Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 22

Roland Barthes (1964)

Elements of Semiology
Source: Elements of Semiology, 1964, publ. Hill and Wang, 196 . !he "irst hal" o" the boo# is reproduced here.

INTRODUCTION
$n his Course in General Linguistics, "irst published in 1916, Saussure postulated the e%istence o" a general science o" signs, or Se&iolog', o" (hich linguistics (ould "or& onl' one part. Se&iolog' there"ore ai&s to ta#e in an' s'ste& o" signs, (hate)er their substance and li&its* i&ages, gestures, &usical sounds, ob+ects, and the co&ple% associations o" all these, (hich "or& the content o" ritual, con)ention or public entertain&ent: these constitute, i" not languages, at least s'ste&s o" signi"ication. !here is no doubt that the de)elop&ent o" &ass co&&unications con"ers particular rele)ance toda' upon the )ast "ield o" signi"'ing &edia, +ust (hen the success o" disciplines such as linguistics, in"or&ation theor', "or&al logic and structural anthropolog' pro)ide se&antic anal'sis (ith ne( instru&ents. !here is at present a #ind o" de&and "or se&iolog', ste&&ing not "ro& the "ads o" a "e( scholars, but "ro& the )er' histor' o" the &odern (orld. !he "act re&ains that, although Saussure,s ideas ha)e &ade great head(a', se&iolog' re&ains a tentati)e science. !he reason "or this &a' (ell be si&ple. Saussure, "ollo(ed in this b' the &ain se&iologists, thought that linguistics &erel' "or&ed a part o" the general science o" signs. -o( it is "ar "ro& certain that in the social li"e o" toda' there are to be "ound an' e%tensi)e s'ste&s o" signs outside hu&an language. Se&iolog' has so "ar concerned itsel" (ith codes o" no &ore than slight interest, such as the High(a' .ode* the &o&ent (e go on to s'ste&s (here the sociological signi"icance is &ore than super"icial, (e are once &ore con"ronted (ith language. it is true that ob+ects, i&ages and patterns o" beha)iour can signi"', and do so on a large scale, but ne)er autono&ousl'* e)er' se&iological s'ste& has its linguistic ad&i%ture. Where there is a )isual substance, "or e%a&ple, the &eaning is con"ir&ed b' being duplicated in a linguistic &essage ((hich happens in the case o" the cine&a, ad)ertising, co&ic strips, press photograph', etc.) so that at least a part o" the iconic &essage is, in ter&s o" structural relationship, either redundant or ta#en up b' the linguistic s'ste&. /s "or collections o" ob+ects (clothes, "ood), the' en+o' the status o" s'ste&s onl' in so "ar as the' pass through the rela' o" language, (hich e%tracts their signi"iers (in the "or& o" no&enclature) and na&es their signi"ieds (in the "or&s o" usages or reasons): (e are, &uch &ore than in "or&er ti&es, and despite the spread o" pictorial illustration, a ci)ilisation o" the (ritten (ord. 0inall', and in &ore general ter&s, it appears increasingl' &ore di""icult to concei)e a s'ste& o" i&ages and ob+ects (hose signifieds can e%ist independentl' o" language: to percei)e (hat a substance signi"ies is ine)itabl' to "all bac# on the indi)iduation o" a language:

there is no &eaning (hich is not designated, and the (orld o" signi"ieds is none other than that o" language. !hus, though (or#ing at the outset on nonlinguistic substances, se&iolog' is re1uired, sooner or later, to "ind language (in the ordinar' sense o" the ter&) in its path, not onl' as a &odel, but also as co&ponent, rela' or signi"ied. 2)en so, such language is not 1uite that o" the linguist: it is a second3order language, (ith its unities no longer &one&es or phone&es, but larger "rag&ents o" discourse re"erring to ob+ects or episodes (hose &eaning underlies language, but can ne)er e%ist independentl' o" it. Se&iolog' is there"ore perhaps destined to be absorbed into a trans-linguistics, the &aterials o" (hich &a' be &'th, narrati)e, +ournalis&, or on the other hand ob+ects o" our ci)ilisation, in so "ar as the' are spoken (through press, prospectus, inter)ie(, con)ersation and perhaps e)en the inner language, (hich is ruled b' the la(s o" i&agination). $n "act, (e &ust no( "ace the possibilit' o" in)erting Saussure,s declaration: linguistics is not a part o" the general science o" signs, e)en a pri)ileged part, it is se&iolog' (hich is a part o" linguistics: to be precise, it is that part co)ering the great signifying unities o" discourse. B' this in)ersion (e &a' e%pect to bring to light the unit' o" the research at present being done in anthropolog', sociolog', ps'choanal'sis and st'listics round the concept o" signi"ication. !hough it (ill doubtless be re1uired so&e da' to change its character, se&iolog' &ust "irst o" all, i" not e%actl' ta#e de"inite shape, at least try itself out, e%plore its possibilities and i&possibilities. !his is "easible onl' on the basis o" preparator' in)estigation. /nd indeed it &ust be ac#no(ledged in ad)ance that such an in)estigation is both di""ident and rash: di""ident because se&iological #no(ledge at present can be onl' a cop' o" linguistic #no(ledge* rash because this #no(ledge &ust be applied "orth(ith, at least as a pro+ect, to non3linguistic ob+ects. The Elements here presented ha)e as their sole ai& the e%traction "ro& linguistics o" anal'tical concepts, (hich (e thin# a priori to be su""icientl' general to start se&iological research on its (a'. $n asse&bling the&, it is not presupposed that the' (ill re&ain intact during the course o" research* nor that se&iolog' (ill al(a's be "orced to "ollo( the linguistic &odel closel'., We are &erel' suggesting and elucidating a ter&inolog' in the hope that it &a' enable an initial (albeit pro)isional) order to be introduced into the heterogeneous &ass o" signi"icant "acts. $n "act (hat (e purport to do is to "urnish a principle o" classi"ication o" the 1uestions. !hese ele&ents o" se&iolog' (ill there"ore be grouped under "our &ain headings borro(ed "ro& structural linguistics: $. Language and Speech. $$. Signified and Signifier. $$$. Syntagm and System. $4. Denotation and Connotation.

$t (ill be seen that these headings appear in dichoto&ic "or&* the reader (ill also notice that the binar' classi"ication o" concepts see&s "re1uent in structural thoughts as i" the &etalanguage o" the linguist reproduced, li#e a &irror, the binar' structure o" the s'ste& it is describing* and (e shall point out, as the occasion arises, that it (ould probabl' be )er' instructi)e to stud' the pre3e&inence o" binar' classi"ication in the discourse o" conte&porar' social sciences. !he ta%ono&' o" these sciences, i" it (ere (ell #no(n, (ould undoubtedl' pro)ide a great deal o" in"or&ation on (hat &ight be called the "ield o" intellectual i&agination in our ti&e.

I. LANGUAGE (LANGUE) AND SPEECH


I.1. IN LINGUISTICS $.1.1 In Saussure !he (dichoto&ic) concept o" language!speech is central in Saussure and (as certainl' a great no)elt' in relation to earlier linguistics (hich sought to "ind the causes o" historical changes in the e)olution o" pronunciation, spontaneous associations and the (or#ing o" analog', and (as there"ore a linguistics o" the indi)idual act. $n (or#ing out this "a&ous dichoto&', Saussure started "ro& the &ulti"or& and heterogeneous, nature o" language, (hich appears at "irst sight as an unclassi"iable realit', the unit' o" (hich cannot be brought to light, since it parta#es at the sa&e ti&e o" the ph'sical, the ph'siological, the &ental, the indi)idual and the social. -o( this disorder disappears i", "ro& this heterogeneous (hole, is e%tracted a purel' social ob+ect, the s'ste&atised set o" con)entions necessar' to co&&unication, indi""erent to the material o" the signals (hich co&pose it, and (hich is a language "langue#$ as opposed to (hich speech "parole# co)ers the purel' indi)idual part o" language (phonation, application o" the rules and contingent co&binations o" signs). $.1.5. The language "langue# % language is there"ore, so to spea#, language &inus speech: it is at the sa&e ti&e a social institution and a s'ste& o" )alues. /s a social institution, it is b' no &eans an act, and it is not sub+ect to an' pre&editation. $t is the social part o" language, the indi)idual cannot b' hi&sel" either create or &odi"' it* it is essentiall' a collecti)e contract (hich one &ust accept in its entiret' i" one (ishes to co&&unicate. 6oreo)er, this social product is autono&ous, li#e a ga&e (ith its o(n rules, "or it can be handled onl' a"ter a period o" learning. /s a s'ste& o" )alues, a language is &ade o" a certain nu&ber o" ele&ents, each one o" (hich is at the sa&e ti&e the e1ui)alent o" a gi)en 1uantit' o" things and a ter& o" a larger "unction, in (hich are "ound, in a di""erential order, other correlati)e )alues: "ro& the point o" )ie( o" the language, the sign is li#e a coin,, (hich has the )alue o" a certain a&ount o" goods (hich it allo(s one to bu', but also has )alue in relation to other coins, in a greater or lesser degree. !he institutional and the s'ste&atic aspect are o" course connected: it is because a language is a s'ste& o" contractual )alues (in part arbitrar', or, &ore e%actl', un&oti)ated) that it resists the &odi"ications co&ing "ro& a single indi)idual, and is conse1uentl' a social institution. $.1.7. Speech "parole# $n contrast to the language, (hich is both institution and s'ste&, speech is essentiall' an indi)idual act o" selection and actualisation* it is &ade in the "irst

