120.6

Mikhail Bakhtin1895-1975

-Mikhail Bakhtin's work is fired by his conviction that language and the fonns it

takes can be properly understood Ol:ily a~ dialogue, as utterances that take placewithin social situations and that at least piirtIy constitute them. He attacks every ap­proach to the study of language orJit~rature that is based on a!Jstract structures orindependent systems of meaning or that focuses on isolated speakers pr texts. He re­jects the position that utterances simply "reflect" something else, such as the psy­chology of the individual or the conditions of society. He proposes instead that dia­logue is the. model for understanding not only literature; but also ideology,psychology; and linguistics. Dialogue, for Bakhtin, occurs both in the literal ex­change of utterances between speakers and in the intentional negotiation of meaningand interpretation between author and reader. He emphasizes the "polyphony" oflanguage seen this way; the "heteroglossia" of speech arid texts that are subject tomultiple interpretations. .

Born into the untitled nobility in 1895 in a to'!"nnot far from Moscow, Bakhtinwas educated in European culture and. studied classics ,at Petersburg University,where he remained through the First World War and the Russian Revolution.Through the revolutionary period, Bakhtin taught school in a provincial town, gavepublic lectures on art and literature (he was considered a dynamic public speaker),cultivated his circle of artists and academics, and worked on his writing. Committedto the Russian Orthodox Church, he also lectured on theology. When Stalin came topower late in the twenties, religion was repressed and Bakhtin was arrested­charged, as Socrates had been, with corrupting youth. In 1930, he was exiled toKazakhstan for six years.

During the twenties, Bakhtin had no official standing in the state intellectual­academic cOIIUllunity and was even somewhat suspect because of his religious lean­ings. For this reason, he apparently felt that he could not find publishers for workson sensitive subjects, such as Marxism, Freud, and literary Formalism. So Bakhtinpublished three books and several articles under the names of other, better­established members of his circle, The books are The Fonnal Merhod in LireraryScholarship (1928), attributed to P. N. Medvedev; Freudianism: A Critical Sketch(1927), attributed to V. N. Voloshinov; and Marxism and the Philosophy of Lan­guage (1929), also attributed to Voloshinov, During this period, Bakhtin publishedProblems of Dostoevsky's Creative Works (1929) under his own name, this topicbeing, presumably, a less sensitive one,

According to Bakhtin's biographers, Katerina Clark and Michael Holquist,Bakhtin's willinguess to publish under his friends' names is typical of his lack ofpersonal ambition, his generosity, and his love of practical jokes-the "carnival"quality he admires in the sixteenth-century French writer Fran~ois Rabelais.' fn-

'Katerina Clark and Michael Holquist, Mikhail Bakhtin (Cambridge: Harvard University Press,1984), p. lSI. See pages 146-70 for a full discussion of the authorship dispute.

MODERN AND ,POSTMODERN RHETORIC

deed, he kept the authorship dispute alive even after he was officially rehabilitatedin the sixties, admitting that he wrote the disputed works but refusing to sigu offi",cia! statements to that effect. Although most scholars seem to have acceptedBakhtin's authorship, some questions about it still remain.

On his return from exile, Bakhtin wrote a dissertation on Rabelais for the GorkyInstitute but, because of the Second World War, was unable to defend it. Even afterthe defense, which occurred in 1947; the faculty at first rejected the dissertation and·then hotly debated it, until finally, in 1951, they accepted it.. In the meantime,Bakhtin taught literature at a provincial university, where he was a popular lecturer.He remained on the fringes ofintellectuallife .until the sixties, when he was discov­ered by yonnger scholars, under whose influence the Dostoevsky book was repub­lished and the dissertation on Rabelais revised and published as Rabelais and HisWorld (1965). Bakhtin was permitted to return to Moscow in 1972, and he diedthere in 1975.

The early works of disputed authorship, particularly The Formal Method in Lit­erary Scholarship and Marxism and the Philosophy of Language, represent themanifesto of the dialogic theory that Bakhtin applies to literary criticism, literarytheory, and language theory. The Formal Method inLiterary Scholarship.discussesthe important but short-lived critical school that we know as Russian Formalism.•Muchlike the New Criticism in its close reading of the style and structure of indkvidual works, Formalism began after the First World War in the discussions ofliter.ature students who were interested in the new ideas about structural linguistics ad­vanced by Ferdinand de Saussure. The Formalists were drawn to Saussure'sdescriptive linguistics because it represented a scientific approach to language thatmight help to put literary criticism, too, on a scientific footing. The Formalistsmaintained that one should describe the literary work as a linguistic object before oreven in preference to any discussion of a work's context or message. Only suchclose attention to the literary object could reveal, they believed, the work's essentialliterariness.

In The Formal Method, Bakhtin acknowledges the virtues of Formalism in at­tending to the linguistic features of the text (in contrast to older methods of histori­cal and biographical criticism) and in combating "vulgar" materialist readings thatviewed literature simplistically, as a reflection of class and economic conditions:But Formalism pursued a programmatic separation of literature from history and so­cial conditions on the principle that literary language is different from practical lan­guage. This position Bakhtin artacks. The advances of Formalism must be main­tained, .he argues, but criticism must also recogrtize that the meaning of literaryworks depends on the historical situation of the text, the social situation of thereader, and the complex interaction thus generated by the act of reading. Literarylanguage is not different from practical language: Literary meaning, unstable andpolysemous, depends on dialogue-that is, on the negotiation of meaning betweentext and interpreting reader-and not on the literariness of the text and its pure per­ception by the reader.

If The Formal Method is the poetic of dialogism, Marxism and the Philosophyof Language (excerpted here) is its rhetoric. In the latter volume, Bakhtin argues

MIKHAIL BAKHTIN n07

noS

that the true basis of linguistics is the study of utterances, of speech acts. Bakhtinaccepts Saussure's semiotic assumption thatcommunication takes place' through ar­bitrary signs: There is, in other words, no inherent meaning in the sounds or sym­bols of language. But Bakhtin objects to Saussure's assumption that the sole pur­pose of linguistics is to analyze how these signs fit together into a system. Bakhtincharges (rather unfairly) that the chief virtue of structural linguistics is to provide anobject of scientific study for linguists. Structural descriptions, he complains, do notexplain the way language is actually used. The individual utterance, in structurallinguistics, is an isolated event rather than a social act.

Structural linguistics regards the utterance (parole) as an individual's selectionof elements (lexical, grammatical, and syntactic) from the larger system of language(langue). Bakhtin argues for a method of analysis that recoguizes the primacy of theutterance-in-context and shows the dynantic relationship between system and utter­ance, a method that he elaborates most effectively in "The Problem of SpeechGenres" (1953: excerpted here). He also rejects Saussure'sassertion that signs arepsychological in nature. Signs are part of material reality, Bakhtin maintains, im­bued with ideological meaning by their use in social situations. They are not, there­fore, simply mental phenomena. The psychologism of Saussure and other linguistsreflects, for Bakhtin, the unacceptable subjectivism that treats language as the ex­pression of the individual's thought. Instead, argues Bakhtin, consciousness shouldbe seen as a social phenomenon that "takes shape and being in the material of signscreated by an orgauizedgroup in the process of its social intercourse,'" He defendsthis position through a discussion of inner speech, borrowing the concept advancedby his contemporary, the great psychologist Lev Vygotsky.

Bakhtin remarks in Marxism and the Philosophy ofLanguage that "the structureof a whole utterance is something linguistics leaves to the competence of other dis~

ciplines-to rhetoric and poetics."3 In "The Problem of Speech Genres," he at­tempts to analyze the "whole utterance" and; in effect, to merge rhetoric and lin­guistics. The genres of speech are, he says, the "relatively stable types" of utterancethat occur in "each sphere in which language is used." They include, for example,the conventions of casual- conversation, the formulas of business documents, andthe well-known literary genres. He wants the idea of speech genres to honor bothconvention and individual choice, to see both constraints and openness in communi~

cation.It would not be wrong to think of the speech genres, then, as rhetorical situations

and to see Bakhtin's argument as a way of extending rhetoric;s gaze to every act ofspeaking or writing. To be sure, Bakhtin takes his approach not from rhetoric butfrom linguistics and semantics. Indeed, Bakhtin slights rhetoric for its traditionallimitation to a few genres and mechanical taxonomies of tropes, just as he criticizesstructural linguistics for its mechanistic approach. Rhetoric scholar Kay Halasekhas suggested, too, that Bakhtin sees rhetoric as too confrontational and monologic

2See p. 1213 in this book.3Mikhail Bakhtin, Marxism and the Philosophy of I..o.nguage (1929; rpt., trans. Ladislav Matejka and

1. R. Titunik, Cambridge: Harvard University Press, 1973), p. 78.

MODERN AND POSTMODERN RHETORIC

in comparison with the approach to the audience implied in literary texts.4 Bakhtin .defines his own dialogic method as a way of recognizing (a) the speaker's orwriter's (rhetorical) intention to move the audience to action and (b) the audience'sactive role in interpreting utterances in order to reply or react, a role that the speakeror writer is well aware of: "The [literary or technical] work, like the rejoinder in di­alogue, is oriented toward the response of the other.'" Bakhtin then distinguishesthe boundaries of the utterance, both spoken and written, and argues that all me­chanical or subjective theories of meaning should be abandoned in favor of a con­textual one. His focus on context aligus well with the social turn in rhetoric andcomposition studies, accounting for the great surge of interest in his work in thesefields since 1985. His emphasis on the dialogic, too, has increased interest in hiswork among feminist scholars looking for ways to theorize the plurality of views intheir own_field.

Selected Bibliography

Marxism and the Philosophy of Language, trans. Ladislav Matejka and I. R. Titunik (NewYork, 1973), gives V. N. Voloshinov as the author. "The Problem of Speech Genres" is pub­lished in Speech Genres and Other Late Essays, trans. Vern W. McGee (Austin, Texas,1986).

Other works by Bakhtin include "Discourse Typology in-Prose," in Readings in RussianPoetics (Formalist and Structuralist Views), ed. L. Matejka and K. Pomorska (Cambridge,Mass., 1971); Frelidianism: A Critical Sketch (by "V. N. Voloshinov"), translated as Freudi­anism: A Marxist Critique by 1. R. Titunik (New York; 1973); The Formal Method in Liter­ary Scholarship: A Critical Introduction to Sociological Poetics (by "P. N. Medvedev"),trans. Albert 1. Wehrle (Baltimore, 1978); Problems in Dostoevsky's Poetics, trans. CarylErp.erson (Minneapolis, 1984); R~belais and His World, trans. Helene Iswolsky (Cambridge,Mass., 1968); and The Dialogic Imagination, ed, Mich~el Holquist, trans. Caryl Emerson andMichaeLflolquist (Austin, Tex., Ig81).

