Cohesion in Russian - A Model For Discourse Analysis [PDF]

Citation preview

American Association of Teachers of Slavic and East European Languages



Cohesion in Russian: A Model for Discourse Analysis Author(s): Cynthia Simmons Source: The Slavic and East European Journal, Vol. 25, No. 2 (Summer, 1981), pp. 64-79 Published by: American Association of Teachers of Slavic and East European Languages Stable URL: http://www.jstor.org/stable/307957 Accessed: 12/09/2010 01:38 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=aatseel. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].



American Association of Teachers of Slavic and East European Languages is collaborating with JSTOR to digitize, preserve and extend access to The Slavic and East European Journal.



http://www.jstor.org



Slavic and East EuropeanJournal



COHESION IN RUSSIAN: A MODEL FOR DISCOURSE ANALYSIS Cynthia Simmons, University of Wisconsin-Madison



In their book Cohesion in English,1 M.A.K. Halliday and Ruqaiya Hasan categorized those items or operations which are available to the speaker of English for the establishment of relationships between sentences-for the creation of a text. By choosing any item or device which presupposes information in another sentence, the speaker (writer) establishes an intersentential cohesive "tie" within the text. Halliday and Hasan attempted to classify all such text-forming mechanisms in English. The implications of these linguists' work are many. For example, any English text can be analyzed definitively as to the number and kind of cohesive ties. Texts can be compared in terms of the types and the degree of cohesion. The student of English composition can learn to recognize and then to provide unity in a text. The categories of cohesion can be applied in an analysis of other languages and thereby, the concept might attain universal significance. In this paper, Halliday and Hasan's description of cohesion will be adapted to Russian.2 This study is only one of many which follow naturally from a reading of Cohesion in English. In Halliday and Hasan's terminology, "cohesion" is a product of certain options in the linguistic system per se. They are to be differentiated from structural choices which must be made as to information structure and topic versus comment.3 Although these latter are textual phenomena and do contribute to coherence in discourse, they do not provide the wide range of options or allow for the same degree of variation between texts as do the "non-structural" devices of cohesion. Halliday and Hasan discovered five types of cohesion in English: Reference, Substitution, Ellipsis, Conjunction, and Lexical Cohesion. For reasons which will be made clear, an analysis of Russian revealed only four of the above categories-substitution is not a viable category of cohesion in Russian. SEEJ, Vol. 25, No. 2 (1981) 64



Cohesion in Russian



65



Before each of the types of cohesion is described, it is necessary for our purposes to define both the "text" and the "sentence" since there is no general consensus on the definitions of these terms. Any string characterized by a definitive intonational contour followed by a full stop can constitute a sentence.4 This definition does not restrict the sentence to any minimal size, as an appropriate intonational contour followed by a full stop can be imposed on any string. This definition can limit the length of a sentence, however, in that it enables the listener/reader to establish sentence boundaries (albeit somewhat impressionistically) in speech or in a text which lacks the expected acoustic or orthographic cues. It is a definition which is useful in the analysis of texts in which the author has purposely disregarded punctuation.



NEUTRAL



SELECTIVE



I



FAR



NEAR I PARTICIPANT



6tot---



tot



tut zdes'



tam



0 PLACE



(ENGLISH "THE") CIRCUMSTANCE



TIME



sejcas teper'



togda



As the name implies, comparative reference involves a comparison of some sort. It may be a matter of likeness or non-likeness (general comparative reference), or it may be a comparison between the quantity or quality of items (particular comparative reference). The table below presents some examples of Russian words which signal comparative reference:



