Barthes, Roland. Text: Essays Selected and Translated by Stephen Heath. NY: Hill & Wang, 1973? Introduction to the Structural Analysis of Narratives: I. The Language of Narrative: 1. Beyond the Sentence/ 2 Levels of Meaning/  II. Functions: 1. The Determination of the Units/ 2. Classes of Units/ 3. Functional Syntax/ III. Actions: 1. Towards a Structural Status of Characters/ 2. The Problem of the Subject/ IV. Narration: 1. Narrative Communication/ 2. Narrative Situation/ V. The System of Narrative: 1. Distortion and Expansion/ 2. Mimesis and Meaning (79-124).

The narrativcs of thc, world ure numberless. Narrative is first and foremost a prodigious variety of genres, themselves distributed amongst different substances - as though any material were fit to re_eive man's stories. Able to be carried by articulated language, spoken or written, fixed or moving images, gestures" and I the ordered mixture of all these substances; narrative is present in myth, legend, fable, tale, novella, epic, hi_tory', tragedy, drama, comedy, mime, painting (think of Carpaccio's Sa;"t Ursllla), stained glass windows, cinema, comics, news item, conversation. _Iore­over, under this almost infinite diversity of forms, narrative is present in every age, in every place, in every society; it begins with the very history of mankind and there nowhere is nor has been a people without narrative. All classes, all human groups, have their narratives, enjoyment or which is very often shared by men with different, even opposing,. cultural backgrounds. Caring nothing ror the division between good and bad literature, narrative is international, transhistorical, transcultural: it is simply there, like life itself.

Must we conclude from this universality that narrative is insignificant 'lis it so general that we can have nothing to say about it except for the modest description of a rew highly individualized varieties, something literary history occasionally undertakes 'l But then how are we to master even these varieties, how are we to justify our right to

1. It must be remembered that this II not the Cue with either poetry or the essay, both of which are dependent on the cultural level of thair consumers.

80

against novella, talc agall1st myth, drama against tragedy (as has been do e a thousand times) wHhput 'reference toa common model I Sucij a model is Implied by every proposi­tion relating to the most' individual, the moSt historical, of narrative forms.! It is _hus legitimate that, 'far from the abandoning of any idea of dealing with narrative on the grounds of its universaiity, there should have been (from Aristotle on) a periodic interest in narrative form and it is normal that the newly developing structuralism should make this form one of its first concerns - is not structuralism's constant aim to master the infinity of utterances [paroles] by describing the 'language' ('lallgllc'] of which they are the products and from which they can be generated. Faced with the infinity of -narrativef, the multiplicity of standpoints ­ historical, psychologica!, sociological, ethnological, aes­thetic, etc. - from which they can be studied, the analyst finds himself in more or less the same situation as Sa us sure confronted by the heterogeneity of language [I all gage] and _eeking to extract a principle of classification and a central focus for description frem the apparent confusion of the individual messages. Keeping simply to modern times, the Russian Formalists, Propp and Levi-Strauss have taught Us to recognize the following dilemma: either a narrative is merely a rambling collection of events, in which case nothing can be said about' it other than by referring back to the storyteller's (the author's) art, talent or genius - all mythical forms of chancel - or else it shares with other narratives a common structure which is open to analysis, no matter how much patience its formulation requires. There is a world of difference between the most complex randomness

1. There does, of course, exist an 'art' of the storyteller, which is the ability to generate narratives (messages) from the structure (the code). This art corresponds to the n >don of performance in Chomsky and is far removed from the 'genius' of the author, romantically conceived as some bardy explicable personal secret.

81

and the most clementhry,combinatory schemc, and it is impossiblc to combin_ (t<) producc) a narrative Wilholl_ reference to an implicit system of units and rules.

Where then are we td look for the structures of narralivc '! Doubtless, in narratives themselves. Each ami el'en' nar­rative? i Many col11ffichtators wh.o accept the ide._ of a narrative structure arclneverthclcss unable to resign thcm­ selves to dissociating literary analysis fnml the exampk _)r

the experimental sciences; nothing daunted, they ask that a purely inductive method be applied to narrative and that one start by studying all the narmtives within a genre, a period, a society. This commonsense view is utopian. linguistics itself', with only some three thousand languages to embrace, cannot manage such a prognunmc and has wisely turned deductive, a step which in fact n1arkeJ its veritable constitution as a science and the beginning of its spectacular progress, it even succeeding in anticipating facts prior to their discovery.. So \Vlmt of narrative analysis, faced as it is with millions or narmtives? Of necessity, it is condemned to a deductive procedure, obliged first to devise a hypothetical model of description (what American linguists call a 'theory') and then gradually to work down fronithis model towards the different narrative species which at once conform to and depart from the model. It is only at the level of these conformities and departHrc:s that analysis will be able to come back to..but now equipped with a single descriptive tool, the plurality of narratives, to t.lei_ historical, geographical and cultural diversity.1

1. See the history of the Hittite a, postulated by Saussurc: and actually discovered fifty years later, as given in Emile: Benveniste, ProbMme$ de IinglJisllqllf gb,,!rale, Paris 1966, p. 35 [Probl':IIIs of General L;ngu;stics, Coral Gables, Florida 1971, p. 32).

2. Let us bear in mind the present conditions of linguistic descrip­tion: '. . . linguistic "structure" is always relative not just to the data or corpus but also to the grammatical theory describing the data' E. Bach, An lnlrodmllon to Transformallonal Grammar!, New York 1964, p. 2_; 'it has been recognized that language must b\: dc:scribcd as a

Tin,s, in onlcr to describe anJ ;Iassify the infinite number of 11:Inati\'cs, a 'theory' (in th'is :)ragmatic sense) is nceded ami the il11mClliate task is that ( f finding it. of starting to defill.. il. Its development can h _ greatly facilitated if one begins from a model able to pro\ ide it with its initial terms and principles. In the current s;ate of research, it seems reasotlable. that the structural analysis of narrative he givcn linguistics itself as foundin:' model.