Katerina Clark and Michael Holquist have written a scholarly, readable biography inMikhail Bakhtin (Cambridge, Mass., 1984). Two essays printed in Marxism and the Philoso~

phy ofLanguage provide an excellent introduction to the issues in Bakhtin's intellectual en­vironment: Ladlslav Matejka's "On the First Russian Prolegomena to Semiotics" and I. R.Titunik's "The Fonnal Method and the Sociological Method (M, M, Bakhtin, P, N.Medvedev, V, N. Voloshinov) in Russian Theory and Study of Literature." Titunik has alsopublished a summary of contemporary" issues in "M. M. Bakhtin (The Bakhtin School) andSoviet Semiotics" (Dispositio 1 [1976]: 327-38).

Bakhtin: Essays and Dialogues on His Work, ed. Gary Saul Morson (Chicago, 1986), is acollection of essays that appeared originally in Critical Inquiry 7 (summer 1981),9 (Septem­ber 1982), 10 (December 1983), and II (June 1985). The essays cover a wide range ofBakhtin's linguistic and literary ideas. Tzvetan Todorov's book-length study, MikhailBakhtin: The Dialogical Principle (I981; trans. Wlad Godzich, Minneapolis, [984), alsocontains several shorter works by Bakhtin. Michael Holquist's Dialogism: Bakhtin andHis World (London, 1990) organizes Bakhtin's work around his concept of "dialogism" and

4Kay Halasek, A Pedagogy ofPossibility: Bakhtinian Perspectives on Composition Studies (Carbon­dale: Southern TIlinois University Press, 1999), pp. x-xi.

SSee p. 1237 in this book.

MIKHAIL BAKHTIN 12°9

focuses on his usefqlness for literary analysis. Dominick LaCapra devotes a chapter. of Re.thinking Intellectual History:- .Texts,. Contexts, Language (lthaca.N.Y., 1983) to "Bakhtin,'Marxism, and the Carnivalesqu~."providing a lucid summary of Bakhtin's major ideas, bothliterary <md linguistic..Feminist ~ork on Bakhtin by literary scholars is collected in Ferni.nism, Bakhtin, "lid the Dialogic (Albany, N.Y., 1991), ed. Dale M. Bauer aud S. JaretMcKinstry. A collection edited by Amy Mandelker, Bakhtin in Contexts: Across the Disci­plines (Evans't~n: -In., ;1995), presents essays that apply Bakhtin's ideas in Slavic studies,classics, American literature, African American studies, sociology, anthropology, linguistics,ana semiotics.

.Landmark Essays on' Bakhtin, Rhetoric, and Writing, ed. Frank. Farmer (Mahwah, N.J.,1998), collects important essays by Charles Schuster, Michael Bernard-DonaIs, Kay Halasek,Don H. Bialostosky, Marilyn Middendorf, Helen Rothschild Ewald, aud other scholars inrhetoric and composition studies. Kay Halasek helpfully surveys this scholarship in A Peda­gogy ofPossibility: Bakhtinian Perspectives on Composition Studies (Carbondale, Ill.,1999),where she articulates Bakhtin's ideas in relation to key composition concepts of the studentwriter, the audience, the topic, and critical reading and writing.

From Marxism and the Philosophy ofLanguage

From Part ICHAPTER ITHE STUDY OF IDEOLOGIES ANDPHILOSOPHY OF LANGUAGE

The problem ofthe ideological sign. The ideologicalsign and consciousness. The word as an ideologicalsign par excellence. The ideological neutrality oftheword. The capacity ofthe word to be an inner sign.Summary. ' '

Problems of the philosophy of language have in .recent times acquired exceptional pertinence andimportance for Marxism. Over a wide range ofthe most vital sectors in its scientific advance, theMarxist method bears directly upon these prob­lems and cannot continue to move ahead produc­tively without special provision for their investi­gation and solution.

First and foremost, the very foundations of aMarxist theory of ideologies - the bases for thestudies of scientific knowledge, literature, reli­gion, ethics, and so forth-are closely bound upwith problems of the philosophy of language.

Any ideological product is not only itself a

Translated by Ladislav Matejka and 1. R. Titunik.

part of a reality (natural or social), just as is auyphysical body, any instrument of production, orany product for consumption, it also, in con­tradistinction to these other phenomena, reflectsand refracts another reality outside itself. Every­thing ideological pOssesses meaning: it repre­sents, depicts, or stands for something lying out­side itself. In other words, it is a sign. Withoutsigns, there is no ideology. A physical bodyequals itself, so to speak; it does not signify any­thing but wholly coincides with its particular,given nature. In this case there is -no question ofideology.

However, any physical body may be perceivedas an image; for instance, the image of natural in­ertia and necessity embodied in that particularthing. Any such artistic-symholic image to whicha particular physical ohject gives rise is already anideological product. The physical ohject is con­verted into a sign. Without ceasing to be a part ofmalerial reality, such an object, to some degree,reflects and refracts another reality.

The same is true of any instrument of produc­tion. A tool by itself is devoid of any specialmeaning; it commands only some designatedfunction - to serve this or that purpose in pro­duction. The tool serves that purpose as the par-

1210 MODERN AND POSTMODERN RHETORIC

From The Problem ofSpeech Genres

I. STATEMENT OF THE PROBLEM ANDDEFINITION OF SPEECH GENRES

All the diverse areas of human activity involvethe use of language. Quite understandably, thenature and forms ofthis use are just as diverse asare the areas of human activity. This, of coursein no way disaffirms the national unity of Ian:guage.' Language is realized in the form of indi­vidual concrete utterances (oral and written) byparttclpants m the various areas of human activ­ity. These utterances reflect the specific condi­tions and goals of each such area not only ,through theIr content (thematic) and linguisticstyle, that IS, the selection of the lexical, phraseo­logIcal, and grammatical resources of the lan­guage, but above all through their compositionalstructure. All three of these aspects -thematiccontent, style, and compositional structure-areinseparably linked to the whole of the utteranceand are equally determined by the specific nature'of the particular sphere of communication. Eachseparate utterance is individual, of course, but~ach sphere !n which language is used developsIts own relatively stable types of these utterances.These we may call speech genres.

The wealth and diversity of speech genres areboundless because the various possibilities ofhuman activity are inexhaustible, and becauseea.ch sphere of activity contains an entire reper­tmre of speech genres that differentiate and growas the particular sphere develops and becomesmore complex. Special emphasis should be placedon the extreme heterogeneity of speech geures(oral and written). In fact, the category of speechgenres should include short rejoinders of daily di­alogue (and these are extremely varied dependingon the subject matter, situation, and participants),

Translated by Vern W. McGee.. <"National unity of language" is a shorthand way ofrefer­~g to the assemblage of linguistic and transtinguistic prac­tIces common to a given region. It is, then, a good example ofwhat ~akhtin ~eans by an open unity. See also,Otto Jesperson,Mankmd. NatiOn, and Individual (Bloomington: Indiana Uni­versity Press, 1964). [Tr.]

everyday narration, writing (in all its variousforms), the brief standard military cornrnand, theelaborate and detailed order, the fairly variegatedrepertoire ofbusiness documents (for the most part.standard), and the diverse world of commentary(m the broad sense of the word: social, political).And we must also include here the diverse formsof scientific statements and all literary genres(from the proverb to the multivolume novel). Itmight seem that speech genres are so heteroge­neous that they do not have and cannot have asingle common level at which they can be studied.For here, on one level of inquiry,appear such het­erogeneous phenomena as the single-word every­day rejoinder and the multivolume novel, the mil­itary command that is standardized even in itsintonation and the profoundly individual lyricalwork, and so on. One might think that such func­tional heterogeneity makes the common featuresof speech genres excessively abstract and empty.This probably explains why the general problemof speecb genres has never really been raised. Lit­erary geures have been studied more than anythingelse, But from antiquity to the present, they havebeen studied in terms of their specific literary andarttstlC features, m terms of the differences thatdistinguish one from the other (within the realm ofliterature), and not as specific types of ulterancesdistinct from other types, but sharing with them acomI~lOn verbal (language) nature. The ,generallingmstlc problem of the utterance and its typeshas hardly been considered at all. Rhetorical gen­res have been studied since antiquity (and notmuch has been added in subsequent epochs to clas­SICal theory). At thattime, more attention was al­ready being devoted to the verbal nature of thesegenres as utterances: for example, to such aspectsas the relation to the listener and his influence onthe ulterance, the specific verbal finalization of theutterance (as distinct from its completeness ofthought), and so forth. But here, too, the specificfe.atures of rhetorical genres (judicial, political)stIll overshadowed their general linguistic nature.Fmally, everyday speech genres have been studied(mainly rejoinders in everyday dialogue), and

BAKHTIN ITHE PROBLEM OF SPEECH GENRES 1227

from a general linguistic standpoint (in the schoolof Saussure and among his later followers-theStructuralists, the American behaviorists, and, ona completely different linguistic basis, the Vossle­rians).' But this line of inquiry could not lead to acorrect determination of the general linguistic na-

2Saussure's teaching is based on a distinction betweenlanguage (la langue)-a system of interconnected signs andfonns that nonnatively determine each individual speech actand are the special object of linguistics-and speech (ia pa­role)-in~vidual instances of language use. Bakhtin dis­cusses Saussure's teachings in Marxism and the Philosophyof Language as one of the two main trends in linguisticthought (the trend of "abstract objectivism") that he uses toshape his own theory of the utterance. See V. N. Voloshinov,Marxism and the Philosophy of Language, tr. Ladislav Mate- .jka and 1. R. Titunik(New York: Seminar Press, 1973), esp.PP·S8-6I.

"Behaviorists" here refers to the school of psychology. in­troduced by the Harvard physiologist J. B. Watson in 1913. Itseeks to explain animal and human behavior entirely in teJ;lllSof observable and measurable responses to external stimuli.Watson, in his insistence that behavior is a physiological re­action to environmentai stimuli, denied the value of intro­spection and of the concept of consciousness. He saw mentalprocesses as bodily movements, even when unperceived, sothat thinking in his view is subvocal speech. There is a strongconnection as well between the behaviorist school of psychol­ogy and the school of American descriptive linguistics, whichis what· Bakhtin is referring to here. The so-called descrip­tivist school was founded by the eminent anthropologistFranz Boas, (1858-1942). Its closeness to behaviorism con­sists in its insistence on careful observation unconditioned bypresuppositions or categories taken from traditional languagestructure. Leonard Bloomfield (1887-1949) was the chiefspokesman for the school and was explicit about his commit­ment to a "mechanist approach" (his term for the behavioristschool of psychology): "Mechanists demand that the facts bepresented without any assumption of such·auxiliary factors[as a version of the mind). I have tried to meet this de­mand...." (Language [New York: Holt, 'Rinehart, and Win­ston, 19331, p. vii). Two prominent linguists sometimes asso­ciated with the descriptivists, Edward Sapir (1884-1939) andhis pupil Benjamin Lee WhOlf (1897-1941), differ fromBloomfield insofar as behaviorism plays a relatively minorrole in their work.