place o" the ,co&bination than#s to (hich the spea#ing sub+ect can use the code o" the language (ith a )ie( to e%pressing his personal thought, (this e%tended speech could be called discourse#, - and secondl' b' the ,ps'cho3ph'sical &echanis&s (hich allo( hi& to e%teriorise these co&binations., $t is certain that phonation, "or instance, cannot he con"used (ith the language* neither the institution nor the s'ste& are altered i" the indi)idual (ho resorts to the& spea#s loudl' or so"tl', (ith slo( or rapid deli)er', etc. !he co&binati)e aspect o" speech is o" course o" capital i&portance, "or it i&plies that speech is constituted b' the recurrence o" identical signs: it is because signs are repeated in successi)e discourses and (ithin one and the sa&e discourse (although the' are co&bined in accordance (ith the in"inite di)ersit' o" )arious people,s speech) that each sign beco&es an ele&ent o" the language* and it is because speech is essentiall' a co&binati)e acti)it' that it corresponds to an indi)idual act and not to a pure creation. $.1.4. The dialectics of language and speech 8anguage and speech: each o" these t(o ter&s o" course achie)es its "ull de"inition onl' in the dialectical process (hich unites one to the other: there is no language (ithout speech, and no speech outside language: it is in this e%change that the real linguistic pra%is is situated, as 6erleau39ont' has pointed out. /nd 4. Brondal (rites, ,/ language is a purel' abstract entit', a nor& (hich stands abo)e indi)iduals, a set o" essential t'pes, (hich speech actualises in an in"inite )ariet' o" (a's.:, 8anguage and speech are there"ore in a relation o" reciprocal co&prehensi)eness. ;n the one hand, the language is ,the treasure deposited b' the practice o" speech, in the sub+ects belonging to the sa&e co&&unit', and, since it is a collecti)e su&&a o" indi)idual i&prints, it &ust re&ain inco&plete at the le)el o" each isolated indi)idual: a language does not e%ist per"ectl' e%cept in the ,spea#ing &ass,* one cannot handle speech e%cept b' dra(ing on the language. But con)ersel', a language is possible onl' starting "ro& speech: historicall', speech pheno&ena al(a's precede language pheno&ena (it is speech (hich &a#es language e)ol)e), and geneticall', a language is constituted in the indi)idual through his learning "ro& the en)iron&ental speech (one does not teach gra&&ar and )ocabular' (hich are, broadl' spea#ing, the language, to babies). !o su&, a language is at the sa&e ti&e the product and the instru&ent o" speech: their relationship is there"ore a genuinel' dialectical one. $t (ill be noticed (an i&portant "act (hen (e co&e to se&iological prospects) that there could not possibl' be (at least according to Saussure) a linguistics o" speech, since an' speech, as soon as it is grasped as a process o" co&&unication, is already part o" the language: the latter onl' can be the ob+ect o" a science. !his disposes o" t(o 1uestions at the outset: it is useless to (onder (hether speech &ust be studied &efore the language: the opposite is i&possible: one can onl' stud' speech straight a(a' inas&uch as it re"lects the language (inas&uch as it is ,glottic,). it is +ust as useless to (onder at the outset ho( to separate the language "ro& speech: this is no preli&inar' operation, but on the contrar' the )er' essence o" linguistic and later se&iological in)estigation: to separate the language "ro& speech &eans ipso facto constituting the proble&atics o" the &eaning. $.1.<. In '(elmsle) H+el&sle) has not thro(n o)er Saussure,s conception o" language!speech, but he has redistributed its ter&s in a &ore "or&al (a'. Within the language itsel" ((hich is still opposed to the act o" speech) H+el&sle) distinguishes three planes: i) the schema, (hich is the language as pure "or& (be"ore choosing this ter&

H+el&sle) hesitated bet(een s'ste&, pattern, or ,"ra&e(or#, "or this plane):= this is Saussure,s langue in the strictest sense o" the (ord. $t &ight &ean, "or instance, the 0rench r as de"ined phonologicall' b' its place in a series o" oppositions* ii) the norm, (hich is the language as &aterial "or&, a"ter it has been de"ined b' so&e degree o" social realisation, but still independent o" this realisation* it (ould &ean the r in oral 0rench, (hiche)er (a' it is pronounced (but not that o" (ritten 0rench)* iii) the usage, (hich is the language as a set o" habits pre)ailing in a gi)en societ': this (ould &ean the r as it is pronounced in so&e regions. !he relations o" deter&ination , bet(een speech, usage, nor& and sche&a are )aried: the nor& deter&ines usage and speech* usage deter&ines speech but is also deter&ined b' it* the sche&a is deter&ined at the sa&e ti&e b' speech, usage and nor&. !hus appear (in "act) t(o "unda&ental planes: i) the schema, the theor' o" (hich &erges (ith that o" the "or&: and o" the linguistic institution* ii) the group norm-usage-speech, the theor' o" (hich &erges (ith that o" the substance, and o" the e%ecution. /s according to H+el&sle) 3 nor& is a pure &ethodical abstraction and speech a single concretion (,a transient docu&ent,), (e "ind in the end a ne( dichotomy schema!usage, Which replaces the couple language!speech. !his redistribution b' H+el&sle) is not (ithout interest, ho(e)er: it is a radical "or&alisation o" the concept o" the language (under the na&e o" schema# and eli&inates concrete speech in "a)our o" a &ore social concept: usage. !his "or&alisation o" the language and socialisation o" speech enables us to put all the ,positi)e, and ,substantial, ele&ents under the heading o" speech, and all the di""erentiating ones under that o" the language, and the ad)antage o" this, as (e shall see presentl', is to re&o)e one o" the contradictions brought about b' Saussure,s distinction bet(een the language and the speech. $.1.6. Some pro&lems Whate)er its use"ulness and its "ecundit', this distinction ne)ertheless brings so&e proble&s in its (a#e. 8et us &ention onl' three. Here is the "irst: is it possible to identi"' the language (ith the code and the speech (ith the &essage> !his identi"ication is i&possible according to H+el&sle),s theor'. 9. ?uiraud re"uses it "or, he sa's, the con)entions o" the code are e%plicit, and those o" the language i&plicit* but it is certainl' acceptable in the Saussurean "ra&e(or#, and /. 6artinet ta#es it up. We encounter an analogous proble& i" (e re"lect on the relations bet(een speech and s'ntag&. Speech, as (e ha)e seen, can be de"ined (outside the )ariations o" intensit' in the phonation) as a ()aried) co&bination o" (recurrent) signs* but at the le)el o" the language itsel", ho(e)er, there alread' e%ist so&e "i%ed s'ntag&s (Saussure cites a co&pound (ord like magnanimus#. !he threshold (hich separates the language "ro& speech &a' there"ore be precarious, since it is here constituted b' ,a certain degree o" co&bination,. !his leads to the 1uestion o" an anal'sis o" those "i%ed s'ntag&s (hose nature is ne)ertheless linguistic (glottic) since the' are treated as one b' paradig&atic )ariation (H+el&sle) calls this anal'sis &orpho3s'nta%). Saussure had noticed this pheno&enon o" transition: ,there is probabl' also a (hole series o" sentences (hich belong to the language, and (hich the indi)idual no longer has to co&bine hi&sel"., $" these stereot'pes belong to the language and no longer to speech, and i" it pro)es true that nu&erous se&iological s'ste&s use the& to a great e%tent, then it is a real linguistics of

the syntagm that (e &ust e%pect, (hich (ill be used "or all strongl' stereot'ped ,&odes o" (riting,. 0inall', the third proble& (e shall indicate concerns the relations o" the language (ith rele)ance (that is to sa', (ith the signi"'ing ele&ent proper in the unit). !he language and rele)ance ha)e so&eti&es been identi"ied (b' !rubet@#o' hi&sel"), thus thrusting outside the language all the non3rele)ant ele&ents, that is, the co&binati)e )ariants. Aet this identi"ication raises a proble&, "or there are co&binati)e )ariants ((hich there"ore at "irst sight are a speech pheno&enon) (hich are ne)ertheless i&posed, that is to sa', arbitrar' : in 0rench, it is re1uired b' the language that the $ should be )oiceless a"ter a )oiceless consonant (oncle) and )oiced a"ter a )oiced consonant (ongle) (ithout these "acts lea)ing the real& o" phonetics to belong to that o" phonolog'. We see the theoretical conse1uences: &ust (e ad&it that, contrar' to Saussure,s a""ir&ation (,in the language there are onl' di""erences,), ele&ents (hich are not di""erentiating can all the sa&e belong to the language (to the institution)> 6artinet thin#s so* 0rei atte&pts to e%tricate Saussure "ro& the contradiction b' localising the di""erences in su&phonemes, so that, "or instance, p could not be di""erentiating in itsel", but onl', in it, the consonantic, occlusi)e )oiceless labial "eatures, etc. We shall not here ta#e sides on this 1uestion* "ro& a se&iological point o" )ie(, (e shall onl' re&e&ber the necessit' o" accepting the e%istence o" s'ntag&s and )ariations (hich are not signi"'ing and are 'et ,glottic,, that is, belonging to the language. !his linguistics, hardl' "oreseen b' Saussure, can assu&e a great i&portance (here)er "i%ed s'ntag&s (or stereot'pes) are "ound in abundance, (hich is probabl' the case in &ass3languages, and e)er' ti&e non3signi"'ing )ariations "or& a second3order corpus o" signi"iers, (hich is the case in strongl' connated languages : the rolled r is a &ere co&binati)e )ariant at the denotati)e le)el, but in the speech o" the theatre, "or instance, it signals a countr' accent and there"ore is a part o" a code, (ithout (hich the &essage o" ,ruralness, could not be either e&itted or percei)ed. $.1.B. The idiolect !o "inish on the sub+ect o" language!speech in linguistics, (e shall indicate t(o appended concepts isolated since Saussure,s da'. !he "irst is that o" the idiolect. !his is ,the language inas&uch as it is spo#en b' a single indi)idual, (6artinet), or again ,the (hole set o" habits o" a single indi)idual at a gi)en &o&ent, (2beling). Ca#obson has 1uestioned the interest o" this notion: the language is al(a's socialised, e)en at the indi)idual le)el, "or in spea#ing to so&ebod' one al(a's tries to spea# &ore or less the other,s language, especiall' as "ar as the )ocabular' is concerned (,pri)ate propert' in the sphere o" language does not e%ist,) : so the idiolect (ould appear to be largel' an illusion. We shall ne)ertheless retain "ro& this notion the idea that it can be use"ul to designate the "ollo(ing realities: i) the language o" the aphasic (ho does not understand other people and does not recei)e a &essage con"or&ing to his o(n )erbal patterns* this language, then, (ould be a pure idiolect (Ca#obson)* ii) the ,st'le, o" a (riter, although this is al(a's per)aded b' certain )erbal patterns co&ing "ro& tradition that is, "ro& the co&&unit'* iii) "inall', (e can openl' broaden the notion, and de"ine the idiolect as the language o" a linguistic co&&unit', that is, o" a group o" persons (ho all interpret in the sa&e (a' all linguistic state&ents: the idiolect (ould then correspond roughl' to (hat (e ha)e atte&pted to describe else(here under the na&e o" ,(riting,.: We can sa' in general that the hesitations in de"ining the concept o" idiolect onl' re"lect