COMPARISON GENERAL



IDENTITY



DETERMINER



ADV



tot le samyj



ravno



odinakovyj



tak



DET



SUBST



to Ze samoe QUANTITY



SIMILARITY



poxotij podobnyj sxodnyj



NON-SIMILARITY



raznyj razlicnyj



drugoj inoj



bol'ge men'se e?te



podobno



raznoobrazno razlicno QUALITY



NON-IDENTITY



A



po-drugomu inate



co com ad



Cohesion in Russian



67



The text is defined here as a spoken or written passage of any length which represents a single unit or whole. The coherence of a passage depends then on its textual unity. The concept of cohesion provides the means for analyzing one aspect of that unity. A series of sentences will constitute a text if each sentence is cohesive with the surrounding (usually preceding) discourse by at least one (cohesive) tie.5 Of course cohesion alone does not insure the unity of the text. There is the question of the text's appropriateness with respect to external reality, its correspondence to the context of situation.6 One must keep in mind, however, that there are texts, belles-lettres being the obvious example, the coherence of which is not dependent primarily or even secondarily on their relationship to external reality.7 There are yet other factors which may lend coherence to a text. These are purely formal devices, such as rhyme, meter, alliteration, syntactic parallelism, and so forth. Literary texts may depend heavily for their coherence on the formal devices just mentioned. In the majority of nonliterary texts, however, such devices are not consciously employed as a source of coherence. The claim here is that concerns of appropriateness and purely structural or purely formal means of coherence are clearly secondary to cohesion in their importance in the text-forming process: the transmission of meaning from sentence to sentence. In fact, it is quite possible that none of these other factors is sufficient to create a text unless it is first cohesive, that is, unified by a series of cohesive ties.8 As stated earlier, only four of the five types of cohesion suggested for English were found in Russian. These will be described briefly; then there will follow a discussion of substitution and why it is not a category of cohesion in Russian. Reference describes the process by which items in the language are interpreted by means of information from the context (accompanying text) or the context of situation (as defined by Malinowski; see Note 7). For reference to be cohesive, the information necessary to interpret a reference item must be located in another sentence within the text. The three sub-types of reference described by Halliday and Hasan, personal, demonstrative, and comparative, are also adequate for an analysis of Russian. Personal reference is signified in Russian by the personal pronouns and possessive determiners. Although virtually all of these are referential, there is a major division between first and second persons and third person as they relate to both the speech situation and cohesion. In the speech situation, there is a speaker and an interlocutor. The third person merely indicates "not one of the above." This meaning can then be extended to "that one" or "it" and can denote objects or abstract



68



Slavic and East European Journal



entities, or it can take on a demonstrative connotation. Because the first and second persons primarily indicate roles in the speech situation, they often refer to empirical reality, not to items within the text, and therefore do not contribute to cohesion. Third-person pronouns and possessive determiners are not fundamental to the speech situation. They indicate other persons or things, usually located in the text, and thus, are characteristically cohesive. This is not to say that first- and secondperson pronouns cannot be cohesive. In a dialogue within a narrative, their ultimate referents are within the text and they provide cohesion. Likewise, third-person pronouns and possessives can indicate a person who is not identified in the text; in this case they would not be cohesive. Items of demonstrative reference convey the proximity of elements in the communication process. They can specify the location of objects in external reality as well as the proximity of items in a text. In the latter case, demonstrative reference effects cohesion. The following table illustrates demonstrative reference in Russian:9 The second type of cohesion, ellipsis, describes the situation where necessary structures or parts of a sentence are omitted, and the missing information can be retrieved from the accompanying text or the context of situation. Again, cohesion results only when the omitted information is to be found in another sentence in the text. Ellipsis, unlike reference, is primarily a grammatical relation. If a word or clause is omitted in a sentence, that omission presupposes a word of the same grammatical class or a clause of the same grammatical type. For this reason, ellipsis is sub-categorized on the basis of a grammatical analysis of the process: Ellipsis A. Nominal Ellipsis 1. Ellipsis of Nominal Phrase with Adjectiveas Head 2. .. .With Post-Determineras Head 3. .. .With Determiner as Head B. Verbal Ellipsis 1. Lexical Elipsis 2. OperatorEllipsis C. Clausal Ellipsis 1. PropositionalEllipsis 2. ModalEllipsis 3. General Ellipsis of the Clause (all elements but one omitted;usually occursin question-answersequence) 4. ZeroEllipsis (entire clause ommitted)



It is not suitable in a paper of this length to describe in detail all the grammatical analyses (of the noun phrase, verb phrase, and clause) upon which the sub-categorization of ellipsis is based. However, since this particular analysis of the verb phrase as consisting of a lexical verb