"Vosslerians" refers to the movement named after theGennan philologist Karl Vossler (1872-1949), whose adher~

eots included Leo Spitzer (1887-1960). For Vosslerians, thereality of language is the continuously creative, constructiveactivity that is prosecuted through speech acts; the creativityof language is likened to artistic creativity, and stylistics be­comes the leading discipline. Style takes precedence overgrammar, and the standpoint of the speaker takes precedenceover that of the listener. In a number of aspects, Bakhtin isclose to the Vosslerians, but differs in his understanding ofthe utterance as the concrete reality of language life. Bakhtin

ture of the utterance either, since it was limited tothe specific features of everyday oral speech,sometimes being directly and deliberately ori­ented toward primitive utterances (American be­haviorists).

The extreme heterogeneity of speech genresand the attendant difficulty of determining thegeneral nature of the utterance should in no waybe underestimated. It is especiallyimportant hereto draw attention to the very siguificant differ­ence between primary (simple) and secondary(complex) speech genres (understood not asafunctional difference). Secondary (complex)speech genres - novels, dramas, all kinds of sci­entific research, major genres of commentary,and so forth:- arise in more complex and com­paratively highly developed and organized cul­tural communication (primarily written) that isartistic, scientific, sociopolitical, and so on. Dur­ing the process of their formation, they absorband digest various primary (simple) genres thathave taken form in unmediated speech commu­nion. These primary genres are altered and as­sume a special character when they enter intocomplex ones. They lose their immediate relationto actual reality and to the real utterances of oth­ers. For example, rejoinders of everyday dialogueor letters found in a novel retain their form andtheir everyday significance only on the plane ofthe novel's content. They enter into actual realityonly via the novel as a whole, that is, as a literary­artistic event and not as everyday life. The novelas a whole is an _~tterance_ just as rejoinders ineveryday dialogue or private letters are (they dohave a common. nature), but unlike these, thenovel is a secondary (complex) utterance.

The difference between primary and sec­ondary (ideological) genres is very great and fun­damental,3 but this is precisely why the nature of

does not, like the Vosslerians, conceive the utterance to be anindividual speech act; rather, he emphasizes the "inner social­ity" in speech communication-an aspect that is objectivelyreinforced in speech genres. The coricept of speech genres iscentral to Bakhtin, then, in that it separates his translinguisticsfrom both Saussureans and Vosslerians in the philosophy oflanguage. [Tf.]

3"Ideology" should not be confused with the politicallyoriented English word. Ideology as it is used here is essen-

1228 MODERN AND POSTMODERN RHETORIC

the utterance should be revealed and definedthrough analysis of both types. Only then can the .definition be adequate to the complex aud pro­found nature of the utterance (and encompass itsmost important facets). A one-sided orientationtoward primary genres inevitably leads to a vul­garization of the entire problem (behaviorist lin­guistics is an extreme example). The very inter­relations between primary and secondary genresand the process of the historical formation of the .latter shed light on the nature of the utterance(and above all on the complex problem of the in­terrelations among language, ideology, and worldview).

A smdy of the nature of the utterance and ofthe diversity of generic forms of utterances invarious spheres of human activity is immenselyimportant to almost all areas of linguistics andphilology. This is because any research whosematerial is concrete language-the history of alanguage,normative granunar, the compilation ofany kind of dictionary, the stylistics of language,and so forth-inevitably deals with concrete ut­terances (written and oral) belonging to variousspheres of human activity and communication:chronicles, contracts, texts of laws, clerical andother documents, various literary, scientific, andcommentarial genres, official and personal let­ters' rejoinders in everyday dialogue (in all oftheir diverse subcategories), and so on. And it ishere that scholars find the language data theyneed. A clear idea of the nature of the utterance·

.in general and of the peculiarities of the various .types of utterances (primary and secondary), thatis, of various speech genres, is necessary, wethink, for research in any special area. To ignorethe nature of the utterance or to fail to considerthe peculiarities of generic subcategories ofspeech in any area of linguistic study leads toperfunctoriness and excessive abstractness, dis­torts the historicity of the research, and weakensthe link between language and life. After all, lan­guage enters life through concrete utterances

tially any system of ideas. But ideology is semiotic in thesense that it involves the concrete exchange of signs in soci­ety and history. Every word/discourse betrays the ideology ofits speaker; every speaker is thus an ideologue and every ut­(erance an ideologeme. [Tr.]

(which manifest language) and life enters lan­guage through concrete utterances as well. Theutterance is an exceptionally important node ofproblems. We shall approach certain areas andproblems of the science of language in this con­text.

First of all, stylistics. Any style is inseparablyrelated to the utterance and to typical forms of ut­terances, that is, speech genres. Any utterance­oral or written, primary or secondary, and in anysphere of communication - is individual andtherefore can reflect the individuality of thespeaker (or writer); that is, it possesses individualstyle. But not all genres are equally conducive toreflecting the individuality of the speaker in thelanguage of the utterance, that is, to an individualstyle. The most conducive genres are those ofartistic literature: here the individual style entersdirectly into the very task of the utterance, andthis is one of its main goals (but even withinartistic literalUre various genres offer differentpossibilities for expressing individuality in lan­guage and various aspects of individuality). Theleast favorable conditions for reflecting individu­ality in language obtain in speech genres that recquire a standard form, for example, many kindsof business documents, military commands, ver­bal signals in industry, and so on. Here one canreflect only the most superficial, almost biologi­cal aspects of individuality (mainly in the oralmanifestation of these standard types of uttercances). In the vast majority of speech genres (ex­cept for literary-artistic ones), the individual styledoes not enter into the intent of the utterance~

does not serve as its only goal, but is, as it were,an epiphenomenon of the utterance, one of itsby-products. Various genres can reveal variouslayers and facets of the individual personality,and individual style can be found in various inter­relations with the national language. The veryproblem of the national and the individual in lan­guage is basically the problem of the utterance(after all, only here, in the utterance, is the na­tional language embodied in individual form).The very determination of style in general, andindividual style in particular, requires deeperstudy of both the nature of the utterance and thediversity of speech genres.

The organic, inseparable link between style

BAKHTIN I THE PROBLEM OF SPEECH GENRES 1229

and genre is clearly revealed also in the problemof language sryles, or functional styles. In es­sence, language, or functional, styles are nothingother than generic styles for certain spheres ofhnman activity and communication. Each spherehas and applies its own genres that correspond toits own specific conditions. There are also par­ticular styles that correspond to these genres. Aparticular function (scientific, technical, com­mentarial, business, everyday) and the particularconditions of speech communication specific foreach sphere give rise to particular genres, thatis,certain relatively stable thematic, compositional,and stylistic types of utterances. Style is insepa­rably linked to particular thematic unities and­what is especially important-to particularcompositional unities: to particular types of con­struction of the whole, types of its completion,and types of relations between the speaker andother participants in speech communication (lis­teners or readers, panners, the other's speech,and so forth). Style enters as one element into thegeneric unity of the utterance. Of course, thisdoes not mean that language style cannot be thesubject of its ·own independent study. Such aslndy, that is, of language stylistics as an inde­pendent discipline, is both feasible and neces­sary. But this study will be correct and produc­tive only if hased on a constant awareness of thegeneric nalnre of language styles, and on a pre­liminary slndy of the subcategories of speechgenres. Up to this poiut the stylistics 01 language:has not had such a basis. Hence its weakness.There is no generally recognized classification oflanguage styles. Those who attempt to create themfrequently fail to meet the fundamental logicalrequirement of the classification: a llnified basis.4

4A unified basis for classifying the enormous diversity ofutteranceR is an obsession of Bakhtin's, one that relates himdirectly to Wilhelm von Humboldt (1767-1835), the first inthe mod~m period to argue systematically that language is thevehicle of thought. He calls language the "labor of the mind"(Arbeit des Geistes) in his famous fannulation "[language] it­self i~ not [mere} work (ergon), but an activity (energeia) ..it is in fact the labor of the mind that otherwise would eter­nally repeat itself to make articulated sound capable of theexpression of thought" (Uber die Verschiedenhcit des me/l­schiichen Sprachbaues, in Werke. vol, 7 [B~rlin: De Gruyter,1968J, p. 46). What is important here i::; that for Bakhtin, asfor von Humboldt, the diversity of languages is itselfofphilo-

Existing tax.onomies are extremely poor and un­differentiated.5 For example, a recently publishedacademy grammar of the Russian language givesthe following stylistic subcategories of language:bookish speech, popular speech, abstract-scien­tific, scientific-technical, journalistic-comrnentar­ial, official-business, and· familiar everydayspeech, as well as vulgar common parlance. Inaddition to these linguistic styles, there are thestylistic subcategories of dialectical words, ar­chaic words, and occupational expressions. Sucha classification of styles is completely random,and at its base lies a variety of principles (orbases) for division into styles. Moreover, thisclassification is both inexhaustive and inade­quately differentiated. All this is a direct result ofan inadequate understanding of the generic na­tureof linguistic styles, and the absence of awellcthought-out classification of speech genresin terms of spheres of human activity (and alsoignorance of the distinction between primary andsecondary genres, which is very important forstylistics).

It is especially harmful to separate style fromgenre when elaborating historical problems. His­torical changes in language styles are inseparablylinked to changes in speech genres. Literary lan­guage is a complex, dynamic system of linguisticstyles. The proportions and interrelations of thesestyles in the system of literary language are con­stantly changing. Literary language, which alsoincludes nonliterary styles, is an even more com~

plex system, and it is organized on differentbases. In order to puzzle out the complex histori­cal dynamics of these systems and move from asimple (and, in the majority of cases, superficial)

sophical significance, for if thought and speech are Ollt:, doesnot each language embody a unique way of thinking? It is herethat Bakhtin also comes vel)' dose to the work of Sapir and,espl'Cially, of Whorl. See Benjamin Lee Whorf, language,Thought, and Reality, ed. John R Carroll (Cambridge, Mass.:MIT Press, 1956), esp. pr. 212-r9 and 239-"45. [Tr.]

5The same kinds of classifications of language styles, im­poverished and lacking clarity, with a fabricated foundation,are given by A. N. Gvozdev in his book Ocherki po slilislik.erussk.ogo jazyka (Essays on the stytistics of the Russian lan­guage) (Moscow, 1952. PP 13-IS). All of these classifi­cations are based on an uncritical assimilation of traditionalideas about language styles. [Au.]