83

arc only more sentences -- having described the Oower, the botanist is not to get involved in describing the bouquct. And yct it is evident that discourse itself (as a set of sen­tences) is organizcd and that. through this organization, it can be seclI as the message of anothcr language, one operat­ing at .. higher level than the language of the linguists.. Dis­course has its units, its rules, its Cgrammar': beyond the sentence, and though consisting solely of sentences, it must naturally form the object of a second linguistics, For a long time indeed, stich a linguistics of discourse bore a glorious name, that of Rhetoric. As a result of a complex historical movement, however, in which Rhetoric went over to belles­lellrcs and the latter was divorced from the study of language, it has recently hel_ome necessary to take up the prohlcm af,csh, The IICW linguistics of discoursc has still to bc developed, hut at least it is being postulated, and by the linguists themselves.2 This last fact is not without significance, for, although constituting an autonomous ohject, discoursc must he studied from the hasis of linguistics. If a working hypothesis is needed for an am,lysiswhose task is immcnse ami whose materials infinite, thcn the most rcasonahle thing is to posit it homological relation between sentence and discourse insofar as it is likely that a similar formal organizat ion ordcrs all semiotic systems, whatever thcir suhstanccs and dimcnsions, A discourse is n long 'sentence' (the units of which arc not 'necessarily sentences), just as a sentence, allowing for certain specifications, is a short cdiscourse'. This hypothesis accords well with a num­ber of propositions put forward in contemporary anthro­

1. It goes without saying, as Jakobson has noted, that between the sentence and what lies beyond the sentence there are transitions; co-ordination. for instance, can work ovcr the limit ortbe sentence.

2, See CSlwchllly: Iknveniste, op. cit., Chapter 10; Z. S. Barris, 'Discourse Analysis', 1,""RI",ge 28, 1952, pp. 18-23 " 474-94; N. RlIwct, 'Analysc strm;turate d'ul1 po_me fran_ais', l.ingll;.tl/rs J, 1964, PI'. 61.83.

85

the homology suggested here have merely a heuristic value: it implies an identity between language and literature (inasmuch as the latter can he seen as a sort of privileged vehicle of narrativc),' It is hardly possible any longer to com:cive of literature us an art that abandons all further relation with language the moment it has used it as an instrument to express ideas, passion or bc_tUty: language never ceases to accompany discourse, holding up to it the mirror of its own structure -- docs not literature, particu­larly toduy, make a language of the very conditions of language'1l

2. Lcvels of mcaning

From the outset, linguistics furnishes the structural analysis of narrativc with a concept which is decisive in that, making explicit immediately what is essential in every system (If meaning, lIamcly its organilation, it allows us both to show how a narrative is not a simple sum of propositi()ns ane.! to classify the enormous mass of elements which go to make up a narrative. This concept is that of …2

A sentence can he described, linguistically, on several levels (phonetic, phonological, gaammatical, contextual) and these levels arc in a hierarchical relationship with one

87

Todorov, rcviving the distinction made by the Russian Formalists, proposes working on two major levels, them­selves suhdivided: Jtory (the argument), comprising a logic of act ions and a 'syntax' of chaructcrs, _nd discollrse, comprising the tenses, aspects and modes (tI\the narrative. I But however many levels are proposed and whatever defini­tion they arc given, there can be no doubt that narrative is a hierarchy of instances. To understand a narrative is not merely to follow the unfolding of the story, it is also to recognize its construct ion in 'storeys', to project the horizontal concatenations of the narrative 'thread' on to an implicitly vertical axis; to read (to listen to) a narrative is not merely to move from one word to the next, it is also to move from one level to the next. Perhaps I may be allowed to oOer a kind of a pologue in t his connection. In T/,e Purloined Letter, Poe gives all acute analysis of the failure of the chief commissioner of the Paris police, powerless to find the letter. Ilis investigations, says Poe, were perfect 'wit/,i"

tlte ,rp"erc (_r "is ,speciality';2 he searched ,everywhere, saturated entirely the level of the 'police search', but in ordcr to find the Ictter, protected by its conspicuousness, it was necessary to shin to another level, to substitute the concealer's principle of relevance for that of the policeman. Similarly, the 'search' carried out over a horizontal set of narrative relations may well be as thorough as possible but must still, to be efrective_ also operate 'vertically': meaning is not 'at the end' of the narrative, it runs across it; just as conspicuous as the purloined Ictter, meaning eludes all unilateral investigation.

I. See T. Todorov, 'Lc_ categories du recitlittcrairc', Communication, 8, 1966 (Todurov's work on narrative is now most easily accessible in two books, Utteratllre l!t Signification, Paris 1967; Poltlq"_ d_ 1(1 prOSl. I}aris 1972. For a short account in English, see 'Structural analysis or narrative', Nnv('/l, 3, 1969, 1'1'. 70-61.

89

Since the Russian Formalists; a unit has been taken as any segment of the story which can be seen as the term of a correlat iOtL The essence of a function is, so to speak, the seed that it sows in thc narrative, planting an clement that will COIIIC to fr uitinn latcr - either on the same level or elsewhcre, on another level. If ill UII Cu'", simple Flaubert at one point tells the reader, seemingly without emphasis, that the daughters of the Sous-Prefet of Pont-I'Eveque owneJ a parrot, it is because this parrot is subsequently to have a great importance in Felicite's life; the statement of this dctail (whatever its linguistic form) thus constitutes a function, or narrative unit.

