UDK 808 George Thomas
McMastrova univerza, Hamilton
THE ROLE OF DIGLOSSIA IN THE DEVELOPMENT OF THE SLAVONIC LITERARY LANGUAGES
Obravnavane so vrste diglosij po slovanskem jezikovnem prostoru - z različnim vplivom na posamezne knjižne jezike - in nakazani možni izidi teh diglosnih stanj oz. potekov.
The rise of Slavonic diglossia is examined, its differential impact across the various literary languages is reassessed, and the possible outcomes of such situations are suggested.
It is no exaggeration to claim that Charles Ferguson's well-known article on diglossia1 has brought about a major reorientation in the study of literary languages. Above all, it has focussed attention on the way in which the literary language shares the socio-communicative functions with other codes of the same diasystem.2 We can no longer be totally satisfied with the hierarchical model which sets the polyvalent literary language above a series of functionally limited dialects, interdi-alects and sociolects. In many speech communities, especially in urban settings, the relationship between the various standard and nonstandard codes is vastly more complicated. The choice of code is determined by such factors as the status of the speakers, the degree of formality in their relationship, the subject-matter, the form of communication (written or oral), and the social setting. In diglossia, there are two codes: 1) Low (L), a natively learned dialect or interdialect, 2) High (H), a strictly codified variety which is learned through formal instruction. These two varieties are, from a socio-communicative point of view, in more or less complementary distribution.
The study of literary languages has a long and honourable tradition in Slavonic scholarship.1 Moreover, it received an enormous boost from the theoretical contributions of the Prague School, whose positions on the subject of literary languages have found support throughout the Slavonic world. Nevertheless, there have so far been few explicit attempts to incorporate the diglossic model into the study of Slavonic literary languages.4
1	Diglossia, Word XV (1959), pp. 325-340.
2	The term diasystem is used in the sense employed by Dalibor Brozovič, as for example in Standardni jezik (Zagreb, 1970).
1 Particularly valuable as indicators of the richness of this scholarship are these four collections: MiloS Weingart (ed.), Slovanské spisovné jazyky v dobê pfitomné, Prague, 1937; Alexander M. Schenker and Edward Stankiewicz (ed.). The Slavic Literary Languages: Formation and Development, New Haven, 1980; Riccardo Picchio and Harvey Goldblatt (ed.). Aspects of the Slavic Language Question, 2 vols., New Haven, 1984; Gerald Stone and Dean Worth (ed.), The Formation of the Slavonic Literary Languages, Columbus, Ohio, 1983. This is, of course, an area in which the late Jakog Rigler was highly active. In this sphere, he will be best remembered, perhaps, for his studies on Trubar's groundwork for establishing a norm of written Slovene (cf. particularly his Začetki slovenskega knjižnega jezika, Ljubljana, 1968, as well as his collection of essays published under the title of Razprave o slovenskem jeziku, Ljubljana, 1986.
4 A measure of the contributions to the study of Slavonic diglossia can be gained from Peter Brang and Monika Ziillig (ed.), Kommentierte Bibliographie zur slavischen Soziolinguistik, 3,
The present paper sets out to examine the rise of Slavonic diglossia, to reassess its differential impact across the various literary languages and suggest the possible outcomes of such situations. Firstly, however, we need to review briefly the refinements and distortions of Ferguson's initial formulation.
The H and L varieties in Ferguson's schema have specialised functions which are controlled by mechanisms of social approval and disapproval. This narrow view of diglossia has been extended by Fishman and others5 to include any vertical switching to codes determined by situational factors. For example, Soviet theoreticians like Krysin and Švejcer discuss diglossia in terms of stylistic variation within sub-codes of the Russian literary language,6 while Tollefson uses the therm to describe switcHing from Slovene to Serbo-Croatian or Slovene to Italian.7 However valuable such analogies with diglossia may be for an understanding of code-switching, they have the disadvantage of obscuring or trivialising the original meaning of the term. Moreover, most scholars have glossed over an important element of Ferguson's definition: the non-native and super-posed character of the H variety. More serious still, some scholars have even ignored the vertical parameter, thus using diglossia to describe situations where there is polycentric codification of a literary language.8
One of the more successful of the recent attempts to rework Ferguson's definition in the light of recent publications,9 is that of Francis Britto,10 who recognises three types of diglossic situation : 1) pseudo-diglossia, where there is no super-position of codes, i.e. group A uses Code 1, group B uses Code 2. 2) user-oriented (or partial) diglossia, where there is partial (i.e.