Cohesion in Russian



69



and perhaps an operator is not so common in a treatment of Russian, a brief explanation would no doubt be helpful. Here "lexical" verb denotes the main verb, that verb which can be itself represent a complete verbal group. The vast majority of Russian verbs fall into this group: sidet', utit'sja, dumat', and so forth. In contrast, verbal operators cannot represent complete verbal groups (unless, of course, there is ellipsis of the lexical verb). Verbal operators are found in analytic (compound) verbal constructions. Examples of verbal operators in Russian are the tense operator byt: "Cto vy budete delat'?" (What will you do?) and the modal operator mo6' as in "Cto ja mogu skazat'?" (What can I say?). There are other modal operators; smet' starat'sja, namerevat'sja, and so forth. Viewing the verb phrase in Russian as comprising these two components, lexical verb and possibly an operator, is useful in describing ellipsis of structures in the verb phrase.10 The third type of cohesion in Russian is called conjuction and results when certain relationships between sentences are made explicit grammatically by the use of some conjunction, adverb, or prepositional phrase. Halliday and Hasan proposed four categories of relationships which hold between sentences. Their analysis, although not indisputable, was accepted on a general level for Russian since it led to an adequate description of Russian "conjunctives." These four types of conjunctive relationships: additive, adversative, causal, and temporal, can be represented in Russian by the words "i," "odnako," "tak," and "potom" respectively. It must be remembered that these conjunctives and others that fall under the categories they represent always operate between sentences. Thus the conjunctive "i"should not be confused with the conjunction "i"which signals coordination within the sentence. 1 The final type of cohesion to be defined is lexical cohesion. This type differs from the others in that it does not involve some grammatical relation but is a matter of vocabulary. That is, a word forms a lexical tie with some other item because of the combined factors of their intrinsic relation through meaning plus their co-occurrence in the same text. Reference items, conjunctives, and structural gaps (ellipsis) in a text presuppose (the preceding) discourse which, if located beyond the sentence boundary, can contribute to cohesion. A word forms a lexical tie only when it occurs in the same text as another word (in a different sentence) which relates to it in some way. There are two major types of lexical cohesion. A word forms a lexical tie only when it occurs in the same text as another word (in a different sentence) which relates to it in some way. There are two major types of lexical cohesion: reiteration and collocation. Reiteration can be sub-categorized according to the degree of identity in meaning that two words share. They can be repetitions,



70



Slavic and East EuropeanJournal



they can be of different forms, but can share the same definition (synonyms/near-synonyms), or they can be in some hierarchical relationship to each other (superordinates). Finally a word of the most general meaning can represent another word even though the words are not members of a specific word class (general words). Some examples of general words in Russian are "eelovek" (human), "vega"' (inanimate concrete count), and "dela" (inanimate abstract). Collocation, the other major sub-type of lexical cohesion, concerns the probability of co-occurrence of two lexemes in a text. Although the words may be quite different in meaning, "6ernyj-belyf' or "janvar'fevral'," they are still related in some lexico-semantic way, as opposites in meaning (similarity) or as members of a series (contiguity). The analysis of collocational ties is to some extent intuitive and in actual textual analysis, an attempt must be made to limit "contamination" of the data, perhaps by rejecting any dubious collocational ties. However, since lexical cohesion is the major source of cohesion in most texts, the fact that some collocational ties might be intentionally overlooked cannot generally affect the overall significance of lexical cohesion in comparison with the other types. It remains to discuss the category of cohesion called substitution which Halliday and Hasan describe for English. It is related to ellipsis in that some noun phrase, verb phrase, or clause is replaced. In ellipsis, it is replaced by nothing, but in the case of substitution, it is replaced by a kind of grammatical marker which holds a structural position "open" for information from the (preceding) text. The nominal substitutes are "one, ones" as in "Does he have a new car?-No, an old one," and "(the) same" as in "He wants a pastrami on rye.-She ordered the same." The verbal substitute is "do (so)"; for example "He promised to submit the report on Friday. He still hasn't done so." The clausal substitutes are "so/not": "Do you think the hurricane will reach New Jersey?-I think so/not." The English substitutes have been reviewed here so that they might be compared linguistically with their translations in Russian. Halliday and Hasan note that the purpose of the nominal substitutes "one, ones" is to prevent any possible confusion as to whether, in an elliptical noun phrase, an attributive adjective is indeed an adjective or a noun. Since for the most part, adjectives and nouns are distinguished morphologically in Russian, there is no need for such a substitute. The Russian counterpart of the example above would be elliptical: "U nego novaja maSina?-Net, staraja." The substantivized phrase "toBesamoe" is accounted for in Russian under comparative reference. Although it seems to function at times