the need "or an inter&ediate entit' bet(een speech and language (as (as alread' pro)ed b' the usage theor' in H+el&sle)), or, i" 'ou li#e, the need "or a speech (hich is alread' institutionalised but not 'et radicall' open to "or&alisation, as the language is. $.1. . Duple* Structures $" (e agree to identi"' language!speech and code!message, (e &ust here &ention a second appended concept (hich Ca#obson has elaborated under the na&e o" duple* structures$ (e shall do so onl' brie"l', "or his e%position o" it has been reprinted. $! We shall &erel' point out that under the na&e +duple* structures+ Ca#obson studies certain special cases o" the general relation code!message t(o cases o" circularit' and t(o cases o" o)erlapping. i) reported speech, or &essages (ithin a &essage (6D6): this is the general case o" indirect st'les. ii) proper na&es: the na&e signi"ies an' person to (ho& this na&e is attributed and the circularit' o" the code is e)ident (.D.): ,ohn means a person named ,ohn$ iii) cases o" auton'&' (,Rat is a s'llable,): the (ord is here used as its o(n designation, the &essage o)erlaps the code (6D.) 3 this structure is i&portant, "or it co)ers the ,elucidating interpretations,, na&el', circu&locutions, s'non'&s and translations "ro& one language into another* i)) the shifters are probabl' the &ost interesting double structure: the &ost read' e%a&ple is that o" the personal pronoun "I, thou# an indicial s'&bol (hich unites (ithin itsel" the con)entional and the e%istential bonds: "or it is onl' b' )irtue o" a con)entional rule that $ represents its ob+ect (so that $ beco&es ego in 8atin, ich in ?er&an, etc.), but on the other hand, since it designates the person (ho utters it, it can onl' re"er e%istentiall' to the utterance (.D6). Ca#obson re&inds us that personal pronouns ha)e long been thought to be the &ost pri&iti)e la'er o" language (Hu&boldt), but that in his )ie(, the' point rather to a co&ple% and adult relationship bet(een the code and the &essage: the personal pronouns are the last ele&ents to be ac1uired in the child,s speech and the "irst to be lost in aphasia* the' are ter&s o" trans"erence (hich are di""icult to handle. !he shifter theor' see&s as 'et to ha)e been little e%ploited* 'et it is, a priori, )er' "ruit"ul to obser)e the code struggling (ith the &essage, so to spea# (the con)erse being &uch &ore co&&onplace)* perhaps (this is onl' a (or#ing h'pothesis) it is on this side, that o" the shifters, (hich are, as (e sa(, indicial s'&bols according to 9eirce,s ter&inolog', that (e should see# the se&iological de"inition o" the &essages (hich stand on the "rontiers o" language, notabl' certain "or&s o" literar' discourse. I. . SE!IOLOGICAL PROSPECTS $.5.1. The language, speech and the social sciences. !he sociological scope o" the language!speech concept is ob)ious. !he &ani"est a""init' o" the language according to Saussure and o" Eur#hei&,s conception o" a collecti)e consciousness independent o" its indi)idual &ani"estations has been e&phasised )er' earl' on. / direct in"luence o" Eur#hei& on Saussure has e)en been postulated, it has been alleged that Saussure had "ollo(ed )er' closel' the debate bet(een Eur#hei& and !arde and that his conception o" the language ca&e "ro& Eur#hei& (hile that o" speech (as a #ind o" concession to !arde,s idea on the indi)idual ele&ent. !his h'pothesis has lost so&e o" its topicalit' because linguistics has chie"l' de)eloped, in the Saussurean idea o" the language, the ,s'ste& o" )alues, aspect, (hich led to acceptance o" the necessit' "or an i&&anent

anal'sis o" the linguistic institution, and this i&&anence is ini&ical to sociological research. 9arado%icall', it is not there"ore in the real& o" sociolog' that the best de)elop&ent o" the notion o" language!speech (ill be "ound* it is in philosoph', (ith 6erleau39ont', (ho (as probabl' one o" the "irst 0rench philosophers to beco&e interested in Saussure. He too# up again the Saussurean distinction as an opposition bet(een speaking speech (a signi"'ing intention in its nascent state) and spoken speech (an ,ac1uired (ealth, o" the language (hich does recall Saussure,s ,treasure,). He also broadened the notion b' postulating that an' process presupposes a system : thus there has been elaborated an opposition bet(een e)ent and structure (hich has beco&e accepted: and (hose "ruit"ulness in histor' is (ell #no(n. Saussure,s notion has, o" course, also been ta#en o)er and elaborated in the "ield o" anthropolog'. !he re"erence to Saussure is too e%plicit in the (hole (or# o" .laude 8F)i3 Strauss "or us to need to insist on it* (e shall si&pl' re&ind the reader o" three "acts: i) !hat the opposition bet(een process and s'ste& (speech and language) is "ound again in a concrete guise in the transition "ro& the e%change o" (o&en to the structures o" #inship* ii) that "or 8F)i3Strauss this opposition has an episte&ological )alue: the stud' o" linguistic pheno&ena is the do&ain o" &echanistic (in 8F)i3Strauss,s sense o" the (ord, na&el', as opposed to ,statistical,) and structural interpretation, and the stud' o" speech pheno&ena is the do&ain o" the theor' o" probabilities (&acrolinguistics)*: iii) "inall', that the unconscious character o" the language in those (ho dra( on it "or their speech, (hich is e%plicitl' postulated b' Saussure, is again "ound in one o" the &ost original and "ruit"ul contentions o" 8F)i3Strauss, (hich states that it is not the contents (hich are unconscious (this is a criticis& o" Cung,s archet'pes) but the "or&s, that is, the s'&bolical "unction. !his idea is a#in to that o" 8acan, according to (ho& the libido itsel" is articulated as a s'ste& o" signi"ications, "ro& (hich there "ollo(s, or (ill ha)e to "ollo(, a ne( t'pe o" description o" the collecti)e "ield o" i&agination, not b' &eans o" its ,the&es,, as has been done until no(, but b' its "or&s and its "unctions. ;r let us sa', &ore broadl' but &ore clearl': b' its signi"iers &ore than b' its signi"ieds. $t can be seen "ro& these brie" indications ho( rich in e%tra3 or &eta3linguistic de)elop&ents the notion language!speech is. We shall there"ore postulate that there e%ists a general categor' language!speech, (hich e&braces all the s'ste&s o" signs* since there are no better ones, (e shall #eep the ter&s language and speech, e)en (hen the' are applied to co&&unications (hose substance is not )erbal. $.5.5. The garment system We sa( that the separation bet(een the language and speech represented the essential "eature o" linguistic anal'sis* it (ould there"ore be "utile to propose to appl' this separation straighta(a' to s'ste&s o" ob+ects, i&ages or beha)iour patterns (hich ha)e not 'et been studied "ro& a se&antic point o" )ie(. We can &erel', in the case o" so&e o" these h'pothetical s'ste&s, "oresee that certain classes o" "acts (ill belong to the categor' o" the language and others to that o" speech, and &a#e it

i&&ediatel' clear that in the course o" its application to se&iolog', Saussure,s distinction is li#el' to undergo &odi"ications (hich it (ill be precisel' our tas# to note. 8et us ta#e the gar&ent s'ste& "or instance* it is probabl' necessar' to subdi)ide it into three di""erent s'ste&s, according to (hich substance is used "or co&&unication. $n clothes as -ritten about, that is to sa' described in a "ashion &aga@ine b' &eans o" articulated language, there is 9racticall' no ,speech,: the gar&ent (hich is described ne)er corresponds to an indi)idual handling o" the rules o" "ashion, it is a s'ste&atised set o" signs and rules: it is a language in its pure state. /ccording to the Saussurean sche&a, a language (ithout speech (ould be i&possible* (hat &a#es the "act acceptable here is, on the one hand, that the language o" "ashion does not e&anate "ro& the ,spea#ing &ass, but "ro& a group (hich &a#es the decisions and deliberatel' elaborates the code, and on the other hand that the abstraction inherent in an' language is here &aterialised as (ritten language: "ashion clothes (as (ritten about) are the language at the le)el o" )esti&entar' co&&unication and speech at the le)el o" )erbal co&&unication. $n clothes as photographed (i" (e suppose, to si&pli"' &atters, that there is no duplication b' )erbal description), the language still issues "ro& the "ashion group, but it is no longer gi)en in a (holl' abstract "or&, "or a photographed gar&ent is al(a's (orn b' an indi)idual (o&an. What is gi)en b' the "ashion photograph is a se&i3"or&alised state o" the gar&ent s'ste&: "or on the one hand, the language o" "ashion &ust here be in"erred "ro& a pseudo3real gar&ent, and on the other, the (earer o" the gar&ent (the photographed &odel) is, so to spea#, a nor&ati)e indi)idual, chosen "or her canonic generalit', and (ho .onse1uentl' represents a ,speech, (hich is "i%ed and de)oid o" all co&binati)e "reedo&. 0inall' in clothes as (orn (or real clothes), as !rubet@#o' had suggested,: (e again "ind the classic distinction bet(een language and speech. !he language, in the gar&ent s'ste&, is &ade i) b' the oppositions o" pieces, parts o" gar&ent and ,details,, the )ariation o" (hich entails a change in &eaning (to (ear a beret or a bo(ler hat does not ha)e the sa&e &eaning)* ii) b' the rules (hich go)ern the association o" the pieces a&ong the&sel)es, either on the length o" the bod' or in depth. Speech, in the gar&ent s'ste&, co&prises all the pheno&ena o" ano&ic "abrication ("e( are still le"t in our societ') or o" indi)idual (a' o" (earing (si@e o" the gar&ent, degree o" cleanliness or (ear, personal 1uir#s, "ree association o" pieces). /s "or the dialectic (hich unites here costu&e (the language) and clothing (speech), it does not rese&ble that o" )erbal language* true, clothing al(a's dra(s on costu&e (e%cept in the case o" eccentricit', (hich, b' the (a', also has its signs), but costu&e, at least toda', precedes clothing, since it co&es "ro& the read'3&ade industr', that is, "ro& a &inorit' group (although &ore anon'&ous than that o" Haute .outure). $.5.7. The food system 8et us no( ta#e another signi"'ing s'ste&: "ood. We shall "ind there (ithout di""icult' Saussure,s distinction. !he ali&entar' language is &ade o" i) rules o" e%clusion (ali&entar' taboos)* ii) signi"'ing oppositions o" units, the t'pe o" (hich re&ains to be deter&ined ("or instance the type sa)oury!s-eet#$ iii) rules o"