123° MODERN AND POSTMODERN RHETORIC

description of styles, which are always in evi,dence and alternating with one another, to a his­torical explanation of these changes, one mustdevelop a special history of speech genres (andnot only secondary, but also primary ones) thatreflects more directly, clearly, and flexibly all thechanges taking place in social life. Utterancesand their types, that is, speech genres, are thedrive belts from the history of society to the his­tory of language. There is not a single new phe­nomenon (phonetic, lexical, or grammatical) thatcan enter the system of language without havingtraversed the long and complicated path ofgeneric-stylistic testing and modification.6

In each epoch certain speech genres set the ,tone for the development of literary language.And these speech genres are not onIy,secondary(literary, commentarial, and scientific), bnt alsoprimary (certain types of oral dialogue-of thesalon, of one's own circle, and other types aswell, such as familiar, family-everyday, sociopo- ,litical, philosophical, and so on). Any expansionof the literary language that resnIts from drawing .on various extraliterary strata of the national lan­guage inevitably entails some degree of penetra­tion into all genres of written language (literary,scientific, commentarial, conversational, and s.oforth) to a greater or lesser degree, and entailsnew generic devices for the construction of thespeech whole, its finalization, the accommoda­tion of the listener or partner, and so forth. Thisleads to a more or less fundamental restructuringand renewal of speech genres. When dealing withthe corresponding extraliterary strata of the na-

.tional lallguage, one inevitably also deals withthe speech genres through which these strata are .manifested. In the majority of cases, these arevarious types of conversatioual-dialogical genres.Hence the more orless distinct dialogization ofsecondary genres, the weakening of their mono­logical composition, the new sense of the listeneras a partner-interlocutor, new forms of finaliza­tion of the whole, and so forth. When there isstyle there is genre. The transfer of style from

6'fhis thesis of ours has nothing in common with theVosslerian idea of the primacy of the stylistic over the gram­matical. Our subsequent exposition will make this completelyclear. [Au.]

one genre to another not only alters the way astyle sounds, under conditions of a genre unnat­ural to it, but also violates or renews the givengenre.

Thus, both individual and general languagestyles govern speech genres. A deeper and broaderstudy of the latter is absolutely imperative for aproductive study of any stylistic problem.

However, both the fundamental and the gen­eral methodological question ofthe interrelationsbetween lexicon and grammar (on the one hand)and stylistics (on the other) rests on the sameproblem of the utterance and of speech genres.

Grammar (and lexicon) is essentially differentfrom stylistics (some even oppose it to stylistics).but at the same time there is not a single gram­matical study that can. do without stylistic obser­vation and excuni:Us. In a large number of casesthe distinction between grammar and stylisticsappears to be completely erased. There are phe­nomena that some scholars include in the area ofgrammar while others include them in the area ofstylistics. The syntagma is an example.

One might say that grammar and stylisticsconverge and diverge in any concrete language.phenomenon. If considered only in the languagesystem. it is a grammatical phenomenon. but ifconsidered in the whole of the individual utter­ance or in a speech genre, it is a stylistic phenom­enon. And this is because the speaker's very se­lection of a particular grammatical form is astylistic act. But these two viewpoints of one andthe same specific linguistic phenomenon shouldnot be impervious to one another and should notsimply replace one another mechanically. Theyshould be organically combined (with, however.the most clear-cut methodological distinction be­tween them) on the basis of the real unity of thelanguage phenomenon. Only a profound under­standing of the nature of the utterance and theparticular features of speech genres can provide acorrect solution to this complex methodologicalproblem.

It seems to us that a study of the nature of theutterance and of speech genres is of fundamentalimportance for overcoming those simplistic no­tions about speech life. about the so-called speechflow, about communication and so forth-ideaswhich are still current in our language studies.

BAKHTIN ITHE PROBLEM OF SPEECH GENRES I23I

Moreover, a study of the utterance as a real unitof speech communion will also make it possible'to understand more correctly the nature of Ian"guage units (as a system): words and sentences.

We shall now tum to this more general prob- .lem.

II. THE UTTERANCE AS A UNITOF SPEECH COMMUNION:THE DIFFERENCE BETWEEN THISUNIT AND UNITS OF LANGUAGE(WORDS AND SENTENCES)

Nineteenth-century linguistics, beginning withWilhelm von Humboldt, while not denying thecommunicative function of language, tried to'place it in the background as something sec­ondary.7 What it foregronnded was the function .of thought emerging independently of communi­cation. The famous Humboldtian formula goeslike this: "Apart from the conununication be­tween one human and another, speech is a neces­sary condition for reflection even in solitude."Others, Vosslerians for example, emphasize theso-called expressive function. With all the vari­ous ways individual theoreticians understand thisfunction, it essentially amounts to the expressionof the speaker's individual discourse. Languagearises from man's need to express himself, to ob­jectify himself. The essence of any form of lan­guage is somehow reduced to' the spiritual' cre­ativity of the individuum. Several other versionsof the function of language have been and arenow being suggested, but it is still typical to un­derestimate, if not altogether ignore, the commucnicative function of language. Language is re­garded from the speaker's standpoint as if therewere only one speaker who does not have anynecessary relation to other participants in speechcommunication. If the role of the other is takeninto account at all, it is the role of a listener, whounderstands the speaker only passively. The ut­terance is adequate to its object (I.e., the contentof the uttered thought) and to the person who ispronouncing the utterance. Language essentially

7See Wilhelm von Humboldt, Linguistic Variability andIntellectual Development (Coral Gables: University of MiamiPress, 1971). (Tr.]

needs only a speaker~one speaker-and an ob­ject for his speech. And if language also serves asa means of communication, this is a secondaryfunction that has nothing to do with its essence.Of course, the language collective, the pluralityof speakers, cannot be ignored when speaking oflanguage, but when defining the essence of lan­guage this aspect is not a necessary one that de­termines the nature of language. Sometimes thelanguage collective is regarded as a kind of col­lective personality, "the spirit of the people," andso forth, and immense significance is attached toit (by representatives of the "psychology of na­tions"),' but even in this case the plurality ofspeakers, and others with respect to each givenspeaker, is denied any real essential significance.

Still current in linguistics are such fictions asthe "listener" and "understander" (partners' of the"speaker"), the "unified speech flow," and s'o on.These fictions produce a completely distortedidea of the complex and multifaceted process ofactive speech communication. Courses in generallinguistics (even serious ones like Saussure's)frequently present graphic-schematic depictionsof the two partners in speech communication­the speaker and the listener (who perceives thespeech).....:...and provide diagrams of the activespeech processes of the speaker and the corre­sponding passive processes of the listener's per­ception and understanding of the speech. Onecannot say that these diagrams are false or thatthey do not correspond to certain aspects of real­ity. But when they are put forth as the actualwhole of speech communication, they become ascientific fiction. The fact is that when the lis­tener perceives and understands the meaning (thelanguage meaning) of speech, he simultaneouslytakes an active, responsive attitude toward it. Heeither agrees or disagrees with it (completely orpartially), augments it, applies it, prepares for its

8The phrase ''psychology of nations" refers to a school or­ganized around the nineteenth-century journal Zeitschrift fiirVolkerpsychologie und Sprachwissenschajt, whose leadingspokesman, Kermann Steinthal. was among the first to intro­duce psychology (especially that of the Kantian biologistHerbart) into language (and vice versa). Steinthal was at~

tractedto von Humboldt's idea of "innere-Sprachfonn" andwas important in Potebnya's attempts to wrestle with innerspeech. [Tr.]

1232 MODERN AND POSTMODERN RHETORIC

execution, and so on. And the listener adopts thisresponsive attitude for the entire duration of theprocess of listening and understanding, fromthe very beginning-sometimes literally fromthe speaker's first word. Any understanding oflive speech, a live utterance, is inherently respon­sive, although the degree of this activity variesextremely. Any understanding is imbued with re­sponse and necessarily elicits it ill one fann oranother: the listener becomes the speaker. A pas"sive understanding of the meaning of perceivedspeech is only an abstract aspect of the actual .whole of actively responsive understanding,which is then actualized in a subsequent responsethat is actually atticulated. Of course, an utter­ance is not always followed immediately by anarticulated response. An actively responsive un­derstanding of what is heard (a command, for ex­ample) can be directly realized in action (the exe­cution of an order or command that has been.understood and accepted for execution), or it canremain, for the time being, a silent responsive un­derstanding (certain speech genres are intendedexclusively for this kind of responsive under­standing, for example, lyrical genres), but this is,so to speak, responsive understanding with a de­layed reaction. Sooner or later what is beard andactively understood will find its response in thesubsequent speech or behavior of the listener. Inmost cases, genres of complex cultural communi­cation are intended precisely for this kind of ac­tively responsive understanding with delayed ac­tion. Everything we have said here also pertainsto written and read speech, with the appropriateadjustments and additions.

Thus, all real and integral understanding is ac­tively responsive, and constitutes nothing otherthan the initial preparatory stage of a response (inwhatever form it may be actualized). And thespeaker himself is oriented precisely toward suchan actively responsive understanding. He doesnot expect passive understanding that, so tospeak, only duplicates his own idea in someoneelse's ntind. Rather, he expects response, agree­ment, sympathy, objection. execution, and soforth (various speech genres presuppose variousintegral orientations and speech plans on the partof the speakers or writers). The desire to makeone's speech understood is only an abstract as-

pect of the speaker's concrete and total speechplan. Moreover, any speaker is himself a responedent to a greater or lesser degree. He is not, afterall, the first speaker, the one who disturbs theeternal silence of the universe. And he presup­poses not only the existence of the language sys"tern he is using, but also the existence of preced­ing utterances-his own arid others' -withwhich his given utterance enters into one kind ofrelation or another (builds on them, polenticizeswith them, or simply presumes that they are aleready known to the listener). Any utterance is alink in a very complexly organized chain of otherutterances.

Thus, the listener who understands passively,who is depicted as the speaker's partner in theschematic diagrams of general linguistics, doesnot correspond to the real participant in speechcommunication. What is represented by the dia"gram is only an abstract aspect of the real totalact of actively responsive understanding, the sortof understanding that evokes a response, and' onethat the speaker anticipates. Such scientific ab­straction is quite justified in itself, but under onecondition: that it is clearly recognized as merelyan abstraction and is not represented as the realconcrete whole of the phenomenon. Otherwise itbecomes a fiction. This is precisely the case inlinguistics, since such abstract schemata, whileperhaps not c1ainting to reftect real speech com­munication, are not accompanied by any indica~

tion of the great complexity of the actual phe"nomenOD. As a result, the schema distorts theactual picture of speech communication, remov­ing precisely its most essential aspects. The ac­tive role of the other in the process of 'speechcommunication is thus reduced to a minimum.