Cohesion in Russian



71



just as the English substitute "(the) same," there are instances where "(the) same" cannot be translated by "to Se samoe" (for example, when co-reference is intended). These apparent counterparts have different distributions. For Russian, "to Se samoe" can be handled satisfactorily under comparative reference where it is categorized along with the related adjectival phrase "tot Se samyj." The English verbal substitute "do" is a later development of the lexical verb "do." As a substitute, it has lost its former meaning and functions only as a grammatical counter. Although "delat"' has a wide range of definitions, it is not a substitute. An English verbal phrase containing the substitute "do"would be expressed in Russian using the intended verb or there would simply be ellipsis of the verb. In the case of clausal substitution, not only is there no evidence of such substitutes in Russian, the question must be raised as to the validity (or elegance) of Halliday and Hasan's treatment of this phenomenon in English. The translation of "I think so/not" in Russian, "Ja dumaju, (Cto) da/net," suggests an alternate explanation of the function of the English "substitutes." If"yes" and "no"can appear in this environment in Russian, why then should the English "so"and "not" be anything but variants in complimentary distribution with "yes" and "no"? These cohesive words are then accounted for in Halliday and Hasan's description under "general ellipsis with 'yes' and 'no."' Although it may be desirable to establish clausal substitution in an analysis where substitution must exist, there is certainly no justification for treating "da" and "net" as clausal substitutes in Russian. Halliday and Hasan prefer to define ellipsis as "substitution by zero." The analysis of cohesion in Russian does not support this hierarchical definition. It would seem that in any language, the omission of information (ellipsis) is a fundamental device of cohesion, and the replacement of that information with a marker (substitution) is an optional development. Thus it would be better to define substitution in terms of ellipsis which is more likely to be a universal category of cohesion. With the cohesive items and operation in the grammar and lexicon so defined, it is possible to devise a coding scheme by which texts can be analyzed and compared for number and types of cohesive ties. The types and sub-types of cohesion can be symbolized by a letter-number code as follows: Reference (R) Personal (1) on/ego ... (entire declension) ona/ee



Rla Rib



72



Slavic and East European Journal



ono/ego onilix ty/tebja vy/vas my/nas Demonstrative (2) Near (oto/Mti, tut, sej6as) Far (to/te, tam, togda) Comparative (3) Identity Similarity Non-Identity Non-Similarity Comparison,quantity Comparison,quality Ellipsis (E) Nominal (1) Determiner as Head Post-determiner as Head Adjective as Head Verbal (2) Lexical Operator Clausal (3) Propositional (can be total or partial-some Complementor Adjunctpresent) Modal General ellipsis of the clause (all elements but one omitted) Zero (entire clause omitted) Conjunction(C) Additive (i) Adversative (odnako) Causal (tak) Temporal (potom) Lexical (L) Same Item Synonym/Near-synonym Superordinate General Word Collocation



Rlc Rid Rle R f Rig R2a R2b R3a R3b R3c R3d R3e R3 f Ela Elb Elc E2a E2b E3a E3b E3c E3d C1 C2 C3 C4 L1 L2 L3 L4 L5



This coding scheme is designed to capture the more general distinctions between two or more texts in a comparative study. If so desired, it For could be expanded in most cases to indicate finer distinctions. be For could classified as to additive example, conjunctives sub-type. "i" could be distinguished from its negative counterpart conjunctives, "ni," and both from the additive conjunctive which connotes an alternative "ili." To illustrate how this model might be used in actual textual analysis, the first eleven sentences of L.N. Tolstoj's "Smert' Ivana Il'ita" have



