How Jakobson's essay on aphasia initiated postmodernist deceits
by James Drake

-- from The Times Literary Supplement (pp. 14-15)
Copyright, September 4, 1998

In the sciences, even questionable examples of research fraud are harshly punished; mere suspicion is enough for funding to be cut off; publicity guarantees that careers can be effectively ended. No such mechanisms exist in the humanities. Why not? Because of the obvious: since much of what humanists call research does not lead to results that are replicable, peer review, as a criterion of reliability, is weaker in the humanities than it is in the sciences. Given the importance of interpretation in historical and literary scholarship, scholars in the humanities are in a position where they can explain away deliberate and even systematic distortions. Forgeries which take the form of pastiches in which the forger intersperses fake and real parts can be defended as mere mistakes or aberrant misreading. Scientists fudging data have no such defences.
Humanists have never been unaware of the dangers lying in waiting for interpretative scholarship, but this awareness has dwindled with the advent of what is called postmodernism. Not only is this a form of scholarship which does not facilitate critical consideration of its doctrines, but it is one whose major thrust is redefining empirical evidence so that all attempts to know the truth become forms of interpretation. Abuses which have always been ignored by scholars in the humanities as "stretching the evidence", or apologized for as "contestable interpretations", are now being justified with a metaphysical doctrine attacking objectivity and truth. This doctrine will be portrayed as being itself a truth objectively arrived at by a scholarly consensus.
More than ten years ago, in Not Saussure (1986), Raymond Tallis, after noting the persistent fraudulent air of postmodernism, characterized it as employing for evidence obiter dicta and "knowing allusions to what has been 'established elsewhere"'. When one tries to find what has been "established elsewhere", it usually turns out to mean one of the following: theoreticians interpreted in the light of other theoreticians; debatable hypotheses made the a priori premisses of arguments; interpretations which, through question-begging terminology, presuppose what is being demonstrated; footnotes, without page references, which cite articles and books whose evidence has been established by doing the very same thing; citations limited to other members of a clique; and jargon employed to give cliches and commonplaces an air of profundity.
To understand how scholarship in the humanities has arrived at this impasse, one should think of postmodernism as a huge inverted triangle. Near its apex is the French "revival of rhetoric" and scholarship consequently derived from it. At the very apex of this inverted triangle is an essay by the linguist Roman Jakobson, "Two Aspects of Language and Two Types of Aphasic Disturbances" (reprinted in Language in Literature, 1987) and his appropriation of aphasia, which is recognized as primarily a naming disorder. Jakobson's essay, originally published in 1956, made the study of tropes central to language. He claimed to have demonstrated, from an analysis of the language of aphasia patients, that semantics and syntax could be reduced to derived and direct expressions of similarity and contiguity. Metaphor and metonymy must therefore be the underlying principles of language. Jakobson quotes the nineteenth-century aphasiologist Hughlings Jackson on metaphoric aphasic language: "To say what a thing is, is to say what it is like", Jackson notes. What Jakobson did not note is that the context of this sentence is a discussion not of an aphasia patient, but of a patient suffering from delirium, one who "takes his nurse to be his wife". Nor does Jakobson mention that he has silently omitted the second half of the sentence which makes this clear. The entire essay has been built in this manner: aphasiologists' words are dragged out of context and given a new context which makes them mean the opposite of what originally was said. Much in the manner of later postmodernist scholars, Jakobson also shuffles abstractions, substituting similarity for "con-currence" and contiguity for "concatenation". Metaphor and metonymy are substituted for them in turn and then proclaimed "the general laws of language". What's concealed by this shell game is that the clinical data of aphasiology, then and now, directly contradict it.
In aphasiology Jakobson's essay has been forgotten, but in the humanities Jakobson's inventions played an important role in the generation of French postmodernist thinking. In his influential essay, "Linguistics and Poetics" (1958), Jakobson added the "poetic function" to reference as one of the major functions of language, while he claimed that "any attempt to reduce the sphere of the poetic function to poetry or to confine poetry to the poetic function would be a delusive oversimplification". Aestheticizing language in this manner, it should be noted, implies an ambitious research programme in which linguistics and literary criticism can be used for cultural criticism.