Is c\crything in a narrative functional? Does everything, down to the slightest detail, have a meHning? Can narrative he <Ii vidce' lip cui irely into funl_tional units'l We .shall see in a 1II01ilent that there arc several kinds of functions, there being several kinds of correlations, but this does not alter the nll:t that a narrative is never made up of anything other than functions: in dincring degrees, everything in it signifies. This is not a matter of art (on the part of the narrator), but of struct lire; ill the realm of discourse, what is noted is by definition notable, Even were a detail to appear irretrievably insignilicant, resistant to all functionality, it would none­theless end up with precisely the meaning of absurdity or uselessncss: everything has a meaning, or nothing has. To put it another way, one could say tlmt art is without noise (as that lerm is cmploycd ill information theory):2 urt is

91

feelings, intentions, motivations, rationalizations of charac­ter).

In the same way, since the 'language' ('lalIgllC'] of narra­tive is IInl the language ('(lIIRIII') or nrticulatetl language ([(",cage arlic,,'e] - though very often vehicled by it - nar­rative units will be substantially independent of linguistic units; they may indeed coincide with the latter, but occa­sionally, not systematically. Functions will be represented sometimes by units higher' tlmn . the sentence (groups of sentences of varying lengths, up to the work in its entirety) and sometimes by lower oncs (syntctgm, word and even, within the word, certain literary clements onlyl). When we are told that - the telephone ringing during night duty at Secret Service headqunrters - Bollel picked "p olle of 'Ire four reCe;l'('r.v, the moneme follr in itself constitutes a functional unit, referring as it does to ct concept necessary to the story (that of a highly developed bureaucratic technology). In fact, the narrative unit in this case is not the linguistic unit (the word) but only its connoted value (linguistically, the wprd /four/ never means 'four'); which explains how certain functional units can be shorter than the sentence without ceasing 10 belong to the order of discourse: such units then extend not beyond the sentence, than which they remain mnterially shorter, but beyond the level of denotation, which, like the sentence. is the province of linguistics pro­perly speaking.     .

2. Classes or units

The functional units must be distributed into a small num­ber of classes. If these classes are to be determined without recourse to the substance of content (psychological substance

93

meaning only on the le\'el of a general typology of the actants (Bond is on the side of order). Indices, because of ,the, in some sort, vertical nature of their relations, are truly semantic units: unlike 'functions' (in the strict sense), they refer to a signified, not to an 'operation', The ratification of indices is 'higher up', sometimes even remaining virtual, outside any explicit syntagm (the 'character' of a narrative agent may very well never be explicitly named while yet being constantly indexed), is a paradigmatic ratification. That of functions, hy contrast, is always 'further on', is a syntagmatic ratification" F"llctio1ls and i1ldices thus overlay another classic distinction: functions involve metonymic relata, indices metaphoric relata; the former correspond to a functionality of doillf!, the latter to a functionality of hcing, 2

These two main classes of units, functions and indices, should already allow a certain classification of narratives, Some narratives lIfe heavily functional (such as folktales), while others on the contrary are heavily indicial (stich as 'psychological' novels); between these two poles lies a whole series of intermediary forms, dependent on history, society, gcme, But we can go further. Within each of the two 111 a ii, classes it is immediately possible to determine two sub-classes of narrative units. Returning to the class of functions, its units arc not all of the same 'importance': some constitute real hinge-points of the narrative (or of a fragment urlhe narrative); others merely 'fill in' the narrative space separating the hinge functions, Let us call the former cardillalfimctio"s (or ""clei) and the laUer, having regard to their complementary nature, ratalJ'sers, For a function to

action), but, so to speak, the risk thcy entail: _anlinal functions are the risky moments of a narrative. Between these points of alternative, these 'dispatchers', the catalysers layout areas of safety, rests, luxuries. Luxuries which arc not, howc,'cr, uselcss: it must be stressed again that from the point of view of the story a catalyser's functionality may be weak but not nil. Were a eatalyser purely redundant (in relation to its nucleus), it would nonetheless participate in the ecollomy of the message; in fact, an apparently merely expletivc noti1tion always h,15 a discursive function: it accelera tes, dclays, gives fresh impetus to the discourse, it summarizes, anticipates and sometimes even leads astray" Since what is noted always appears as being notable, the catalyser ceaselessly revives the semantic tension of the discourse, says ceaselessly that there has been, that there is going to be, meaning. Thus, in the linal amalysis, the catu­Iyser has a constant function which is, to use Jakobson's term, a pllatic one:l it maintains the contact between narrator ami addressee. A nucleus cannot he deleted without ultering Ihe story, hut neither can it catalyst witho"t altering the discourse.

As for the other main dass of units, the indices, an inte­grational class, its units have in common that they can only he saturated (complcted) on the levCl of characters or on the level of narration. They are thus part of a pm'ometr;col relation3 whose second - implicit - term is continuous, extended over an episode, a character or the whole work.

97

certain ..tllltlsphere (modernity, rehaxatiol1, reminiscell_t:, etc.), In other words, certain units can be mixed. giving a play of pnssibilitics in the narrative economy, In the novel (;o/djillgc,., Bond, having to search his adversary's bedroom, is givcn ,I master-key by his associate: the notation is a pure (cardinal) function, In the film, this detail is altered and Rond laughingly takes a set of keys from a willing chamher-maid: the notation is no longer simply functiomtl hut also indicial, referring to lJond's character (his easy charm aud success with women). Secondly, it should be noted (thi_ will be taken up again later) that the four, classes jm,l tlcserihell can he distributed in tl different way which is nlOrCOver closer to the linguistic model, Catalyscrs, indices and informants have it Ctunmon characteristic: in relation to nuclei, they arc cxpa"siOlu, Nuclei (as will be seen in a uHHl1cnl) form finite sets grouping a small number of tcrms, are governed by tl logic, arc at once necessary ami sullicient. Once the framework they provide is given, the other units fill it out according to a mode of proliferation in principle illfinite. As we know, this is what happens in the case oflhe sentcncc, which is made up of simple propositions endlessly complicated with duplications, paddings, em­beddings and so on. So great an importance did Mallarme attach to this type of structure that from it he constructed JtlIlIa;J 1111 coup (Ie ties, a poem which with its 'nodes' and 'loops', its 'nuclcus-words' and itS 'Iace_words" can well he regarded as the, emblem of every narrative -- of every la taguagc.

