1. Introduction

One should expect the reflection about the interplay of social and linguistic changes to touch upon quite fundamental questions. Still, if we do not want to bet blindly on the chances of the paralogic known from old pioneers' accounts and from detective stories, where false hypotheses lead to true discoveries, we should keep in mind how insidious any kind of not avowed "evidence" may be. We are ready to accept as a truism that, when a major social transformation occurs, linguistic changes are likely to happen as well. Yet a closer scrutiny beyond such common sense does not fail to blur the picture. To keep the epistemological background of this paper sufficiently open, let me just indicate that there is no arguable need to assign to the "social" any priority over the linguistic; nor is there any reason to expect some kind of automatic causation relation between a social transformation and the language change.

Against an eventual socio-centric temptation, the opposite thesis can quite plausibly be argued: linguistic operations in speech can "cause" social changes, and speech-induced systemic linguistic changes can precede socio-systemic transformation.Since the so-called social change is an outcome of complex human interaction, it is in many ways always - already linguistically mediated: what may later become a major social transformation, may have originated in a purely verbal operation, like calling a governmental decree illegal or the regime itself illegitimate.I have myself been involved in an effort where a lexical substitution may have had consequences I still feel reluctant to assume: an appeal to the "sovereignity of the Republic of Slovenia", initially meant as a tactical side-move offering to the local authorities a common ground for an alliance against the political aspirations of the Federal Army, has been irresistibly transformed in the public use into the slogan "sovereignity of Slovene nation",which was to become the rallying cry of a process that eventually lead to secession and to war. Incidentally, the affair was partially linked to the linguistic rights, since one of the issues was the right of the accused at a mounted trial at a military court to defend themselves in their mother tongue, which happened to be the official language of the Republic where the trial was being held. These rights we were defending, not knowing that they were already socially conceived as collective rights, and functioning as part of a parcel of collective national rights that were to become the most controversial issue of this end of the century. -So linguistic matters may well precede and, in appropriate conditions, even trigger quite unexpected social changes.

On the other hand, there does not seem to be any necessity that the language changes if and when the "society" changes. It is one of the definitional properties of the human language that it can be put to new uses, geared to new situations - even more, that it is itself capable of creating new uses and new situations. There is usually something ridiculous in the attempts of nouveaux regimes to emphasise their novelty by re-baptisation campaigns; there is also something scary about these preoccupations: one cannot resist the feeling that names, words etc. are being changed in order that other things do not change.

2. Some prima facie observations

2.1. Selective corpus permeability

It is remarkable how easily the language has shed off most of the lexical baggage of the previous political jargon. It is not that this was an "unnatural" or "unorganic" part of the corpus (although it is true that barbarism was one of its favoured figures; but other barbarisms have come, so that formally this does not seem to be characteristic of a particular regime): the survival of certain para-systemic formal procedures (we will consider one such procedure in the sequel) warns us against unwarranted simplifications, to say nothing about the "survival"¹ of argumentation procedures. Our hypothesis would rather be that the reproduction of the previous social relations (of domination and repression) depended to a large extent upon the readiness of speakers to use these terms in a non-problematic quasi-spontaneous way. These relations having disappeared en bloc², there neither is any need to use the terms that were their vehicle, nor are these terms neutral enough to be reactivated in some revised usage. An example may be the marked preference to designate with the native term stranka the political formation that can synonymously be referred to by the long-ago naturalised partija; the later term has been too contaminated by its long use to designate the only party.³

Correlatively, relations of the liberal economy being imported in a ready-made fashion (or so, at least, most people perceive their introduction), the corresponding vocabulary is permeating the lexicon in a similarly bound-trade manner (the term together with the referent). Let me only quote a commercial operation called faktoring (noun, m, declinabile) (daily Delo, May 25, 1992) and a financial operation, based upon an institution designated by the semi-domesticated term fransiza, and simply fransizing (noun, m, declinabile; a senior member of what may become a paradigm still conserves the trace of its origin: leasing) (Delo, July 9, 1993).