Long before his acquaintance with the work of Jakobson, Roland Barthes had actually been "reading" cultural products in the same way as literary critics "read" a poem. Yet when an advertising agency, impressed by Barthes's work, hired him in the early 1960s to improve car sales, they were already considering advertising with an approach which, in the words of his biographer Louis-Jean Calvet, "owed as much to the work of Jakobson as that of Barthes". One could argue that Jakobson's doctrines merely give linguistics' stamp of approval to the sort of cultural criticism to be found in Barthes's Mythologies (1957). Later, Barthes was to attribute his initiation into linguistics to his friend Julien Greimas and their translating Jakobson's essay together.
Jakobson's influence on French thinkers was principally mediated through Jacques Lacan and Claude Lévi-Strauss. Sean Burke, in The Death and Return of the Author (1992), has described how the effect of Lacan's and Lévi-Strauss's publications, among French philosophers, was the dethroning of phenomenology. The linguist Emile Benveniste's essay absorbing subjectivity into language is also an important link between Jakobson, Barthes and present forms of postmodernism. Like Lacan and Lévi-Strauss, he had personal ties with Jakobson. Most scholars would agree with Burke, however, that had Levi Strauss not become Visiting Professor at the New School for Social Research in New York in the years 1941-5 where he worked with Jakobson, "the development of critical theory may have taken significantly different routes . . .". Structuralists such as Foucault and Althusser were influenced by Lévi-Strauss's adaptations of Jakobson's theories for anthropological research. For example, Lacan's earlier notion of the "Symbolic order" had been primarily derived from Levi-Strauss's anthropological version of underlying laws of the unconscious.
The Lacan who is famous is the one who combines Freud with terminology derived from Jakobson's essay on aphasic language, and who makes remarks such as 'The symptom is a metaphor whether one likes it not, as desire is a metonymy, however funny people may find the idea." Lacan's essay "The Agency of the Letter in the Unconscious or Reason since Freud" (1957) even begins by announcing that it is the works of Roman Jakobson "to which I owe much of this formulation; works to which a psychoanalyst can constantly refer in order to structure his own experience, and which render superfluous the 'personal communications' of which I could boast as much as the next fellow". When Lacan adapts rhetorical figures such as periphrasis, hyperbaton, etc. as "Tropes . . . the most proper for the labelling" of psychological defence mechanisms, Jakobson's reduction of language to tropes is transformed into Lacan's claim that "the unconscious is structured like a language".
In The Savage Mind (1962), Levi-Strauss himself added to his Jakobsonian structuralism a Jakobsonian account of myth in terms of metaphor and metonomy. By the late 1960s, structuralist and post-structuralist ideas derived from Jakobson were available from many sources. According to Barthes's biographer, he was at this time reading Benveniste, Jakobson in the French translation of his work by Nicolas Ruwert, and Jacques Derrida. At just about the same time, Barthes began to use the term "writing", in Derrida's manner, to refer to autonomonous conceptual structures. One can find "discourse", the metaphor/metonymy opposition and Derridean "writing" all in Barthes's famous extravaganza S/Z.
The strong relationship between postmodernist doctrines and Jakobson's ideas can also be found in Derrida's work, particularly in his famous essay, "White Mythology" (1971). The title refers to "Metaphysics -the white mythology which reassembles and reflects the culture of the West: the white man takes his own mythology, Indo-European mythology, his own logos, that is, the mythos of his idiom, for the universal form of that he must still wish to call Reason". One might call this the academic version of a guerrilla action; one undermines the "West" by attacking the belief that there is such a thing as the grounds of knowledge. To do this, Derrida cites a new "attentiveness to metaphorical activity in theoretical or philosophical discussion". Along with references to Nietzsche, we get references to Lacan, Benveniste's Lacanian "Remarks on the Function of Language in Freudian Discovery", and, of course, Jakobson's essay on aphasic language.
For Derrida's purposes, adapting an imperialistic account of figurative language to philosophical activity is not quite radical enough. So he develops an extremely elaborate analysis of Aristotle's notion of metaphor, one which turns what Derrida calls the "Aristotelian tradition" of rhetorical analysis into something which goes beyond Jakobson by claiming that Aristotle's notion of metaphor, in his Poetics, includes not just all of language, but all of thought. When we get to Derrida's claim that "Metaphor is less in the philosophical text . . . than the philosophical text is within metaphor", Jakobson has long since disappeared, and so has the "Aristotelian tradition" of identifying metaphor with words. Of course, these are just props for a larger argument; he concludes by citing Gaston Bachelard and Georges Canguilhem on the use of metaphor in science. According to Derrida, these scholars have demonstrated that the "domain of metaphor is extended even beyond language", Bachelard in particular showing the need for a "psychoanalysis of objective knowledge". Urgently Derrida asks, "Are not all metaphors, strictly speaking, concepts, and is there any sense in setting metaphor against concept?" We are especially urged to ask if one can "transport into the philosophical field the Bachelardian program of a metapoetics". This programme, of course, looks very much like the sort of programme implied by Jakobson in "Linguistics and Poetics" (1958), except that Derrida has now expanded Jakobson's claims for the "poetic function" to all of thought.