This disregard for the active role of the other inthe process of speech communication, and the de_sire generally to bypass this process, are mani­fested in the imprecise and ambiguous use of suchterms as "speech" or "speech ftow." These deliber­ately indefinite terms are usually intended to desig­nate something that can be divided into languageunits, which are then interpreted as segments oflanguage: phonetic (phoneme, syllable, speechrhythm [takt]) and lexical (sentence and word)."The speech flow can be broken down ..."; "OUfspeech is divided ..." This is the way those

BAKHTIN I THE PROBLEM OF SPEECH GENRES 1233

sections of grammars devoted to the study ofsuch language units are usually introduced intogeneral courses in linguistics and granunar, andalso into special research on phonetics and lexi­cology. Unfortunately, even our recently pub- .lished ~cademy grammar uses the same indefiniteand ambiguous tenn "our speech." Here is howthe section on phonetics is introduced:. "Ourspeech is basically divided into sentences, whichin turn can be broken down into phrases .andwords. The word is clearly divided into smallsound units-syllables . .. syllables are dividedinto individual speech sounds or phonemes ..."9

But what sort of thing is this "speech lIow" andwhat is meant by "our speech"? What is the natureoftheir duration? Do they have a beginning and anend? If their length is indefinite, which of theirsegments do we use when we break them downinto units? These questions have not been raised ordefined at all. Linguists have not yet transformedthe imprecise word "speech"- which can desig­nate language, the speech process (i.e., speaking),the individual utterance, an entire long indefiniteseries of snch utterances, or a particular speechgeme ("he gave a speech")-into a definite (de­fined) term with clear-cut semantic boundaries(similar situations also exist in other languages).This can be explained by the almost complete lack'of -research into the problem of the utterance andspeech gemes (and, consequently, of speech coo\­munion as well). What we almost always find is aconfused play with all these meaniugs (except forthe last). Most frequeutly the expression "ourspeech," simply means any utterances of any per­son. But this meaning is never consistently sus­tained throughout. W

9Grammatika russkogo jazYka (Grammar of the Russianlanguage) (Moscow, 1952), vol. I, p. 51. rTr.]

lOAnd it cannot be sustained. For example, such an utter­ance as "Ah!" (a rejoinder in dialogue) cannot be brokendown into sentences, phrases, or syllables. Consequently! notjust any utterance will do. Further, they divide up the utter­ance (speech) and obtain units of language. Frequently thesentence.is then defined as the simplest utterance and, conse­quently, it cannot be a unit of the utterance. It is tacitly as­sumed that there is only one speaker. and dialogical overtonesare thus ignored.

As compared to the boundaries of the utterance, all otherboundaries (between sentences, phrases, syntagmic units. andwords) are relative and arbitrary. [Au.]

And if it is indefinite and unclear just what itis that is divided and broken down into units oflanguage, this lack of definition and confusionalso spread to these units themselves.

The t~nninologicalimprecision and confusionin this methodologically central point of linguis­tic thinking result from ignoring the real unit ofspeech communication: the utterance. For speechcan exist in reality only in the form of concreteutterances of individual speaking people, speechsubjects. Speech is always cast in the form of anutterance belonging to a particular speaking sub­ject, and outside this form it cannot exist. Re­gardless of how varied utterances may be interms of their length, their content, and theircompositional structure, they have commonstructural features as units of speech communica­tion and, above all, quite clear-cut boundaries.Since these boundaries are so essential and fun­damental they must be discussed in detail.

The boundaries of each concrete utterance as aunit of speech communication are determined bya change ofspeaking subjects, that is, a change ofspeakers. Any utterance-from a short (single­word) rejoinder in everyday dialogue to the largenovel or scientific treatise - has, so to speak, anabsolute beginning and an absolute end: its be­ginning is preceded by the utterances of others,and its end is followed by the responsive utter­ances of others (or, although it may be silent, oth­ers' active responsive understanding, or, finally,a responsive action based on this understanding).The speaker ends his utterance in order to relin­quish the lIoor to the other or to make room forthe other's active responsive understanding. Theutterance is not a conventional unit, but a realunit, clearly delimited by the change of speakingsubjects, which ends by relinquishing the floor tcthe other, as if with a silent dixi, perceived by thelisteners (as a sign) that the speaker has finished.

This change of speaking subjects, which cre­ates clear-cut boundaries of the utterance, variesin nature and acquires different forms in the het­erogeneous spheres of human activity and life,depending on the functions of language and onthe conditions and situations of communication.One observes this change of speaking subjectsmost simply and clearly in actual dialogue wherethe utterances of the interlocutors or partners in

1234 MODERN AND POSTMODERN RHETORIC

dialogue (which we shall call rejoinders) alter~

nate. Because of its simplicity and clarity, dia­logue is a classic fann of speech communication~

Each rejoinder, regardless of how brief andabrupt, has a specific, quality of completion thatexpresses a particnlar position of the speaker, towhich one may respond or may assume, with re­spect to it, a responsive position. We shall dis~

cuss further this specific quality of completion ofthe utterance, one of its main markers. But at thesame time rejoinders are all linked to one an­other. And the sort of relations that exist amongrejoinders of dialogue-relations between ques­tion and answer, assertion and objection, asser­tion and agreement, suggestion and acceptance,order and execution, and so forth - are impos­sible among units of language (words and sen­tences), either in the system of language (in thevertical cross section) or within the utterance (onthe horizontal plane). These specific relationsamong rejoinders in a dialogue are only subcate­gories of specific relations among whole utter­ances in the process of speech ,communication.These relations are possible only among utter­ances of different speech subjects; they presup~

pose other (with respect to the speaker) partici­pants in speech communication. The relationsamong whole utterances cannot be treated gram~

matically since, we repeat, such relations are im­possible among units of language, and not onlyin the system of language, but within the utter­ance as well.

In secondary speech genres, especially rhe­torical ones, we encounter phenomena that ap­parently contradict this tenet. Quite frequentlywithin the boundaries of his own utterance thespeaker (or writer) raises questions, answersthem himself, raises objections to his own ideas,responds to his own objections, and so on. Butthese phenomena are nothing other than a .con­ventional playing out of speech communicati()nand primary speech genres. " This kind of playingout is typical of rhetorical genres (in the broadsense, which would include certain kinds of sci­entific popularization), but other secondary gen­res (artistic and scholarly) also nse various formssuch as this tointroduce primary speech genres

I (The seam of boundaries in secondary genres. [Au.]

and relations among them into the constructionof the utterance (and here they are altered to agreater or lesser degree, for the speaking sllbjectdoes not really change). Snch is the nature of sec­ondary genres. But the relations among the repro­duced primary genres cannot be treated grammat~

ically in any of these phenomena, even thoughthey appear within a single utterance. Within theutterance they retain their own specific nature,which is essentially different from the nature ofrelations among words and sentences (and otherlanguage units, i.e., phrases and so forth).

Here, drawing on material from dialogue andthe rejoinders that comprise it, we must provi­sionally pose the problem of the sentence as aunit of language, as distinct from the utterance asa unit of speech communication.

(The question of the nature of the sentence isone of the most <;omplicated and difficult in lin­guistics. The clash of opinions regarding thisquestion continues in our scholarship to this day.Of course, the task we set for ourselves here doesnot include an investigation of this problemin allits complexity; we intend to mention only one ofits aspects. But it seems to us that this aspect isessential to the entire problem. It is important forus to define precisely the relationship betweenthe sentence and the utterance. This will give us aclearer picture of both the utterance and the sen-.tence.)

But this will come later. Here,we shall simplynote that the boundaries of the sentence as a unitof language are never determined by a change ofspeaking subjects. Such a change, framing thesentence on both sides, transforms the sentenceinto an entire utterance. Such a sentence assumesnew qualities and is perceived quite differentlyfrom the way it would be if it were framed byother sentences within the single utterance of oneand the same speaker. The sentence is a rela­tively complete thought, directly correlated withthe other thoughts of a single speaker within hisutterance as a whole. The speaker pauses at theend of a sentence in order then to move on to hisown next thought, continuing, supplementing,and substantiating the preceding one. The contextof the sentence is the speech of one speaking sub­ject (speaker). The sentence itself is not corre­lated directly or personally with the extraverbal

BAKHTIN ITHE PROBLEM OF SPEECH GENRES 1235

context of reality (siloation, setting, prehistory)or with the utterances of other speakers; thistakes place only indirectly, through its entire sur'rounding context, that is, through the utterancea.a whole. Aud if the sentence is not surrounded bya context of speech of the same speaker, that is, ifit constilotes an entire completed utterance (a re­joinder in dialogue), then it (itself) directly con­fronts reality (the extraverbal context of thespeech) and the different utterances of others. Itis not followed by a pause that the speaker him­self designates and interprets. (Any pause that isgrammatical, calculated, or interpreted is pos­sible only within the speech of a single speaker,i.e., within a single utterance. Pauses between ut':'terances are, of course, not grammatical but real.Such real pauses-psychological, or promptedby some external circumstances-can also inter­rupt a single utterance. In secondary artistic gen­res such pauses are calculated by the artist, direc­tor, or actor. But these pauses differ essentiallyfrom both grammatical and stylistic pauses-forexample, among syntagmas-within the utter'ance.) Oue expects them to be followed by a re­sponse or a responsive understanding on the partof another speaker. Such a sentence, having be­come an entire utterance, acquires a special se­mantic fullness of value. One can assume a re­sponsive position with respect to it; one canagree or disagree with it, execute it, evaluate it,and so on. But a sentence in context cannot elicita response. It acquires this capability (or, rather,assimilates to it) only in the entirety of the wholeutterance.

All these completely new qualities and pecu­liarities belong not to the sentence that has bi­come a whole utterance, but precisely to the ut'terance itself. They reflect the nature of theutterance; not the natnre of the sentence. They at­tach themselves to the sentence, augmenting ituntil it is a complete utterance. The sentence as alanguage unit lacks all of these properties; it isnot demarcated on either side by a change ofspeaking subjects; it has neither direct contactwith reality (with an extraverbal siloation) nor adirect relation to others' utterances; it does nothave semantic fullness of value; and it has no ca­pacity to determine directly the responsive posi­tion of the other speaker, that is, it cannot evoke

a response. The sentence as a language unit isgrammatical in nature. It has grammatical bound­aries and grammatical completedness and unity.(Regarded in the whole of the utterance and fromthe standpoint of this whole, it acquires stylisticproperties.) When the sentence figures as a wholeutterance, it is as though it has been placed in aframe made of quite a different material. Whenone forgets this in analyzing a sentence, one dis;:,torts the nature of the sentence (and simultane­ously the nature of the utterance as well, by treat­ing it grammatically). A great many linguists andlinguistic schools (in the area of syntax) are heldcaptive by this confusion, and what they study asa sentence is in essence a kind of hybrid of thesentence (unit of language) and the utterance(unit of speech communication). One does notexchange sentences any more than one ex­changes words (in the strict linguistic sense) orphrases. One exchanges utterances that are con­sttucted from language units: words, phrases, andsentences. And an utterance can be consttuctedboth from one sentence and from one word, so tospeak, from one speech unit (mainly a rejoinderin dialogue), but this does not transform a lan­guage unit into a unit of speech communication.

The lack of a well-developed theory of the ut­terance as a unit of speech communication leadsto an imprecise distinction between the' sentenceand the utterance, and frequently to a completeconfusion of the two.

Let us return to real-life dialogue. As we havesaid, this is the simplest and ihe most classicform of speech communication. The change ofspeaking subjects (speakers) that detemunes theboundaries of the utterance is especially clearhere. But in other spheres of speech communica­tion as well, including areas of complexly orga­nized culloral communication (scientific andartistic), the nalore of the boundaries of the utter­ance remains the same.