3, I_uncti()nal syntax

lIow, Clcconling to what 'grammar', &ue the different units strung togclhcr along the narrative synlagm'l Whut are the rules of the functional combinatory system " Informants and indkes can combine frcely together: as for example in the

99

Analysis tmlay tcnds to \Iechronologize' the narrative continuum and to Crclogici1.e' it, to make it dependent on what Mallarmc called with regard to the French language 'the primifirc tJu",clerboft.<; of logic';1 or rather, more exactly (such al Icast is our wish), the task is to _ucceed in giving a structural description of the chronological illusion - it is for narrative logic to account for narrative time. To put it anothcr way, one could say that temporality is only a structural cntegory of narrative (of discourse), just as in language [fatIgue] temporality only exists in the form of a system; from the point of view of narrative, what we call time docs not exist, or at least only exists functionally, as an element of a semiotic system. Time belongs not to dis­course strictly speaking but to the referent; both narrative and language know only a semiotic time, Ctrue' time being n 'rcalist', referential illusion, as Propp's commentary shows. It is as stich that structural analysis must deal with it.1

What then is the logic which regulates the principal narrativ_ functions 'lit is this that current work is actively trying to establish and that has son... been the Imajor focus of debate, Three main directions of research can be seen. The first (Urcmond) is more properly logical in approach: it aims to reconstitute the syntax of human behaviour utilized in naHat ivc, to retrace the course of the 'choices' which inevitably faceJ the individual character at every point in

101

suHicicl1tly do_e as to aCl:ount for all the nan\&tive units, fur the smallest narrativc segments. We must remember that cardinal functions cannot he determined by their 'import­ance', only hy the (doubly implicative) nature of their rela­lions. A 'telephone cnll', no matter how futile it may seem, on the one hand itself comprises some few cardinal fun_tions (telephone ringing, picking up the receiver, speaking, putting down the receiver), while on the other, taken as a whole, it must he linkable -- at thc very least proceeding step by step­to the major articulations of the anecdote. The functional covering of the narrativc necessitates .111 organization of relays thc basic unit of which can only he a small group of functions, hereafter referred to (following Dremond) as a ,'tel/llenec.

A sequence is a logical succession of nuclei bound together by a rclat ion of solidarity:1 Ihe sequence opens when one of its Icrms has no solidary antecedent ami doses when another of its tl'rms has no consequent. To t..kc a deliberately trivial example, the different functions ordcr a drink, obtain it, drink it, pay for it, constitute an obviously c_osed sequencc, it being impossible to put any thins before the ordcr or after the payment without moving out of the homo­geneous group' /liU';ng a drink'. The sequence indeed is always nameable. Determining the major functions of the f\)lktalc, Propp ;tnd subsequently. Oremond have been led to name them (fi'cllld, Bctrayal, Struggle, ColJlract, Seduc­t;oll, elc.); the naming operation is equally inevitable in the case oft livial sequences, the 'micro-sequences' which often form the finest grain of the narnttive tissue. Are these namings solely the province of the analyst '1 In other words, arc they purely rnetalinguistic'l No doubt they are, dealing as they do with the code of narrative. Yet at the same time they can be imagined as forming part of an inner meta­

quel1l:e. Ilerc, for example, is a l11icro-seqttcllce: /ralld held 011', /rand .\11l1k ('II, I1a/l(1 relcaJctl. This Grcetillg then becomes II simplc funclion: 011 the one hand, it =ISSllmes the Tole of an indice (nahhiness of nil Pont, BOIul's distaste); 011 the other, it font1S globally a term in a I:uger sequence, with the name Afcctillg, whose other tenus (approach, halt, ;lItcr­pellatiOlI, sitting cloII"') can themselves be micro-sequences, A whole network of subrogations structures the narrative in this way, from the smallest matrices to the largest func­tions. What is in question here, of course, is a hierarchy that remains within the functional level: it is only when it has been possible tn widen the narrative out step by step, from Dli Pont's cigarette to Bond's battle against Goldfinger, that functional analysis is over - the pyramid of functions then touches the ncxt level (that of the Actions), There is both :I sy"t:tx within the sequences and a (subrogating) syntax hetwecn the sequences together, The first episode of Golclfil/ger thus takes on a 'sternl11atic' aspect:

Obviously this representation is' analytical; the reader perceives a linear succession of terms. What needs to be noted, however, is that the terms from several sequences can easily he imbricated in one another: a sequence is not yet completed when already, cutting in, the first term of a new sequence I11:tY appear, Sequences move in counterpoint;' functionally, the structurc of narrative is fugued: thus it

I. This counterpoint was recognized by the Russian Formalists who outlincd il_ typology; it is not without recallinl the principal 'Intricate' str\ltttltc_ (If the sentence (see below V.I.).

stoppcd heing subordinate to thc'action, embodied immedi­ately psychological essences: which essences could be drawn, up into lists, as can be seen in its purest form in the list of 'chal'acter parts' in bourgeois theatre (the coquette, the noble father, etc.). From its vcry outset, structural analysis has shown the utmost reluctance to treal the character as an essence, even merely for purposes or classification; Tomachevski went so far as to deny the character any narrative importaOl:e, a poillt of view he subsequently modificd, Without Icaving characters out of the analysis altogcthn, Propp I'educcd them to a simple typology ba_ctl not on p_YLhology hut on thc unity of the actions a_signcd them by thc narrativc (DI)flO" c_r (I 1II.,giccll.agc"t, lIe/pa, Villa;", etc,).