Within the segments where a domastic tradition exists, old terms are simply being revived, and there is even some limited creativity, like tecajnica, the stock-exchange list as opposed to the currency-exchange list, tecajna lista.

On one hand, the corpus is revivifying some of its hitherto silent layers, where, together with the neighbouring active layers, shows some marginal innovativeness; on the other hand, and more importantly, the corpus shows a remarkable permeability to the foreign (English) lore in the domains of the newly imposed social relations and their ideological glamour. This propensity for borrowing spreads beyond the specialist areas where the vocabulary is really lacking; thus, under the title "What is modern glamour?" (a striking shift after years of post-modernist frenzy when "modern" was not fashionable at all), the term glamour is being domesticated by being declined as a native word, although consistently spelled as in English, up to the point to form a home adjective, immediately put into superlative: najglamuroznejsi /the most glamouros/ (Pepita, a fashion and life-style monthly, July 1992).

Similar snobbery may be prompting the preference to use parlament instead of skupscina. Here analogous ideological reflexes as in the case of partija/stranka led to the opposite result-to the preference for the foreign word. Additionally, the term parlament seems to be perceived as having particularly "democratic" connotations, and is used abusively even where no such institution exists (as in Slovenia), or where the original home term has sometimes been used (sobranje, sabor, skupptina, sejm).

This localised over-permeability within the domains of certain functional codes, and the tendency for borrowing in contexts of established life-style, attitudes preferred by the dominating ideology, etc., contrast with the linguistic attitude of the previous establishment and its ideology which was rather purist, while the lexical borrowing was mostly limited to subcultures, which sometimes engaged in a conscious effort at vulgarisation-vernacularisation of more important terms (pank for "punk", rok, roker for "rock", "rocker"). As linguistic phenomena, they are also in contradiction with a kind of generalised (non-localised) impermeability-filtration on the level of ideological systems.⁴

2.2. Filtration

These examples will illustrate the point:

A.

(1) Alf: "I am feeling a productive member of society again".

Translation on Slovene TV: "Spet se pocutim ustvarjalnega."

/...creative.../ (1990)

The petrified syntagm, seemingly too reminiscent of the formulas of (Communist?) solidarity was replaced by a piece taken from another ideological inventory, that of romanticism. Since part of Alf's humour is to play with native terrestrial stereotypes, the comic effect has been sacrificed for what seems to be the new ideological "purity" of liberal individualism. This may be a case of censorship in the Freudian sense, for we can assume that the translator did not willingly destroy Alf's witticism. If the borrowing of the foreign (English) lexical items is too large even for a tolerant vie, than the ideological "filtration" is too narrow even under the assumption that reflexive anti-communism may wipe out also some left-sounding phraseology (in (1), it was a "solidarity" expression that was erased, with no particular left/right ideological inclination).^4a

(2) Paul Veyne (in: L'elegie erotique romaine): "..l'auto-ironie (chez les gauchistes et les taoistes).".

Translation into Slovene: "..samoironija (pri skrajnih levicarjih in taoistih)..."

/...with extreme leftists.../ (1992)

This transposition seems like a systematic deformation generated by some kind of geometrical projection: since the space into which the French historian's expression is being transferred has shifted to the right, his "leftists" become "extreme leftists". Correlatively, right-wing parties call themselves "centre-right", and are accordingly referred to by the media. The adjective is desno-sredinski, formed by the now popular -o+A construction, which, in this case, is disconcerting for the fact that the NP from which this AP is derived, is not desna sredina, but desni center. This may be a kind off hyper-correctness, otherwise typical of the traditional political right (so that the extreme right is "centre-right" at least in its linguistic practise).

In these examples, a recognisable "filter" operates by substitution or shifting; since the two operations can also be interprated as Freudian mechanisms of Verdichtung (1) and Verschiebung (2), we may be sure that the filter is ideological stricto sensu, i.e. it pertains to a particular ideology. The A-type of deformation is thus an intra-ideological shift.

But similar phenomena occur both on lower, infra-ideological, and on higher, supra-ideological levels.

B.