Derrida would not be Derrida if he just left it at that, and so he announces that "We come back to our questions: can one transport into the philosophical field the Bachelardian program of a metapoetics?" It should be recalled that Jakobson in his aphasia essay, from which Derrida's own essay takes off, associates metaphor with semantics and metonymy with syntax, for Derrida's discussion, without directly alluding to it, now returns to this theory in order to question allowing meaning to dominate "the play of syntax". Derrida explains that "We have tried to demonstrate above that this subordination of the syntactic was inscribed in the most invariable characteristics of the concept of metaphor, and tried to show elsewhere the essential limits of such a thematism".
Derrida faces the very same problem as Jakobson originally faced in his aphasia essay, for if language in some way is metaphoric, then how can this statement itself be literally true? Instead of adopting Jakobson's ploy of talking about heuristic models, Derrida escapes self contradiction by talking about meaning as if it is also circular in nature: "metaphor is dangerous and foreign, . . . but it is in complicity with what it endangers, is necessary to it to the extent to which the de-tour is a re-turn guided by the function of resemblance (mimesis or homoisis), under the law of the same." What this verbiage is supposed to mean is suggested by what Derrida "co-ordinates" with manifestations of truth: "the reappropriation of a full language without syntax, with the vocation of a pure nomination". This turns out "to use metonymic abbreviations here", to be what Derrida calls "the heliotrope". Ignoring his obfuscation, a better description would be that philosophy has been rescued from metaphor, while Jakobson's theory of the metaphoric basis of language has been rescued by a reinterpretation of its metonymic component.
Language reduced to a fancy philosophical version of metaphor and metonymy is also the conclusion drawn by Derrida's Of Grammatology (1967). Exiled to the footnotes, Jakobson's theory is relaunched by means of a theory about what Derrida calls "writing". This new terminology is necessary, he explains, because "writing" means more than just inscription, but signs, and thus is "metaphoricity" itself. As in "White Mythology", where analogies are confused with metaphor, signs are now to be confused with metaphor because they "stand for" something else. It is an argument which makes the basic distinctions of logic - predication, identity, equivalence, membership and class inclusion - disappear. Just as Jakobson's aphasia essay seems to provide scientific evidence that language in its essential nature is not referential, Derrida seems to provide evidence that language does not work well because its foundation is metaphor. As one of Derrida's commentators, Sollace Mitchell has noted, "a prosaic metaphorical relation" can be found in Derrida's "tacit theory of meaning", for by reducing words to signs Derrida can claim that something is always being substituted for something else. In Of Grammatology, this expansion of Aristotle's much panned substitution theory of metaphor to all of language is strengthened by misinterpretations of the semiotician C. S. Peirce. Supposedly he claimed that "the thing itself is a sign", while the linguist Ferdinand Saussure is misrepresented, so that Derrida can equate signs with words and then both with concepts, and thus arrive at his famous dictum "There is no outside-of-text". Such collages, which mimic the very theory which excuses them, are also reminiscent of Jakobson's collage of clinical materials to give a scientific air to impressionistic literary criticism.
Derrida's terminological revolution, which arises from the introduction of the arbitrariness of the sign into language, actually was inevitable once Jakobson insisted that metonymy had something to do with contiguity and then claimed that naming could be reduced to similarity and contiguity. Jakobson's premise remained the Saussurean reduction of language to signs, and so, while he himself had no use for Saussurean arbitrariness, by rooting similarity and contiguity in the structure of the linguistic sign he unintentionally reaffirmed it. Already implied is that causality and the veracity of science are inexplicable, while metonymy interpreted in terms of contiguity anticipates Derrida's reduction of language to a "prosaic metaphorical relation". Derrida's originality lies not in his ideas, then, but in his having blended Jakobson with Heidegger.