Complexly structnred and specialized worksof various scientific and artistic genres, in spiteof all the ways in which they differ from rejoin­ders in dialogue, are by nalore the same kind ofunits of speech communication. They, too, areclearly demarcated by a change of speaking sub­jects, and these boundaries, while retaining theirexternal clarity, acquire here a special internal

MODERN AND POSTMODERN RHETORIC

aspect because the speaking subject- iu thiscase, the author of the work-manifests his ownindividuality iu his style, his world view, and inall aspects of the design of his work. This imprintof individuality marking the work also createsspecial internal boundaries that distinguish thiswork from other wQrks connected with it in theoverall processes of speech communication inthat particular cultural sphere: from the works ofpredecessors on whom the author relies, fromother works of the same school, from the worksof opposing schools with which the author iscontending. and so all.

The work, like the rejoinder in dialogue, isoriented toward the response of the other (oth­ers), toward his active responsive understanding,which can· assume various forms: educationalinfluence on the readers. persuasion of them,critical responses, influence on followers andsuccessors, 'and so on. It can detennine others'responsive positions under the complex condi­tions of speech communication in a particularcultural sphere. The work is a link in the chain ofspeech communion. Like the rejoinder in a dia­logue, it is related to other work-utterances: boththose to which it responds and those that respondto it. At. the same time, like the rejoinder in adialogue, it is separated from them by the abso­lute boundaries created by a change of speakingsnbjects.

Thus, the change of speaking subjects, byframing the utterance and creating for it a stablemass that is sharply delimited from other relatedutterances, is the first constitutive feature of theutterance as a unit of speech communication, afeature distinguishing it from units of language.Let us turn to this second feature, which is insep­arably linked to the first. This second feature isthe specific finalization of the utterance.

The finalization of the utterance is, if you will,the inner side of the change of speech subjects.Tbis change can only take place because thespeaker has said (or written) everything hewishes to say at a particular moment or underparticular circumstances. When hearing or read­ing, we clearly sense the end of the utterance, asif we hear the speaker's concluding dixi. This fi­nalization is specific and is determined by specialcriteria. The first and foremost criterion for the fi-

nalization of the utterance is the possibility of re­sponding to it or, more precisely and broadly, ofassuming a responsive attitude toward it (for ex­ample, executing an order). This criterion is metby a short everyday question, for example,"What time is it?" (one may respond to it), aneveryday request that one mayor may not fulfill,a scientific statement with which one may agreeor disagree (partially or completely), or a novel,which can be evaluated as a Whole. Some kind offinalization is necessary to be able to react to anutterance. It is not enough for the utterance to beunderstood in terms of language. An absolutelyunderstood and completed sentence, if it is a sen­tence and not an utterance comprised of one sen­tence, cannot evoke a responsive reaction: it iscomprehensible, but ilis still not all. This all~the indicator of the wholeness of the utterance­is subject neither to grammatical nor to abstractsemantic definition.

This finalized wholeness of the utterance,guaranteeing the possibility of a response (or ofresponsive understandiug), is determined bythree aspects (or factors) that are inseparablylinked in the organic whole of the utterance: (1)semantic exhaustiveness of the theme; (2) thespeaker's plan or speech will; (3) typical compo'sitional and generic forms of finalization.

The first aspect- the referential and semanticexhaustiveness of the theme of the utterance­differs profoundly iu various spheres of commu­nication. This exhaustiveness can be almost com­plete in certain spheres of everyday life(questions that are purely factual and similarlyfactual responses to .them, requests, orders. andso forth), in certain business circles, in the sphereof military and industrial commands and orders,that is, in those spheres where speech genres aremaximally standard by nature and where the ere,.ative aspect is almost completely lacking. Con­versely, in creative spheres (especially, of course,in scientific ones), the semantic exhaustivenessof the theme may be only relative. Here one canspeak only of a certain minimum of finalizationmaking it possible to occupy a responsive posi­tion. We do not objectively exhaust the subject,but, by hecoming the theme of the utterance (i.e.,of a scientific work) the subject achieves a rela­tive finalization under certain conditions, when

BAKHTINI THE PROBLEM OF SPEECH GENRES 1237

the problem is posed in a particular way, on thebasis of particular material, with particular aimsset by the author, that is, already within theboundaries of a specific authorial intent. Thus,we inevitably come to the second aspect, whichis inseparably linked to the first.

In each utterance- from the single-word;everyday rejoinder to large, complex works ofscience or literature- we embrace~ understand,:and sense the speaker's speech plan or speechwill, which determines the entire utterance; itslength and boundaries. We imagine to ourselveswhat the speaker wishes to say. And we also usethis speech plan, this speech will (as we under­stand it), to measure the finalization of the utter­ance. This plan determines both the choice ofthesubject itself (under certain conditions of speechcommunication, in necessary connection withpreceding utterances), as well as its boundariesand its semantic exhaustiveness. It also deter­mines,- of course, the choice of a generic fonn inwhich the utterance will be constructed (this isalready the third aspect, to which we shall turnnext). This plan- the subjective aspect of the ut­terance-combines in an inseparable unity withthe objective referentially semantic aspect, limit­ing thelalter by relating it to a concrete (individ­ual) situation of speech communication with allits individual circumstances, its personal partici­pants, and the statement-utterances that precededit. Therefore, the immediate participants in com­munication, orienting themselves with respect tothe situation and the preceding utterances, easilyand quickly grasp the speaker's speech plan, hisspeech will. And from the very beginning of hiswords they sense the developing whole of the llt­terance.

Let us turn to the third and, for us, most im­portant aspect: the stable generic forms of the ut­terance, The speaker's speech will is manifestedprimarily in the choice of a particular speechgenre. This choice is determined by the specificnature of the given sphere of speech communica­tion, semantic (thematic) considerations, the con­crete situation of the speech communication, thepersonal composition of its participants, and soon. And when the speaker's speech plan with allits individuality and subjectivity is applied andadapted to a chosen genre, it is shaped and devel-

oped within a certain generic form. Such genresexist above all in the great and multifarioussphere of everyday oral communication, includ­ing the most familiar and the most intimate.

We speak only in definite speech genres, thatis, all our utterances have definite and relativelystable typical fonns ofconstruction of the whole:Our repertoire of oral (and written) speechgenresis rich. We use them confidently and skillfully inpractice, and it is quite possible for us not evento suspect their existence in theory. LikeMoliere's Monsieur Jourdain who, when speak­ing in prose, had no idea that was what he wasdoing, we speak in diverse genres without sus­pecting that they exist. Even in the most free, themost ,unconstrained conversation, we cast ourspeech in definite generic forms, sometimes rigidand trite ones, sometimes more flexible, plastic,and creative ones (everyday communication alsohas creative genres at its disposal). We are giventhese speech genres in almost the same way thatwe are given OUf native language, which we mas­ter fluently long before we begin to study gram­mar. We know our native language-its lexicalcomposition and grammatical structure-not fromdictionaries and granunars but from concrete ut­terances that we hear and that we ourselves repro-­duce in live speech communication with peoplearound us. We assimilate forms of language onlyin fonus of utterances and in conjunction withthese forms. The forms of language and the typi­cal forms of utterances, that is, speech genres,enter our experience and our consciousness to­gether, and ,in close connection with one another.To learn to speak means to learn to constructut­terances (because we speak in utterances and notin individual sentences, and, of course, not in in­dividual words). Speech genres organize ourspeech in almost the sarne way as granunatical(syntactical) forms do. We learn to cast ourspeech in generic forms and, when hearing oth­ers' speech, we guess its genre from the very firstwords; we predict a certain length (that is, the ap­proximate length of the speech whole) and a cer­tain compositional structure; we foresee the end;that is, from the very beginning we have a senseof the speech whole, which is only later differen­tiated during the speech process. If speech genresdid not exist and we had not mastered them, if we

MODERN AND POSTMODERN RHETORIC

had to originate them during the speech processand conslructeach utterance at will for the firsttime, speech communication would be almostimpossible.

The generic forms in which we cast ourspeech, of course, differ essentially from lan'goage forms. The latter are stable and compul­sory (normative) for the speaker, while genericforms are much more flexible, plastic, and free;Speech genres are very diverse in this respect. Alarge number of genres that are widespread· ineveryday life are so standard that the speaker'sindividual speech will is manifested only in itschoice of. a particular genre, and, perhaps, in itsexpressive intonation. Such, for example, are thevarious everyday genres of greetings, farewells;congratulations, all kinds of wishes, informationabout health, business, and so forth. These genresare so diverse because they differ depending onthe situation, social position, and personal inter­relations of the participants in the communica-

, lion. These genres have high, strictly official, re'spectiul forms as well as familiar ones." Andthere are forms with varying degrees of familiar_ity, as well as intimate forms (wbich differ fromfamiliar ones). These genres also require a cer­tain tone; their structure includes a certain ex­pressive intonation. These genres, particularlythe high and official ones, are compulsory andextremely stable. The speech will,is usually lim­ited here to a choice of a particular genre. Andouly slight nuances of expressive intonation (onecan take a drier or more respectful tone, a colderor wanner one; one can introduc.e the intonationof joy, and so forth) can express the speaker's in­dividuality (his emotional speech intent). Buteven here it is generally possible to -reaccentuategenres. This is typical of speech communication:thus, for example, the generic form of greetingcan move from the official sphere into the sphereof familiar communication, that is, it can be usedwith parodic-ironic reaccentuation.,- To a similar

l:>.These and other phenomena have interested linguists(mainly language historians) in the purely stylistic level as areflection in language of historically changed forms of, eti~quette, courtesy, and hospitality. See, for example, F. Brunat,'Histoire de la langue fraTlfaise des origines a1900, 10 vals.(Paris: A. Colin, I90S). [Au.)

end, one can deliberately mix genres from vari­ous spheres.

In addition to these standard genres, of course,freer and more creative genres of oral speechcommunication have existed and still exist: gen~

res of salon conversations about everyday. social,aesthetic, and other subjects, genres of table can'versation. intimate conversations'among friends;intimate conversations within the family, and soon. (No. list of oral speech genres yet exists, oreven a principle on which such a list might bebased.) The majority of these genres are subjectto free creative reformulation (like artistic generes, and some, perhaps, to a greater degree). Butto use a genre freely and creatively is not thesame as to create a genre from the beginning;genres must be fully mastered in order to be rna'nipulated freely.

Many people who have an excellent commandof a language often feel quite helpless in certainspheres of communication precisely because theydo not have a practical command of the genericforms used in the given spheres. Frequently Iiperson who has an excellent command ofspeechin some areas of cultural communication, who isable to read a scholarly paper or engage in ascholarly discussion, who speaks very well onsocial questions. is silent or very awkward in so­cial conversation.-Here it is not a matter of an im­poverished vocabulary or of style, taken ab­stractly: this is entirely a matter of the inability tocommand a repertoire of genres of social conver­sation, the lack of a sufficient supply of thoseideas about the whole of the utterance that help tocast one's speech quickly and naturally iu certaincompositional and stylistic forms, the inability tograsp a word promptly, to begin and end cor­rectly (composition is very uncomplicated inthese genres).