Cohesion in Russian



73



been reproduced below. In this brief excerpt, each major category of cohesion is represented by at least one tie. There is a listing beneath each sentence of the number and type of cohesive ties found therein, the cohesive words themselves, and those words (or sentences) which they presuppose. In order to locate within the text the item presupposed by the cohesive word, a system of notation is used which was suggested by Halliday and Hasan in Cohesion in English. Under the column entitled "Distance," a "O" indicates that the presupposed word (or sentence) is immediately preceding. A number preceded by "N" (non-mediated) denotes that the presupposed word is to be found that many sentences backward in the text. If an "M" (mediated) precedes the number, it signifies that the presupposed word is mediated by another cohesive word that many sentences back in the text. it should be noted that unless otherwise stated, these letter-number codes direct the reader to information in the preceding text. From "Smert' Ivana Il'ita" Number of Ties



Cohesive Item



Type



Distance



Presupposed Item



1. B 6onbLUoM 3AaHMMcyAe6HblX y'pe)KAeHM



BO BpeMsl nepepblBa M B Ka6 no MenbBMHCKI4X HneHbl Aeny npoKypop COwLnMCb 3aceAlaHmR WHeTe IBaHa EroposB4a WUe6eK, H 3awuen pa3roBop o 3HaMeHMTOM



KpacOBCKOM Flene. (During the interval in the Melvinski trial in the large building of the Law Courts the members and public prosecutor met in Ivan Egorovich Shebek's private room, where the conversation turned on the celebrated Krasovski case. 12 o



2. QCeAop BacMnbeBMs pa3ropswlHncs, OKa3bBiaRHenofcyLHOCTb, He BcTynuB laBaH EropoeMH cTORn Ha CBoeM, neTp Ke IIBaHOBsHL, cHaLiana B cnop, He npMHMManB HeM y4aCTMAMnpocMaTpMBan TOJlbKO iTO nolaHHble "BeAOMOCT.,,> (Fedor Vasilievich warmly maintained that it was not subject to their jurisdiction, Ivan Egorovich maintained the contrary, while Peter Ivanovich, not having entered into the discussion at the start, took no part in it but looked through the Gazette which had just been handed in.) 3



HenofcyFHOCTb MBaHEropoeBw B cnop



L5 LI L2



0 0



VosaHa EropoBw4a



0



pa3roeop



/Aeno



74



Slavic and East European Journal



3. rocnoAa! - CKa3anJOH,- IaBaH l/nbIbll-TOyMep. ("Gentlemen,"



he said,



"Ivan Ilych has died!") 2



rocnoAa



L4



0



$eAop etc. BacMnbeBMH,



OH



Rla



0



LBaHOBMH neTp



E3c



0



Sentence 3



4. Hey>Kenw?("You don't say!") 1



5. BOT, HMTaiTe, - CKa3an OH c)eAopy BacMnbeBMHy,nolAaBBaL eMy cBeKMM,naxyqMr eque HOMep.("Here, read it yourself," replied Peter Ivanovich, handing Fedor Vasilievich the paper still damp from the press.) 3



OH



Rla



N.1+ M.



OH neTpVBaHoBMH



L1 L3



N.2 N.2



BacHnbieBH-



(Delop



ceA?opy BacmnbieBH4y HoMep



( BeAoMOCTWM,



6. B HepHOMo6OAKe 6bino HaneHaTaHO:"nfpacKOBbqceAopOBHa ronOBsHa c ywUJeBHblM npHcKop6He M H3BseaeT pOAlHblX1 3HaKOMblXO KOHHMHeBO3njo6neHHoro cynpyra CBoero, LneHa CyAe6HoA nanaTbl,



IBaHa lnbu4la ronoBHHa, nocneAoeaBuweR 4-ro cpeBpanFlcero 1882



roAqa.(Surroundedby a black borderwere the words:"Praskovya Fedorovna Golovina, with profoundsorrow, informs relatives and friendsof the demiseof her belovedhusbandIvanIlychGolovin,Member of the Courtof Justice,whichoccurredon Februarythe 4th of this year 1882.) 6



L5



N.2



yMep



npMcKOp64eM



L5



N.2



yMep



OKoHHHHe qneHa Cyqe6HoR lBaHaMnbMwa



L5 L1



N.2 N.4



yMep lneHbI



L5



N.3



L1



N.2



HepHoM



HenofcyAHOCTb HIBaH nbWqI



7. BblHOCTena B nflTHI4Ly,B 1 Hac nonony/HM.1 (The funeral will take



placeon Fridayat oneo'clockin the afternoon."). 2



8.