association, either si&ultaneous (at the le)el o" a dish) or successi)e (at the le)el o" a &enu)* i)) rituals o" use (hich "unction, perhaps, as a #ind o" ali&entar' rhetoric. /s "or ali&entar' ,speech,, (hich is )er' rich, it co&prises all the personal (or "a&il') )ariations o" preparation and association (one &ight consider coo#er' (ithin one "a&il', (hich is sub+ect to a nu&ber o" habits, as an idiolect). !he menu, "or instance, illustrates )er' (ell this relationship bet(een the language and speech: an' &enu is concocted (ith re"erence to a structure ((hich is both national 3 or regional 3 and social)* but this structure is "illed di""erentl' according to the da's and the users, +ust as a linguistic ,"or&, is "illed b' the "ree )ariations and co&binations (hich a spea#er needs "or a particular &essage. !he relationship bet(een the language and speech (ould here be "airl' si&ilar to that (hich is "ound in )erbal language: broadl', it is usage, that is to sa', a sort o" sedi&entation o" &an' people,s speech, (hich &a#es up the ali&entar' language* ho(e)er, pheno&ena o" indi)idual inno)ation can ac1uire an institutional )alue (ithin it. What is &issing, in an' case, contrar' to (hat happened in the gar&ent s'ste&, is the action o" a deciding group: the ali&entar' language is e)ol)ed onl' "ro& a broadl' collecti)e usage, or "ro& a purel' indi)idual speech. $.5.4. The car system, the furniture system !o bring to a close, so&e(hat arbitraril', this 1uestion o" the prospects opened up b' the language!speech distinction, (e shall &ention a "e( &ore suggestions concerning t(o s'ste&s o" ob+ects, )er' di""erent, it is true, but (hich ha)e in co&&on a dependence in each case on a deciding and &anu"acturing group: cars and "urniture. $n the car s'ste&, the language is &ade up b' a (hole set o" "or&s and details, the structure o" (hich is established di""erentiall' b' co&paring the protot'pes to each other (independentl' o" the nu&ber o" their ,copies,)* the scope o" ,speech, is )er' narro( because, "or a gi)en status o" bu'er, "reedo& in choosing a &odel is )er' restricted: it can in)ol)e onl' t(o or three &odels, and (ithin each &odel, colour and "ittings. But perhaps (e should here e%change the notion o" cars as o&(ects "or that o" cars as sociological "acts* (e (ould then "ind in the dri)ing o" cars the )ariations in usage o" the ob+ect (hich usuall' &a#e up the plane o" speech. 0or the user cannot in this instance ha)e a direct action on the &odel and co&bine its units* his "reedo& o" interpretation is "ound in the usage de)eloped in ti&e and (ithin (hich the ,"or&s, issuing "ro& the language &ust, in order to beco&e actual, be rela'ed b' certain practices. 0inall', the last s'ste& about (hich (e should li#e to sa' a (ord, that o" "urniture, is also a se&antic ob+ect: the ,language, is "or&ed both b' the oppositions o" "unctionall' identical pieces (t(o t'pes o" (ardrobe, t(o t'pes o" bed, etc), each o" (hich, according to its ,st'le,, re"ers to a di""erent &eaning, and b' the rules o" association o" the di""erent units at the le)el o" a roo& (,"urnishing,)* the ,speech, is here "or&ed either b' the insigni"icant )ariations (hich the user can introduce into one unit (b' tin#ering (ith one ele&ent, "or instance), or b' "reedo& in associating pieces o" "urniture together. $.5.<. Comple* systems !he &ost interesting s'ste&s, at least a&ong those (hich belong to the pro)ince o" &ass3co&&unications, are co&ple% s'ste&s in (hich di""erent substances are engaged. $n cine&a, tele)ision and ad)ertising, the senses are sub+ected to

the concerted action o" a collection o" i&ages, sounds and (ritten (ords. $t (ill, there"ore, be pre&ature to decide, in their case, (hich "acts belong to the language and (hich belong to speech, on the one hand as long as one has not disco)ered (hether the ,language, o" each o" these co&ple% s'ste&s is original or onl' co&pounded o" the subsidiar' ,languages, (hich ha)e their, places in the&, and on the other hand as long as these subsidiar' languages ha)e not been anal'sed ((e #no( the linguistic ,language,, but not that o" i&ages or that o" &usic). /s "or the 9ress, (hich can be reasonabl' considered as an autono&ous signi"'ing s'ste&, e)en i" (e con"ine oursel)es to its (ritten ele&ents onl', (e are still al&ost entirel' ignorant o" a linguistic pheno&enon (hich see&s to pla' an essential part in it: connotation, that is, the de)elop&ent o" a s'ste& o" second3order &eanings, (hich are so to spea# parasitic on the language proper . !his second order s'ste& is also a ,language,, (ithin (hich there de)elop speech3pheno&ena, idiolects and duple% structures. $n the case o" such co&ple% or connoted s'ste&s (both characteristics are not &utuall' e%clusi)e), it is there"ore no longer possible to predeter&ine, e)en in global and h'pothetical "ashion, (hat belongs to the language and (hat belongs to speech. $.5.6. .ro&lems "I# - the origin of the )arious signifyings systems !he se&iological e%tension o" the language!speech notion brings (ith it so&e proble&s, (hich o" course coincide (ith the points (here the linguistic &odel can no longer be "ollo(ed and &ust be altered. !he "irst proble& concerns the origin o" the )arious s'ste&s, and thus touches on the )er' dialectics o" language and speech. $n the linguistic &odel, nothing enters the language (ithout ha)ing been tried in speech, but con)ersel' no speech is possible (that is, "ul"ils its "unction o" co&&unication) i" it is not dra(n "ro& the ,treasure, o" the language. !his process is still, at least partiall', "ound in a s'ste& li#e that o" "ood, although indi)idual inno)ations brought into it can beco&e language pheno&ena. But in &ost other se&iological s'ste&s, the language is elaborated not b' the ,spea#ing &ass, but b' a deciding group. $n this sense, it can be held that in &ost se&iological languages, the sign is reall' and trul' ,arbitrar':, since it is "ounded in arti"icial "ashion b' a unilateral decision* these in "act are "abricated languages, ,logo3techni1ues,. !he user "ollo(s these languages, dra(s &essages (or ,speech,) "ro& the& but has no part in their elaboration. !he deciding group (hich is at the origin o" the s'ste& (and o" its changes) can be &ore or less narro(* it can be a highl' 1uali"ied technocrac' ("ashion, &otor industr')* it can also be a &ore di""use and anon'&ous group (the production o" standardised "urniture, the &iddle reaches o" read'3to3(ear). $", ho(e)er, this arti"icial character does not alter the institutional nature o" the co&&unication and preser)es so&e a&ount o" dialectical pla' bet(een the s'ste& and usage, it is because, in the "irst place, although i&posed on the users, the signi"'ing ,contract, is no less obser)ed b' the great &a+orit' o" the& (other(ise the user is marked (ith a certain ,asociabilit',: he can no longer co&&unicate an'thing e%cept his eccentricit')* and because, &oreo)er, languages elaborated as the outco&e o" a decision are not entirel' "ree (,arbitrar',). !he' are sub+ect to the deter&ination o" the co&&unit', at least through the "ollo(ing agencies: i) (hen ne( needs are born, "ollo(ing the de)elop&ent o" societies (the &o)e to se&i32uropean clothing in conte&porar' /"rican countries, the birth o" ne( patterns o" 1uic# "eeding in industrial and urban societies)* ii) (hen econo&ic re1uire&ents bring about the

