Articulation and the limits of metaphor

Ernesto Laclau
I
In a well-known essay
1
Grard Genette discusses the question of the interdependence
between metaphor and metonymy in the structuration of roust narrati!e" #ollowin$ the
pathbreakin$ work of %tephen &llmann'
(
he shows how' on top of the central role traditionally
$ranted to metaphor in rousts work' there are other semantic mo!ements of a typical
metonymic nature whose presence is' howe!er' necessary' for metaphor to succeed in its
fi$ural effects" A hypalla$e such scheresse brune des cheveux ) instead of scheresse
des cheveux bruns ) would be a typical e*ample of such metonymical displacements"
Genette' howe!er' insists from the !ery be$innin$' that it is not a simple question of
reco$nisin$ the coe*istence of both metaphor and metonymy in the roustian te*t' but of
showin$ how they require each other' how without the one shadin$ into the other neither of
them could play the specific role which is e*pected from them in the constitution of a
narrati!e economy" In his words+ ,far from bein$ anta$onistic and incompatible' metaphor
and metonymy sustain and interpenetrate each other' and to $i!e its proper place to the
second will not consist in drawin$ a concurrent list opposed to that of metaphors' but rather
in showin$ the relations of -coe*istence. within the relation of analo$y itself+ the role of
metonymy within metaphor"
/
Genette $i!es se!eral e*amples of such interconnection" 0hus' he refers to the numerous
cases in which ,bell tower 1clocher2 is metaphorically 1analo$ically2 related to ,ear 1pis2' or
to ,fish' dependin$ on the en!ironment of the church ) rural in the first case' and maritime in
the second" 0his means that the spatial relation of conti$uity is the source of metaphoric
analo$ical effects" ,Ear ) bell tower 1or glise ) meule2 in the middle of the fields' ,fish ) bell
tower near the sea' ,purple ) bell tower o!er the !ineyards' ,brioche ) bell tower at the time
of the sweets' ,pillow ) bell ) tower at the be$innin$ of the ni$ht' there is clearly in roust a
recurrent' almost stereotyped stylistic scheme' which one could call cameleo ) bell tower
1clocher-camleo2" 0hus there is a sort of resemblance by conta$ion" 0he metaphor finds its
support in a metonymy" 3uotin$ 4ean 5icardou' Genette enounces the principle+ ,qui se
ressemble sassemble 1et rciproquement2"
6
7any more e*amples of this essential solidarity between conti$uity and analo$y are $i!en+
that between autoctonous dishes and vin de pays8 between peintures and their $eo$raphical
framework8 between the desire for peasant women and their rural milieu8 between relati!es8
1
between ima$es succeedin$ each other in die$etic metaphors8 between landscapes and
their reflection in the $lass doors of a bookshelf' etc" In all these cases we see that' without
the mutual implication between metaphor and metonymy' it would be impossible to ensure
the unity of a discursi!e space" roust himself was only partially aware of this mutual
implication and tended to pri!ile$e its metaphorical side" As Genette says+ ,0he
indestructible solidarity of writin$' whose ma$ic formula roust seems to be lookin$ for 1-only
metaphor can $i!e a sort of eternity to style.' he will say in his article on #laubert2 cannot
only result from the hori9ontal link established by the metonymical tra:ectory8 but one cannot
see how could it result from :ust the !ertical link of the metaphoric relation either" ;nly the
crossin$ of one by the other can subtract the ob:ect of the description' and the description
itself from -times contin$encies.' that is' from all contin$ency8 only the mutual crossin$ of a
metonymic net and a metaphoric chain ensures the coherence' the necessary cohesion of
text"
<
Let us see how this crossin$ takes place" =entral to it is the structure of ,in!oluntary
memory" Apparently we ha!e' in the mechanism of reminiscence' the case of a pure
metaphor' de!oid of any metonymic contamination 1the taste of the 7adeleine' the position of
the foot in the une!en pa!ement' etc2" >ut the punctual character of that analo$ical memory
is immediately o!erflown" As Genette shows' it is only retroacti!ely that the analysis finds
that reminiscence starts from an analo$y which it would isolate as its ,cause" ,In fact' the
real e*perience be$ins' not by $raspin$ an identity of sensation' but by a feelin$ of -pleasure.
of -happiness.' which first appears without the notion of its cause"
?
Althou$h the e*amples
in Swann and in Le Temps Rtrouv differ in their unfoldin$' the essential point is' in both
cases' that the chain of reminiscences $oes' in a metonymic way' far beyond the ori$inal
analo$y 1in %wann' the cup leads to the reminiscence of the room' from the room to the
house' then to the !illa$e and from there to the whole re$ion2" ,102he essential here is to
note that this first e*plosion @the analo$ic detonatorA is always accompanied also and
necessarily' by a kind of chain reaction which proceeds' not by analo$y but by conti$uity' and
which is !ery precisely the moment in which the metonymic conta$ion 1or' to use roust
term' the irradiation2 substitutes the metaphoric e!ocation"
B
#or Genette it is this crossin$ between metaphor and metonymy that ensures that there is a
narrati!e" If we had only had the metaphoric dimension' A la recherch du temps perdu
would not ha!e been a no!el but a succession of lyrical moments without any temporal
chainin$" %o he concludes+ ,Cithout metaphor roust 1appro*imately2 says' there are no
true memories8 we add for him 1and for e!erybody2+ without metonymy' there is no chainin$
of memories' no history' no no!el" #or it is metaphor that retrie!es lost 0ime' but it is
2
metonymy which reanimates it' that puts it back in mo!ement+ which returns it to itself and to
its true -essence.' which is its own escape and its own %earch" %o here' only here ) throu$h
metaphor but within metonymy ) it is here that the Darrati!e 1Rcit2 be$ins"
E
A few remarks before takin$ lea!e of Genette" Fe has illuminated !ery well the relation of
mutual implication between metaphor and metonymy which alone creates the unity of the
te*t" 0hat mutual implication has' thus' totaliing effects" Fe quotes' for e*ample' the
followin$ passa$e from roust+
,4e me :etais sur mon lit8 et' comme si :a!ais t sur la chouchette dun de ces bateau*
que :e !oyais asse9 prGs de moi et que la nuit on stonnerer de !oir se dplacer
lentement dans lobscurit' comme des cy$nes assombris et silentieu* mais qui ne
dorment pas' :tais entour de tous cHts des ima$es de la mer 1I' pEI62
And Genette comments+ ,;ne remarks here the e*plicit concurrence of the metaphoric
relation 1comme si2 and of the metonymic one 1pr!s de moi28 and the second metaphor is
also itself metonymic' $rafted into the first 1navires " cygnes2"
J
0he question which remains' howe!er' to be posed' is that concernin$ the kind of unity that
the articulation metaphorKmetonymy mana$es to constitute" Grantin$ ) as I think it should be
) that such a unity is !ital to the coherence of a te*t' there are se!eral possibilities as to how
to concei!e the interaction between these two dimensions" Genette does not' certainly'
su$$est that such an interaction should be concei!ed as the ad:ustment of the pieces of a
clockwork mechanism' and the !ery terms that he uses 1recoupement' croise2 su$$est that
he has somethin$ considerably more comple* in mind" Fe does not' howe!er' ad!ance !ery
much in determinin$ the specific nature of that recoupement' lar$ely' I think' because his
main concern is to show the presence of both tropes in the roustian te*t" Liscussin$
4akobsons distinction between metonymy as the prosaic dimension of discourse and
metaphor as the poetic one' he asserts that ,one should consider roustian writin$ as the
most e*treme attempt towards this mi*ed sta$e' fully assumin$ and acti!atin$ the two a*es
of lan$ua$e' which it would certainly be lau$hable to call -poem in prose. or -poetic prose.'