Tena Tena



L5 L4



0 0



KOHL*HHe



VBaH /nb14q



IBaHLlonbMH 6bln coTOBapHu4 rocnoA,MBce no6wn4 co6paBWumxcF



ero. (IvanIlychhadbeena colleagueof the gentlemenpresentandwas liked by them all.)



Cohesion in Russian 4



9. OH 6onen



MBaHM1nbM4 co6paBWIUxcS rocnoA nKo6Min



L1 L2 L1 L2



N.1 N.6 N.4 N.



75



VBaHaVlnbMa cowJnWCb



rocnoAa Bo3nio6jeHHoro



y)Ke HecKOJnbKOHeAen; roBopn4M, HTO 60ne3Hb



ero



HeM3nequMa.(He had been ill for some weeks with an illness said to be incurable.) 2



OH,ero (2)



Rla



0



VIBaHlnbuW4



10. MecTo OCTaBanOCb3a HMM,HO6bino coo6pa)KeHmeo TOM,HTOB AneKceeB MOKeT6blTb Ha3HaleH Ha ero MeCTO,Ha cny4ae ero CMepTM MecTe Ke AneKceeBa - unw BIHHIKOB, MnM UWTa6enb. (His post had been



kept open for him, but there had been conjectures that in case of his death Alexeev might receive his appointment, and that either Vinnikov or Shtabel would succeed Alexeev.) 4



3a HHM,ero ero (3) cMepTM



Rla



M. 1



L5



N.2



OH,ero mBaHMlAJbMH Teno



11. TaK HTO,ycnbixaB o CMepT M1BaHaMlnbwHa, nepBaSi MblCnbKa)KAoro B 0 6blna M3 rocnog, Ka6wHeTe, TOM, KaKoe 3HaHeHMe co6paBWiuXC5: CaMmXHneHOB MoKeT MMeTbCMepTb Ha nepeMeu4eHU1l MJ nOBbwBeHMI4 HnIMX1 3HaKOMblx. (So on receiving the news of Ivan Ilych's death the first thought of each of the gentlemen in that private room was of the changes and promotions it might occasion among themselves or their acquaintances.) 10



TaK4TO o cMepT4 (2x) MsaHa MlnbLHa Mblcnb t3 rocnoA co6paBsUJxcF B Ka6HHeTe qneHOB 3HaKOMblX



C3 LI LI L2 L1 L1 L1 L1 L1



0 0 N.2 0 N.2 N.2 N.9 N.4 N.4



Sentence 10 cMepTH lMaH M1nbH4 coo6paKeHme rocnoA co6paBuWLxc5



Ka6HHeT ,neHa 3HaKOMblX



It is impossible to draw any conclusions concerning the nature of Tolstoj's discourse on the basis of a few lines from the story. The observations which can be made concern texts in general. Upon reviewing just a few lines from "Smert' Ivana Il'i,a," the reader will notice the predominance of lexical ties. This is to be expected since it is words such as these which denote material reality (or in this case, the reality of the novel) which are the main vehicle transmitting meaning in discourse. Because