Bern/Frankfurt am Main, 1981. Of the 28 items listed under the rubric of diglossia, 7 are treatments of non-Slavonic diglossic situations, 5 are theoretical, one deals with the impact of bilingualism on language planning in Slovenia, 7 deal with instances of diglossia in Serbo-Croatian, and 8 treat the Russian-Church Slavonic symbiosis, and one is comparative.
5	Joshua A. Fishman, Bilingualism with and without Diglossia: Diglossia with and without Bilingualism (Journal of Social Issues, XXIII (1967), No. 2, 29-38; Jack Fellman, On diglossia, Language Sciences, XXXIV, Feb., 38-39; Heinz Kloss, Bilingualism and Nationalism, Journal of Social Issues, XXIII (1967), No. 2, 39-47; ibid., Ueber Diglossie, Deutsche Sprache, Jahrgang 1976, 313-323.
6	L. P. Krysin, Vladenie raźnymi podsistemami jazyka kak javlenie diglossii in: L. P. Krysin and D. N. Šmelev (ed.), SociaPno-lingvističeskie issledovanija, Moscow, 1976, 62-69; ibid., K voprosu o vnutrijazykovoj diglossii in: A. N. Baskakov (ed.), Metody bilingvističeskix issledovanij, Moscow, 1976, 61-7; A. D. Švejcer, Bilingvizm i diglossija in: Sovremennaja sociolingvistika. Teorija, problemv, metody, Moscow, 1977, 114-31.
7	James W. Tollefson, Diglossia and Language Policy with Special Reference to Slovenia (Ph. D. Dissertation, Stanford University, California, 1978), 346 pp. It is perhaps instructive that when the dissertation appeared in book form as The Language Situation and Language Policy in Slovenia, Washington, 1981; all reference in the title to diglossia was removed. Indeed, apart from a theoretical discussion of diglossia, little attention is paid to diglossia in the sense originally used by Ferguson - all the more surprising when one considers the fact that the latter was one of Tollefson's advisors.
8	For example, Valter Tauli, The Theory of Language Planning in: Joshua A. Fishman (ed.), Advances in Language Planning, The Hague, 1974, 49-68 characterises the relationship of the Croatian and Serbian standards as diglossic.
9	For a critical summary of this literature, see Ralph Fasold.The Sociolinguistics of Society, Oxford, 1984.
10	Diglossia: A Study of the Theory with Application to Tamil, Washington, D. C., 1986.
acquisitional and functional) super-position, i.e. there are some native speakers of the H variety, who might (at least potentially) enjoy social advantages as a result. 3) use-oriented (or total) d i gloss i a ,n where there is total super-position, i. e. all members of the speech community use each code according to a generally accepted pattern of social functions. - Against this he identifies three positions along a scale of variation: 1) optimal, pertaining to varieties where the codes are optimally distant for Ferguson's narrow definition of diglossia, i.e. diglossia sensu stricto. 2) sub-optimal, pertaining to varieties which are insufficiently differentiated for Ferguson's definition, i.e. stylistic variation within a single code. 3) super-optimal, pertaining to varieties which are over-differentiated for Ferguson's definition, i.e. bilingualism. - Two further features of the H language need to be distinguished: 1) It may be "dead" like classical Arabic, Latin or Greek or "living" like German or French.12 2) It may originate within or without the same diasystem to which the L variety belongs. The first type (e.g. German and Schwyzerdütsch) we shall call endoglossic; the second type (e.g. the adoption of Danish by Norwegian-speakers), exoglossic.13 Another parameter of diglossia is its geographical extent: it may be total, i.e. with identical territorial distribution of L and H varieties as in the Greek situation, locally differentiated, i.e. the L or H codes show variation throughout the language as in Arabic, or localised, i.e. limited to one or more identifiable areas of the diasystem as in German.