Since Propp, the charadeI' has l'unstantly set the struclUral analysb of narrative the S.tII1C problem, On the one Imnd, thc chal'ackl's (whatever l)ne calls them - drama/is persoIJae or (/ctalltJ) fl}rm a nccc_sary plane of description, outside of whid, the slightest reported 'actions' cease to be intel­ligiblc; so that it can be said that there is not a single narrative in th_ world without 'characters'. 1 or at least without agcnts. Yet on the other hand, these - extremely numcrous -- 'agcnts' can be neither described nor classified in terms of 'pcrson_' - whether the 'person' be considered as' a purdy Ilistorical form, limited to certain genres (those most familiar h) us it is tru_), in whkh case it is nccess..ry to leave out of account the very large number of narratives

107

an actant defines a class, it can be filled by different actors, mobilized according to rules of multiplication, substitution or replacement.

These three conceptions, have many points in common. The most important, it must be stressed again, is the defini­tion of the character according to participation in a sphere of actions, these spheres being few in number, typical and classi­fiable; which is why t_lis second level of description, despite its being that of thc characters, has here been called the level of Actions; the word actionf is not to be understood in the sense of the trining acts which form the tissue of the first level hut in that of the IIInjor articulations of praxis (desire, comnHlnil:ation, struggle),

2. The problem of the subject

The prohlcms raised by a dassification of the characters of narrative are not as yet satisfactorily resolved. Certainly there is ready agrcement on the fact that the innumerable ch:tTactcls of narrative can be brought under rules of sub­stitution and that, c"en within the one work, a single figure can absorb different characters,l Again, the actantialmodel proposed by Grcimas (and adopted hy Todorov in another pcrspcl:ti\l') seems to sland the test of a large number of narratives. Like any structural model, its value lies less in its canonic form (a matrix of six actants) than in the regulated transformations (rcplaccmcnts, confusions, duplications, substitutions) to which it lends i,tself, thus holding,out the hope of an actantial typology of narratives.2 A difficulty.

howl.:'.'cr, is that when th_ matrix has a high classificational power (as is the case with Greimas's actants) it fails aoe­quatdy to account for the multiplicity of participations as S(h III itS t hese arc analyscd in tefl'lS of perspectives and that \\ IlL"n these perspectives are rc_ pected (as in Bremond's docril>tion) the system of characters remains, .too frag­IIIcntrd, The reduction proposed by Todorov avoids both pit fan:; but has so far only been a('plied to one narrative. All this, ' seems, can be quickly af1d harmoniously resolved. Tile Il'al dinicllity posed hy the .:Iassification of characters is tilL place (and hent'ethe existc'lCC) of the .'iII1_icCI in any actantlal matrix, whatever its formulation. IVlm is the subject (the bero) of it narrative? Is thel'e - or not - a privileged class of actors? The novel has ac,::ustomed us to emphasize in 011(; way or another - sometilres in a devious (negative) way one character in particula. But such privileging is far fr"111 extending over the wiKle of narralive literature. Man} narratives, for example, sd t \vo adversaries in con­nkt (".er some stake; the subjcc_ is then truly double, not rcJunble further by substitution. Indeed, this is eyen perhaps a COl11illOn archaic form, as though narrative, after the f_lshion ()f ccr:ain languages, had also l nown a dual of persons. This dual is all the more interestin:_ in that it relates narrative to the structures of certain (very modern) games in which two c,.tlal °PI'mncnts try to gain r ossession of an object put into circulation by a referee; a schema which recalls the acLtluial matrix proposed by Grcimas, and there is nothing slIrpr:"ing in this if one is willil1g to allow that a game, heing .l language, depends on th.: same symbolic structure as is 1', be found in language and narrative: a game too is

Just as there is within narrative a major function ofexchange (set out between a donor and a beneficiary), so, homologi­cally, narrative as object is the point of a communication: there is a donor of the narrative and a receiver of the narrative. J n linguistic communication, je and IU (1 and }'Oll) are absolutely presupposed by one another; similarly, there can be no narrative without a narrator and a listener (or reader). Banal perhaps, but still little developed. Cer­tainly the role of the sender has been abundantly enlarged upon (much study of the 'author' of a novel, though

109

a sentence.. If therefore a privileged class of actors is retained (the subject of the quest, of the desire, oCthe action), it needs at least to be made more flexible by bringing that actant under the very categories of the grammatical (and not psychological) person. Once again, it will be necessary to look towards. linguistics for the possibility of describing and classifying the personal (jeltu, first person/second person) or apersonal (ii, third person), singular, dual or plural, instance of the action. ] t will- perhaps - be the grammatical categories of the person (accessible in our pronouns) which will provide the key to the actional level; but since these categories can only be defined in relation to the instance of  discourse, not to that of reality,2 characters, as units of the actionallevel, find their meaning (their intelligibility) only if integrated in the third level of description, here called the level of Narration (as oppossed to Functions and Actions).

IV. Narration

1. Narrative communication

\"it hollt any t:(msideration of whetler he really is the cnar­rah)r'); \Vhen it comes to the reader, however, literary thC\HY i_; much more modest. In fa< t, the problem is not to illtrospect the n"otives of the nanator or the effects the lI:tlTatiol! produces on the reader, i:, is to describe. the code by which narrator and reader arc _ignified throughout the narrative itself. At first sight, the signs of the narrator appear mure evident and more I1tIf1_erous than those of the rcader (a narrative more frequently says I than you); in actual fact, the latter arc sil11plynore oblique than the funner. Thus, each time the narraf or stops crepresenting' and reports details which he knows perfectly well but which arc unknown to the reader, there occurs, by signifying failure, it sign of reading, for there would be no sense in the nar; ator giving himself a )iece of information. Leo lras flU! OU"'CT oj the joint, t we are told in a first-person novel: a _ign of the reader, close to what Jakobson calls the nmative function of communication. Lacking an inventory however; we shall leave aside for the moment these signs of n:L'cptiOl_ (though they are of equal importance) and say a few wC\rds concerning the signs of narration. 2