and which constitute' absolutely and in the full sense of the term' the 0e*t"
1I
#or the issues
that we are $oin$ to discuss in this essay it is crucial to precisely determine the lo$ics
in!ol!ed in the articulation of a*es of that ,mi*ed sta$e"
II
Genette is clearly conscious that his use of the cate$ories ,metaphor and ,metonymy is
somewhat idiosyncratic' for it $oes beyond what canonical rhetoric would ha!e ascribed to
3
them" 0here is in roust' for instance' a marked preference for ,continuous metaphors
1metaphores suivies2" ,0here are !ery rare in his work those ful$urant rapprochements
su$$ested by a sin$le word' the only ones for which classical rhetoric reser!ed the name
metaphor"
11
In many cases the analo$ical comparisons take place in a continuous way'
occupyin$ se!eral pa$es of the te*t" >ut also' it could seem abusi!e to call metonymy a
conti$uity of memories which does not in!ol!e any relation of substitution" Fowe!er' as
Genette points out' ,it is the nature of the semantic relation what is at stake' and not the form
of the fi$ure M roust himself has $i!en an e*ample of such an abuse by callin$ metaphor a
fi$ure which' in his work' is most frequently a comparison e*plicit and without substitution' so
that the effects of conta$ion to which we ha!e referred are nearly the equi!alent' on the a*is
of conti$uity' of what roustian metaphors are in the a*is of analo$y ) and are' in relation to
metonymy stricto sensu' what roustian metaphors are !is-N-!is classical metaphors M 0he
si$nal-sensation becomes !ery quickly in roust a sort of e#uivalent of the conte*t to which it
is associated' as the -petite phrase. of Ointeuil has become' for %wann and ;dette' -as the
national air of their lo!e.+ that is' its emblem"
1(
0his passa$e is crucial" Genette speaks' on the one hand' of an ,abusi!e use of rhetorical
cate$ories8 but' on the other' he describes such an abuse as a trans$ression in!ol!in$ a
mo!ement from the $orm of the fi$ure to a semantic relation which' while implicit in that form'
$oes clearly beyond those formal limits" %o the followin$ questions arise+
12 If the semantic relations underlyin$ both metaphor and metonymy transcend their
rhetorical form' are not those relations anchored in si$nification as such' beyond
classical rhetorical limits' or' alternati!ely' could not si$nification be seen as a
$eneralised rhetoric ) ie" that ,rhetoricity could be seen not as an abuse but as
constituti!e 1in the transcendental sense2 of si$nificationP
(2 In that case' is it enou$h to concei!e that ,beyond the rhetorical form as simply
,semantic ) which would necessarily attach it to the le!el of the si$nifiedP Could not
the relationship si$nifierKsi$nified in!ol!e a dialectic which takes us beyond
semantics' to a materiality of the si$nifier which inscribes rhetorical displacements in
the !ery structure of the si$nP 1Let us think in #reuds ,!erbal brid$es2"
/2 Chy are those displacements rhetorical in nature ) ie" dominated by the basic
opposition metaphorKmetonymyP
62 Fow to concei!e of that oppositionP Loes it in!ol!e a relation of complementarity or'
rather' a mutual limitation of their effects' so that metonymy establishes the limits of
metaphor and !ice !ersaP
4
;ne way of dealin$ with these questions would be to turn our attention to a theoretical
approach which e*plicitly tries to link rhetorical cate$ories to the structural dimensions of
si$nification as such" I am referrin$ to the famous essay by 5oman 4akobson ,0wo aspects
of lan$ua$e and two types of aphasic disturbances"
1/
4akobsons startin$ point is that
aphasia' bein$ a disturbance in lan$ua$e use' ,must be$in with the question of what aspects
of lan$ua$e are impaired in the !arious species of such a disorder 1p?J2 " %uch
interro$ation could not be answered ,without the participation of professional lin$uists familiar
with the patternin$ and functionin$ of lan$ua$e Jp?J2"
As 4akobson points out' any lin$uistic si$n presupposes its arran$ement throu$h two
different operations+ combination and contexture' by which the si$n $ets its location' in
accordance with syntactic rules in an orderly succession with other si$ns8 and selection and
substitution' by which a si$n can be replaced by others in any $i!en structural location" 0his
distinction corresponds to the two a*es of lan$ua$e identified by %aussure+ the synta$matic
and the paradi$matic 1which he called associati!e2" =ombination and substitution were' for
%aussure' the only two kinds of operation re$ulatin$ the relations between si$ns" %tartin$
from these two dimensions' 4akobson identifies two aphasic disturbances+ the first' the
similarity disorder' is related to the impossibility of substitutin$ terms' while the ability of
combinin$ them remains impaired8 in the second ) the contiguity disorder ) is that ability to
combine words what is affected" 3uite apart from aphasic disorders there is' accordin$ to
4akobson' a propensity in each lan$ua$e user to primordially rely on one or the other pole of
lan$ua$e" ,In a well known psycholo$ical test' children are confronted with some noun and
told to utter the first !erbal response that comes into their heads" In this e*periment two
opposite lin$uistic predilections are in!ariably e*hibited+ the response is intended either as a
substitute for' or as a complement to' the stimulus M 0o the stimulus hut one response was
burnt out8 another' is a poor little house" >oth reactions are predicati!e8 but the first creates
a purely narrati!e conte*t' while in the second there is a double connection with the sub:ect
hut+ on the one hand' a positional 1namely' syntactic2 conti$uity' and on the other a semantic
similarity 1pJIKJ12"
#rom these two a*es of lan$ua$e ) the paradi$matic and the synta$matic' substitution and
combination ) 4akobson mo!es to the rhetorical field+ metonymy would correspond to
combination and metaphor to substitution" And this alternati!e is not purely re$ional' but
re$ulates human beha!iour as a whole+ ,In manipulatin$ these two kinds of connection
1similarity and conti$uity2 in both their aspects 1positional and semantic2 ) selectin$'
combinin$ and rankin$ them ) an indi!idual e*hibits his personal style' his !erbal
predilections and preferences 1pJ12" ,0he bipolar structure of lan$ua$e 1or other semiotic
5