(3) Clinton: If we are asked /to military intervene in Bosnia/, we will give it a good consideration".

Slovene TV: "...se bomo odzvali."

/we will respond/ (August 1993)

This seems a case of "wishful translating": since the TV channel doing the translation favours military intervention, they put it into the President's mouth. The mis-translation is "infra"-ideological in the sense that it appeals to no particular ideology, although it is possible only within a certain ideological horizon.

C.

(4) The Setup (film): "You are going to picket, to sing ‘We shall overcome', and to support liberal causes again".

Slovene TV: "...bop podpiral liberalce."

/...support the Liberals/ (June 1993)

(5) CNN: "..the break down of order and law".

Croatian TV: "kriminal"

/the crime/ (November 1992)

These two examples operate on the level of what we would ad hoc dub the vocabulary, i.e. on the level of device which assigns to lexical items their "domains of distribution" that befall to them under the conditions imposed by the "valid", "accepted" background beliefs. Since the specific beliefs that form the background of the C-type deformations importantly restrict the lexical potential of the core-items, they may have, in the long run, a para-linguistic effect and impose an at least temporary limitation on the use of certain terms:

- there is no lexical constraint prohibiting the application of the term liberal to entities other than political parties; still (4) procedes as if it were self-evident that it can only apply to a political party;

- there is a semiotically hierarchical difference between the disappearance of the demarcation line distinguishing the right or at least the non-prohibited from the wrong - and the simple appearance of the wrong, of the "crime".

If, in (4), it is the restricted application-domain of a predicate that is assumed as "evident", in (5), it is the very demarcation line whose existence has been denied in the original, that is presupposed as "evidently given". In (4), it is the scope of a certain intellectual operation that is restricted; in (5), it is the intellectual horizon itself that is limited: a higher intellectual operation-questioning of an axiological axis - is tabooed.

We consider the C-type as supra-ideological, since it does not directly pertain to any ideology in particular (as does the A-type), but defines the general horizon within which different ideologies are alowed to compete; needless to say that some others are thereby excluded from "legitimate" competition.⁵

3. Discussion of an apparent contradiction

A superficial observation has discovered two opposite trends in recent linguistic practices:

- on one side (2.1), a remarkable permeability of the corpus in certain sectors held to be of crucial importance (liberal economy, finances...), and in sectors linked to fashionable, established or expected life-style, ways of behaviour, culture of the "every-day-life" etc.; this results in unnecessary borrowings that sometimes neglect morphological and morphogenetic considerations;

- on the other side (2.2), a vague impermeability of the local semiotic universe, an isolationist tendency that impedes communication with other semiotic universes.

Albeit contrary, the two trends are not exactly contradictory, since they are located on two different levels.

Ad 2.1. A historical parallel may shed some light on the problem of lexical borrowings. Radio and kino /cinema/, naturalised a long time ago, are masculina in Slovene, although, being of the same type as okno, oko etc., they should, gramatically, be neutra (as they are in Croatian, Serbian, Macedonian and Bulgarian); the same spontaneous grammatical gender ascription happened later with stereo, video. It took an institutional intervention to make finale /finals/ neutrum, as it is in Latin and as it should be in Slovene. It seems that in these cases, the ascription of grammatical gender has been influenced by the dominating cultural gender-semiotics which wants the masculine to be the "neutral" or the "zero" gender. If this is the case, then we have here an instance of intrusion of ideological system ("secondary system") into the linguistic system ("primary system").

Likewise, the facility of some recent borrowings may be ideologically motivated. This is all the more probable since ideological factors always intervene in corpus changes.⁶ Specific ideological pressures may bend corpus changes one way or the other - but the change as such cannot occur if it is not backed by some factor of "authority, credibility" etc. (cf. J.A.Fishman, "Language and nationalism - Two Intergrative Essays"). The necessary formal mechanism of transference. The transference of Eastern Societies to the "West", "Europe" or, simply, the prevailing respective mythologies, are apparently so strong as to warrant linguistically unwarranted borrowings. There may even be some undercurrent of a magic creed that the borrowed terms make their referents function the better, the closer they are to their "originals". This magic approach can also be detected in the preference of new enterprises to adopt English, or "English-sounding" names. This Zeitgeist has appropriately been caught (or has it been co-produced?) by the advertisement one can spot around in Eastern Europe: "Test the West!".