The history of Slavic diglossia begins with the introduction of Old Church Slavonic as a super-posed variety in the Orthodox world. Except in Bulgaria, where the canonical codification of OCS took place, the diglossia involving Church Slavonic may be described as exoglossic, locally differentiated (i.e. accomodated to some features of the local L variety) and super-optimal. In the East Slavonic area there also came to exist written forms of the L variety leading to a form of literary diglossia.14 On Russian territory, despite periods of intensified differentiation of the two literary codes (the so-called Second and Third South Slavonic Influences), this literary diglossia came to resemble an optimal, endoglossic situation. In the course of the eighteenth century this gave way to user-oriented, sub-optimal diglossia, i.e. stylistic variation within a single code acquired by the elite as its first language.15 This situation - with a broadening of the elite and a partial or passive
11	The source of this terminology is the work of the British pragmatists H a 11 i d ay and S t re ve n s.
12	This important distinction is made, for example, by H. and R . K a h a n e , Decline and Survival of Western Prestige Languages, Language, LV (1979), 183-198.
13	This term is taken from J. Cobarrubias, Ethical Issues in Status Planning in: J. Cobarrubias and Joshua A. Fishman (eds.), Progress in Language Planning, Berlin/New York/Amsterdam, 1983, 41-85.
14	The precise nature of this diglossia is controversial: Gerta Xjutl'-Fol'ter, Diglossija v Drevnej Rusi, Wiener Slavistisches Jahrbuch, XXIV (1978), 108-123; Ladislav Matejka, Diglossia in the Oldest Preserved Legal Codex of Novgorod (Papers in Slavic Philology, I (1977), 186-197, Dean S. Worth, On 'diglossia' in Medieval Russia, Die Welt der Slaven, XXIII (1978), No. 2, 371-393.
15	A. V. I sače n ko, Kak sformirovalsja russkij literaturnyj jazyk?, (Wiener Slavistisches Jahrbuch, XXIV (1978), 124-136; ibid., Mythen und Tatsachen über die Entstehung der russischen Literatursprache, Vienna, 1975; J. Lothe, Russe et slavon: la controverse de 1822, Annuaire de
acquisition of the H variety by the majority of the population - has continued until the present day.
Elsewhere in East Slavonic territory, the L variety receded in written use in the seventeenth and eighteenth centuries. When the Belorussian and Ukrainian vernaculars were revived in the nineteenth century, the earlier literary diglossia was repudiated, and the L variety, now virtually free from admixtures from H, became the new H variety. Thus was born a typical "standard with dialects" situation
-	complicated in the Ukrainian case by the existence of two centres of literary codification (the east and Galicia).16 Until the establishment of a single codification of literary Ukrainian, therefore, it is legitimate to speak of the existence of pseudo-diglossia in the Ukraine. In Subcarpathian Rus', however, the introduction of literary Ukrainian as the H variety to an area where the L variety was highly divergent was bound to create a situation of optimal, use-oriented diglossia. The separate codification of Ukrainian promoted in some quarters,17 whatever its nationalist or regionalist motivation, would have had the force of replacing this form of diglossia by literary pseudo-diglossia.
In Bulgaria and Serbia, literary diglossia did not develop. In the course of the eighteenth century, however, Serbian Church Slavonic was subjected to influence from both Russian Church Slavonic and the Serbian vernacular. The highly chaotic and unstable literary situation in Serbian gave way under the pressure of the Vukovian reforms to the rise of a vernacular-based literary language.18 In present-day Serbia, diglossia is characteristic only of the speech of Niš with its Torlak dialectal base.19
The Catholic Slavs, with some important exceptions, had Latin as their cult language, while other exoglossic codes - Middle High German, Middle Low German or Italian - carried out many other important social functions. Inevitably, as some of the Slav peoples - notably the Czechs, Poles and Dalmatian Croats
-	began to use their own vernaculars for written purposes, there was a struggle for ascendancy in carrying out the various socio-communicative functions: the higher social functions tended to be filled by exoglossic languages, the lower ones by Slavonic vernaculars. There is no question, however, that the domains of use of these Slavonic vernaculars were steadily expanding. Moreover, with the Reformation, the Sorbs and the Slovenes also began writing in their own vernaculars. In
l'institut de philologie et d'histoire orientales et slaves. Université libre de Bruxelles, XIX (1968), 709-728; L. L. Kutina, Poslednji period slavjano-russkogo dvujazyčija v Rossii in : Slavjanskoe jazykoznanie. VIII Meidunarodnyj s'ezd slavistov Zagreb-Ljubljana 1978 g., Moscow, 1978, 241-64.