\Vlto L Ihe donor of the narrative? So far, three concep­tions sec'.' to havc been formulated, The first holds that a narrative eman:\tes from a person (ilI" the fully psycho­logical sense or the term). This person has a name, the author, in whom there is an endless exchange between the 'per­sonality' :qu) the 'art' of a perfectly identified individual who periodic:l!ly takes up his pen to writ,_ a story: the narrative (notably lhe novel) then being simply the expression or an I

Structural Analysis oj Narrath'es I J 11

external to it. The second conception regards the narrator as a sort of omniscient, apparently impersonal, conscious­ness that tells the story from a superior point of view, that of God:l the narrator is at once inside his characters (since he knows everything that goes on in them) and outside them (since he never identifies with anyone more than another). The third and most recent conception (Henry James, Sartre) decrees that the narrator must limit his narrative to what the characters can observe or know, everything proceeding as if each of the characters in turn were the sender of the narrative. All three conceptions are equally difficult in that they seem to consider narrator and characters as real - Cliving' - people (the unfailing power of this literary myth is well known), as though a narrative were originally determined at its referential level (it is a matter of equally crealist' conceptions). Narrator and characters, however, at least from our perspective, are essentially 'paper beings'; the (material) author of a narra­tive is in no way to be confused with the narrator of that narrative.2 The signs of the narrator are immanent to the narrative and hence readily accessible to a semiological analysis; but in order to conclude that the author himself (whether declared, hidden or withdrawn) has 'signs' at his disposal which he sprinkles through his work, it is necessary to assume the existence between this 'person' and his language of a straight descriptive relation which makes the author a full subject and and the narrative the instrumental ex pression of that fullness. Structural analysis is unwilling to accept such an assumption: wllo speaks (in the narrative)

I. 'When will someone write from the point of view of a superior joke, that is as God sees things from above?' Flaubert, Prlface d In vie d'Ccrim;n, ed, G. Boll_me, Paris 1965, p. 91.            ,

2. A distinction all the more necessary, liven the sCale at which we arc working, in that historically a large mass of narratives ue without authors (oral narratives, folktales, epics entrusted to bards, reciters, . etc.).      i i         .

113

are to be found (and of the most common kinds) which mix together in extremely rapid succession, often within the limits of a single sentence, the personal and the apersonal; as for instance this sentence from Golclfinger:

personal apersonal

His eyes, grey-blue, looked into those of Mr Du Pont who did not know what face to put on

for this look held a mixture of candour, irony and self-deprecation,   apersoIJa/

The mixing of the systems is clearly felt as a facility and this facility can go as far ns trick effects, A detective novel by Agatha Christie (The Sittaforcl AJystery) only keeps the enigma going by cheating on the person of the narration: a character is described from within when he is already the murderer. - as if in a single person there were the con­sciousness of a witness, immanent to the discourse, and the consciousness of a murderer, immanent to the referent, with the dishonest tourniquet of the two systems alone producing the enigma, Hence it is understandable that at the other pole of literature the choice of a rigorous system should have been made a necessary condition of a work ­ without it always being easy fully to meet that condition.

Rigour of this kind - the aim of certain contemporary writers - is not necessarily an aest_etic imperative. What is called the psychological novel usually shows a mixture of the two systems, successively mobilizing the signs of non­ person and those of person; 'psychology', that is, paradoxically, cannot accomm_date i_self to a pure system,! for by bringi.ng the whole n_rrative down to the .s_le jn_tanc of the discourse - or, I,e one prefers, to th_ ,Iocutlo_ary i.1 \ ji

I. Personal mode: ',It e,ve_ __cmed t_

114

ad -- it i_j the vcry content of the perS()11 which is threatened: t h_ psydiOlogical person (of refenntial order) bears no It:blion 10 the linguistic person, the latter never defined by Sl.ltcs of mind, intentions or traits or character but only by it:> (l:oocd) place in discourse.. It is t:lis formal person that "'Iiters today arc attempting to spe<Jk and such an attempt rcp..esen(_i all important subversion (the public 11101eOVer has the impression that 'novcls' are LO longcr being written) for it aims to transpose narrative r'om the purely constativc plane, which it has occupied lInt:1 now, to the performa­ti\c pIal}(, whereby the meaning of an utterance is the very act hy whic_l it is uttcrcd:l today, writing is not 'telling' bllt saying that one is telling and a!-.signing all the referent ('what one says') to this act of locution; which is why part (If contemporary literature is no longer descriptive, but transitivc. striving to accomplish so pure a present in its language that the whole of the disct'HJrse is identified with the act of its delivery, the whole logos being brought down. - or extended - to a Icxi.r;.2

2. Narrativc situation

The narmtional level is thus occupied by the signs of nar­ralivity, _he set of operators whicL reintegrate functions and actions in the narrative commu lication articulated on i IS donO! anJ its addressee. Some of Iliese signs have already received _"udy; we are familiar in oral literatures with certain codes of recitation (mctrical fc.rmulae,. conventional prcscntat:on protocols) and we knovl that here the 'author' is not the person who invents the fine:,t stories but the person

\I. On the perforl11ati\'c, !iee Todorov's 'r.CS categories till recit liucmirc'. The cJa5sic example of a perfnrl11alive i5 the statement 1 elrc!art' 1m' which neither 'con!ilates' ncr 'describes' anything but exhausts it:. meaning in the act or its utl_rance (by contrast 10 the S';ltcmcnt ,II(' king ,Iedarrd war, which con:itales, describes).

2. For the opposition logos/Iexis, see Gcnelte, 'Fronticres till rccit'.

115

who best masters the code which is practised equally by his listeners: in such literatures the narrational level is so clearly defined, its rules so binding, that it is difficult to conceive of a 'talc' devoid of the coded signs of narrative ('once "pOll a time', etc.). In our written literatures, the 'forms of discourse' (which are in fact signs of narrativity) were early identified: classification of the modes of authorial inter­vention (outlined by Plato and developed by Diomedesl), coding of the beginnings and endings of narratives, defini­tion of the different styles of representation (oralio directa, oralio ;Ildirccta with its inquit, oralio tecta), 2 study of 'points of view' and so on. All these clements form part of the narrational level, to which must obviously be added the writing as a whole, its role being not to 'transmit' the nar­rative but to display it.

