, what he has in fact done is to ask a metaphysical question about the nature of language. And the consequence of this is that he misconstrues what Wittgenstein would call the very grammar of our expressions. The role of the philosopher, Wittgenstein said, is "to bring words back from their metaphysical to their everyday use" (Investigations 1.116), and that is just what has been attempted here. This imperative, of bringing words back from their metaphysical use to everyday role, has even greater bearing on exploring certain aspects of contemporary literary theory, one that requires another and far more extensive treatment.
(1) Jakobson, follows the Polish linguist Mikolaj Kruszewki, who observed the existence of the two poles of metaphor and metonymy operating in language, and treats these poles as "fundamental forces acting in language as well as in all forms of art" (Jakobson and Pomorska 123). He made this point as early as 1919 in an essay "Futurism" and returned to it again in essays on the cinema and on Pasternak (125).
(2) See Rene Wellek's criticism: 372-76.
Jakobson, Roman. "Two Aspects of Language and Two Types of Aphasic Disturbances." 1956. Fundamentals of Language. By Jakobson and Morris Halle. 2nd ed, rev. The Hague: Mouton, 1971.69-96.
--, and Krystyna Pomorska, Dialogues. Cambridge: MIT PP, 1980.
Wellek, Rene. A History of Modern Criticism, 1750-1950, vol. 7: German, Russian, and Eastern European Criticism, 1900-1950. New Haven: Yale UP, 1991.
Wittgenstein, Ludwig, The Blue and Brown Books. 1958. Oxford: Basil Blackwell, 1964.
--. Philosophical Investigations. Trans. G. E. M. Anscombe. New York: Macmillan, 1953.
University of Arizona
SURESH RAVAL, Jakobson, Method, and Metaphor: A Wittgensteinian Critique