Diglossia as a Sociolinguistic Situation

Harold F. Schiffman
South Asian Regional Studies
University of Pennsylvania


The sociolinguistic condition known as diglossia has attracted wide attention since the publication of Ferguson's seminal article (1959).[*] Despite its occurrence in many non-western contexts, it is not simply a phenomenon of exotic third-world cultures, but characterizes a number of languages found in various parts of the world, including western Europe.[*]

Power and Prestige.

Diglossic languages (and diglossic language situations) are usually described as consisting of two (or more) varieties that coexist in a speech community; the domains of linguistic behavior are parceled out in a kind of complementary distribution. These domains are usually ranked in a kind of hierarchy, from highly valued (H) to less valued (L); when the two varieties are recognized (or tacitly accepted) as genetically related, the H domains are usually the reserve of the more conservative form of the language, which is usually the literary dialect if there is a written form. `Formal' domains such as public speaking, religious texts and practice, education, and other prestigious kinds of usage are dominated by the H norm; the L norm is used for informal conversation, jokes, street and market, the telephone, and any other domains (e.g. letter writing, cinema, television) not reserved for the H norm. For diglossic situations involving two different (genetically unrelated) linguistic codes, (sometimes referred to as `extended' diglossia) the one dominating the H domains has the greater international prestige or is the language of the local power elite or the dominant religious community and/or its priesthood. In such cases the H-variety language is clearly the language of the more powerful section of the society, however power is defined.

Thus in French Canada, English occupies the H-variety niche because it has the greatest prestige in North America (and perhaps internationally as well), its population even within Canada is numerically greater than the community of French speakers, and its speech community is economically dominant, both in English Canada and in French Canada. Conversely, in France, French is the H-variety in diglossic situations involving other languages or dialects, such as Breton or Alsatian, where these varieties are only used as L-variety spoken vehicles in the home, on the street, in the construction trades, etc.

It remains to be seen whether the same kind of imbalance of power exhibited in non-genetic diglossia can be said to exist with regard to classical or genetic diglossia. In many diglossic situations, only a minority or elite control the H domain successfully, so those who know only L are at a disadvantage.

Ferguson's original formulation.

Ferguson originally summarized diglossia (1959: 435) as follows:

DIGLOSSIA is a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any section of the community for ordinary conversation.

The notion that diglossia could also be used to characterize other multilingual situations where the H and L varieties were not genetically related, such as Sanskrit (as H) and Kannada (as L) in India, was developed by Fishman (1967) and research on diglossias since have focused to a great extent, though not entirely, on characterizing various kinds of extended diglossias.

Post-1959 research on diglossia has concentrated on a number of variables and important questions: function, prestige, literary heritage, acquisition, standardization, stability, grammar, lexicon, phonology, the difference between diglossia and standard-with-dialects, extent of distribution in space, time, and in various language families, and finally what engenders diglossia and what conditions favor its development.


Function. The functional differentiation of discrepant varieties in a diglossia is fundamental, thus distinguishing it from bilingualism. H and L are used for different purposes, and native speakers of the community would find it odd (even ludicrous, outrageous) if anyone used H in an L domain, or L in an H domain.


Prestige: in most diglossias examined, H was more highly valued (had greater prestige) than was L. The H variety is that of `great' literature, canonical religious texts, ancient poetry, of public speaking, of pomp and circumstance. The L-variety is felt to be less worthy, corrupt, `broken', vulgar, undignified, etc.


Literary Heritage: In most diglossic languages, the literature is all in H-variety; no written uses of L exist, except for `dialect' poetry, advertising, or `low' restricted genres.[*] In most diglossic languages, the H-variety is thought to be the language; the L-variety is sometimes denied to exist, or is claimed to be only spoken by lesser mortals (servants, women, children). In some traditions (e.g. Shakespeare's plays), L-variety would be used to show certain characters as rustic, comical, uneducated, etc.


Acquisition: L-variety is the variety learned first; it is the mother tongue, the language of the home. H-variety is acquired through schooling. Where linguists would therefore insist that the L-variety is primary, native scholars see only the H-variety as the language.


Standardization: H is strictly standardized; grammars, dictionaries, canonical texts, etc. exist for it, written by native grammarians. L is rarely standardized in the traditional sense, or if grammars exist, are written by outsiders.