Ad 2.2. The opposite, isolationist trend operates through two types of mechanisms:

- Shifts along the horizontal axis: (1) is a shift from one (ideological) paradigm to another, but both are situated on the same level; (2) is a shift within the same paradigm; (4) restricts the paradigmatic potential of a lexic item.

- Substitutions on the vertical axis: (3) substitutes a statement on a meta-discursive level with the object of the stated meta-discursive operation; (5) substitutes the level where a distinction is drawn with the result of such a distinction on the lower (the object-discursive) level.

Horizontal shifts impose a certain international organisation upon the semiotic universe. Vertical substitutions assign the horizon, the external border to the semiotic universe.

Vertical operations draw the line beyond which it is, so to say, impossible to think. Horizontal operations determine the premisses from which it is possible to think.

2.1. and 2.2. Together. The failure of the inter-cultural communication is thus a symptom of intra-cultural pressures to impose a certain reshaping of the over-all "internal" semiotic organisation. Correlatively, intra-cultural corpus changes indicate a certain selective inter-cultural orientation.

Both dimensions are thus intimately interlaced. We may say that the major ideological conflict is being fought over the issue which ideology is going to over-determine the social consensus about the necessary "integration" into the world system.⁷ Since we were taking our examples from the national hegemonic media, it seems that presently dominant ideology is the one that:

-Promotes an accelerated "integration" whose socio-linguistic symptom is a regional colonisation of the language corpus.

-Tries to block the channels of other possible ways of integration, with the effect of important impoverishment of the local "semiotic universe".

A scarry symptom of this cultural isolation, this time blocking the access to the society's own past, could be heard on Slovene Television when the speaker, reporting on French debates about the involvement of the Vichy regime in antisemitic razzias, repeatedly pronounced Vichy as /v-iki/, accent on the first syllable.

4. Sociological Comment

Two general sociological features of this particular ideology deserve special attention: this ideology is inwardly conflictual and illusionary. It thus fails to perform the two basic functions of a dominant ideology: it is unable to secure internal integration of the society and likewise unable to define its external delimination.

By a set of ideological mechanisms that cannot be analysed in this paper⁸ (and which include ethno-chauvinism, racism, conspiracy-speeking campaigns etc.), this ideology compensates for its inherent insufficiency, and produces important discourses of hatred. In the final section of this paper, we will try to analyse some linguistically pertinent features of this type of speech.

5. Intrusion of rhetorical turns into the language-system

We will now examine a case where the language-use (utterance-level, "parole" ) produces language-systemic effects.⁹

One of the most succesful inovations of this type was introduced by the Croatian Television:

srbo-cetnicke formacije,
jugo-srbo-komunisticka armija, etc.

The -0+A construction is a kind of quantitative or qualitative specification of the adjective, and is semantically ambiguous: it can be a specification proper, i.e. a limitative determination, defining a sub-set of the extension-set of the adjective; or else it can be a conjunctive determination, defining a union of the extension-sets of the two adjectives, one of which is adverbalised into the -o form (this type may be a borrowed construction, a sort of morphological calque). It seems that the linguistically-systemic (language-specific) construction is always semantically limitative, while the frequent use of the borrowed conjunctive construction in the political jargon, journalistic discourse etc. has achieved that it is no more felt as foreign or imported.