The better our command of genres, the morefreely we employ them, the more fully andclearly we reveal our own individuality in them(where this is possible and necessary), the moreflexibly and precisely we reflect the unrepeatablesituation of communication-in a: word, .themore perfectly we implement our free speechplan.

Thus, a speaker is given not only mandatoryforms of the national language (lexical composi-

BAKHTINITHE PROBLEM OF SPEECH GENRES 1239

tion and grammatical structure), but also forms ofutterances that are mandatory, that is, speechgenres. The latter are just as necessary for mutualunderstanding as are forms of language. Speechgenres are much more changeable, flexible, andplastic than language forms are, but they have. anormative significance for the speaking indieviduum, and they are not created by him but aregiven to him. Therefore, the single utterance,with all its individuality and creativity, can in noway be regarded as a completely free combina­tion of forms of language, as is supposed, for ex,ample, by Saussure (and by many other linguistsafter him), who juxtaposed the utterance (la pa­role), as a purely individual act, to the system oflanguage as a phenomenon that is purely socialand mandatory for the individuum. '3 The vastmajority of linguists hold the same position, intheory if not in practice. They see in the utteranceonly an individual combination of purelylinguis­tic (lexical and grammatical) forms and they nei­ther uncover nor study any of the other normativefonns the utterance acquires in practice.

Ignoring speech genres as relatively stable andnormative forms of the utterance inevitably led tothe confusion we have already pointed out be~

tween the utterance and the sentence. and it hadto lead them to the position (which, to.be sure,was never consistently defended) that our speechis cast solely in stable sentence forms that aregiven to us; and the number of these interrelatedsentences we speak in a row and when we stop(end)-this is completely subject to the individ­ual speech will of the speaker or to the caprice ofthe mythical "speech flow."

When we select a particular type of sentence,we do so not for the sentence itself; but out ofconsideration for what we wish to express withthis one given sentence. We select the type of

13Saussure defines the utterance (La parole) as an "indi­vidual act. It is willful and intellectual. Within the act, weshould distinguish between (I) the combinations by which thespeaker uses the language code for expressing his ownthought; and (2) the psychological mechanism that allowshim to exteriorize those combinations" (Course in GeneralLinguistics [New York: McGraw-Hill, 19661, p. 14). Thus,Saussure ignores the fact that in addition to fonus of languagethere are also Jonns ofcombinations of these fonns. that is, heignores speech genres. [Au. J

sentence from the standpoint of the whole utter­ance. which is transmitted in advance ta ourspeech imagination and whicb determines ourchoice. The idea of the form of the whole utter­ance, that is, of a particular speech genre, guidesus in the process of our speaking. The plan of theutterance as a whole may require only one sen~

tence far its implementation, but it may also re­quire a large number of them. The chosen genrepredetermines for us their type and their compo­sitionallinks.

One reason why fonns of utterances are ig­nored in linguistics is that these fanns are ex­tremely diverse in compositional structure, par­ticularly in size (speech length)-from thesingle-word rejoinder to a large novel. There isalso a great range of sizes in oral speech genres.Thus, speech genres appear incommensurableand unacceptable as units of speech.

This is wby many linguists (mainly those in­vestigating syntax) try to find special forms thatlie somewhere between the sentence and the ut­terance, forms with the completeness of the utter­ance and at the same time the commensurabilityof the sentence. Such are the "phrase" (i.e., inKartsevsky) and "communication" (in Shakbma­tOY and others).14 There is no common under-

14S. D. Kartsevsky. Russian linguist of the Geneva Schoolwho also participated in the Prague Linguistic Circle. He ar­gued that the "phrase" should be used as a different kind oflanguage unit.from that of the sentence. Unlike the sentence,the phrase "does not have its own grammatical structure. Butit has its own phonetic structure, which consists in its intona~

tion. It is intonation that fonns the phrase" (S. Karcewski,"Sur la phonologie de 1a phrase," in Travaux du Cercle lin­guistique de Prague 4 [1931), 190). 'The sentence. in orderto be realized, must be given the intonation of the phrase. ...The phrase is a function of dialogue. It is a unit of exchangeamong conversing parties...." (S. Karcewski, "Sur laparataxe et la syntaxe en russe," in Cahiers Ferdinand deSaussure, no. 7 [1948J. 34)·

Aleksey Shakhmatov (I864-1920), linguist and academi­cian whose most important works were devoted to the historyof the Russian language, modern Russian, and comparativestudies of the grammars of different Slavic languages. "Com­munication" has a rather distinctive meaning for Shakhma~

tov: it refers to the act of thinking. this being the psychologi­cal basis of the sentence. the mediating link "between thepsyche of the speaker and its manifestation in the discoursetoward which it strives" (A. Shakhmatov, Sintaksis rnsskogojazyka [SyntaX of the Russian language] [Leningrad, £941],pp. 19-20). [Tr.]

1240 MODERN AND POSTMODERN RHETORIC

standing of these units among researchers whouse them because no definite and clearly delim"ited reality corresponds to them in the life of lan­guage. All these artificial and conventional unitsneglect the change of speech subjects that takesplace in any real live -speech communication, andtherefore the most essential boundaries areerased in all spheres of language activity: bound"aries between utterances.' Hence (in consequenceof this) one also forfeits· the main criterion for thefinalization of the utterance as a true unit ofspeech communication: the capability of deter­mining the active responsive position of the otherparticipants in the communication.

We shall conclude this section with a fewmore remarks about the sentence (and return todiscuss this issue in detail in the summary of ouressay).

The sentence as a unit of language lacks thecapability of determining the directly active re"sponsive position of the speaker. Only after be­coming a complete utterance does the individualsentence acqnire this capability. Any sentencecan act as a complete utterance, but then, as', weknow, it is augmented by a number of very es­sential nongrammatical aspects that change itradically. And this circumstance also causes aspecial syntactic aberration. When the individualsentence is analyzed separately from its context,it is interpreted to the point of becoming a wholeutterance. As a result. it acquires that degree offinalization that makes a response possible.

The sentence, like the word, is a signifyingunit oflanguage. Therefore, each individual sen"tence, for example, "The sun has risen," is com­pletely comprehensible, that is, we understand itslanguage meaning, its possible role in an utter­ance. But in no way can we assume a responsiveposition with respect to this individual sentenceunless we know that with this sentence .thespeaker has said everything he wishes to say, thatthis sentence is neither preceded nor followed byother sentences of the same speaker. But thenthis is no longer a sentence, but a full-fledged ut­terance consisting of one sentence. It is framedand delimited by a change of speech subjects andit directly reflects an extraverbal reality (situa­tion). It is possible to respond to such an utter­ance.

But if this sentence were surrounded by con.,text, then it would acquire a fullness of its ownsense only in this context, that is, only in thewhole of the utterance, and one could respondonly to this entire utterance whose siguifying ele­ment is the given sentence, The utterance, for ex..,ample, can be thus: "The sun has risen. It's timeto get up." The responsive understanding (or ar­ticulated response): "Yes, it really is time." But itcan also be thus: "The sun has risen. But it's stillvery early. Let's get some more sleep." Here thesense of the utterance and the responsive reactionto it will be different. Such a sentence can alsoenter into the composition of an artistic work asan element of landscape. Here the responsive re­action - the artistic-ideological impression andevaluation-can pertain only to the entire land­scape. In the context of another work this sen­tence can acquire symbolic significance. In allsuch cases the sentence is a signifying element ofthe whole utterance, which acquires its finalmeaning only in this whole.

If our sentence figures as a completed utter­ance, then it acquires its own integral sense underthe particular concrete circumstances of speechcommunication. Thus, it can be a response to an·other's question: "Has the sun risen?" (of course,under the particular circumstances that justifythis question). -Here this utterance is an assertionof a particular fact, an assertion that can be trueor false, with which one can agree or disagree·. Asentence that is assertive in its form becomes areal assertion in the context of a particular utter­ance.

When this individual sentence is analyzed, itis.· usually perceived as a completed utterance insome extremely simplified situation: the sun re­ally has risen and the speaker states: "The sunhas risen." The speaker sees that the grass isgreen and anoounces: "The grass is green." Suchsenseless "communications" are often directly re~

garded as classic examples of the sentence. Butin reality any communication like that, addressedto someone or evoking something, has a. parti­cular purpose, that is, it is a real link in the chainof speech communion in a particular sphere ofbuman activity or everyday life.

The sentence, like the word, has a finality ofmeaning and a finality of grammatical form, but

BAKHTIN I-THE PROBLEM OF SPEECH GENRES 1241

this finality of meaning is abstract by natore andthis is precisely why it is so clear-cnt: this is thefinality of an element, but not of the whole. Thesentence as a unit oflanguage, like the word, hasno author. Like the word, it belongs to nobody,and only by functioning as a whole utterancedoes it become an expression of the positioncofsomeone speaking individually in aconcrete situ..:ation of speech commnnication. This leads us toa new, third feature of the utterance-the rela'tion of the utterance to the speaker himself (theauthor of the utterance) and to the other partici­pants in speech communication.

A1ly utterance is a link in the chain of speechcommunion. It is the active position of thespeaker in one referentially semantic sphere oranother. Therefore, each utterance is character'ized primarily by a particular referentially se­mantic content. The choice- of linguistic meansand speech genre is detennined primarily by thereferentially semantic assignments (plan) of thespeech snbject (or author). This is the first aspectof the utterance that detennines its compositionaland stylistic featores.

The second aspect of the utterance that deter­mines its composition and style is the expressiveaspect, that is, the speaker's subjective emotionalevaluation of the referentially semantic contentof his utterance. The expressive aspect has vary­ing significance and varying degrees of force invarious spheres of speech communication, but itexists everywhere. There can be no such thing asan absolutely neutral utterance. The speaker'sevaluative attitude toward the subject of hisspeech (regardless of what his subject may be)also determines the choice of lexical, grammatbcal, and compositional means of the utterance.The individual style of the utterance is deter­mined primarily by its expressive aspect. This isgenerallyrecognized in the area of stylistics. Cer­tain investigators even reduce style directly to theemotionally evaluative aspect of speech.

Can the expressive aspect of speech be re­garded as a phenomenon of language as a sys­tem? Can one speak of the expressive aspect oflanguage units, that is, words and sentences? Theanswer to these questions must be a categorical"no." Language as a system has, of course, a richarsenal of langnage tools -lexical, morphologi-

cal, and syntactic~for expressing the speaker'semotionally evaluative position, but all thesetools as language tools are absolutely neutralwith respect to any particular real evaluation. Theword udarling-" - which. is affectionate in ..boththe meaning of its root and its suffix-is in·it­self, as a language unit, just as neutral as theword "distance." J5 It· is only a language, tool forthe possible expression ofan emotionally evalua­tive attitude toward reality, but it is not applied toanyparticullrr reality;.and this application, that is,the actual evaluation, can be accomplished onlyby the speaker in his concrete utterance, Wordsbelong to nobody, and in themselves they evalu~

ate nothing. But they can serve any speaker andbe used for the most varied and directly contra­dictory evaluations on the part of the· speakers.