It is indeed precisely in a display of the narrative that the units of the lower levels find integration: the ultimate form of the narrative, as narrative, transcends its contents and its strictly narrative forms (functions and actions). This explains why the narrational code should be the final level attainable by our analysis, other than by going outside of the narrative-object, other, that is, than by transgressing the rule of immanence on which the analysis is based. Narration can 0' 'y receive its meaning from the world which makes use of it: beyond the narrational level begins the world, other systems (social, ,economic, ideological) whose terms are no longer simply narratives but elements of. a different substance (historical fncts, determinations, behaviours, etc.). Just as linguistics stops at the sentence, so narrative analysis stops at discourse - from there it is

1l2 I iMAl.iE - MOSIC - TEXT

i_ 111:1 Irho U'ritc.',5 (in real life) and who writes is not ,1'110 is.'

In I"act, narration strictly speaking (the code of the nar­LthH}, Ii ke language, knows only two systems of signs: pnsoIi1al and apersonal. These two narrational systems do lIul ..cccssarily present the linguistic marks attached to pCr_ll!1 (1) and Bon-person (he): there arc narratives or at least arrative episodes, for example, which though written in th third person nevertheless have as their true instance the fii _t person. lIow can we tell? J t sunlces to rewrite the narw:ive (or the passage) from he to I: so long as the rcwriling entails no alteration of the discourse other than this changc of the grammntical pronouns, we can be slire that I\C arc dealing with a personal system. The whole of the b.:ginning of Goldfinger, thJlIgh written' in the third person, is in fact 'spoken' by James Bond. For the instance to change, rewriting 11 list become impossible; thus the scnte:l_e 'he saw a man in his fifties, still young-looking, , " is perf_ctly personal despite the II( ('I, James Bond, saw, , . '), but tbe narrative statement 'the tinkling of the ice against the t'ass appeared to give Do"HI a sudden inspiration' CIIII)II' be personal on accoun[ of the verb 'appeared', it (and not the he) becoming a sign of the apersonal. 'J here is no Joubt that the apersonal is the traditional mode of /larrative, language having Jeveloped a whole tense sy_tcJl1 peculiar to narrative (based on the aorist2), designed to wipt.: out the present of the speaker, As Benveniste puts it: 'I n narrativc, nO one speaks.' The personal instance (under more or less disguised forllJs) has, however, gradually invaded narrative, the narration being referred to the hic ('{ I/lI//(' of the locutionary net (whkh is the definition of the l>LrSonal system), Thus it is that today many narratives

116

necessary to shift to another i semiotics. Linguistics is acquainted with suc,h boundari,es which it has already postulated - if not explored - ulI(_er the n_lIne of situations. lIaliiday defines the 'situation' (in relation to a sentence) as 'the associated non-linguistic factors', I Prieto as 'the set of facts known by the receiver at the momcnt of the semic act and independently of this act'.2 In the same way, one can say that every narrative i) dependent on a 'narrative situation" the set of protocols ac\_ording to which the narrative is 'consumed'. In so-caliLI 'archaic' societies, the narrative situation is heavily coded} nowadays, avant-garde literature alone still dreams of reading protocols - spectacu­lar in the case of Mallarmc who wanted the book to be recited in public according to a precise combinatory scheme, typographical in that of Dutor who tries to provide the book with its own specific signs. Generally, however, our society takes the' greatest rains to conjure away the cooing of the narrative situation: there is no counting the number of narratio.lal devices which seek to naturalize the subsequent narrative by feigning to make it the outcome of some natural circumstance and thus, as it were, 'dis­inaugurating' it: epistolary no,'ct", supposedly rediscovered manu!o.cripts, author who met. he narrator, films which begin the story before the credits. The reluctance to declare its c()des characterizes bourgecis society and the mass culture issuing from it: both demand signs which do not look like signs. Yet this is only. so to speak, a. structural epiphenomenon: however fa mil.ia r, however casual may today be the act of opening a nt>vel or a newspaper or of turning on the telcvisio!l, no(hin._ can prevent that humble act from installing in us, all at once and in its entirety, the narrativc code we are going to need. Hence the narra­tional level has an ambiguous role: contiguous to the nar­rative situation (and sometimes even including it), it gives on to the world i"f1 _hich the narrative is undone (consumed), while at the samei time, capping the preceding levels, it closes the narrative, constitutes it definitively as utterance . of a language [langue] which provides for and bears along           its own metalanguage.

V. The System of Narrative

Language [langue] proper can be defined by the concurrence 'of two fundamental proccsses: articulation, or segmentation, which produces units (this being what Benveniste calls form), and integration, which gathers these units into units of a higher rank (this being meaning). This dual process can be found in the language of narrative ria langue dll rt.?cit] which also has an articulation and an integration, a form and a meaning.

1. Distortion and expansion

The form of narrative is essentially characterized by two powers: th_t of distending its signs over the length of the story and that of inserting unforeseeable expansions into these distortions. The two powers appear to be points of freedom but the nature of narrative is precisely to include these .deviations' within its language.1

The distortion of signs exists in linguistic language [langue] and was studied by Bally with reference to French

1. Valery: 'Formally the novel is close to the dream; both can be defined by considcration of this curious property: all their deviations /OTl" paff 0/ ,hem.'            .;