The postmodernist term "discourse" probably entered French philosophy by means of Alexandre Kojève's famous seminar on Hegel's Phenemonology of Mind and Kojève's integration of Heidegger's theory of "discourse" with Hegel's philosophy of language. Thus in Lacan's early Discours de Rome (1953, published 1955), the word "discourse" is a constant refrain, and there are plentiful references to Hegel along with an acknowledgement of Kojève's influence. According to the Lacanian scholar Anthony Wilden, Heidegger has been the most influential exponent in our century of a philosophical theory of "discourse" which matches the more technically oriented views of a number of linguists. Wilden cites the proto-structural linguist Edward Sapir, but linguists working in the field of text linguistics assume that use of the term "discourse" to refer to cohesive units beyond the sentence can tee traced to Jakobson's essay "Linguistics and Poetics". In their introduction to Jakobson's linguistic papers, the editors, Linda R. Waugh and Monique Monville-Burston, particularly cite his work on "shifters" (pronouns) as having inspired the "linkage of language to culture". Supposedly, if larger units of meaning beyond the sentence did not exist, we would not always know what pronouns refer to. The extremes to which this can be taken can be seen in Benveniste's essay "Subjectivity in Language" (1958). Following up on Jakobson's account of pronouns, Beneveniste claims that the pronoun "I" is a term which must be considered "an instance of discourse and that has only a momentary reference". Here the ability of pronouns to shift their reference becomes an attack on phenomenological philosophy. This same blend of phenomenology and Jakobsonian linguistics is to be found in Derrida's claim that all of thought can be reduced, if not to metaphor, then to metonymy in the guise of "contiguity" and the "play of differences". Only what is attacked is no longer phenomenology, but reference.
At this point, someone might be moved to protest that discussing the relationship of Derrida's work with that of Roman Jakobson gives a very one-sided account of postmodernism. If, with Derrida, we are one level up from Jakobson's frauds in the inverted triangle of postmodernism, we still need to show who occupies the other corner of this inverted triangle. That Derrida belongs in one corner is evident from the notoriety which he has achieved for his distortions of the views of others. In Derrida's postmodernist manifesto, "Structure, Sign and Play in the Discourse of the Human Sciences" (1966), he even announces that scholars ought to be "bricoleurs". This term, picked up from Lévi Strauss's The Savage Mind, refers to someone making use of the various materials at hand. Collage, according to Derrida, is not "menaced" by the faults of "empiricism". The Jakobsonian - now Derridean - doctrine that the "signifier" floats free of referential restraints even hints that science and scholarship ought to be collage; Derrida's systematic use of misrepresentations suggests that this description is based on his own practices. Derrida himself, in his manifesto, compares science to literary criticism.
While Derrida's place in the annals of literary criticism (not scholarship) is secure, a demonstration of how widespread the influence of Jakobson has been is still required in order for it to account for postmodernism. A good place to look is the work of Michel Foucault. This is indicated in two recent books: Keith Windschuttle's The Killing of History, and The Social Misconception of Reality by Richard F. Hamilton. These historians have both demonstrated that Foucault's misrepresentations are the result of an arbitrary methodology in which he selectively rearranges his materials. But what about the relationship between Foucault's ideas and Jakobson's frauds?
Here is one clue. In the essay which popularized the term "discourse" among American academics, "What is an Author" (1969), Foucault mentions Jakobson's term "shifters". The beginning and ending of the essay also echo an earlier essay of Barthes on "the death of the author", and we know that there was a close relationship between Barthes and Foucault, to the extent, according to one of Foucault's biographers, of their being, in the early 1960s, "occasional lovers". Since Barthes, in the preface to his Critical Essays (1964), cited Jakobson's discussion of pronouns as evidence for the existence of "discourses", then later praised Benveniste for the way he "considerably widens the notion of shifter, which Jakobson advanced with such brig", we can assume that Foucault, in some way or another, was acquainted with Jakobson's theories about pronouns and probably also with Benveniste's linguistic "proof" of the existence of "discourses".
We need now to ask a simple question. In a, earlier treatment, "Aphasia as Linguistic Topic (1953), Jakobson announced that "Aphasic regression has proved to be a mirror of the child's acquisition of speech sounds; it shows the child's development in reverse", and then the "The search for acquisitions and losses for the general laws of implications cannot be confined to the phonetic pattern but must be extended also to the grammatical system." This is repeated i Jakobson's essay on pronouns, and Barthe echoed it a number of times, claiming that the personal pronoun I "is a shifter, a translator of all the signs it is the most difficult to use, since the child acquires it last of all and the aphasic loses it first". Is this true? That aphasia patients first lose pronouns and that this is the last part of speech to be acquired by children?