disappearance or pro&otion o" certain &aterials (arti"icial te%tiles)* iii) (hen ideolog' li&its the in)ention o" "or&s, sub+ects it to taboos and reduces, so to spea#, the &argins o" the ,nor&al,. $n a (ider sense, (e can sa' that the elaborations o" deciding groups, na&el' the logo3techni1ues, are the&sel)es onl' the ter&s o" an e)er3(idening "unction, (hich is the collecti)e "ield o" i&agination o" the epoch: thus indi)idual inno)ation is transcended b' a sociological deter&ination ("ro& restricted groups), but these sociological deter&inations re"er in turn to a "inal &eaning, (hich is anthropological. $.5.B. .ro&lems "II# - the proportion &et-een +language+ and +speech+ in the )arious systems The second proble& presented b' the se&iological e%tension o" the language!speech notion is centred on the proportion, in the &atter o" )olu&e, (hich can be established bet(een the ,language, and the corresponding ,speech, in an' s'ste&. $n )erbal language there is a )er' great disproportion bet(een the language, (hich is a "inite set o" rules, and speech, (hich co&es under the heading o" these rules and is practicall' unli&ited in its )ariet'. $t can be presu&ed that the "ood s'ste& still o""ers an i&portant di""erence in the )olu&e o" each, since (ithin the culinar' ,"or&s,, the &odalities and co&binations in interpretation are nu&erous. But (e ha)e seen that in the car or the "urniture s'ste& the scope "or co&binati)e )ariations and "ree associations is s&all: there is )er' little &argin 3 at least o" the sort (hich is ac#no(ledged b' the institution itsel" 3 bet(een the &odel and its ,e%ecution,: these are s'ste&s in (hich ,speech, is poor. $n a particular s'ste&, that o" (ritten "ashion, speech is e)en al&ost non3e%istent, so that (e are dealing here, parado%icall', (ith a language (ithout speech ((hich is possible, as (e ha)e seen, onl' because this language is upheld b' linguistic speech). !he "act re&ains that i" it is true that there are languages (ithout speech or (ith a )er' restricted speech, (e shall ha)e to re)ise the Saussurean theor' (hich states that a language is nothing but a s'ste& o" di""erences (in (hich case, being entirel' negati)e, it cannot be grasped outside speech). and co&plete the couple language!speech (ith a third, presigni"'ing ele&ent, a &atter or substance pro)iding the (necessar') support o" signi"ication. $n a phrase li#e a long or short dress, the ,dress, is onl' the support o" a )ariant "long!short# -hich does "ull' belong to the gar&ent language 3 a distinction (hich is un#no(n in ordinar' language, in (hich, since the sound is considered as immediately signi"icant, it cannot be deco&posed into an inert and a se&antic ele&ent. !his (ould lead us to recognise in (non3linguistic) se&iological s'ste&s three (and not t(o) planes: that o" the &atter, that o" the language and that o" the usage. !his o" course allo(s us to account "or s'ste&s (ithout ,e%ecution,, since the "irst ele&ent ensures that there is a &aterialit' o" the language* and such a &odi"ication is all the &ore plausible since it can be e%plained geneticall': i", in such s'ste&s, the ,language, needs a ,&atter, (and no longer a ,speech,), it is because unli#e that o" hu&an language their origin is in general utilitarian, and not signi"'ing.

II. SIGNI"IER AND SIGNI"IED


II.1. THE SIGN

The classification of signs !he signi"ied and the signi"ier, in Saussurean ter&inolog', are the co&ponents o" the sign. -o( this ter&, sign, (hich is "ound in )er' di""erent )ocabularies ("ro& that o" theolog' to that o" &edicine), and (hose histor' is )er' rich (running "ro& the ?ospels:, to c'bernetics), is "or these )er' reasons )er' a&biguous* so be"ore (e co&e bac# to the Saussurean acceptance o" the (ord, (e &ust sa' a (ord about the notional "ield in (hich it occupies a place, albeit i&precise, as (ill be seen. 0or, according to the arbitrar' choice o" )arious authors, the sign is placed in a series o" ter&s (hich ha)e a""inities and dissi&ilarities (ith it: signal, inde*, icon, sym&ol, allegory, are the chie" ri)als o" sign. 8et us "irst state the ele&ent (hich is co&&on to all these ter&s: the' all necessaril' re"er us to a relation bet(een t(o relata. !his "eature cannot there"ore be used to distinguish an' o" the ter&s in the series* to "ind a )ariation in &eaning, (e shall ha)e to resort to other "eatures, (hich (ill be e%pressed here in the "or& o" an alternati)e "presences a&sence# i) the relation i&plies, or does not i&pl', the &ental representation o" one o" the relata$ ii) the relation i&plies, or does not i&pl', an analog' bet(een the relata$ iii) the lin# bet(een the t(o relata (the sti&ulus and its response) is i&&ediate or is not* i)) the relata e%actl' coincide or, on the contrar', one o)erruns the other* )) the relation i&plies, or does not i&pl', an e%istential connection (ith the user. Whether these "eatures are positi)e or negati)e (&ar#ed or un&ar#ed), each ter& in the "ield is di""erentiated "ro& its neighbours. $t &ust be added that the distribution o" the "ield )aries "ro& one author to another, a "act (hich produces ter&inological contradictions* these (ill be easil' seen at a glance "ro& a table o" the incidence o" "eatures and ter&s in "our di""erent authors: Hegel, 9eirce, Cung and Wallon (the re"erence to so&e "eatures, (hether &ar#ed or un&ar#ed, &a' be absent in so&e authors). We see that the ter&inological contradiction bears essentiall' on inde% ("or 9eirce, the inde% is e%istential, "or Wallon, it is not) and on sym&ol ("or Hegel and Wallon there is a relation o" analog' 3 or o" ,&oti)ation, 3 bet(een the t(o relata of the s'&bol, but not "or 9eirce* &oreo)er, "or 9eirce, the s'&bol is not e%istential, (hereas it is "or Cung). But (e see also that these contradictions 3 (hich in this table are read )erticall' 3 are )er' (ell e%plained, or rather, that the' co&pensate each other through trans"ers o" &eaning "ro& ter& to ter& in the sa&e author. !hese trans"ers can here be read hori@ontall': "or instance, the s'&bol is analogical in Hegel as opposed to the sign (hich is not* and i" it is not in 9eirce, it is because the icon can absorb that "eature. /ll this &eans, to su& up and tal# in se&iological ter&s (this being the point o" this brie" anal'sis (hich re"lects, li#e a &irror, the sub+ect and &ethods o" our stud'), that the (ords in the "ield deri)e their &eaning onl' "ro& their opposition to one another (usuall' in pairs), and that i" these oppositions are preser)ed, the &eaning is una&biguous. $n particular, signal and inde*, sym&ol and sign, are the ter&s o" t(o di""erent "unctions, (hich can the&sel)es be opposed3as a (hole, as the' do in Wallon, (hose ter&inolog' is the clearest and the &ost co&plete "icon and allegory are con"ined to the )ocabular' o" 9eirce and Cung). We shall there"ore sa', (ith Wallon, that the signal and the inde% "or& a group o" relata de)oid o" &ental representation, (hereas in the opposite group, that o" sym&ol and sign, this representation e%ists* "urther&ore, the signal is i&&ediate and e%istential, (hereas the inde% is not (it is onl' a trace)* "inall', that in the sym&ol the representation is analogical and inade1uate (.hristianit' ,outruns, the cross), (hereas in the sign the relation is un&oti)ated and e%act (there is no analog' bet(een the (ord o% and the i&age o" an o%, (hich is per"ectl' co)ered b' its relatum#.

$$.1.5. The linguistic sign $n linguistics, the notion o" sign does not gi)e rise to an' co&petition bet(een neighbouring ter&s. When he sought to designate the signi"'ing relationship, Saussure i&&ediatel' eli&inated s'&bol (because the ter& i&plied the idea o" &oti)ation) in "a)our o" sign (hich he de"ined as the union o" a signi"ier and a signi"ied (in the "ashion o" the recto and )erso o" a sheet o" paper), or else o" an acoustic i&age and a concept. Gntil he "ound the -ords signifier and signified, ho(e)er, sign re&ained a&biguous, "or it tended to beco&e identi"ied (ith the signi"ier onl', (hich Saussure (anted at all costs to a)oid* a"ter ha)ing hesitated bet(een s/me and same, form and idea, image and concept, Saussure settled upon signifier and signified, the union o" (hich "or&s the sign. !his is a para&ount proposition, (hich one &ust al(a's bear in &ind, "or there is a tendenc' to interpret sign as signi"ier, (hereas this is a t(o3 sided Canus3li#e entit'. !he (i&portant) conse1uence is that, "or Saussure, H+el&sle) and 0rei at least, since the signi"ieds are signs a&ong others, se&antics &ust be a part o" structural linguistics, (hereas "or the /&erican &echanists the signi"ieds are substances (hich &ust be e%pelled "ro& linguistics and le"t to ps'cholog'. Since Saussure, the theor' o" the linguistic sign has been enriched b' the dou&le articulation principle, the i&portance o" (hich has been sho(n b' 6artinet, to the e%tent that he &ade it the criterion (hich de"ines language. 0or a&ong linguistic signs, (e &ust distinguish bet(een the significant units, each one o" (hich is endo(ed (ith one &eaning (the ,(ords,, or to be e%act, the &one&es,) and (hich "or& the "irst articulation, and the distincti)e units, (hich are part o" the "or& but do not ha)e a direct &eaning (,the sounds,, or rather the phone&es), and (hich constitute the second articulation. $t is this double articulation (hich accounts "or the econo&' o" hu&an language* "or it is a po(er"ul gearing3do(n (hich allo(s, "or instance, /&erican Spanish to produce, (ith onl' 5$ distincti)e units, 1HH,HHH signi"icant units. $$.1.7. 0orm and su&stance.- !he sign is there"ore a co&pound o" a signi"ier and a signi"ied. !he plane o" the signi"iers constitutes the plane of e*pression and that o" the signi"ieds the plane of content. Within each o" these t(o planes, H+el&sle) has introduced a distinction (hich &a' be i&portant "or the stud' o" the se&iological (and no longer onl' linguistic) sign. /ccording to hi&, each plane co&prises t(o strata form and su&stance$ (e &ust insist on the ne( de"inition o" these t(o ter&s, "or each o" the& has a (eight' le%ical past. !he form is (hat can be described e%hausti)el', si&pl' and coherentl' (episte&ological criteria) b' linguistics (ithout resorting to an' e%tra3 linguistic pre&ise* the su&stance is the (hole set o" aspects o" linguistic pheno&ena (hich cannot be described (ithout resorting to e%tra3linguistic pre&ises. Since both strata e%ist on the plane o" e%pression and the plane o" content, (e there"ore ha)e: i) a substance o" e%pression: "or instance the phonic, articulator', non3"unctional substance (hich is the "ield o" phonetics, not phonolog'* ii) a "or& o" e%pression, &ade o" the paradig&atic and s'ntactic rules (let us note that the sa&e "or& can ha)e t(o di""erent substances, one phonic, the other graphic)* iii) a substance o" content: this includes, "or instance, the e&otional, ideological, or si&pl' notional aspects o" the signi"ied, its ,positi)e, &eaning* i)) a "or& o" content: it is the "or&al organisation o" the signi"ied a&ong the&sel)es through the absence or presence o" a se&antic &ar#. !his last notion is di""icult to grasp, because o" the i&possibilit' o" separating the signi"iers "ro& the signi"ieds in hu&an language* but "or this )er' reason the subdi)ision form!su&stance can