16	See G. Y. Shevelov, Die ukrainische Schriftsprache 1798-1965, Wiesbaden, 1966; for Galicia, see P. R. Magocsi, The Language Question in Nineteenth-Century Galicia in: Picchio/ Goldblatt, op. cit., 49-64.
17	See P. R. Magocsi, The Shaping of a National Identity. Subcarpathian Rus', 1848-1948, Cambridge, Mass., 1978, 130-144); ibid., The Language Question Among the Subcarpathian Rusyns in: Picchio/Goldblatt, op. cit.. Vol. 1, 65-86.
18	The term schizoglossia, coined by Einar Haugen, Language Conflict and Language Planning: The Case of Modern Norwegian, Cambridge, Mass., 1966, p. 280, seems appropriate to describe the confused and unstable periods of literary usage in both Serbia und Subcarpathian Rus'.
lv See Thomas F. M agn e r, A Century of the Niš Dialect, Papers in Slavic Philology, V ( 1984), 133-145.
some cases - notably in Croatian and Slovene - these written vernaculars displayed considerable regional variation, while in others, like Czech and Polish, a single, centralised, standardised language began to crystalise around the usage of major urban centres. In neither case, however, did any clear-cut stratification of communicative functions develop. As a result of a complex interplay of factors affecting the sociolinguistic situation of central Europe, the seventeenth century saw a sharp decline in the written use of all the Slavonic vernaculars except in Poland and, to some extent, in Dalmatia. Consequently, by the second half of the eighteenth century, the codified Slavonic vernaculars (excepting Polish) were never fully acquired by more than a handful of the population.
The revival of the Slavonic written languages from the end of the eighteenth century saw as its goal the transference to them of all socio-communicative functions (i.e. the end of super-optimal exoglossic diglossia). Moreover, the solution reached by the elite in each community to the problem of selecting a prestigious literary norm was bound to have important consequences for diglossia. In Czech, for example, the adoption of the old sixteenth-century norm meant that there were no longer any native speakers of the standard language. As a result, there developed a use-oriented, optimal diglossia involving the literary language as the H variety and obecna češtinaas the L variety. This situation has persisted until the present day except that intermediary forms incorporating elements from both H and L varieties now share the communicative functions.20
Czech had also served as the basis for the written language in Slovakia. This situation persisting well into the nineteenth century may be described as super-optimal, user-oriented diglossia. With the rise of a standard language based on the central Slovak dialects this diglossia was replaced by a "standard with dialects" situation.
In Croatia, only one prominent member of the Illyrian Movement - Vjekoslav Babukić - was a native speaker of štokavian, the dialect on which the new literary language was based. From 1836 onwards, therefore, the language situation in Zagreb and other kajkavian cities was characterised by use-oriented, optimal diglossia.21 For similar reasons, diglossia has also prevailed in the major čakavian urban centres (Split, Šibenik, Zadar).22 A feature of both these diglossias is the
2(1 The exact nature of these intermediary forms is a matter of dispute. In the opinion expressed by the American Slavist Henry KucerainhisA Phonology of Czech, The Hague, 1%1, is that there is a single, stable but variable form hovorovâ čeština. The position adopted by most linguists in Prague, however, is that hovorovâ čeština is the colloquial variant of the literary language (akin to the spoken form of codified Russian), while a widely used spoken non-codified idiom known as bčžnč mluvenâ čeština (somewhat akin to the German Umgangssprache and Russian razgovor-naja reč') now fills many of the L functions. For the relations between these various forms of Czech usage, see for example Jaromir Bčlič, Sedm kapitol o češtinč (Prague, 1955); ibid., Vznik hovorové češtiny a jeji pomër k češtinč spisovné in: Československe prednśśky pro IV Mezinârodnt sjezd slavistu v Moskvč, Prague, 1958, 59-71; Jaromir Bčlič, Bohuslav Havrânek et al., K otâzce obecné češtiny a jejiho pomčru k češtinč spisovné. Slovo a Slovesnost, XXII (1961), 98-107.