systems2 and' in aphasia' the fi*ation of one of these two poles to the e*clusion of the other'
require systematic comparati!e study" 0he retention of either of these alternati!es in the two
types of aphasia must be confronted with the predominance of the same pole in certain
styles' personal habits' current fashions' etc" 1pJ/2" 0his ar$ument is' for 4akobson at the
basis of a wider cultural interpretation" In !erbal art we ha!e that in poetry' lyrics pri!ile$es
the metaphorical a*is' as in romanticism and symbolism' while in realist art' whose epitome
is the no!el' metonymic displacements pre!ail" Ce ha!e here a$ain' in different terms' the
ar$ument that we had already found in Genette+ rousts ma:or work is a no!el and not a
paratactic succession of lyrical moments' because metaphors are $rounded in metonymic
connections" #or 4akobson this alternati!e applies equally to non-!erbal art+ in cubism' the
succession of synecdoches is essentially metonymic' while in surrealism the quasi-alle$orical
ima$es lean towards metaphor" And' in film' the plurality of an$les and close-ups in Griffins
production is metonymic in nature' while in =harlie =haplin and Eisenstein a metaphoric
substitution of ima$es structure the narrati!e" Indeed' any semiotic system can' for
4akobson' be understood in terms of the metaphoricKmetonymic alternati!e"
0he $reat merit of 4akobsons analysis is to ha!e brou$ht rhetorical cate$ories to their
specific location within lin$uistic structure' that is' to ha!e shown that it is the latter which is
at the root of all fi$ural mo!ements" 7etaphor and metonymy' in that sense' are not :ust
some fi$ures amon$ many' but the two fundamental matrices around which all other fi$ures
and tropes should be ordered" %o the classification of rhetorical fi$ures ceases to be a
heteroclite enumeration of forms and presents a clear structure anchored in their
dependence on the fundamental dimensions of lan$ua$e" 0he transition from these
dimensions to their specific rhetorical in!estment requires' howe!er' some further
considerations which I will summarise in the ne*t few pa$es"
12 0here is' in the first place' the question of the transition from the a*is of combination )
the synta$matic dimension ) to metonymy" >ecause' althou$h a tropolo$ical mo!ement
alon$ that dimension can only be concei!ed in metonymical terms' there is nothin$ in
combination' considered in isolation' requirin$ that such a mo!ement should take place" ;ne
can perfectly ima$ine a combination of terms followin$ syntactic rules which would not
in!ol!e any metonymic displacement" 0here is a 9ero-de$ree of the tropolo$ical as far as
combination is concerned" I can perfectly say ,scheresse des cheveux bruns instead of
,scheresse brune des cheveux" If so' the fi$ural would be somethin$ added to si$nification
from outside' not an inte$ral part of si$nification' and we would be back to the classical !ision
of the rhetorical as an adornment of lan$ua$e" %o if we want to establish a more intimate
6
connection between tropes and si$nification' we ha!e to find a way of underminin$ the !ery
possibility of a rhetorically neutral 9ero de$ree"
(2 0his way is quickly found once we mo!e from ,combination to the second a*is+
,substitutionKselection" #or here' on the difference with the a*is of combination' there is no
9ero-de$ree+ substitution 1a$ain' considered in isolation2 is not submitted to any a priori
syntactic rule" %aussure himself says it+ ,Chile a synta$m immediately calls the idea of an
order of succession and of a determinate number of elements' the terms of an associati!e
family do not present themsel!es in either a definite number or in a determinate order"
16
%o
the a*is of substitution' which is also constitutive o$ language' sub!erts the !ery principle of
structural locations on which the synta$matic succession is $rounded" %aussures dia$ram
of the ensemble of possibilities opened by substitution is most re!ealin$+
;ne of these possibilities is particularly important for our ar$ument+ the impossibility of
confinin$ substitution 1and' as a result' tropolo$ical trans$ression2 to the order of the
si$nified" %aussure asserts+ ,0here is either double community of sense and form' or
community of only either sense or form" Any word can e!oke anythin$ susceptible of bein$
associated with it one way or another"
1<
0his is why we asserted before that the ,beyond the
rhetorical form cannot be confined to semantic associations" ;ne possibility is that rhetorical
mo!ements do not only take place at the le!el of the si$nified but also at that of the si$nifier"
1In #reuds ,rat man' there is displacement from ,rat to Spielratten 1$amblin$2 and thus the
father 1a $ambler2 is incorporated into the ,rat comple*2"
7
/2 Chere do these considerations lea!e us as far as the relationship
metaphorKmetonymy is concernedP 0he main conclusion is that the notions of ,analo$y and
,conti$uity which are' respecti!ely' the definin$ $rounds of the two tropes' far from bein$
entirely different in nature' tend' on the contrary' to shade one into the other" Chy soP
>ecause both of them are trans$ressions of the same principle' which is the differential lo$ic
associated to the synta$matic a*is of the si$nifyin$ system" 0he only distinction which it
possible to establish between both fi$ures is that' in the case of metonymy' the trans$ression
of the structural locations that define the relations of combination is fully !isible' while in
metaphor' analo$y entirely i$nores those structural differentiations ) associations' as
%aussure shows' can mo!e into the most different directions" In one sense it can be said
that metaphor is the telos of metonymy' the moment in which trans$ression of the rules of
combination has reached its point of no return+ a new entity has come into e*istence which
makes us for$et the trans$ressi!e practices on which it is $rounded" >ut without those
trans$ressi!e practices which are essentially metonymic' the new metaphoric entity could not
ha!e emer$ed" As Genette shows in the case of roust' analo$y is always $rounded in an
ori$inary conti$uity"
Ce can draw here a conclusion which will be important for our political analysis+ conti$uity
and analo$y are not essentially different from each other but the two poles of a continuum"
Let us $i!e an e*ample which I ha!e discussed elsewhere"
1?