be &ade &ore use"ul and easier to handle in se&iolog', in the "ollo(ing cases: i) (hen (e deal (ith a s'ste& in (hich the signi"ieds are substanti"ied in a substance other than that o" their o(n s'ste& (this is, as (e ha)e seen, the case (ith "ashion as it is (ritten about)* ii) (hen a s'ste& o" ob+ects includes a substance (hich is not i&&ediatel' and "unctionall' signi"icant, but can be, at a certain le)el, si&pl' utilitarian: the "unction o" a dish can be to signi"' a situation and also to ser)e as "ood. $$.1.4. The semiological sign !his perhaps allo(s us to "oresee the nature o" the se&iological sign in relation to the linguistic sign. !he se&iological sign is also, li#e its &odel, co&pounded o" a signi"ier and a signi"ied (the colour o" a light, "or instance, is an order to &o)e on, in the High(a' .ode), but it di""ers "ro& it at the le)el o" its substances. 6an' se&iological s'ste&s (ob+ects, gestures, pictorial i&ages) ha)e a substance o" e%pression (hose essence is not to signi"'* o"ten, the' are ob+ects o" e)er'da' use, used b' societ' in a deri)ati)e (a', to signi"' so&ething: clothes are used "or protection and "ood "or nourish&ent e)en i" the' are also used as signs. We propose to call these se&iological signs, (hose origin is utilitarian and "unctional, sign-functions. !he sign3"unction bears (itness to a double &o)e&ent, (hich &ust be ta#en apart. $n a "irst stage (this anal'sis is purel' operati)e and does not i&pl' real te&poralit') the "unction beco&es per)aded (ith &eaning. !his se&antisation is ine)itable: as soon as there is a society, e)ery usage is con)erted into a sign of itself$ the use o" a raincoat is to gi)e protection "ro& the rain, but this use cannot be dissociated "ro& the )er' signs o" an at&ospheric situation. Since our societ' produces onl' standardised, nor&alised ob+ects, these ob+ects are una)oidabl' realisations o" a &odel, the speech o" a language, the substances o" a signi"icant "or&. !o redisco)er a non3signi"'ing ob+ect, one (ould ha)e to i&agine a utensil absolutel' i&pro)ised and (ith no si&ilarit' to an e%isting &odel (8F)i3Strauss has sho(n to (hat e%tent tin#ering about is itsel" the search "or a &eaning): a h'pothesis (hich is )irtuall' i&possible to )eri"' in an' societ'. !his uni)ersal se&antisation o" the usages is crucial: it e%presses the "act that there is no realit' e%cept (hen it is intelligible, and should e)entuall' lead to the &erging o" sociolog' (ith sociological But once the sign is constituted, societ' can )er' (ell re"unctionalise it, and spea# about it as i" it (ere an ob+ect &ade "or use: a "ur3coat (ill be described as i" it ser)ed onl' to protect "ro& the cold. !his recurrent "unctionalisation, (hich needs, in order to e%ist, a second3order language, is b' no &eans the sa&e as the "irst (and indeed purel' ideal) "unctionalisation: "or the "unction (hich is re3presented does in "act correspond to a second (disguised) se&antic institutionalisation, (hich is o" the order o" connotation. !he sign3"unction there"ore has (probabl') an anthropological )alue, since it is the )er' unit (here the relations o" the technical and the signi"icant are (o)en together. II. . THE SIGNI"IED $$.5.1. 1ature of the signified: $n linguistics, the nature o" the signi"ied has gi)en rise to discussions (hich ha)e centred chie"l' on its degree o" ,realit',* all agree, ho(e)er, on e&phasising the "act that the signi"ied is not ,a thing, but a &ental representation o" the ,thing,. We ha)e seen that in the de"inition o" the sign b' Wallon, this representati)e character (as a rele)ant "eature o" the sign and the s'&bol (as opposed to the inde% and

the signal). Saussure hi&sel" has clearl' &ar#ed the &ental nature o" the signi"ied b' calling it a concept the signi"ied o" the (ord o% is not the ani&al o%, but its &ental i&age (this (ill pro)e i&portant in the subse1uent discussion on the nature o" the sign). !hese discussions, ho(e)er, still bear the sta&p o" ps'chologis&, so the anal'sis o" the Stoics (ill perhaps be thought pre"erable. !he' care"ull' distinguished the phantasia logiki (the &ental representation), the tinganon (the real thing) and the lekton (the utterable). !he signi"ied is neither the phantasia nor the tinganon but rather the lekton* being neither an act o" consciousness, nor a real thing, it can be de"ined onl' (ithin the signi"'ing process, in a 1uasi3tautological (a': it is this ,so&ething, (hich is &eant b' the person (ho uses the sign. $n this (a' (e are bac# again to a purel' "unctional de"inition: the signi"ied is one o" the t(o relata o" the sign* the onl' di""erence (hich opposes it to the signi"ied is that the latter is a &ediator. !he situation could not be essentiall' di""erent in se&iolog', (here ob+ects, i&ages, gestures, etc., inas&uch as the' are signi"icant, re"er bac# to so&ething (hich can be e%pressed onl' through the&, e%cept that the se&iological signi"ied can be ta#en up b' the linguistic signs. ;ne can sa', "or instance, that a certain s(eater &eans long autumn -alks in the -oods$ in this case, the signi"ied is &ediated not onl' b' its )esti&entar' signi"ier (the s(eater), but also b' a "rag&ent o" speech ((hich greatl' helps in handling it). We could gi)e the na&e o" isology to the pheno&enon (hereb' language (ields its signi"iers and signi"ieds so that it is i&possible to dissociate and di""erentiate the&, in order to set aside the case o" the non3isologic s'ste&s ((hich are ine)itabl' co&ple%), in (hich the signi"ied can be si&pl' (u*taposed (ith its signi"ier. $$.5.5. Classification of the linguistic signifieds Ho( can (e classi"' the signi"ieds> We #no( that in se&iolog' this operation is "unda&ental, since it a&ounts to isolating the form "ro& the content. /s "ar as linguistic signi"iers are concerned, t(o sorts o" classi"ication can be concei)ed. !he "irst is e%ternal, and &a#es use o" the ,positi)e, (and not purel' di""erential) content o" concepts: this is the case in the &ethodical groupings o" Hallig and Wartburg, and in the &ore con)incing notional "ields o" !rier and le%icological "ields o" 6atorF. But "ro& a structural point o" )ie(, this classi"ication (especiall' those o" Hallig and Wartburg) ha)e the de"ect o" resting still too &uch on the (ideological) su&stance o" the signi"ieds, and not on their form. !o succeed in establishing a reall' "or&al classi"ication, one (ould ha)e to succeed in reconstituting oppositions o" signi"ieds, and in isolating, (ithin each one o" these, a rele)ant co&&utati)e "eature: this &ethod has been ad)ocated b' H+el&sle), SIrensen, 9rieto and ?rei&as. H+el&sle), "or instance, deco&poses a &one&e li#e ,&are, into t(o s&aller signi"icant units: ,Horse, J ,"e&ale,, and these units can be co&&utated and there"ore used to reconstitute ne( &one&es (,pig,, J ,"e&ale, K ,so(,, ,horse, J ,&ale, K ,stallion,)* 9rieto sees in ,)ir, t(o co&&utable "eatures ,ho&o, J ,&asculus,* SIrensen reduces the le%icon o" #inship to a co&bination o" ,pri&iti)es, (,"ather, K &ale parent, ,parent, K "irst ascendant). -one o" these anal'ses has 'et been de)eloped . 0inall', (e &ust re&ind the reader that according to so&e linguists, the signi"ieds are not a part o" linguistics, (hich is concerned onl' (ith signi"iers, and that se&antic classi"ication lies outside the "ield o" linguistics.:

$$.5.7. The semiological signifieds Structural linguistics, ho(e)er ad)anced, has not 'et elaborated a se&antics, that is to sa' a classi"ication o" the forms o" the )erbal signi"ied. ;ne &a' there"ore easil' i&agine that it is at present i&possible to put "or(ard a classi"ication o" se&iological signi"ieds, unless (e choose to "all bac# on to #no(n notional "ields. We shall )enture three obser)ations onl'. !he "irst concerns the &ode o" actualisation o" se&iological signi"ieds. !hese can occur either isologicall' or not* in the latter case, the' are ta#en up, through articulated language, either b' a (ord "-eek-end# or b' a group o" (ords "long -alks in the country#$ the' are thereb' easier to handle, since the anal'st is not "orced to i&pose on the& his o(n &etalanguage, but also &ore dangerous, since the' ceaselessl' re"er bac# to the se&antic classi"ication o" the language itsel" ((hich is itsel" un#no(n), and not to a classi"ication ha)ing its bases in the s'ste& under obser)ation. !he signi"ieds o" the "ashion gar&ent, e)en i" the' are &ediated b' the speech o" the &aga@ine, are not necessaril' distributed li#e the signi"ieds o" the language, since the' do not al(a's ha)e the sa&e ,length, (here a (ord, there a sentence). $n the "irst case, that o" the isologic s'ste&s, the signi"ied has no &aterialisation other than its t'pical signi"ier* one cannot there"ore handle it e%cept b' i&posing on it a &etalanguage. ;ne can "or instance as# so&e sub+ects about the &eaning the' attribute to a piece o" &usic b' sub&itting to the& a list o" )erbalised signi"ieds "anguished, stormy, som&re, tormented, etc.)*: (hereas in "act all these )erbal signs "or a single &usical signi"ied, (hich ought to be designated b' one single cipher, (hich (ould i&pl' no )erbal dissection and no &etaphorical s&all change. !hese &etalanguages, issuing "ro& the anal'st in the "or&er case, and the s'ste& itsel" in the latter, are probabl' ine)itable, and this is (hat still &a#es the anal'sis o" the signi"ieds, or ideological anal'sis, proble&atical* its place (ithin the se&iological pro+ect (ill at least ha)e to be de"ined in theor'. ;ur second re&ar# concerns the e%tension o" the se&iological signi"ieds. !he (hole o" the signi"ieds o" a s'ste& (once "or&alised) constitutes a great "unction* no( it is probable that "ro& one s'ste& to the other, the great se&iological "unctions not onl' co&&unicate, but also partl' o)erlap* the "or& o" the signi"ied in the gar&ent s'ste& is probabl' partl' the sa&e as that o" the signi"ied in the "ood s'ste&, being, as the' are, both articulated on the large3scale opposition o" (or# and "esti)it', acti)it' and leisure. ;ne &ust there"ore "oresee a total ideological description, co&&on to all the s'ste&s o" a gi)en s'nchron'. 0inall' 3 and this (ill be our third re&ar# 3 (e &a' consider that to each s'ste& o" &agni"iers (le%icons) there corresponds, on the plane o" the signi"ieds, a corpus o" practices and techni1ues* these collections o" signi"ieds i&pl' on the part o" s'ste& consu&ers (o" ,readers,, that is to sa'), di""erent degrees o" #no(ledge (according to di""erences in their ,culture,), (hich e%plains ho( the sa&e ,le%ie, (or large unit o" reading) can be deciphered di""erentl' according to the indi)iduals concerned, (ithout ceasing to belong to a gi)en ,language,. Se)eral le%icons3and conse1uentl' se)eral bodies o" signi"ieds 3 can coe%ist (ithin the sa&e indi)idual, deter&ining in each one &ore or less ,deep, readings.