Stability: Diglossias are generally stable, persisting for centuries or even millennia. Occasionally L-varieties gain domains and displace the H-variety, but H only displaces L if H is the mother tongue of an elite, usually in a neighboring polity.


Grammar: The grammars of H are more complex than the grammars of L-variety. They have more complex tense systems, gender systems, agreement, syntax than L-variety.


Lexicon: Lexicon is often somewhat shared, but generally there is differentiation; H has vocabulary that L lacks, and vice-versa.


Phonology: Two kinds of systems are discerned. One is where H and L share the same phonological elements, but H may have more complicated morphophonemics. Or, H is a special subset of the L-variety inventory. (But speakers often fail to keep the two systems separate.)

A second type is one where H has contrasts that L lacks, systematically substituting some other phoneme for the lacking contrast; but L may `borrow' elements as tatsamas,[*] using the H-variety contrast in that particular item.


Difference between Diglossia and Standard-with-dialects. In diglossia, no-one speaks the H-variety as a mother tongue, only the L-variety. In the Standard-with-dialects situation, some speakers speak H as a mother tongue, while others speak L-varieties as a mother tongue and acquire H as a second system.


Distribution of diglossia in language-families, space, and time. Diglossia is not limited to any geographical area or language family, and diglossias have existed for centuries or millennia (Arabic, South Asia). Most diglossias involve literacy, but oral diglossias are conceivable.[*]


What engenders diglossia and under what conditions.


Existence of an ancient or prestigious literature, composed in the H-variety, which the linguistic culture wishes to preserve as such.


Literacy is usually a condition, but is usually restricted to a small elite. When conditions require universal literacy in H, pedagogical problems ensue.


Diglossias do not spring up overnight; they take time to develop

These three factors, perhaps linked with religion, make diglossia extremely stable in Arabic and other linguistic cultures such as South Asia.

Extended diglossia (Fishman 1967)

Given the extensive research on diglossia and the many recent attempts to both refine and extend it, a review of some of these studies, especially those pertaining to the socio-economic conditions in which diglossic languages are usually embedded, seems to be warranted. It should be noted, however, that diglossia is a gradient, variable phenomenon, which cannot easily be boxed into an either-or, binary system of categorization. And as Ferguson himself recently pointed out (Ferguson 1991, in Hudson 1991a), his original formulation of diglossia was not meant to encompass all instances of multilingualism or functional differentiation of languages. Thus the many attempts to `refine' or `extend' diglossia, or to discern whether such and such is or is not a case of diglossia, may be barking up the wrong tree.

Fishman (1967) introduced the notion that diglossia could be extended to situations found in many societies where forms of two genetically unrelated (or at least historically distant ) languages[*] occupy the H and L niches, such that one of the languages (e.g. Latin in medieval Europe), is used for religious, educational, literacy and other such prestigious domains, while another language (in the case of medieval Europe, the vernacular languages of that era) is rarely used for such purposes, being only employed for more informal, primarily spoken domains.

Diglossia and Language Shift.

Diglossia has often been noted as a factor in language shift, especially in speech communities where a minority language is in a diglossic relationship with a majority language. Fishman (1967: 36) had previously noted that

``[B]ilingualism without diglossia tends to be transitional both in terms of the linguistic repertoires of speech communities as well as in terms of the speech varieties involved per se. Without separate though complementary norms and values to establish and maintain functional separatism of the speech varieties, that language or variety which is fortunate enough to be associated with the predominant drift of social forces tends to displace the other(s)."

Classical and Extended Diglossia.

Various scholars have proposed terminologies for a taxonomy of diglossias. For what here is referred to as `classical' (Ferguson 1959) and `extended' (Fishman 1967) diglossia, Kloss has proposed the terms `in-diglossia' (for the kind where the two varieties are closely related) and `out-diglossia' (for situations where the two languages are unrelated or at best distantly related) (Kloss, 1966: 138.) A classicist might prefer something like `endo-diglossia' and `exo-diglossia', i.e. prefixes that fit better with the original Greek roots of the terms. But it is clear to some researchers that there are important differences in the dynamics of societies characterized by the (at least) two basic kinds of diglossia. Fishman has also proposed a useful distinction between `consensually different languages' and `consensual dialects', since there is an unresolved debate as to whether Caribbean English (for example, but any Creole language/dialect could be used) is in fact genetically descended from English, i.e. is consensually a dialect of English, or is consensually (agreed to be classified as) a separate language. This would also be useful in situations found in South Asia, where some L-varieties are associated with H-varieties that are not in fact their closest genetic ancestor; for example, eastern varieties of Hindi (Bihari dialects, etc.) that have long been noted to have descended from eastern apabhramsas but are treated by their speakers as being dialects of standard Hindi; one could make the case that Sri Lanka Tamil may also be more closely related to Malayalam than it is to Tamil, but not in the minds of its speakers. And it seems to be the case that Swiss German was once consensually agreed to be in a diglossic hierarchy with Standard German, but that this consensus is now breaking down.