76



Slavic and East European Journal



of the number of such ties and of the complexity of the relationships expressed by them, lexical ties are at once the richest source of stylistic variation and the most dificult category of cohesion to analyze. One possible way to sub-categorize lexical ties might be on the basis of thematic lexical "chains." Even in this small sample from Tolstoj's story, the theme of death is already represented by seven lexical ties. If the entire story were analyzed for cohesive lexical ties, other themes could be identified and compared to this early-developing theme of death in terms of their lexical representation, development and location in the text, and so on. There are, of course, other ways of studying lexical cohesion or what is often termed generally as "repetitions" elsewhere in research on discourse. What an analysis of textual cohesion offers to discourse studies is a more reliable method of locating and classifying (grammatically and semantically) lexical "repetitions" of all sorts. Another general observation which can be made from a study of even the brief excerpt above is that cohesive ties are overwhelmingly anaphoric, interpretable from information in the preceding text. Although a cohesive tie can indicate an item which is located in the ensuing text, it is a relatively rare occurrence. A complete analysis of the number, type, and distribution of cohesive ties in "Smert' Ivana MI'ita"would lead to many other conclusions concerning Tolstoj's particular style of discourse as well as the nature of discourse in general. It would be seen that a relatively large number of personal referential ties (in this case, masculine personal reference) is an indication of third person narration. It would also become apparent that certain kinds of cohesive ties are much less common than others. These and other observations could then be compared to those resulting in an analysis of a literary work which differs dramatically from "Smert' Ivana Il'ita." Such a comparative study could lend support to the generalizations concerning cohesion in discourse or it could serve as a basis for investigating stylistic variation in the works under consideration. Limitations of space preclude a discussion of a comparative study of the sort just suggested from which the above analysis of an excerpt from "Smert'Ivana Il'iUa"was taken (See Note 2). Nonetheless a few results of that investigation should be mentioned since they illustrate some ways in which the analysis of textual cohesion can contribute to comparative studies in literature. One would expect that the obvious differences between Tolstoj's "Smert' Ivana Il'iUa" (henceforward "Smert"') and Sasa Sokolov's Skola dlja durakov (Skola) would be revealed by almost any method of comparative analysis, including that of the texts' patterns of cohesion. In general this proved to be true. The excerpts varied significantly in the



Cohesion in Russian



77



number of occurrences of personal reference, especially of masculine personal reference. The greater number of masculine personal referential ties in "Smert"' was attributed, as mentioned above, to the thirdperson narration in the excerpt and, in addition, to frequent reference to the central figure Ivan Il'ic. The excerpt from Skola, on the other hand, represents a dialogue between two personalities of a schizophrenic boy, and there is none of the personal reference (third person) which is characteristic of omniscient narration. Nor is there, in this excerpt, any recurring reference to other characters which is another possible source of personal reference. The texts exhibited the same number of occurrences of ellipsis. Ellipsis is most common in conversation, and nearly all of the cases of ellipsis in "Smert"' occur in the dialogue at the beginning of the story. The entire excerpt from Skola is a dialogue and the elliptical ties are fairly evenly distributed throughout the text. There were twice as many conjunctives in the excerpt from Skola than in that from "Smert'." These were predominantly internal conjunctives, an aspect of conjunction not discussed in this paper, but which is related to the representation of direct speech in a text. The greater use of internal conjunctives in Skola was interpreted as being a function of the narrator's (narrators') effort to impose lucidity of a less-than-lucid discourse. Since lexical cohesive ties (L1-L5) are overwhelmingly the most frequent type of tie in most discourse, it is this category which offered the best means of characterizing and differentiating the texts. Lexical chains in "Smert"' identified the major theme of death in the excerpt, the major character Ivan Il'ic, a minor theme, religion, and others. It is difficult to view the major lexical chains in Skola (for example, "the station," "the pond," "the bicycle") as dominating ideas or figures in the novel. Moreover, the chains in Skola seem to be related not only paradigmatically in the text, but also syntagmatically or metonymically according to their proximity in the world of the novel. Just how lexical chains are to be interpreted is a complex question, and in the study under consideration, the results are, for the most part, merely suggestive. Still, as concerns Skola, the nature of the lexical chains in the excerpt seem to represent a general problem in the novel; that is, the eccentric perceptions of the narrator(s) and their confrontation with the reader's expectations. An analysis of the lexical cohesive ties in "Smert' Ivana Il'ica" and Skola dlja durakov is a logical starting point for a comparative study of the works. Presumably an analysis of cohesive ties would provide just as fruitful a method for comparative literary analyses in general.