Let us suppose that there is a
nei$hbourhood where there is racist !iolence and the only force capable of confrontin$ it in
that area are the trade unions" Ce would think that' normally' opposin$ racism is not the
natural task of the trade unions' and that if it is taken up by them in that place it is by a
contin$ent constellation of social forces" 0hat is' that such a ,takin$ up deri!es from a
relation of conti$uity ) ie" that its nature is metonymic" Let us howe!er think that this ,takin$
up continues for a lon$ period of time ) in that case people would $et accustomed to it and
would tend to think that it is a normal part of trade union practices" %o what was a case of
contin$ent articulation becomes a part of the central meanin$ of the term ,trade union'
,conti$uity shades into ,analo$y' ,metonymy into ,metaphor" Anticipatin$ what we will
discuss presently' we can say that this is inherent to the central political operation that we call
,he$emony+ the mo!ement from metonymy to metaphor' from contingent articulation to
essential belon$in$" 0he name ) of a social mo!ement' of an ideolo$y' of a political
institution ) is always the metaphorical crystalli9ation of contents whose analo$ical links
result from concealin$ the contin$ent conti$uity of their metonymical ori$ins" =on!ersely' the
dissolution of a he$emonic formation in!ol!es the reacti!ation of that contin$ency+ the return
from a ,sublime metaphoric fi*ation to a humble metonymic association"
8
62 Cith this conclusion' howe!er' we ha!e only established+ a2 that the
metaphoricKmetonymic distinction has a matricial priority o!er other tropes ) which it is
possible' one way or the other' to reduce them to that matri*8 and b2 that such a matricial
distinction does not simply refer to opposites but to the two poles of a continuum" >ut to
assert that rhetoricity is inherent to si$nification requires one more step+ to show that without
a tropolo$ical displacement si$nification could not find its own $round" I ha!e tried to pro!e
this point elsewhere and I will not repeat it here"
1B
Let us :ust say that this proof requires
showin$ that si$nification' to be possible' requires its own closure' and that such a closure'
because it in!ol!es the representation of an ob:ect which is both impossible and necessary'
leads to the discursi!e production of empty si$nifiers" An empty si$nifier' as I ha!e tried to
show' is not :ust a si$nifier without a si$nified ) which' as such' would be outside si$nification
) but one si$nifyin$ the blind spot inherent to si$nification' the point where si$nification finds
its own limits and which howe!er' if it is $oin$ to be possible at all' has to be represented as
the meanin$less precondition of meanin$" In psychoanalytic terms' it would be the moment
of the 5eal ) the moment of distortion of the %ymbolic' which is the precondition for the
symbolic to constitute itself as totality" Dow if the representation of somethin$
irrepresentable is the !ery condition of representation as such' this means that the 1distorted2
representation of this condition in!ol!es a substitution' that is' it can only be tropolo$ical in
nature" And it is not a substitution to be concei!ed as a replacement of positi!e terms+ it will
in!ol!e $i!in$ a name to somethin$ which is essentially ,nameless' to an empty place" 0hat
is what $i!es its centrality to catachresis" And as any fi$ural mo!ement in!ol!es sayin$
somethin$ more than what can be said throu$h a literal term' catachresis is inherent to the
fi$ural as such' it becomes the trade mark of ,rhetoricity as such"
<2 Let us $o back' at this point' to the question of the $round of the
metaphoricKmetonymic continuum" %uch a $round is $i!en' as we ha!e indicated' by the
opposition of any tropolo$ical mo!ement to the differential lo$ic of combination inherent in
the synta$matic pole of si$nification" 0he difference between analo$y and conti$uity is that
althou$h both' throu$h their substitutions' sub!ert such a differential lo$ic' the !isibility of
what is sub!erted is !ery much present in the case of metonymy' while it tends to disappear
in the case of metaphor" >ut if this sub!ersion of combinatorial locations is inherent to
rhetoricity' and rhetoricity is one of the dimensions of si$nification' this means that the latter
can only be concei!ed as an endless process of successi!e institutions and sub!ersions of
different locations" 0hat is why structuralism of strict obser!ance has always tended to
emphasise the synta$matic pole of lan$ua$e at the e*pense of the paradi$matic one" >ut
the ambi$uity created by the operation of the two opposed lo$ics of combination and
substitution did not $o entirely unnoticed' e!en in the work of %aussure" As 4oan =op:ec has
9
pointed out+ ,Emphasi9in$ the -synchronic perspecti!e. of the lin$uist and his community'
%aussure e!entually decided to $i!e priority to the contemporaneous system of si$nifiers
operatin$ at some 1hypothetical2 moment+ the present" #or$ettin$ for his own purposes his
important stipulation that meanin$ must be determined retroacti!ely' that is' for$ettin$ the
diachronic nature of meanin$' he ultimately founded the science of lin$uistics on the
systematic totality of lan$ua$e" 0hus' the structuralist ar$ument ceased to be that the final
si$nifier %
(
determines that which has come before' %
1
and became instead that %
(

determines %
1
and %
1
determines %
(
8 that is reciprocal oppositions stabli9e meanin$s
between coe*istent terms8 and differential relations no lon$er threatened the trans!aluation
of all precedin$ si$nifiers"
1E
If we incorporate' howe!er' the diachronic perspecti!e that %aussure himself enounced but
for$ot about' the consequence is clear+ %
(
can be the $round of the system only as far as it
does not ha!e a precise' particular location within it" 0he same ar$ument can be presented
in terms of set theory+ what names the set cannot be part of it" Chat the rhetorical turn
would add to this ar$ument is that the term namin$ the set would be one of the particular
elements of that set which splits its own identity between its own particularity and its role of
si$nifyin$ the totality" It is this double role which is at the root of all tropolo$ical displacement"
?2 5hetoricity' as a dimension of si$nification' has no limits in its field of operation" It is
co-terminous with the !ery structure of ob:ecti!ity" 0his is' first of all' connected with the
notion of ,discourse that we ha!e used in our work' which is not e*clusi!ely or primarily
linked to speech or writin$' but to any si$nifyin$ practice" 0his in!ol!es that it is equi!alent to
the social production of meanin$' that is' to the !ery fabric of social life" 0here is no
possibility of any strict separation between si$nification and action" E!en the most purely
constati!e of assertions has a performati!e dimension' and' con!ersely' there is no action
which is not embedded in si$nification" #or the same reason' there cannot be any stark
separation between si$nification and affect' $i!en that the latter is only constituted throu$h
differentially cathectin$ the !arious components of a si$nifyin$ chain" As in Citt$ensteins
,lan$ua$e $ames' words and actions 1to which we should add affects2' are part of an
interdependent network2" 0his means that lin$uistic cate$ories such as the si$nifierKsi$nified
and synta$mKparadi$m distinctions ) if properly theori9ed ) cease to belon$ to a re$ional
discipline and come to define relations operatin$ in the !ery terrain of a $eneral ontolo$y"
>ut' secondly if si$nification could close itself in synta$matic terms ) ie" if paradi$matic
relations of substitution could themsel!es be reabsorbed by combinatorial rules ) the role of
rhetoric could not be ontolo$ically constituti!e" 0he structuralist closure of the relation of
10
mutual determination between %
1
and %
(
could be achie!ed without any tropolo$ical de!ice
bein$ brou$ht into the picture' and so rhetoric would be rele$ated to its traditional role as
adornment of lan$ua$e" >ut it is here that our remarks concernin$ the impossibility of