21	Thomas F. Magner, A Zagreb Kajkavian Dialect, University Park, Pennsylvania, 1966.
22	Thomas F. M ag n e r, The Dialect of Split in: Xenia Slavica (The Hague, 1975), pp. 125-32; ibid., Zapažanja o sadašnjem splitskom govoru, Čakavska Rič, VI ( 1976), No. 2, 83-92; ibid., Diglossia in Split, Folia Slavica I (1978), 400-436.
relaxation of rigid distinctions between the two codes: as well as "textbook" Croatian (the H variety) there is a commonly used formal Zagreb and Split usage which incorporates some features of the local dialect; in addition to the "pure" dialect form, there is a spoken variety which has lost some of the features which would otherwise distinguish the H variety from the L.23
In addition to this localised, use-oriented optimal Croatian diglossia, mention should be made of the relationship between the various regional and national codifications of standard neoštokavian.24 Within Britto's framework this polycen-tric codification may be described as a complicated instance of pseudo-diglossia. Bearing in mind the mutual interference between the Serbian and Croatian poles of this polycentric codification and the existence of usage in Bosnia-Hercegovina which incorporates in free variation elements from each,25 this pseudo-diglossia may also be characterised as "leaky" or "overlapping".
The rejection by the majority of the Slovene elite of the Illyrian solution to their communicative needs rescued Slovenia from a potential super-optimal, exoglossic, use-oriented diglossia with standard Serbo-Croatian and the Slovene dialects as the H and L varieties respectively. Like the Slovaks, the Slovenes adopted a "standard with dialects" situation, with the important difference, however, that dialectal variation in Slovene was and remains very much greater. Moreover, the Slovene literary language was subject to Slavization, purification and archaization which distanced it somewhat from gorenjsko and dolenjsko, on which it was ostensibly based.26 As a result, even in Ljubljana speakers of the super-posed literary language are usually native speakers of a local idiom, which serves them as the L variety. Outside Carniola, where the distance between the standard language and the local koiné or dialect is still greater, the existence of optimal, use-oriented diglossia is still more evident. Moreover, in some parts of the Slovene-speaking area (e.g. the Littoral, Carinthia, Styria and Prekmurje), the diglossic situation is accompanied by command of two codes of another language (Italian, German or Hungarian), i.e. we have bilingualism with double diglossia.27 A similar situation is found in the area of the Sorbian languages with the added complication that here, as in the Serbo-Croatian situation, there is interference between the H codes of Upper and Lower Sorbian.
Standard Polish now services the whole Lechitic-speaking territory. In Pomerania, where an independent written vernacular - Kashubian - had developed, the spread of Polish has created a super-optimal, use-oriented, exoglossic diglossia
with a Pomeranian koiné or dialect as the L and standard Polish as the H variety.28
* * *
23	This form of diglossia has been termed by Britto as "leaky" or "overlapping".
24	For a discussion of the viability of this term, see D a 1 i b o r B r o z o v i ć, Contemporary Standard Slovene - A Complex Linguistic Problem, Slovene Studies (forthcoming).
25	George Thomas, The Role of the Lexical Variants in the present-day language situation in Bosnia-Hercegovina, Language Problems and Language Planning, VI (1982), 29^14.
26	See Rado L. Le nee k , The Structure and History of the Slovene Language, Columbus, Ohio, 1982, 272-278.
27	For the situation in Carinthia. for example, see Tom Priestly, Cultural Consciousness and Political Nationalism: Determinants of Language Choice in Carinthia, Canadian Review of Studies in Nationalism, (forthcoming).