II.#. THE SIGNI"IER $$.7.1. 1ature of the signaller. !he nature o" the signi"ier suggests roughl' the sa&e re&ar#s as that o" the signi"ied: it is purel' a relatum, (hose de"inition cannot be separated "ro& that o" the signi"ied. !he onl' di""erence is that the &agni"ier is a &ediator: so&e &atter is necessar' to it. But on the one hand it is not su""icient to it, and on the other, in se&iolog', the signi"ier can, too, be rela'ed b' a certain &atter: that o" (ords. !his &aterialit' o" the signi"ier &a#es it once &ore necessar' to distinguish clearl' matter "ro& su&stance a substance can be i&&aterial (in the case o" the substance o" the content)* there"ore, all one can sa' is that the substance o" the signi"ier is al(a's &aterial (sounds, ob+ects, i&ages). $n se&iolog', (here (e shall ha)e to deal (ith &i%ed s'ste&s in (hich di""erent #inds o" &atter are in)ol)ed (sound and i&age, ob+ect and (riting, etc.), it &a' be appropriate to collect together all the signs, inasmuch as they are home &y one and the same matter, under the concept o" the typical sign the )erbal sign, the graphic sign, the iconic sign, the gestural sign are all t'pical signs. $$.7.5. Classification of the signifiers !he clari"ication o" the signi"iers is nothing but the structuralisation proper o" the s'ste&. What has to be done is to cut up the ,endless, &essage constituted b' the (hole o" the &essages e&itted at the le)el o" the studied corpus, into &ini&al signi"icant units b' &eans o" the co&&utation test,: then to group these units into paradig&atic classes, and "inall' to classi"' the s'ntag&atic relations (hich lin# these units. !hese operations constitute an i&portant part o" the se&iological underta#ing (hich (ill be dealt (ith in chapter 111* (e anticipate the point in &entioning it here. II.$. THE SIGNI"ICATION $$.4.1. The significant correlation !he sign is a (t(o3"aced) slice o" sonorit', )isualit', etc. !he signification can be concei)ed as a process* it is the act (hich binds the signi"ier and the signi"ied, an act (hose product is the sign. !his distinction has, o" course, onl' a classi"'ing (and not pheno&enological) )alue: "irstl', because the union o" signi"ier and signi"ied, as (e shall see, does not e%haust the se&antic act, "or the sign deri)es its )alue also "ro& its surroundings* secondl', because, probabl', the &ind does not proceed, in the se&antic process, b' con+unction but b' car)ing out. /nd indeed the signi"ication "semiosis# does not unite unilateral entities, it does not con+oin t(o ter&s, "or the )er' good reason that signi"ier and signi"ied are both at once ter& and relation. !his a&biguit' &a#es an' graphic representation o" the signi"ication so&e(hat clu&s', 'et this operation is necessar' "or an' se&iological discourse. ;n this point, let us &ention the "ollo(ing atte&pts: 1) SrDSd: $n Saussure, the sign appears, in his de&onstration, as the )ertical e%tension o" a situation in depth in the language, the signi"ied is, as it (ere, &ehind the signi"ier, and can be reached onl' through it, although, on the one hand, these e%cessi)el' spatial &etaphors &iss the dialectical nature o" the signi"ication, and on the other hand the ,closed, character o" the sign is acceptable onl' "or the "ran#l' discontinuous s'ste&s, such as that o" the language.

5) 2R.: H+el&sle) has chosen in pre"erence a purel' graphic representation: there is a relation (R) bet(een the plane o" e%pression (2) and the plane o" content (.). !his "or&ula enables us to account econo&icall' and (ithout &etaphorical "alsi"ication, "or the &etalanguages or deri)ati)e s'ste&s 2 R (2R.). 7) SDS: 8acan, "ollo(ed b' 8aplanche and 8eclaire, uses a spatialised (riting (hich, ho(e)er, di""ers "ro& Saussure,s representation on t(o points: i) the signi"ier (S) is global, &ade up o" a &ulti3le)elled chain (&etaphorical chain): signi"ier and signi"ied ha)e onl' a "loating relationship and coincide onl' at certain anchorage points* ii) the line bet(een the signi"ier (S) and the signi"ied (s) has its o(n )alue ((hich o" course it had not in Saussure): it represents the repression o" the signi"ied. 4) Sr K Sd: 0inall', in non3isologic s'ste&s (that is, those in (hich the signi"ieds are &aterialised through another s'ste&), it is o" course legiti&ate to e%tend the relation in the "or& o" an e1ui)alence but not o" an identit'. $$.4.5. The ar&itrary and the moti)ated in linguistics We ha)e seen that all that could be said about the signi"ier is that it (as a (&aterial) &ediator o" the signi"ied. What is the nature o" this &ediation> $n linguistics, this proble& has pro)o#ed so&e discussion, chie"l' about ter&inolog', "or all is "airl' clear about the &ain issues (this (ill perhaps not be the case (ith se&iolog'). Starting "ro& the "act that in hu&an language the choice o" sounds is not i&posed on us b' the &eaning itsel" (the o% does not deter&ine the sound o%, since in an' case the sound is di""erent in other languages), Saussure had spo#en o" an ar&itrary relation bet(een signi"ier and signi"ied. Ben)eniste has 1uestioned the aptness o" this (ord: (hat is arbitrar' is the relation bet(een the signi"ier and the ,thing, (hich is signi"ied (o" the sound o% and the ani&al the o%). But, as (e ha)e seen, e)en "or Saussure, the sign is not the ,thing,, but the &ental representation o" the thing (concept)* the association o" sound and representation is the outco&e o" a collecti)e training ("or instance the learning o" the 0rench tongue)* this association 3 (hich is the signi"ication 3 is b' no &eans arbitrar' ("or no 0rench person is "ree to &odi"' it), indeed it is, on the contrar', necessar'. $t (as there"ore suggested to sa' that in linguistics the signi"ication is unmoti)ated. !his lac# o" &oti)ation, is, b' the (a', onl' partial (Saussure spea#s o" a relati)e analog'): "ro& signi"ied to signi"ier, there is a certain &oti)ation in the (restricted) case o" ono&atopoeia, as (e shall see shortl', and also e)er' ti&e a series o" signs is created b' the tongue through the i&itation o" a certain protot'pe o" co&position or deri)ation: this is the case (ith so3called proportional signs: pommier, poirer, a&ricotier, etc., once the lac# o" &oti)ation in their roots and their su""i% is established, sho( an analog' in their co&position. We shall there"ore sa' in general ter&s that in the language the lin# bet(een signi"ier and signi"ied is contractual in its principle, but that this contract is collecti)e, inscribed in a long te&poralit' (Saussure sa's that ,a language is al(a's a legac',), and that conse1uentl' it is, as it (ere, naturalised$ in the sa&e (a', 8e)i3Strauss speci"ied that the linguistic sign is arbitrar' a priori but non3arbitrar' a posteriori. !his discussion leads us to #eep t(o di""erent ter&s, (hich (ill be use"ul during the se&iological e%tension. We shall sa' that a s'ste& is arbitrar' (hen its signs are "ounded not b' con)ention, but b' unilateral decision: the sign is not arbitrar' in the language but it is in "ashion* and (e shall sa' that a sign is