achie!in$ any closure of a si$nifyin$ system without representin$ the irrepresentable become
rele!ant" ;nce the centrality of catachresis is fully accepted' rhetoricity becomes a condition
of si$nification and' as a result' of ob:ecti!ity"
0hirdly' once the status of rhetoric has been reco$nised in its true ontolo$ical $enerality'
relations that in this essay we ha!e approached in a strictly tropolo$ical terminolo$y' are
likely to be reproduced at different le!els of analysis of human reality' e!en when the
rhetorical nature of the distinction introduced is not percei!ed or reco$nised" In
sychoanalysis' to $i!e the most ob!ious e*ample' the rhetorical character of the workin$s of
the unconscious has been e*plicitly reco$nised a lon$ time a$o" =ondensation has been
assimilated to metaphor and displacement to metonymy" 0he lo$ic of the ob:ect a in!ol!es
precisely an in!estment by which an ordinary ob:ect becomes a substitute for the
unreachable 0hin$" In Lacans terms8 sublimation is to ele!ate an ob:ect to the di$nity of the
0hin$" 0his operation of in!estment is catachrestical throu$h and throu$h" And =op:ec' in
her film studies' has shown how the close ups are not a part within the whole but a part
which functions as the !ery condition of the whole' as its name' leadin$ to that contamination
between particularity and totality which' as we ha!e seen' is at the heart of all tropolo$ical
mo!ement"
In the rest of this essay I will try to show the operation of those distinctions that we ha!e
been discussin$' within the political field" I will ar$ue that the tensions that we ha!e detected
alon$ the continuum metaphorKmetonymy' can be seen as fully operatin$ in the structuration
of political spaces" I will discuss two cases" In the first' we will see an almost complete
unilaterali9ation of the metaphoric operation" In the second' a systematic blocka$e of the
transition from metonymy to metaphor ) ie" the pre!ention that conti$uity shades into
analo$y" 0he first possibility I will illustrate with the lo$ic of the $eneral strike in %orel8 the
second' with the political strate$y of Leninism"
III
Ce ha!e to $i!e some precise theoretical status to the operation in which we en$a$e
oursel!es when tryin$ to see the way rhetorical cate$ories are 1implicitly2 present in those
lo$ics $o!ernin$ the distinctions which structure areas different from those in which rhetoric
was ori$inally thou$ht to be operati!e" Ce should basically a!oid two temptations" 0he first
is to make of rhetorical cate$ories the locus of a hard transcendentality' that is' of a le!el in
11
which all pertinent theoretical distinctions would be formulated and which would reduce the
terrains of their ,application to the empiricity of ,case studies" >ut we should also a!oid the
other e*treme' consistin$ of seein$ the two le!els as fully enclosed uni!erses' whose mutual
relations could only be concei!ed in terms of purely e*ternal homolo$ies" 0he question of
the comparison itself between re$ions and le!els should be concei!ed in tropolo$ical terms+
no le!el has a transcendental priority o!er the other' so that their !ery interaction should be
seen as an area of displacements blurrin$ the frontiers between the empirical and the
transcendental" Each should theoretically enrich the understandin$ of the other in an
interte*tuality which has no ultimate anchorin$ point"
If we try to think those or$anisin$ cate$ories of the political field which make possible a
comparison with out rhetorical analysis' we could ad!ance the followin$ thesis+ politics is
articulation of hetero$eneous elements' and such an articulation is essentially tropolo$ical'
for it presupposes the duality between institution and sub!ersion of differential positions
which we found as definin$ a rhetorical inter!ention" %ocial or$anisation is not' howe!er'
e*clusi!ely political8 to a lar$e e*tent it consists of differential positions which are not
challen$ed by any confrontation between $roups" It is only throu$h this confrontation that the
specifically political moment emer$es' for it shows the contin$ent nature of articulations"
&sin$ a Fusserlian distinction' we could say that the social is equi!alent to a sedimented
order' while the political would in!ol!e the moment of reactivation" =ontemporary forms of
technocratism would e*press this dissolution of the political and the reduction of the
mana$ement of the community to a mere question of e*pertise" It is the replacement of
politics by knowled$e' whose earliest formulation we find in lato"
Ce ha!e here the basis for a comparison between this duality politicsKadministration and the
two a*es of si$nification ) that of combinations and that of substitutions" 0he more social
order is stable and unchallen$ed' the more institutional forms will pre!ail and will or$anise
themsel!es in a synta$matic system of differential positions" 0he more the confrontations
between $roups defines the social scene' the more society will be di!ided into two camps+ at
the limit' there will be a total dichotomisation of the social space around only two synta$matic
positions+ ,us and ,them" All social elements would ha!e to locate their identities around
either of these two poles' whose internal components would be in a mere relation of
equi!alence" Chile in an institutionalist political discourse there is a multiplication of
differential positions in a relation of combination with each other' in an anta$onistic discourse
of rupture the number of synta$matic differential positions is radically restricted' and all
identities establish paradi$matic relations of substitution with all the others in each of the two
poles" In my work I ha!e called these two opposed political lo$ics' lo$ic of difference and of
12
equi!alence' respecti!ely" Gi!en that the equi!alential chain establishes a paratactic
succession between its component links' none of them can ha!e a position of centrality
founded in a combinatorial lo$ic of a hypotactic nature" %o if the unity of the equi!alential
chain is $oin$ to be or$anised around a pri!ile$ed si$nifier' such a pri!ile$e cannot be
deri!ed from a differential structural position' but from a cathectic in!estment of a radical
kind" 0he symbols of %olidarnosc in oland $ot their success not from any structural
centrality of the Lenin shipyards in the country' but from the fact that they e*pressed radical
anti-status quo feelin$s at the moment in which many other social demands were frustrated
for not findin$ institutional channels of e*pression within the e*istin$ political system" 0his
process by which identities cease to be purely immanent to a system and require an
identification with a point transcendent to that system ) which is the same as sayin$+ when a
particularity becomes the name of an absent uni!ersality ) is what we call hegemony" Its
lo$ic is identical with the lo$ic of the ob:ect a' which we ha!e already referred to' and' for the
reasons that we ha!e $i!en' it is essentially catachrestical 1Q rhetorical2"
;ne last point requires our consideration" A he$emonic operation is essentially tropolo$ical'
but requires !ery particular strate$ic mo!es to be performed within the
metaphoricKmetonymic continuum" ;ther mo!es' howe!er' are equally possible' $i!en that
the continuum does not prescribe a priori either the direction that inter!entions in it should
take' or the different forms of articulation between its two e*treme poles" Genette presents
the decision by roust which made possible the e*istence of a narrati!e' as precisely that+ a
decision" >ut he himself points out that other decisions would ha!e been equally possible' in
which case we would not ha!e had a no!el but' for instance' a succession of lyrical
moments" In the same way' the emer$ence of a he$emonic lo$ic in Gramscis political
thou$ht takes place a$ainst the back$round of !arious different ways of concei!in$ politics in
the 7ar*ist tradition which' while still bein$ describable in terms of the possibilities opened by
the metaphoricKmetonymic distinction' are different from the he$emonic turn" It is to that