Ferguson claims that diglossia arises when there exists a sizeable body of prestigious literature in a language closely related to the natural language of an area, when literacy is confined to a small elite, and a period of time passes between these two stages.29 Several of the Slavonic diglossic situations fit neatly into his scheme (e.f. medieval Rus', Subcarpathian Rus', Pomerania and Bohemia). In the Croatian situation, however, there was no time lapse between the adoption of štokavian and its acquisition by a larger body of the population. Diglossia arose in those areas where štokavian was not the native dialect. Indeed, the existence of four competing Croatian literary traditions (ikavian and jekavian štokavian, čaka-vian and kajkavian), each with its own geographical distribution meant that, whichever of the idioms was chosen as the basis of the new Croatian standard, the rise of some localised diglossia was inevitable.
More seriously, Ferguson's scheme does not offer a satisfactory explanation for the rise of Slovene and Sorbian diglossia. If we examine each of these situations for common features, we find that in both cases an exoglossic super-optimal diglossia (most often with German as the super-posed variety) has given way to an endoglossic optimal diglossia (with a standardised form of the native language as the super-posed variety). While, in most of the Slovene-speaking area this process is now complete, in Carinthia as also in Lusatia it has never been fully carried out. As a result, Slovene-speaking Carinthia and Lusatia are characterised by a fourfold distribution of communicative functions - the respective standards sharing the H functions, the respective dialects sharing the L functions.
It is significant that several other Slavonic diglossias have arisen in circumstances similar to those obtaining in Slovenia and Lusatia, cf. the communicative role of standard German and Austro-Bavarian dialect in Bohemia and civil Croatia, of standard German and plattdeutsch in Pomerania, of standard Italian and Venetian dialect in Dalmatia. Can it be that this transition offers a cogent, alternative explanation for the rise of optimal diglossia in situations other than the sort discussed by Ferguson? It is also noteworthy that in all of these Slavonic diglossias the L variety contains features taken from the exoglossic languages (usually the L variety), which purism has removed from the corresponding Slavonic H variety.
While stressing their basic stability, Ferguson tentatively suggests three possible outcomes for diglossic situations:301) If the H variety is already serving some other community with which the diglossic community for whatever reasons merges, it will succeed in establishing itself as the standard. This is essentially the situation in Pomerania, where Kashubian has come to be regarded as a dialect of Polish and written Kashubian has almost disappeared. 2) Conversely, unless the H variety serves as a standard language elsewhere, it will fade away or become a learned or a purely cultic language (as happened to Church Slavonic), in which case a form of
28 Zuzanna Topolińska, Kashubian in: Schenker/Stankiewicz, op. cit., 183-94. If Kashubian is to be regarded from a functional point of view as a dialect of Polish, the diglossia may be described as optimal, use-oriented and endoglossic.
24 Op. cit., 338.
30 Op. cit., 339-40.
the L variety (which we may call outcome 2a) or a hybrid of the two (i.e. outcome 2b) will become the H variety. In Russian a hybrid has emerged, but elsewhere in Orthodox Slavdom the H functions have been entrusted to a former L variety. In Slovakia, however, the L variety has replaced Czech as the H variety, even though Czech performs these functions in Moravia and Bohemia. 3) If the important communicative centres are concentrated in one dialect area, then a single new H variety will emerge as in the case of Croatian, Slovene, Slovak and Belorussian (outcome 3a); if these centres are scattered through several dialects with none of them paramount then several new H varieties will emerge (e.g. Bulgarian and Macedonian, Upper and Lower Sorbian) (outcome 3b). In Ukrainian a unified standard has finally emerged only after a period of co-existing local H varieties.
It is more often the case, however, that there is a redistribution of functions between the H and L varieties. For example, the growth of literacy and the increased prestige of the standard language, which are characteristic of all the Slavonic language situations in this century, have resulted in a shift to the H variety for certain functions formerly performed by the L variety. Similarly, the in-migration into a diglossic area of speakers who do not share the same L variety may lead to a reluctance on the part of the original population to use the L variety even in informal settings when addressing speakers of another dialect; if the number of those in-migrating is sufficiently high, use of the L variety may disappear altogether in this particular setting. In the dialectally heterogeneous populations of modern cities like Szczecin, Niš, Zagreb or Split the local L variety is therefore inclined to be used in fewer and fewer functions and with fewer and fewer individuals.31 Conversely, the democratising tendencies which characterise the polities where the Slavonic languages are spoken have led to some relaxation - at least in theory - of the requirement that the H variety be used in all formal, public settings.