120 I IMAGE - MUSIC - TI:XT

What can be separated can also be tilled. Distended, the functional nllcld furnish intercalating spaces which Can be packrJ out almost infinitely; the interstices can be filled in ' wilh a very large number of carr Iysers. Here, however. a ilL'\\' typnlogy comes in. for the freedom to catalyse can be regulated according both to .the content of the functions (l'crtain functions arc more apt than others for catalysing ­as for example 'Vailingl) and to the substance of the nar­rative (writing contains possibilities of diaeresis - and so of catalysing - far superior to those of film: a gesture related

linguistically can be 'cut up' mud1 more easily than the _ame g('sture visualizcd2). The catalystic power of narrative has for corollary its elliptic power. Firstly, a function (he lwei (/ good meal) can economize on all the potential cata­Iy_crs it covers over (the details of the meal)3; secondly, it is possible to reduce a sequence to its nuclei and a hierarchy ofsrq\lcncc:s to its higher terms witl\()ut altering the meaning of the story: a narrative can he ilfentified even if its total syntagm be reduced to its actants and its main functions as these result from the progressivt: upwar_s integration of' il_ functional units.4ln other words, narrative lends itsclfto .HIIJlmmy (what used to be called the argument). At first sight this i_ truc or any discourse. but ea.:h discourse hasits own kind of summary. A lyric poem, for example, is simply the'

Strllctural Analysis of Narratives I 121

vast metaphor of a singlesignifiedl and to summarize it is thus to give this signified, an operation so drastic that it eliminates the poem's identity (summarized, lyric poems come down to the signifieds Lot'e and Death) - hence the conviction that poems cannot be summarized. By contrast, the summary of a narrative (if conducted according to structural criteria) preserves the individuality of the message; narrative, in other words, is translatablewithout fundamental damage. What is untranslatable is determined only at the last, narrational, level. The signifiers of narrativity, for instance, are not readily transferable from novel to film, the latter utilizing the personal mode of-treatment only very exceptionally;2 while the last layer oc the narrational level, namely the wri ting, resists transference Crom one language to another (or transfers very badly). ThetranslatabHity of narrative is a result of the structure of its language, so that it would be possible, proceeding in reverse, to determine this structure by identifying and classifying the (varyingly) translatable and untranslatable elements of a narrative. The existence (now) of different and concurrent semiotics (literature, cinema, comics, radio-television) would greatly facilitate this kind of analysis.

1. Logically Waiti"K has only two nu:lci: I. the wait established

2. the wait rewarded or disappointed; the first, however. can be c:\lcnsivcly catalyscd, occasionally even indefinitely (IVai/bIg for Godo/): )'ct another game - this time extreme - w:th structure.

2. \'a/_r}': 'Proust divides up - and gives us the feeling of being able tu divide up indefinitely - what other writers arc in the habit of passing ovcr.'

3. Here again. there arc qualifications according to substance: literature has an unrivalled clliptic powcr - which cinema lacks.

4. This rcduction does not necessarily correspond to the division of the blluk into chapters; on the contrary, it seems that increasingly chaplers havc the role of inlroducing brea:{s, points of suspense (serial tcchl1iqll_).

2. Mimesis and meaning

The second important process in the language oC narrative

I. N. Ruwet: 'A poem can be understood as the outcome or a series of transformations applied to the proposition "1 love you".' 'Analyse structurale d'un po_me fran_ais', Linguistics 3, 1964, p. 82, Ruwct here refers r::cisely to the analysis of paranoiac delirium given by Freud in connection with President Schreber ('Psychoanalytic Notes on an Autobiographical Account of a Case of Paranoia', Standard Elli/ ;011 Vol. 12).      i

2. Once again, there is no relation between the grammatical 'person' ofth,e narrator and the 'personality' (or subjectivity) that a film director puts into his way of presenting a story: the CQmera-l (continuously identified with the vision of;a particUlar character) is exceptional in '

the history of cinema. , '"' ; . ': I; , I

124

Narrative does not show, docs not 'mitate; the passion which may excite us in reading

a novd i:: not that of a 'vision' (in actual fact, wc do not '5ce' anything). Rather it is that of meaning, that of a higher order of relfltion which also has its emotions, its hopes,. its dangers, its triumphs. 'V/hat takes place' in a narrative is from the referential Coeality) point of view likrally IIDllting;1 'what happens is language alone, the aovclilure of language, the unceasing celebration of its coming. Although we know scan:cly more about the origins of narrative than we do about :he origins  of language, it can reasonably. be suggested tha_. narrative is contem­porancous with monologue, a crcati'Jn seemingly posterior to that of dialoguc. At all eve Its, wilhout wanting to strain the phylogenetic hypothesis, it may b{'. significant that it is at the samc momcnt (around the age (.f three) that the little human 'invents' at ollce sentence, narrativc, and the Oedipus. 1966

reported dial,)guc (cr. 'fwnticrcs du recit'); yet even dialoguc always contains a function of intelligibility, not of ;nimcsis.

The Struggle with the Angel Text ual analysis of Genesis 32: 22-32

(22) And he rose up that night, and took his two wives, and his two women servants, and his eleven sons, and passed over the ford Jabbok. (23) And he took them, and sent them over the brook, and sent over that he had. (24) And Jacob was left alone; and there wrestled a man with him until the breaking of the day. (25) And when he saw that he prevailed not against him, he touched the hollow of his thigh;, and the hollow of Jacob's thigh was out of joint as he wrestled with him. (26) And he said, Let me go, for the day breaketh. And he said, I will not let thee go, except thou bless me. (27) And he said unto him. What is thy name? And he said, Jacob. (28) And hc said, Thy name shall be called no more Jacob, but Israel: for as a prince hast thou power with God and with men, and hast prevailed. (29) And Jacob asked him, and said, Tell me, I pray thee, thy: name. And he said, Wherefore is it thou dost ask after my name? And he blessed him there. (30) And Jacob called the name of the place Peniel: for I have seen God face to face, and my life is preserved. (31) And as he passed over Penuel the sun rose upon him, and he halted: upon his thigh. (32) Therefore the children of Israel eat not of the $inew which shrank, which is upon the hollow of the thigh, unto this day: because he touched the hollow of Jacob's thigh in the sinew that shrank. (Authorized Version).