The construction therefore has a particularly vague semantic effect, and is producing a sort of "semantic haze" that offers itself to ideological manipulation. The construction was popular in the old regime jargon, and particularly the expression drustveno-politicki or, in Slovene, druzbeno-politicen /socio-political/ had a high frequency and a wide distribution, although not as clear a meaning. We think that this construction functions in much the same manner in Croatian, Serbian and Slovene, so that most of our examples will be from Slovene use, and some from the Croatian. Although the construction was used even in the Constitution, its semantics remained rather undecidable (e.g., we could ask ourselves what would be the paradigmatic opposition to "societal politics"; or else, what would be the union of the sets defined by the predicates "the political" and "the social"). The construction was frequently used to label phantasmatic "internal enemies": the anarho-liberals (a recent re-appearance: in a letter to the Editor, Delo, August 28, 1993: anarholiberali in udbomafijci /anarcho-liberals and secret-state-political-police Mafiosil, techno-managerial elites, techno-liberal positions; the composed adjective was frequently substantivated. In the 1930ies, when the Communists had stricter linguistic notions, the positions of those who presumably professed "national nihilism" and whom they called "national nihilists", were labeled nacionalno-nihilisticne pozicije.

Let us try to analyse this time-honoured construction with special affinity to political jargons, which has never really made it into more cultivated kinds of speech, but which seems to be cherished by the central media who had been promoting its usage already in the past, and have now only lexicalised it according to the new circumstances. (So we read in a letter to the Editor in Delo, August 28, 1993: mladostarokomunisticni ideologi /young-old-Communist-ideologues/.)

(For this occasion, we will not specially analyse the "contradiction" srbo from srpsko and jugo from jugoslovensko.The contradiction has a rhetorical effect that presently functions as degrading connotation.)

1. AP-> QA : temno moder /deep blue/

We leave to the grammarians the problem that the specifying category Q is here a "qualifier", rather than a "quantifier". Let us just note that the construction SpecA is the same as with a standard quantifier in zelo lep /very beautiful/ or as with the ambiguous quantifier/qualifier hudo neumen /mighty foolish/ or as with a qualifier nebesno moder /sky-blue/, lisnato zelen /leaf-green/. The Spec component Q answers the question "How blue?". Semantically, this is a limitative determination.

This kind of AP quite trivially enters into the NP constructions of the type:

/  / QA/ N/

np ap ap np

2. Morphological transformation:

/ AN/ -> / QA/

np np    ap ap

With adjectivation of noun phrases of the AN type, the A becomes a specifier of the Q sort (an adverb):

Srednja Evropa -> srednjeevropski /Central Europe -> Centroeuropean/

zunanja politika -> zunanjepoliticen /foreign policy/

Semantically, this is a limitative determination; it enters into constructions of the same type as (1)

/  / QA/ N/

np ap ap np

3. Morphological transformation

/ A / A N / / -> /  / Q A/ N /

np  np    np np  np ap   ap  np

mali obmejni promet -> maloobmejni promet / litt. "smally border communication"/

This is a limitative determination, constructed in manner that old rhetoricans may have called hypallage (cf.DuMarsais) or enallage (Fontanier): A is originally specifying the A N group, and is then transformed into Q as if it were specifying A. On the surface, the transformation yields a construction of the type (1). (The construction itself is rather complicated, and we skip a more detailed discussion.)

Without going into more detailed discussion, let us note that -o+A construction is semantically open and grammatically versatile. It lends itself to figurative turns that easily become syntagmes figes or even starts to behave systematically on the basis of the figurative turns that later get "forgotten".

E.g., what had first been a hypallagic syntagme fige, gave rise to a whole paradigm: maloobmejna prepustnica, maloobmejni promet, maloobmejno povezovanje, etc. /smally border pass, smally border crossing, smally border intergration/.

This seems a case of intrusion of a rhetorical procedure into the linguistic system. Since the background knowledge firmly anchors the uvnderstanding that might otherwize be derouted by the systemic faultiness of the construction, what is, in principle, a semantic "haze", has here been productively used for linguistic innovation.

4. The conjunctive construction may be a calque from the type "le conflit franco-britannique", "the Anglo-American juridic system", "judeo-christian world" etc. The simplest form of transformation would then be:

/  / A Conj. A /  N/ -> /  / Q A / N/

np ap          ap np    np ap    ap np

Since the same surface structure may be generated from different deep structures, or from no deep structure at all, in which case it is a "petrified", i.e. linguistically accepted, albeit "faulty" or at least "problematic" rhetorical turn, the effect of a neologism moulded in this type of construction is a manipulable semantic indecidability. Specifically, it is at the first sight difficult to see whether the construction entails a limitative or a conjunctive determination. The construction is therefore prone to produce all kinds of ideological, emotional etc. effects.