The sentence as a unit of language is also·neu­tral and in itself has no expressive aspect. It ac­quires this expressive aspect (more precisely,joins itself to it) only in a concrete utterance. Thesame aberration is possible here. A sentence like"He. died" obviously embodies a certainexpres­siveness, and a sentence like "What joy!" does soto an even greater degree. But in fact we perceivesentences of this kind as entire utterances, and ina typical situation, that is,. as kinds of speech gem­res that embody typical expression. As sentencesthey lack-this expressiveness and are neutral. De­pending on'the' context of the utterance, the sen-'.tence "He died" can also reflect a positive, joyful,even a rejoicing expression. And the. sentence"What joy!" in the context of the particular utter­ance can assume ,an ironic or bitterly sarcastictone.

One of the means of expressing the speaker'semotionally evaluative attitude toward the sub­ject of his speech is expressive intonation, whichresounds clearly in oral speech.'6 Expressive in­tonation is a constitutive marker of the utterance.It does not exist in the system of language assuch, that is, outside the utterance. Both the wordand the sentence as language units are devoid of

ISThe Russian word· Bakhtin uses here (milenkiJ) is adiminutive. of milyj, itself a term of endeannent meaning"nice" Dr "sweet." [Tf.]

16Qf course, intonation is recognized by us and exists as astylistic factor even with silent reading of written speech.[Au.]

MODERN AND POSTMODERN RHETORIC

expressive intonation. If an individual word ispronounced with expressive intonation it -is _nolonger a word, but a completed utterance ex­pressed.by one word (there is no need to developit into a sentence). Fairly standard types ofevalu,ative. utterances. are very widespread in speechcommunication, that is, evaluative speech ,genresthat .express praise, approval, rapture, reproof, orabuse:. '~Excenent!" ,~'Good for, you!" "Charm­ing'" "Shame!"· "Revolting!" "Blockhead'" andso forth. Words that acquire special weight underparticular conditions of sociopolitical life be­come, expressive exclamatory utterances:"Peace!" "Fre.edom!".andso forth. (These consti­tute a special. sociopolitical speech geure.) In aparticular situation a word can acquire a pro­foundly expressive.meaning in the form of an ex­clamatory utterance: ."Thalassa, Thalassa!" [Thesea! The sea!] (exclaimed 10,000 Greeks inXenophon).

In each of these cases we.are dealing not withthe individual·word as a unit of language and notwith the meaning of this word but with a com­plete utteranc.e and·with a specific sense-thecontent of a given utterance}? Here the meaningof the word pertains to a particttiar actual realityand.particular real conditions of speech commu­nication. Therefore here we do not understandthe meaning of a given word simply as a word'ofa language; rather. we assume an active respon­sive position with respect to it (sympathy, agree­ment or disagreement; stimulus to action). Thus,expressive intonation belongs to the utteranceand not to .the word. But still it is very difficult toabandon the notion that each word of a languageitself has-or can have, an "emotional tone," "emo­tionaJ..coloring," an-,"evaluative aspect," a "stylis:­tic aura." and.so forth, and, consequently, also anexpressive intonation that is inherent in the wordas such. After all, one might thi.nk that when se­lecting a word for an utterance we are guided byan emotional tone inherent in the individual

171n Marxism and the Philosophy of Language. the spe­cific sense of an utterance is defined as its theme (tema): 'Thetheme, of.an utterance is essentially individual and unrepeat­able, like the utterance itself.. . The theme of the utteranceis ~ssentially' indivisible: The significance of the utterance; onthe contrary, breaks down into a number of significances thatare included in its linguistic' elements" (pp. 101-2). (Tr.l

word:_ we select those that in their tone, corre;spond to the expression of our utterance and wereject others. Poets themselves' describe. theirwork on the word in precisely this way; and thisis ,precisely ,the way this process is interpreted instylistics (see Peshkovsky's "stylistic experi­menf').I8

But still this is not what really happens. IUsthat same, already familiar aberration. Whense­lecting words we proceed from. the plannedwhole of our utterance, '9 and this whole that wehave planned .and created is always expressive,The utterance is what radiates its expression(rather, our expression) to the word we have se­lected, which is to say, invests the word with theexpression of the whole. And we select the wordbecause of its meaning, which. is not in itself ex­pressive but which can accommodate or not" ac­commodate our expressive goals in combinationwith other words, that is, in combination with thewhole of our utterance. The neutral meaning ofthe word applied to a particular actual realityunder particular real conditions of speech com­munication creates a spark of. expression. And1,

after all, this is precisely what takes place.in theprocess of creating an utterance. We repeat, only,the contact between the language meaning andthe concrete reality that takes place in the utter­ance can create the spark of expression. It exists·neither in the system of language nor in the ob,jective reality surrounding us.

Thus, emotion, evaluation,. and expressionare foreign to the word of language and are bornonly in the' process of its live usage in· a con-.crete utterance. The meaning of a wordin itself

,8Aleksandr Peshkovsky ([878-[933), Soviet linguistspecializing in grammar and stylistics in the schools. His,"stylistic experiment" consisted in artificially devising stylis·tic variants of the text, a device he used for analyzing.artisticspeech. See A. M. Peshkovsky, Voprosy metodiki rodnogojazyka. lingvistiki i stilistiki (Proble;ms in the methodolog:yof folk language, linguistics, and stylistics) (Moscov.:­Leningrad, 1930), p. 133. ,[Tr.]

19When we construct oUf'speech, we are always aware ofthe whole of our utterance: both in the fonn of a' particulargeneric plan and in the form of an individual speech plan. Wedo nm string words together smoothly and we do not proceedfrom word to word; ra'ther, it is as though we fill in the wholewith the necessary words. Words are strung together only inthe first stage of the study of a foreign language, and -thenonly when the methodological guidance is poor. [Au.]

BAKHTIN -I TaB PROBLEM OF SPEECH GENRES 1243

(umelated to actual reality) is, as we have alreadysaid, out of the range of emotion. There arewords that specifically designate emotionsand evaluations: "joy," "sorrow," "wonderful~"

'lcheerful," "sad," and so forth. But these mean..ings are just as neutral as are all the others. Theyacquire their expressive coloring only in the ut­terance, and this coloring is independent of theirmeaning taken individually and abstractly. Forexample: "Any joy is now-only bitterness to me."Here the word "joy" is given an expressive into­nation that resists its own·meaning, as it were.

But the- above far from exhausts the question.The matter is considerably more complicated.When we select words in the process of con­structing an utterance, we by no means alwaystake them from the system of language in theirneuttal, dictionary form. We usually take themfrom other utterances, and mainly from utter­ances that are kindred to ours in geme, that is, intheme, composition, or style. Consequently, wechoose words according to their generic specifi­cations. A speech genre is not a form of lan­guage, but a typical fonn of utterance; as suchthe geure also includes a certain typical kind ofexpression that inheres in it. In the genre theword acquires a particular typical expression.Genres correspond to typical situations of speechcommunication, typical themes, and, conse­quently, also to particular contacts between themeanings of words and actual concrete realityunder certain typical circumstances. Hence alsothe possibility of typical expressions that seem toadhere to words. This typical expression (and thetypical intonation that corresponds to it) does nothave that force of compulsoriness rhat languagefonns have. The generic normative quality isfreer. In our example, "Any joy is now bitternessto me," the expressive tone of the word "joy" asdetermined by the context is, of course, not typi­cal of this word. Speech genres in general submitfairly easily to reaccentuation, the sad can bemade jocular and gay, but as a result somethingnew is achieved (for example, the geme of comi­cal epitaphs).

This typical (generic) expression can be re­garded as the word's "stylistic aura," but thisaura belongs not to the word of language as suchbut to that genre in which the given word usually

functions. It is· an echo of the generic whole thatresounds in the word,

The word's generic expression-and itsgeneric expressive intonation-are impersonal,as speech genres themselves are impersonal (forthey are typical fonns of individual utterances,but not the utterances themselves). But words canenter our speech from others' individual utter­ances; thereby retaining to a greater or lesser de­gree the tones and echoes of individual utter­ances.

The words of a language belong to nobody,bur still we hear those words only in particularindividual utterances, we read them in particularindividual works, and in such tases the words al­ready have not only a typical, but also (depend­ing on the geme) a more or less clearly reflectedindividual context of the utterance.

Neutral dictionary meanings of the words of alanguage ensure their common features and guar­antee that all speakers of a given language willunderstand one another, but the use of words inlive speech communication is always individualand contextual in nature. Therefore, one can saythat any word exists for the speaker in three as­pects: as a neutral word of a language, -belongingto nobody; as an other's word, which belongs-toanother person and is filled with echoes of theother's utterance; and, finally, as my word, for,since I am dealing with it in a particular situation,with a particular speech plan, it is already im­bued with my expression. In both of the latter as­pects, the word is expressive, but, we repeat, thisexpression does not inhere in the word itself. Itoriginates at the point of contact between theword and actual reality, under the conditions ofthat real situation articulated by the individual ut­terance. In this case the word appears as an ex­pression of some evaluative position of an indi­vidual person (authority, writer, scientist, father,mother, friend, teacher, and so forth), as an ab­breviation of the utterance.

In each epoch, in each social circle, in eachsmall world of family, friends, acquaintances,and comrades in which a human being grows andlives, there are always authoritative utterancesthat set the tone-artistic, scientific, and journal­istic works on which one relies, to which onerefers, which are cited, imitated, and followed. In

1244 MODERN AND POSTMODERN -RHETORIC

each epoch, in all areas of life and activity, thereare particular traditions that are expressed and re­tained in verbal vestments: in written works, inutteninces, in sayings, and so' forth. There are al- 'w.ays some v;erbally expressed leading ideas ofthe "masters of thought" of a .given epoch, some

,basic tasks; slogans~' and so forth, I am not evenspeaking abont those examples from school read'ers with which children study dieir native Ian:'guageandwhich, of course, are always expres'sive.

This is why the unique speech experience ofeach individual is shaped and developed; in con-

tinuou~ahd canst'ant interaCtidn with others' in­dividual utterances. This experience can be char­acterized to some degree as the process of assim­ilation-more or less creative-of others',words (and not the words of a language). Ourspeech, that is, all our utterances (including cre­ative works), is filled with others' words, varyingdegrees of otherness or varying degrees of "our­own-ness," 'varying degrees' of awareness anddetachment. These words' of others carty withthem their own expression, their own evaluative

, tone, which we assimilate, rework, and reaccen-tuate....

BAKHTIN ITHE PRoBLEM OF SPEECH GENRES 1Z45