history that we ha!e now to address our attention"
IO
Ce ha!e spoken about a 9ero-de$ree of the rhetorical' whose attainment would ideally
require that the synta$matic differential lo$ic is able to dominate the whole field of
si$nification 1in the e*panded sense that we ha!e $i!en to this last term2" 0he pre-requisite
for attainin$ such a 9ero de$ree would be' of course' the ability of the synta$matic lo$ic to
fully control paradi$matic substitutions 1an ability which we ha!e $ood reasons to be rather
sceptical about2" Fowe!er' we ha!e so far limited the question of the 9ero de$ree to its
structuralist !ersion ) ie" to a purely synchronic system ) while identifyin$ the notion of
13
diachrony to a retroacti!e fi*ationKtrans$ression which would operate from ,outside the
structural ,inside" Is this' howe!er' the only true alternati!eP Is it necessary that a purely
synta$maticKcombinatorial space is or$anised in a synchronic wayP I think it is not" As far as
diachrony is not concei!ed as a contin$ent' e*ternal inter!ention' but as structured by a
teleolo$y' a diachronic succession is perfectly compatible with a 9ero de$ree of the
tropolo$ical" ure differentiality 1our 9ero)de$ree2 is not necessarily linked to either
simultaneity or succession"
It is from this point that we ha!e to start in our consideration of the 7ar*ist tradition" #or at
the root of this tradition there is a discourse anchored in Fe$elian teleolo$y" Ce know the
definin$ features of the latter+ the essential determinations of any entity are to be found in its
conceptual specificity8 the conceptual contradictions inherent in this specificity force us to
mo!e to a new entity embodyin$ a new conceptual sta$e' etc" 7ar* did not chan$e thin$s in
the least with his ,in!ersion of Fe$elian dialectics+ if the $round is ,matter rather than the
,Idea' but matter has inner laws of mo!ement which are conceptually specifiable' 7ar*s
materialism is as idealistic as Fe$els" ;ntolo$ically speakin$ they are not' actually' different
from each other"
0he important point for our sub:ect is that in the !ision of Fistory which emer$es from this
diachrony' the different sta$es in the succession are not concei!ed as interruptions of what
preceded them but as teleological $ul$ilments" Ce are dealin$ with a pure combination in
which each actor and task has an assi$ned place in a secular eschatolo$y $rounded in the
,necessary laws of Fistory" It comes as no surprise that the main political consequence of
this approach is to pri!ile$e ,strate$y o!er ,tactics" Lon$ term strate$ic calculations were
considered to be possible because the teleolo$ism of the premises opened the way to
historical predictions' e!en if they were only ,morpholo$ical predictions' to use the words of
Antonio Labriola" And any unfulfilment of those predictions could be dismissed as a
temporary aberration to be superseded once the ,necessary laws reasserted their lon$ term
!alidity"
0he most e*treme !ersions of this teleolo$ism are to be found' of course' in the orthodo*
currents of the %econd International' but it is enou$h to read the ,reface to the %riti#ue o$
&olitical 'conomy to realise that' althou$h in less crude ways' it impre$nates the whole of the
7ar*ist tradition" 0hat is why we can speak of a rhetorical 9ero de$ree+ in this synta$matic
succession there is no place for either metonymic displacements or metaphoric
rea$$re$ations" ;ne could' howe!er' ask oneself+ but is it not precisely alon$ the
combinatorial succession of differential positions that metonymy operatesP 0he answer is
14
yes' but metonymy' as we know' in!ol!es a sub!ersion of the principle of differentiality
throu$h substitutions $rounded in conti$uity' and it is precisely these substitutions that
synta$matic literalism tends to block"
7ar*ist literalism required the reduction of the process of historical de!elopment to a
mechanism which had to be conceptually apprehensible as far as its laws of mo!ement are
concerned" >ut that conceptual apprehensibility also required that anythin$ escapin$ what is
specifiable by those laws should be discarded as historically irrele!ant" ,0he chan$es in the
economic foundation lead sooner or later to the transformation of the whole immense
superstructure" In studyin$ such transformation it is always necessary to distin$uish
between the material transformation of the economic conditions of production' which can be
distin$uished with the precision of natural science' and the le$al' political' reli$ious' artistic or
philosophical ) in short' ideolo$ical forms in which men become conscious of this conflict and
fi$ht it out" 4ust as one does not :ud$e an indi!idual by what he thinks about himself' so one
cannot :ud$e such a period of transformation by its consciousness' but' on the contrary' this
consciousness must be e*plained from the contradictions of material life' from the conflict
e*istin$ between the social forces of production and the relations of production"
1J
Dow' it is precisely this sharp distinction between what is rele!ant and what is not' that is
blurred durin$ the first ,crisis of 7ar*ism at the end of the RIRth century" =apitalism
reco!ered after a lon$ period of depression' and the transition to the monopolistic phase and
to imperialism started" In such a situation the socialist faith in the collapse of the system as a
result of its internal contradictions was shaken" Fistorical de!elopment had re!ealed itself to
be far more comple* than had been assumed' and such a comple*ity took the form of a
contamination between social le!els which' accordin$ to the classical theory should ha!e
remained distinct" 1,;r$anised capitalism ceased to be e*plainable by pure market laws and
an element of conscious re$ulation inter!ened at the !ery le!el of the infrastructure8
imperialism led to the emer$ence of a ,workin$ class aristocracy and consequently to an
attenuation of class conflicts' etc"2 0he consequence for our analysis is that the !ery terrain
which had made accessible the 9ero de$ree of the tropolo$ical was shattered and rhetorical
mo!ements became hi$hly important both in a metaphoric and a metonymid direction"
0his tropolo$ical turn' howe!er' took a !ariety of forms and directions" As we anticipated' the
first e*ample that well refer to is the later work of Geor$es %orel" As many other socialist
thinkers of his time' %orel' at the time of writin$ the Re$lections on (iolence' had lost faith in
the perspecti!e of capitalism brin$in$ about its own collapse as a result of purely economic
laws" %o in order to keep ali!e the re!olutionary !ocation of the workin$ class' it was
15
necessary to appeal to somethin$ different from economic determinism" %ome kind of
sub:ecti!e principle had to be brou$ht into the picture" It is important to realise that' for %orel'
his support for the proletarian stru$$le was not $rounded in the :ustice of the workers
demands but in his belief that the proletariat was the only force in society capable or
pre!entin$ bour$eois decadence" #or' the prospect facin$ contemporary societies was a
$eneral decline of ci!ilisation" 0he principle capable of keepin$ the purity of proletarian
identity was violence" #or this purpose' it was essential that the workin$ class did not
inter!ene in politics' for that would co-opt it into the mechanisms of the bour$eois %tate" Fe
opposed ,proletarian !iolence to ,political !iolence ) the latter bein$ epitomised by
4acobinism"
roletarian !iolence had to be or$anised around a myth" ,@7Aen participatin$ in $reat social
mo!ements represent to themsel!es their immediate action under the form of ima$es of
battles ensurin$ the triumph of their cause" I propose to call myths these constructions
whose knowled$e is so important for the historian+ the syndicalists $eneral strike and 7ar*s
catastrophic re!olution are myths"
(I
Fe counterposes ,myth and ,utopia" Chile the latter is
a pure intellectual construction' the blueprint of a future or ideal society' myth is :ust a set of
ima$es capable of $al!anisin$ the masses ima$ination and pro:ectin$ them into historical