So much for the outcomes of past diglossic situations. But what prognostications do the dynamics of the present Slavonic instances of diglossia allow? In Lusatia, the pressures of the German H and L varieties threaten to crowd out the Sorbian H and L varieties despite the efforts to promote Sorbian by the Government of the German Democratic Republic. A similar threat faces Carinthia despite the moral support provided by the existence of a vibrant Slovene-speaking community in neighbouring Yugoslavia. In both cases it seems likely that eventually the Slavonic L variety will disappear and the Slavonic H variety will be forced to share an ever diminishing set of communicative functions for a steadily decreasing population. On the basis of these two examples, we can propose the following fourth outcome for diglossia: 4) In diglossic situations co-existing with bilingualism, the H and/or L varieties may be replaced in some or all of their respective functions by the exoglossic language.
For the Czech-speaking and the remainder of the Slovene-speaking territory the situation is probably stable, except that we can expect a further development of mixed usage employing features from the H and L varieties until we reach a continuum of stylistic variation, i.e. a movement from optimal to sub-optimal
31 For the situation in Zagreb, see A. Šojat, O zagrebačkom kajkavskom govoru, (Rasprave Zavoda za jezik Instituta za filologiju i folkloristiku, IV/V (1979), 125-34.
diglossia. In Croatia, however, the situation is somewhat more complex: mass literacy is bringing ever greater exposure to the H variety; urbanization and migration in search of work is greatly increasing the amount of contact between speakers of different dialects; furthermore, there are signs of the growth a Croatian interdialect based on the informal H variety used in Zagreb.32 If present trends continue, the situation in Croatia will come to resemble that of France or England, i.e. stylistic variation with some local pronunciation and usage.33 These.speculations lead us to formulate a fifth outcome for diglossia: 5) If the barriers between the H and L varieties are relaxed (i.e. there is a diminution in elitist purism34), intermediate varieties will develop. The fluidity may eventually lead to a stylistic continuum between the parameters set by the previous H and L varieties. Some phonologically insignificant local phonetic features may be preserved.
These thoughts about the rise and fall of diglossia have emerged from taking a distant overview of the historical development of Slavonic literary usage. Their validity needs to be tested against a series of more detailed studies. Nevertheless, the diglossic model for looking at the development of the Slavonic literary languages allows us to focus on their socio-communicative functions in relation to other native and non-native codes while at the same time providing a framework for understanding the impact on them of such factors as power and solidarity, social and geographical mobility, urbanization and mass literacy. Furthermore, because of the explicit link between super-optimal diglossia (bilingualism) and optimal diglossia (diglossia proper), the five outcomes of diglossia proposed above may offer some insight into problems surrounding closely related Slavonic languages which share the socio-communicative functions in certain polities.35
32	For a discussion of the effects of these and other sociolinguistic factors in the present-day language situation in Croatia, see A. Šojat, Standardni jezik i dialekt u urbanim sredinama SR Hrvatske, Rasprave Zavoda za jezik Instituta za filologiju i folkloristiku, IV/V (1979), 119-122.
33	For this term, see George Thomas, Towards a Typology of Lexical Purism in the Slavic Literary Languages, Canadian Slavonic Papers, XXX (1988), pp. 95—III, pp. 103-4.
34	The present-day Serbo-Croatian situation is characterised by both centrapetal and centrifugal forces, see George Thomas, "The Origin and Nature of Lexical Purism in the Croatian Variant of Serbo-Croatian" (Canadian Slavonic Papers, XX (1978), pp. 405-420).
35	I have in mind here the possibility of a shift in the socio-communicative functions of Slovak vis-à-vis Czech, Slovene vis-à-vis Serbo-Croatian, and Belorussian and Ukrainian vis-à-vis Russian. In each of these cases, the re-distribution of these functions could conceivably lead to any one of the outcomes 1, 2, 4 or 5 discussed above. Arguably, the threat posed by outcomes 1, 4 or 5 is every bit as great as the diminishing language-loyalty of the native speakers themselves.