The answer is no. It has long been known the pronouns are among the first items of speech which children learn, while for some time now it has been known that pronouns are the last, not the first parts of speech which an aphasia patient will lose. The theory itself is a gross exaggeration of one about the development of the components of phonemes. "Higher layers are eliminated before lower ones. Aphasic impairments reproduce in reverse order the child's language acquisition." This is from an essay which is a later summary of Jakobson's Child Language, Aphasia, and Phonological Universals (1941). In neurolinguistics, this iron law of phonological development is at best a controversial proposal, but whatever its status, Jakobson's expansion of his thesis from certain aspects of phonemes to grammar remains unsupported. Some forms of segmental phonology do develop before syntax takes shape, but the later stages of phonology and syntax develop more or less independently.
What has happened is that Jakobson has seized on the appearance of errors as proof of grammar's absence. It is the same false logic he uses to materialize figurative language into existence from "language deficits". While it is usually acknowledged that postmodernism has been heavily influenced by linguistics, what is not acknowledged is that the influence has all come from linguistics pretending to be a formal science like mathematics. Also not noticed is that most of this formal "science" rests upon exactly this kind of pseudologic: "If Y, then X before Y". This is certainly what lies behind the structuralism that poststructuralism is a reaction against.
The only actual evidence for the universality of linguistic laws which this fallacy has brought into existence is in Jakobson's early monograph Child Language, Aphasia, and Phonological Universals. What do we find in it? The answer is a discussion of the components of phonemes; nothing about grammar, and phonological detail which is invented. In the passages Jakobson cites from the aphasiologist Henry Head, for instance, Head mentions only "slurring"; nothing is said about phonemes; nothing about their loss in any particular order. Among the hundreds of examples of aphasic speech which Head quotes, Jakobson seizes on phrases which seem to support his theories, while ignoring those which contradict them. The same creativity is to be found in Jakobson's citation of cases. Evidence turns out to be "mentally diseased patients who have been treated with insulin" and "a schizophrenic in the process of awakening". Jakobson also uses phrases from aphasiologists, in order to turn them into supporters of this theories. His idea of an aphasiologist is Pierre Marie, who even then was controversial for his claim that there was no such thing as subtypes of aphasia and that in the one true aphasia comprehension of language was disturbed because of a deficit in general intelligence. Jakobson announces that Marie had showed that aphasia had its roots in the "conceptual, 'semiotic' sphere, an 'intellectual deficiency relating specifically to language', according to the formulation of this scholar", then promptly adds, referring to speech sounds, "Not the perception as such, but rather its linguistic value is impaired". Just as Jakobson later was to rename naming itself metaphor and metonymy, he gives "intellectual deficiency" a new name: "linguistic value". The manoeuvre ought to be familiar, for Jakobson's spate of renaming included calling syntax contiguity, contiguity metonymy, and pronouns "shifters". Then, of course, Benveniste gave "shifters" a new name - "discourse" - by means of similar reflections on "linguistic value".
This sort of thing may account for pronoun-lacking children and metaphor-lacking aphasia patients, but preconceived notions are taking leave of the facts. We are just getting the "If Y, then X before Y" argument packed into a term - as when Benveniste tells us that a pronoun ought to be renamed an "instance of discourse" because "the reality to which it refers is the reality of the discourse". The same shift is discernible in Foucault's substitution of "discourse" for his earlier term "episteme". At the beginning of Foucault's The Order of Things, a description of people suffering from visual agnosia, explained as "aphasiacs", is cited as evidence for the existence of concept-controlling "epistemes". Later we get no such arguments; we get "discourse", especially about the "discourse of science".
Allowing question-begging terminology and citation chains to substitute for serious research, scholars influenced by postmodernism no longer even bother with the presence that evidence exists for their doctrines. Empirical evidence is no longer needed, because theoretical perspectives, relied on in the same way as their inventors relied on Roman Jakobson, are considered sufficient evidence that the insights and perceptions of science are the result of metaphorical and linguistic interpretations of the world. However, the emphasis on metaphor and linguistics directly echoes Jakobson, and since the validity of Jakobson's theories derives entirely from misrepresentations, lies and distortions, while a similar mendacity has played a role in their postmodernist manifestations, no aspect of postmodernist scholarship is proof of anything except the decline of some university departments into a condition where, like aphasia patients, they wallow in diseased naming.
Back to Topics Page