action"
0he myth around which proletarian identity should be or$anised is that of the general stri)e"
,I understand that this myth of the $eneral strike horrifies @$roisseA many wise people because
of its character of infinitude8 the present world is !ery much inclined to return to the opinion of
the ancients and to subordinate morals to the $ood mana$ement of public affairs' which
leads to locate !irtue in a :ust middle" As far as socialism remains a doctrine entirely
presented through words' it is easy to make it de!iate towards a :ust middle8 but this
transformation is clearly impossible once one introduces the myth of the $eneral strike' which
in!ol!es an absolute re!olution"
(1
And' a$ain+ ,0oday re!olutionary myths are almost pure'
they make possible to understand the acti!ities' feelin$s and ideas of the popular masses
preparin$ themsel!es to enter into a decisi!e stru$$le' they are not descriptions of thin$s' but
e*pressions of wills"
((
In a myth' the infinitude of the task $oes to$ether with the paucity of its contents" Its function
is' precisely' to separate the militant from the concrete aim of his particular action" Let us
suppose that a $roup of workers participate in a strike for hi$her wa$es" If the strike is
successful' and its only aim was that particular demand' success leads to demobili9ation and
to the inte$ration of the workers into the status quo" Fowe!er' if participation in that concrete
16
action is seen as a simple episode' educatin$ the proletariat for the final aim' the meanin$ of
the particular stru$$le chan$es alto$ether" >ut' for this' the myth of the $eneral strike has to
be operatin$ from the !ery be$innin$" 0his e*plains the in$initude of the task' to which %orel
refers+ it cannot be identified with any particular aim" And it e*plains also the po!erty of its
contents which is actually more than po!erty' for' as the name of an infinite task' it ne$ates
the !ery possibility of any content 1which would necessarily ha!e to be finite2" 0he %orelian
myth is one of the purest e*amples of what we ha!e called ,empty si$nifiers" It does not
matter whether the $eneral strike is an e!ent which could happen or not" Althou$h %orel is
not entirely e*plicit in this respect' I think that the !ery lo$ic of his ar$ument leads towards a
ne$ati!e response' for any finite fulfilment would compromise the infinitude of the task" Its
status approaches that of Sants re$ulati!e idea"
Fow' howe!er' to read this set of displacements that %orel brin$s about a$ainst the
sequence of cate$ories of classical 7ar*ismP Chere and how e*actly does the tropolo$ical
turn take placeP 0o start with' in %orel there is not any synta$matic plurality of places of
enunciation because they all con!er$e in reinforcin$ a unique proletarian identity" Chether
we are dealin$ with a strike' a demonstration' a factory occupation' they are simply
occasions for the rehearsal of a unique ,future e!ent+ the $eneral strike" 0hese occasions
are certainly plural' but their plurality is present only to eclipse itself as a mere support of the
sin$le e!ent which speaks throu$h all of them" 0hat is' we are faced with a pure
metaphorical rea$$re$ation which is not interrupted by any metonymical plurality" 0here is
nothin$ to displace' because the sites of the metaphorical e!ent are there :ust in order to be
ne$ated by the latter" 0o put it in clear terms+ the re!olutionary break does not proceed
throu$h equi!alence but throu$h absolute identity" %o' in some way %orel is the symmetrical
re!erse side of the ,rhetorics 9ero de$ree of the %econd International" #or the latter' there
was no room for any tropolo$ical mo!ement in the determination of the emancipatory
sub:ect" #or %orel' such a determination could only proceed throu$h an e*treme form of that
tropolo$ical mo!ement' namely' a pure metaphor which has eliminated all traces of its
metonymical $roundin$" Analo$y unconceals an essence which has broken all links with
conti$uity" Equi!alence is replaced by pure identity" 1As this identity' howe!er' is constructed
around an empty place ) the $eneral strike ) whose discursi!e effects depend on its lack of
content' its assertion is close to nihilism" Dot surprisin$ly' %orelianism fed !ery different
currents of thou$ht' from radical communism and ultra-leftism' to fascism2" Ce can $o back
here to Genettes analysis of roust" Accordin$ to him' as we ha!e seen' there is narrati!e in
roust only because metaphors are inscribed in a metonymic mo!ement8 otherwise we will
only ha!e a succession of lyrical moments" Cell' this last possibility is what %orels te*t
enacts" Each re!olutionary act does not find its meanin$ in a succession endowin$ it with its
17
raison d*tre within the series' but' rather' each of them is the e*pression of some sort of
repetition dri!e constantly reinstatin$' in a %ysifus way' a sin$le identity" 0hat is why
Genettes notion of a succession of lyrical moments as an alternati!e to rousts narrati!e )
ie" pure metaphorical flashes not inscribed in any metonymical succession ) applies so well
to %orels !ision of politics" And' also' why there can be no %orelian strate$y based in a lon$
term calculation" Chile for a Santsky or a lekhano! such calculation was based in
supposedly known laws of history' for %orel the mere idea of a lon$ term prediction makes no
sense" 0he assertion of a re!olutionary sub:ecti!ity lar$ely escapes strate$ical
considerations"
O
If %orels discourse is structured in a terrain in which political sub:ecti!ity can only operate
throu$h a total metaphor which conceals e!en the traces of its metonymic $round' the
e*perience of Leninism is different+ the metonymic sub!ersion of the differential space of
7ar*ist teleolo$y has to remain !isible' to the point of makin$ impossible the mo!ement
towards its metaphoric telos" Leninism emer$es as a political answer to an anomaly in
historical de!elopment" 5ussia was supposed to follow the pattern of the classical
bour$eois-democratic re!olutions of the Cest" 0he task ahead was the o!erthrowin$ of
0sarism and the openin$ of a lon$ period of capitalist democracy' so that socialism was only
a lon$-term prospect' to be achie!ed as a result of the contradictions of a fully fled$ed
capitalist society" In that democratic re!olution the bour$eoisie was supposed to be the
,natural leadin$ force" 0asks and forces were assi$ned roles accordin$ to a pre-ordained
succession" 0he anomaly was that the autoctonous 5ussian >our$eoisie had arri!ed too late
to the historical scene' when a world capitalist market was already well established' and as a
result it was too weak to carry out its own democratic re!olution" =apitalism' howe!er' was
rapidly de!elopin$ in 5ussia as a result of forei$n in!estments' so that there was the
parado*ical situation - ,anomalous re$ardin$ the canonical pattern ) of a country which was
mature for a democratic re!olution and in which' howe!er' the ,natural a$ent of that historical
transformation was incapable of carryin$ out its task"
As a result of capitalist de!elopment' howe!er' a robust workin$ class was emer$in$' which
had none of the limitations of the indi$enous bour$eoisie' and so ) this was the thesis of the
5ussian social-democrats ) it had to take up the historical task of leadin$ the democratic
re!olution 1in alliance with the peasantry' in the Leninist !ersion2 which its natural a$ent' the
bour$eoisie' had left unfulfilled" 0his anomalous takin$ up of a task by a force which was not
its natural a$ent is what the 5ussian social-democrats called ,he$emony" %o we ha!e a
fracture in historical de!elopment' a discontinuity in the sequence of its cate$ories" 0he
18
takin$ up of the democratic tasks by the workin$ class was an e!ent politically e*plainable by
a set of historical circumstances' but not insertable as one of the necessary links of the
canonical paradi$m" It was an ,e*ceptionality' to use the terminolo$y of the time"
Dow if we study the structure of this e*ceptionality' we immediately see that it was the