POVZETEK
Eden izmed poskusov, da bi znano Fergusonovo shemo diglosije (»visoki« in »nizki« kod) dodelali v skladu z novejšo literaturo o tem problemu, razločuje: (1) lažno diglosijo, kjer skupina A uporablja kod 1, skupina B pa kod 2, (2) delno diglosijo, kjer nekaj pripadnikov jezika zaradi pričakovanih družbenih ugodnosti uporablja tudi visoki kod, (3) popolno diglosijo, kjer vsi pripadniki uporabljajo oba koda po splošno sprejetem vzorcu družbenih vlog. Diglosija je dalje lahko (1) optimalna, če sta koda drug od drugega oddaljena optimalno za Fergusonovo ozko definicijo diglosije (diglosija v strogem pomenu besede), (2) suboptimalna, če sta koda nezadostno diferencirana za Fergusonovo definicijo (stilistična variantnost znotraj istega koda), (3) superoptimalna, če sta koda preveč diferencirana (dvojezičnost). Visoki kod je ob tem lahko mrtev jezik, npr. klasična arabščina ali latinščina, ali pa živ jezik, npr. francoščina ali nemščina; nastal pa je lahko znotraj istega diasistema kakor nizki kod ali pa ravno ne: prvi tip je istojezikovni (endoglosni, npr. nemščina in švicarska nemščina), drugi pa raznojezikovni (eksoglosni, npr. prevzem danščine od Norvežanov). Še en parameter je zemljepisna razprostranjenost: (1) pri celoviti diglosiji imata visoki in nizki kod enako ozemeljsko razvrstitev (grška situacija), (2) pri spreminjavi diglosiji visoki ali nizki kod variirata po vsem jezikovnem ozemlju (arabščina), (3) pri lokalizirani diglosiji pa gre za omejitev na eno ali več področij diasistema (nemščina).
Na primer: Ko so Slovenci zavrnili ilirizem, so se izognili potencialni superoptimalni raznojezični delni diglosiji (pri kateri bi bila visoki kod zborna srbohrvaščina, nizki pa slovenska narečja) in so, kakor Slovaki, dosegli položaj »zborni jezik z narečji«: pokrajinska (nad)narečja imajo vlogo nizkega, knjižni jezik pa visokega koda, oddaljenega od narečij tudi zaradi slaviziranosti, očiščenosti tujejezičnih prvin in arhaiziranosti. Zunaj Kranjske, kjer je ta oddaljenost še večja, se bolj kaže tudi obstoj optimalne delne diglosije. Še več: ponekod (Primorska, Koroška, Prekmurje) spremlja tako stanje še znanje dveh kodov drugega jezika (italijanščine, nemščine, madžarščine), tako da imamo dvojezičnost z dvojno diglosijo.
Ferguson je ponudil tri možne izide diglosnih situacij: (1) Če visoki kod že rabi kaki drugi skupnosti, s katero se diglosna skupnost spoji, se bo vzpostavil kot zborni jezik (kašubščina je npr. »postala« narečje poljščine). (2) Če visoki kod drugod ni rabljen kot zborni jezik, bo izginil oz. postal obredni jezik (cerkvena slovanščina), pri čemer bo nizki kod ali pa križanec med obema kodoma (za slednje prim. Rusijo, za prvo ostali pravoslavni slovanski svet) prevzel vlogo visokega koda. (3) Če so pomembna komunikacijska središča skoncentrirana na enem narečnem območju, bo nastal en nov visoki kod (slovenščina, hrvaščina, slovaščina, beloruščina), sicer pa več njih (bolgarščina in makedonš-čina, gornja in spodnja lužiška srbščina). - Dodati je mogoče še dve predvidevanji: (4) Če diglosija soobstaja z dvojezičnostjo, utegne visoko in/ali nizko različico v vseh ali nekaterih njunih vlogah zamenjati eksoglosni jezik (to grozi npr. na Koroškem v Avstriji). (5) Če se pregraje med visoko in nizko različico sproščajo, tj. zmanjša se elitistični purizem, se bodo razvile vmesne različice (v to smer gre npr. položaj na Hrvaškem).