78



Slavic and East European Journal



Despite the difficulties involved in a study of textual cohesion; for example, the copious amount of data which is produced and must be analyzed or the degree of subjectivity which is certainly introduced by the person classifying the ties, the schema introduced by Halliday and Hasan and presented here in its Russian variant offers a fairly analytical method of investigating the relationships between sentences in a text. This information is of obvious interest to the linguist concerned with discourse analysis. Perhaps even more exciting, however, are the implications of this approach for defining stylistic variation in literature. NOTES 1 For a more detailed explanation and analysis of this conceptin English, the reader should consult M.A.K. Halliday and Ruqaiya Hasan, Cohesionin English (London: Longman, 1976). 2 The investigation which led to the results presentedin this paperis presentedin full detail in Cynthia F. Simmons, Cohesionin Russian: The MajorResourceof Textual Unity (Dissertation, Brown University, 1979). Another extension of the study of textual cohesion is the comparativeanalysis of cohesionin literary texts as a device for locating some aspects of stylistic variation. In my dissertation, I examine the cohesion in excerpts from two works of Russian literature: L.N. Tolstoj's"Smert' Ivana Il'ica"and Sasa Sokolov'sSkola dlja durakov. 3 In his article "LanguageStructure and Language Function,"John Lyons ed., New Horizons in Linguistics (Baltimore:Penguin, 1970), Halliday illustrates the distinction between topic and comment and old and new information and locates these (along with cohesion)within the so-called"textualcomponent"of language. 4 This definition of a sentence is not a new one and has been proposed,in varying terminology, elsewhere in the literature. For one argument, see Bohumil Palek, Cross-Reference:Study from Hyper-Syntax(Praha:Universita Karlova, 1968), 135, on the significance of the full stop. 5 A passage consisting of only one sentence naturally constitutes a text since it is unified by grammatical constituent structure. Still, cohesion does occur below the level of the sentence (with the use of conjunctions,referentialpronouns,wordrepetitions, and so forth), although it is of more interest in textual analysis as a hypersentential phenomenon. 6 Since Malinowski first introducedthe term "contextof situation"in "TheProblemof Meaning in Primitive Languages"in C.K. Ogdenand I.A. Richards,The Meaningof Meaning (New York:Harcourt,Brace, 1923),the problemsof (external)referenceand pragmatics have received increasingly more attention from modern linguists. A number of analyses have been offeredto describethe relationshipbetween language and environment;see for example, the schema in Halliday, McIntosh,and Strevens, The Linguistic Sciences and Language Teaching(London:Longman,1964). 7 At times it may be demandedthat the text be "appropriate" with respectto empirical reality (perhapsforpolitical reasons),but such a demandintrudeson the autonomyof the aesthetic text. On the other hand, it is possible to focus on the relationship between the aesthetic text and its imaginaryworld;yet it is usually assumedthat the work of art is compatiblewith the worldit has created.



Cohesion in Russian 8



9



10



11



12



79



An analysis of the cohesion in those literary works which rely mainly on formal devices (rhyme, meter, syntactic parallelism, and so on) for their coherence might be one way to determine whether or not the works are successful; that is, constitute a text. The terms "participant" and "circumstance" are Halliday's and concern transitivity functions. See the reference in Note 3. Also, the arrow from "otot"to "tot"symbolizes the fact that in contemporary standard Russian, the distinction between near and far in the demonstrative determiner is neutralized. Etot is ambiguous as to proximity, while tot is marked "+ far." The dichotomy of lexical verbs versus verbal operators is more common in the descriptions of other languages, like English, where analytic verbal constructions are more prevalent. This kind of analysis is nonetheless preferable in dealing with the Russian language; not only because it facilitates a description of verbal ellipsis, but also because it distinguishes between verbs which can occur with infinitives (catenatives) and verbs which must do so (operators). This distinction is missed in the discussion of"subjunctive infinitives" in Grammatika sovremmenogo russkogo literaturnogojazyka (Moskva: Nauka, 1970), 514. As a source of cohesion, intonation is included by Halliday and Hasan under the category of conjunction. Due to the obvious complexity involved in analyzing the cohesive effect of intonation, and since intonation is not usually of concern in textual analysis, their treatment of this aspect of cohesion was brief. For the same reasons, no attempt was made to categorize the various cohesive sentence intonations in Russian. Translations by Aylmer Maude.