presence of the workin$ class at the centre of historical e!ents at a moment in which the
country was mature for a democratic re!olution which assi$ned it to that role" It was a
relation of contiguity" %o we are dealin$ with the construction of a new link between task and
a$ent which can only be concei!ed as a metonymic displacement"
Ce know' howe!er' that any metonymy has a natural tendency to shade into a metaphor' the
relation of conti$uity to become' throu$h continuous association' one of analo$y" %o we
could normally e*pect that the nature of the democratic task chan$ed when taken up by the
proletariat' and that the class nature of the latter was also altered as a result of takin$ up a
democratic task" Fowe!er' nothin$ of the kind happened" 0he whole Leninist strate$y was
desi$ned to pre!ent that the e*ceptional task became the site of the construction of a new
political sub:ecti!ity" 0he class nature of the proletariat had to remain unchan$ed" 0he
Leninist motto was+ ,to strike to$ether and to march separately" Chy soP Oarious reasons
conspire to it' but the main one was that for 5ussian re!olutionists ) the >olshe!iks included
) 5ussian e*ceptionality was e*actly that+ an e*ception and' on top of that' one which was
$oin$ to be short li!ed" Deither 0rotsky' nor Lenin ) e!en after the ,April 0heses ) thou$ht
that a proletarian power in 5ussia' $i!en its backwardness' had any prospect unless it found
its natural continuity in a re!olution in Germany and the main other hi$hly de!eloped
capitalist countries in the Cest" If that had been the case' the 5ussian ,e*ceptionality would
ha!e been quickly inte$rated into a ,normal process of historical de!elopment"
If we consider the matter retrospecti!ely' we find here the root of the double discourse which
will be inscribed in the =ommunist e*perience of the years to come" 0he canonical
sequence of cate$ories had to be maintained as an ultimate unsurpassable hori9on ) the
7ar*ist synta$m was ne!er formally questioned ) but' as a counterpart' actual politics was
$oin$ to be dominated increasin$ly by an empiricism of e*ceptionalities which eluded any
theorisation" %talins Realpoliti) was the e*treme e*pression of this di!orce between theory
and practice' but in more attenuated forms it is $oin$ to dominate the whole of =ommunist
e*perience" 0he way in which both le!els were combined can perhaps be seen at its best in
the case of 0rotsky" 0he whole lo$ic of ,permanent re!olution is only thinkable if the
empiricism of the e*ceptionalities is articulated to the discourse of the ,normal synta$matic
de!elopment" 0he ar$ument runs as follows" 5ussia was mature for a democratico-
19
bour$eois re!olution in which the bour$eoisie ) 0rotsky accepted the point ) was incapable
of playin$ the leadin$ role" 0his would result in a democratic re!olution led by the proletariat"
>ut ) 0rotsky added ) the bour$eoisie would not tolerate proletarian power ) e!en if confined
to democratic limits ) and would respond with a massi!e lockout" 0he result would be that
the workers mo!ement' in order to consolidate its power' would ha!e to ad!ance in a
socialist direction" 5e!olutions always start with democratic banners' but their stabilisation
and consolidation requires their transition to the socialist sta$e" 0his model will be repeated
ad nauseam by 0rotskyists in all ima$inable historical conte*ts" 0he classic ,sta$eism'
althou$h interrupted by an ,e*ceptionality is in full operation+ the class nature of social
a$ents is unquestioned as well as that of the tasks and of the succession of phases"
%o the metonymic moment had to be fro9en' pre!entin$ the construction of new identities
throu$h metaphoric rea$$re$ations" Fere we see the difference with %orel" #or him there is
no narrati!e' only the sequence of metaphoric moments throu$h which proletarian identity is
constantly reinforced" #or Leninism' the interaction between the two discursi!e le!els' forces
it to en$a$e in a permanent narrati!e' so that the metonymic moment is ne!er abandoned" It
is for that reason that Leninism is an eminently strate$ic type of discourse' whose difference
with the strate$y of the %econd International is' howe!er' !isible+ for the latter' strate$ic
reflection was based on a historical prediction $rounded in the necessary laws of history8
while for Leninism' $i!en the operation of e*ceptionalities' strate$ies ha!e more the
character of con:unctural analyses"
0his notion of con:unctural analysis forces us' howe!er' to mo!e beyond Leninist fro9en
metonymies and' indeed' beyond the historical hori9on of 7ar*ism" #or the question is+ how
e*ceptional are the e*ceptionsP Accordin$ to Lenin' world capitalist market is not only an
economic but also a political reality+ it is structured as an imperialist chain" =rises can take
place in one point of it which result ) $i!en that the chain is broken by its weakest link ) in
dislocations of the relations of forces in other points of the chain" 0his makes possible a
sei9ure of power e!en if the ,ob:ecti!e material conditions ha!e not been met" In such
situations there is no lon$er any question of either a pure combination of sta$es ) as the one
postulated by the theory of combined and une!en de!elopment ) nor of a necessary class
belon$in$ of social a$ents' for what is at stake is the constitution of comple* social identities
constructed on the basis of practices homo$enei9in$ the hetero$eneous" 0hat is' we are
dealin$ with metaphoric rea$$re$ations" #ro9en Leninist metonymies no lon$er do the trick"
I think that Gramscis notion of ,collecti!e wills should be read in this li$ht" >ut this
incorporation of the metaphoric dimension does not lead us back to %orels camp either" #or
%orel this is a unilateralisation of metaphor' because the proletarian identity that he tries to
20
consolidate is $i!en in ad!ance" Do question for him of incorporatin$ hetero$eneous
elements into a wider social identity" 0hat could only lead' in his !iew' to undermine the class
consciousness of the proletariat" Fowe!er' once the political process is seen not only as a
reassertion of an identity but as its construction ) as in Gramscis ,war of position ) the
metonymic dimension cannot be i$nored" Fe$emony means the passa$e from metonymy to
metaphor' from a ,conti$uous startin$ point to its consolidation in ,analo$y" >ut with this we
are !ery close to the relationship metaphorKmetonymy which Genette finds in rousts te*t"
0ranslatin$ it into political lan$ua$e' we could say that because there is Darrati!e 1Rcit2
there is strate$y" >ut as the identity of the a$ents of that strate$y is not $i!en beforehand'
we will always ha!e short term strate$ic mo!ements' not anchored in any eschatolo$y" 0hey
will e*actly operate at the point in which metaphor and metonymy cross each other and limit
their mutual effects"
21
1
,7tonymie che9 roust' in Grard Genette' +igures ,,,' aris' Editions du %ueil' 1JB(' p61-?/"
2
%tephen &llmann' Style in the +rench -ovel' =ambrid$e' 1J<B"
3
Genette' op cit' p6(
4
.p cit' p6<
5
.p cit' p?I
6
.p cit' p<?
7
.p cit' p<?
8
.p cit' p?/
9
.p cit' p<1' note <
10
.p cit' p?1
11
.p cit' p<<
12
.p cit' p<E
13
5 4akobson' ,0wo aspects of lan$ua$e and two types of aphasic disturbances' in Falle and 4akobson'
+undamentals o$ Language' 0he Fa$ue' 1J<E" a$es of the quotations are indicated in the te*t"
14
# de %aussure' %ours de linguisti#ue gnrale' aris' ayot' 1JEI' p1B6
15
.p cit' p1B6
16
E Laclau' ,0he olitics of 5hetoric' in /aterial 'vents0 &aul de /an and the A$terli$e o$ Theory' edited by 0om
=ohen' 4 Fillis 7iller' Andr9e: Carminski and >arbara =ohen' 7innesota &ni!ersity ress (II1' pp((J-(</
17
%ee 'mancipation1s2' London' Oerso' 1JJ? 1essay on ,Chy do Empty %i$nifiers 7atter to oliticsP2 and .n
&opulist Reason' London' Oerso' (II<' chapter 6
18
4oan =op:ec' ,%e* and Euthanasia of 5eason in Read my 3esire' =ambrid$e 17ass2' 7I0 ress' 1JJ<'
pp(I<-(I?
19
S 7ar*' A %ontribution to the %riti#ue o$ &olitical 'conomy' London' 1JB1' p (6
20
Geor$es %orel' Re$lexions sur la violence' aris' %euil' 1JJI' p(1
21
.p cit' p(<
22
.p cit' p(JK/I