A most important feature is that this type of linguistic innovation operates within the area of language-specific systemic features: which means it plays with the "mother-tongue" competence, and elicits spontaneous "evidences" that, although ideological by nature, seem to procede from the innocent functioning of the language system.

(1) srbo-cetnicka vojska / the Serbo-cetnik Army/

Surface type:

/  / QA/ N/

np ap  ap np

Since cetnici are an extremist nationalist Serbian movement, all cetnici are analytically Serbian. The initial determination in (1) is thus redundant and the whole expression pleonastic. The function (cf. Jakobson 1958) of (1) is therefore not referential, it does not define an object of the world, but some other one: our hypothesis is that expression (1) and its kind are specifically designed to structure the speech situation, i.e. the relationship between the addresser and the addressee; (1) is performing on the level of Jakobson"s emotive and conative "functions of language".

(1) cannot entail a semantically limitative determination, since the cetnici are a (tiny) sub-set of the set Srbi ; neither can (1) vehiculate a conjunctive determination, for it does not concern two otherwise independent and hierarchically equivalent sets. The optimal interpretational strategy, propelled by the general principles of communicational economy (Grice; Mocnik), would then be to take (1) as its face value: the limitative and the conjunctive options being blocked by the background "knowledge" (or better: by the background beliefs), this means that a spontaneous interpretation would read (1) as what "it is", as a pleonasm, that is, as an equation.

With its grammatical-rhetorical mechanism, (1) thus performs a fantastic ideological operation; from the premise:

"All cetnici are Serbs",

validated by the background knowledge or maybe even already featuring in the lexicon as the "meaning" of cetnik, it deduces the suggested conclusion

"All Serbs are cetnici".

"validated" by the effect of pleonasm, equivalence, synonymy.

By the same token, and on another level, (1) helps to transfer the connotational charge and the emotional cathexes burdening the expression cetnici in different sorts of discourses within the "discursive universe", onto the adverbial praefix serbo-, and further on to its radical.

2) jugobeograjski prijemi

/ Yugo-Belgrade methods/

(Letter to the Editor, Delo, July 15, 1993)

Surface type:

/  / QA/ N/

np ap  ap np

The -o component can neither be a limitative determination (or a specification) of the core-A, since A is already a specification of the -o component's root adjective; nor can it, for the same raeson, be a conjunctive determination. "Grammatically", (2) is a pleonasm of the kind philosophers use to ridiculise certain logical errors: "fruito-plum jam".

Rhetorically, (2) can than operate as an emphasis; traditionally (Fontanier, 1827), it might have been classified as the figure of metabole or synonymy. But the two terms involved are synonymous for contrary reasons - the first (jugo-) by the token of "totum pro parte" (meaning: "centralism"), the other (beograjske) by the token of "pars pro toto" (meaning: "(Serbian) hegemony"). By this double synechdochal device, the two components are contrived to "coincide" in that part of their meaning which suggests something bad and evil.

It is so a semantically flawed usage of an otherwise regular construction that produces the semantic effect synonymy, neutralising other possible meanings of both components.

An even more striking case of discursive overdetermination, i.e. speech overdetermination or the language-systemic transformations is another Croatian TV expression:

(3) jugokomunisticke snage

/Yugo-Communist forces/

(Same surface type)

The AP is an adjectivised NP, derived by the canonic transformation we discussed under "2". If (3) were used in an "internationalist" discursive universe, it would be a regular limitative determination, specifying the core-A (the adjectivated noun) with its national characterisation in the -o form. When used in the context of the "Serbo-Croatian" war, this construction is "semantically" problematic, and produces a similar effect as (2): it contaminates the adverbalised A (the root-A of the -o form) with the adjectivated N, and vice versa: again, the overall effect is a synonymy "in the evil".