moti)ated (hen the relation bet(een its signi"ied and its signi"ier is analogical (Bu'ssens has put "or(ard, as suitable ter&s, intrinsic semes "or &oti)ated signs, and e*trinsic semes "or un&oti)ated ones). $t (ill there"ore be possible to ha)e s'ste&s (hich are arbitrar' and &oti)ated, and others (hich are non3arbitrar' and un&oti)ated. $$.4.7. The ar&itrary and the moti)ated in semiology $n linguistics, &oti)ation is li&ited to the partial plane o" deri)ation or co&position* in se&iolog', on the contrar', it (ill put to us &ore general proble&s. ;n the one hand, it is possible that outside language s'ste&s &a' be "ound, in (hich &oti)ation pla's a great part. We shall then ha)e to establish in (hat (a' analog' is co&patible (ith the discontinuous character (hich up to no( has see&ed necessar' to signi"ication* and a"ter(ards ho( paradig&atic series (that is, in (hich the ter&s are "e( and discrete) can be established (hen the signi"iers are analogs this (ill probabl' be the case o" ,i&ages,, the se&iolog' o" (hich is, "or these reasons, "ar "ro& being established. ;n the other hand, it is highl' probable that a se&iological in)entor' (ill re)eal the e%istence o" i&pure s'ste&s, co&prising either )er' loose &oti)ations, or &oti)ations per)aded, so to spea#, (ith secondar' non3 &oti)ations, as i", o"ten, the sign lent itsel" to a #ind o" con"lict bet(een the &oti)ated and the un&oti)ated. !his is alread' to so&e e%tent the case o" the &ost ,&oti)ated, @one o" language, that o" ono&atopoeia. 6artinet has pointed out, that the ono&atopoeic &oti)ation (as acco&panied b' a loss o" the double articulation (ouch, (hich depends onl' on the second articulation, replaces the doubl' articulated s'ntag& +it hurts+#$ 'et the ono&atopoeia (hich e%presses pain is not e%actl' the sa&e in 0rench "aie# and in Eanish (au), "or instance. !his is because in "act &oti)ation here sub&its, as it (ere, to phonological &odels (hich o" course )ar (ith di""erent languages: there is an i&pregnation o" the analogical b' the digital. ;utside language, proble&atic s'ste&s, li#e the ,language, o" the bees, sho( the sa&e a&biguit': the hone'3gathering dances ha)e a )aguel' analogical )alue* that at the entrance o" the hi)e is "ran#l' &oti)ated (b' the direction o" the "ood), but the (riggl' dance in a "igure o" eight is 1uite un&oti)ated (it re"ers to a distance).3: 0inall', and as a last e%a&ple o" such ill3de"ined areas, certain trade3&ar#s used in ad)ertising consist o" purel' abstract, (non3analogical) shapes* the' can, ho(e)er, e%press, a certain i&pression ("or instance one o" ,po(er,) (hich has a relation o" a""init' (ith the signi"ied. !he trade3&ar# o" the Berliet lorries (a circle (ith a thic# arro( across it) does not in an' (a' ,cop', po(er 3 indeed, ho( could one ,cop', po(er> 3 and 'et suggests it through a latent analog'* the sa&e a&biguit' is to be "ound in the signs o" so&e ideographic (ritings (.hinese, "or instance). !he coe%istence o" the analogical and the non3analogical there"ore see&s un1uestionable, e)en (ithin a single s'ste&. Aet se&iolog' cannot be content (ith a description ac#no(ledging this co&pro&ise (ithout tr'ing to s'ste&atise it, "or it cannot ad&it a continuous di""erential since, as (e shall see, &eaning is articulation. !hese proble&s ha)e not 'et been studied in detail, and it (ould be i&possible to gi)e a general sur)e' o" the&. !he outline o" an econo&' o" signi"ication (at the anthropological le)el) can, ho(e)er, be percei)ed: in the language, "or instance, the (relati)e) &oti)ation introduces a certain order at the le)el o" the "irst (signi"icant) articulation : the ,contract, is there"ore in this case underpinned b' a certain naturalisation o" this a priori arbitrariness (hich 8F)i3Strauss tal#s about* other s'ste&s, on the contrar', can go "ro& &oti)ation to non3

&oti)ation: "or instance the set o" the ritual puppets o" initiation o" the Senou"o, cited b' 8F)i3Strauss in The Sa)age 2ind. $t is there"ore probable that at the le)el o" the &ost general se&iolog', (hich &erges (ith anthropolog', there co&es into being a sort o" circularit' bet(een the analogical and the un&oti)ated: there is a double tendenc' (each aspect being co&ple&entar' to the other) to naturalise the un&oti)ated and to intellectualise the &oti)ated (that is to sa', to culturalise it). 0inall', so&e authors are con"ident that digitalis&, (hich is the ri)al o" the analogical, is itsel" in its purest "or& 3 binaris& 3 a ,reproduction, o" certain ph'siological processes, i" it is true that sight and hearing, in the last anal'sis, "unction b' alternati)e selections. II.%. &ALUE $$.<.1. 3alue in linguistics We ha)e said, or at least hinted, that to treat the sign ,in itsel",, as the onl' lin# bet(een signi"ier and signi"ied, is a "airl' arbitrar' (although ine)itable) abstraction. We &ust, to conclude, tac#le the sign, no longer b' (a' o" its ,co&position,, but o" its ,setting,: this is the proble& o" )alue. Saussure did not see the i&portance o" this notion at the outset, but e)en as earl' as his second Course in General Linguistics, he increasingl' concentrated on it, and )alue beca&e an essential concept "or hi&, and e)entuall' &ore i&portant than that o" signi"ication ((ith (hich it is not co3e%tensi)e). 4alue bears a close relation to the notion o" the language (as opposed to speech)* its e""ect is to de3ps'chologise linguistics and to bring it closer to econo&ics* it is there"ore central to structural linguistics. $n &ost sciences, Saussure obser)es, there is no coe%istence o" s'nchron' and diachron': astrono&' is a s'nchronic science (although the hea)enl' bodies alter)* geolog' is a diachronic science (although it can stud' "i%ed states3)* histor' is &ainl' diachronic (a succession o" e)ents), although it can linger o)er so&e ,pictures,. Aet there is a science in (hich these t(o aspects ha)e an e1ual share: econo&ics ((hich include econo&ics proper, and econo&ic histor')* the sa&e applies to linguistics, Saussure goes on to sa'. !his is because in both cases (e are dealing (ith a s'ste& o" e1ui)alence bet(een t(o di""erent things: (or# and re(ard, a signi"ier and a signi"ied (this is the pheno&enon (hich (e ha)e up to no( called signification#. Aet, in linguistics as (ell as in econo&ics, this e1ui)alence is not isolated, "or i" (e alter one o" its ter&s, the (hole s'ste& changes b' degrees. 0or a sign (or an econo&ic ,)alue,) to e%ist, it &ust there"ore be possible, on the one hand, to e*change dissi&ilar things ((or# and (age, signi"ier and signi"ied), and on the other, to compare si&ilar things (ith each other. ;ne can e%change a "i)e3"ranc note "or bread, soap or a cine&a tic#et, but one can also co&pare this ban#note (ith ten3 or "i"t'3"ranc notes, etc.* in the sa&e (a', a ,(ord, can be ,e%changed, "or an idea (that is, "or so&ething dissi&ilar), but it can also be co&pared (ith other (ords (that is, so&ething si&ilar): in 2nglish the (ord &utton deri)es its )alue onl' "ro& its coe%istence (ith sheep$ the &eaning is trul' "i%ed onl' at the end o" this double deter&ination: signi"ication and )alue. 4alue, there"ore, is not signi"ication* it co&es, Saussure sa's, ,"ro& the reciprocal situation o" the pieces o" the language,. $t is e)en &ore i&portant than signi"ication: ,(hat 1uantit' o" idea or phonic &atter a sign contains is o" less i&port than (hat there is around it in the other signs,:3 a prophetic sentence, i" one realises that it alread' (as the "oundation o" 8F)i3Strauss,s ho&olog' and o" the principle o" ta%ono&ies. Ha)ing thus care"ull' distinguished, (ith Saussure, signi"ication and )alue, (e i&&ediatel' see that i" (e return to H+e&sle),s

strata (substance and "or&), the signi"ication parta#es o" the substance o" the content, and )alue, o" that o" its "or& "mutton and sheep are in a paradig&atic relation as signifieds and not, o" course, as signi"iers). $$.<.5. The articulation $n order to account "or the double pheno&enon o" signi"ication and )alue, Saussure used the analog' o" a sheet o" paper: i" (e cut out shapes in it, on the one hand (e get )arious pieces (/, B, .), each o" (hich has a )alue in relation to its neighbours, and, on the other, each o" these pieces has a recto and a )erso -hich ha)e &een cut out at the same time (/3/,, B3B,, .3.,): this is the signi"ication. !his co&parison is use"ul because it leads us to an original conception o" the production o" &eaning: no longer as the &ere correlation o" a signi"ier and a signi"ied, but perhaps &ore essentiall' as an act of simultaneously cutting out t(o a&orphous &asses, t(o ,"loating #ingdo&s, as Saussure sa's. 0or Saussure i&agines that at the (entirel' theoretical) origin o" &eaning, ideas and sounds "or& t(o "loating, labile, continuous and parallel &asses o" substances* &eaning inter)enes (hen one cuts at the sa&e ti&e and at a single stro#e into these t(o &asses. !he signs (thus produced) are there"ore articuli* &eaning is there"ore an order (ith chaos on either side, but this order is essentiall' a di)ision. !he language is an inter&ediate ob+ect bet(een sound and thought: it consists in uniting &oth -hile simultaneously decomposing them. /nd Saussure suggests a ne( si&ile: signi"ier and signi"ied are li#e t(o superi&posed la'ers, one o" air, the other o" (ater* (hen the at&ospheric pressure changes, the la'er o"3 (ater di)ides into (a)es3. in the sa&e (a', the signi"ier is di)ided into articuli. !hese i&ages, o" the sheet o" paper as (ell as o" the (a)es, enable us to e&phasise a "act (hich is o" the ut&ost i&portance "or the "uture o" se&iological anal'sis: that language is the do&ain o" articulations, and the &eaning is abo)e all a cutting3out o" shapes. $t "ollo(s that the "uture tas# o" se&iolog' is "ar less to establish le%icons o" ob+ects than to redisco)er the articulations (hich &en i&pose on realit'* loo#ing into the distant and perhaps ideal "uture, (e &ight sa' that se&iolog' and ta%ono&', although the' are not 'et born, are perhaps &eant to be &erged into a ne( science, arthrolog', na&el', the science o" apportion&ent.

0urther Reading: Biograph' L 6ar%ist 9s'cholog' and 8inguistics L Saussure L Eur#hei& Ca#obson L 8F)i3Strauss L Eerrida 6ar%ist 8iterar' .riticis& L 0rance Sub+ect /rchi)e 9hilosoph' /rchi)e M &ar%ists.org

You might also like