In our preliminary discussion of the -o+A construction, we noted that, as a surface construction, it either is ambiguous (for it can be generated along the type 2 (limitative determination) or along the type 4 (conjunctive determination, derivation) or is hypallagic, i.e. it violates a linguistic rule by some creative rhetorical procedure. In our case, "hypallagic" means that the construction violates what generativists (after Chomsky"s Current Issues, 1964) call A-over-A Principle :

a construction of the type

  NP
 /  \
A   NP
   /  \
  A    N

is treated as the construction

    NP
   /  \
  AP   N
 / | \  \
A Con A  N

so as to yield

   NP
  /  \
 AP   N
 |    |
QA    N

The hypallagic construction produces a phantomatic AP, whose surface realisation is -o+A. It is this phantomatic nature of this type of QA, coupled with the fact that even a linguistically correct -o+A is ambiguous, that requires further semantic information which is taken contextually infused, or, after a while, pumped from the background beliefs, into the linguistically "deficient" construction.

Or, if viewed from another perspective: a phantomatic linguistic structure, produced on the locale of a surface-ambiguous construction, creates a systemic vacuum that attracts supplementary contextual or background information which is then conveniently provided.

The semantic equivalence-creating process is further enhanced by a purely formal device, a sort of "secondary" or dummy-constraint that permits only those As that figure in an -o+A construction to assume the -o form in such a construction.

Because of the restricted vocabulary of this kind of discourse, the "secondariness" of the constraint can hardly be detected or its rhetorical nature perceived; its "uptake" is quasi-natural. On the other hand, within the lexicon of the language as a whole, the constraint creates a lexical island, a limited inventory of terms relating to the enemy and behaving in a specific way. The so-called "image of the enemy" is, in this way, formally constructed, and the appropriate semantic contents can, secondarily, be safely deposited upon this infra-structure. We think that, generally, it is this kind of formal procedures that can ultimately guarantee a successful promotion of the more obvious "semantic" imaginery.

There is another important effect to this formal isolation of the enemy. We have seen that the corpus expansion by direct borrowing of foreign (English) lexical items betrys a self-deprecating linguistic ideology, and, by forcing upon the language what would "normally" be barbarisms, somehow "de-universalises" the language. By reserving a special treatment to a cluster of terms referring to or qualifying the enemy, the enemy is parochialised, and the aim of any war propaganda - the ideological superiority over the adversary - is achieved. What cannot be achieved by demonstrating one's own "universality", is achieved by degrading, parochialising, "exoticising" the enemy.¹⁰

That a figure of the enemy comes to fill in the lack produced in the "imagined community" by its own self-deprecating ideology of (linguistic, economic, social... general?) inferiority, is certainly a topic worth nothing, although it cannot be further elaborated here.

The end-stop of this morpho-genetic chain may be contracted NP construction, tolerated in commercial names, but otherwise ostracised by the purists; its historical precedents are from the hard Stalinist era - histomat for "historical materialism", diamat for dialectical materialism; it appeared on the front page of Delo, the daily with the widest circulation in Slovenia:

(4) Jugomina ubila teritorialca

/Yugo-mine killed teritorial defence soldier/

(August 2, 1993)

This is a splendid condensation of equivalence-producing procedures of pleonasm (as in (1)) and synonymy (as in (2) and (3)). Since the rhetorical mechanism is occulted by the quasi-linguistic construction, this type of trick is much more efficient than more overt, linguistically more correct and cultivated rhetorical procedures:

- of pleonasm: "conditions south of southern Slovenia border" (President of the Republic of Slovenia Kucan at the talks with the Austrian President Klestil, according to Delo , July 7, 1993);

- of suggested synonymy: "the bodies of 38 Croats, mostly civilians, massacrated by Izetbegovic's criminals" (Croatian TV report from Doljane, B&H, August 3, 1993).

Webmaster: Slobodan Markovic (twiddle@eunet.yu)

All rights reserved. Belgrade Circle Journal encourages the reproduction of material appearing on its pages, provided that the source and the author are cited, except in cases where this would constitute violation of copyright held by other organisations or individuals.