Scotton (1986) proposes the terms `narrow' for Ferguson's 1959 version of diglossia, and `broad' (or `diglossia extended') to refer to Fishman's expansion of the discussion. According to Scotton, few truly diglossic (in the 1959 sense) communities actually exist, because to meet the criteria, two conditions must hold: ``(1) Everyone ...speaks the Low variety as a mother tongue." and ``(2) The High variety is never used ...in informal conversations." Unambiguous examples of these are Tamil, Lëtzebuergesch, and Swiss German. Britto (1986) proposes the terms `Use-oriented' (or diatypical) and `User-oriented' (or dialectal) diglossia to refer roughly to the same dichotomy others have also attempted to define.[*]

Fishman's 1980 taxonomy of ``kinds of linguistic relationships between H's and L's" is worth stating in full:

(a) H as classical, L as vernacular, the two being genetically related, e.g. classical and vernacular Arabic, classical or classicized Greek (Katarevusa) and demotiki, Latin and French among francophone scholars and clergy in earlier centuries, classical and vernacular Tamil, classical and vernacular Sinhalese, Sanscrit and Hindi, classical Mandarin and modern Pekinese, etc.[*]

(b) H as classical, L as vernacular, the two not being genetically related, e.g. Loshn koydesh (textual Hebrew/Aramaic) and Yiddish (Fishman, 1976) (or any one of the several dozen other non-semitic Jewish L's, as long as the latter operate in vernacular functions rather than in traditional literacy-related ones (Weinreich, 1980).

(c) H as written/formal-spoken and L as vernacular, the two being genetically unrelated to each other; e.g. Spanish and Guaranʬ in Paraguay (Rubin, 1972), English (or French) and various vernaculars in post-colonial areas throughout the world ....

(d) H as written/formal-spoken and L as vernacular, the two being genetically related to each other. Here only significantly discrepant written/formal-spoken and informal-spoken varieties will be admitted, such that without schooling the written/formal-spoken cannot even be understood (otherwise every dialect-standard situation in the world would qualify within this rubric), e.g. High German and Swiss German, standard spoken Pekinese [Putonghua] and Cantonese, Standard English and Caribbean Creole. (Fishman, 1980: 4).[*]

These differences range beyond the obvious ones of genetic vs. non-genetic relationship, and in fact have to do primarily with power relationships in the societies characterized by them. Various scholars have proposed that extended diglossia is usually unstable, unless certain conditions having to do with power are not met. Classical diglossia, usually thought to be more stable than extended diglossia, can also be shown to be unstable under certain conditions. It may also be the case that the type of diglossia in question may also itself change, i.e. a narrow kind of diglossia may be replaced by a broad form without much overt awareness on the part of the speech community.

Diglossia as a Continuum.

Classical Fergusonian genetic endo-diglossia (Fishman's (a) and (d) types) characterizes a number of linguistic situations that have already received much attention in the literature. Fishman distinguishes usefully between classical and related and written/formal-spoken; he places Tamil in the former situation, whereas one could just as easily put it in the latter; one might say that Tamil actually has three norms (a classical, i.e. Sangam or Pandit style, modern Literary/formal-spoken, and educated colloquial, not to mention local dialects). Between these styles there are shadings from one style into another, i.e. it is possible to write modern Literary Tamil with an archaic lexicon but with non-archaic grammar; it is also possible to swing in the other direction and make modern LT more spoken, or to make educated colloquial more literary in flavor, or more non-standardly colloquial. In any event, though linguistic cultures think of diglossia as either-or, it is often a gradient cline, with one variant shading into another.

Diglossia and the Linguistic Culture that maintains it.

Speech communities have belief systems about their language--origin myths, beliefs about `good' and `bad' language, taboos, shibboleths, and so on. These beliefs are part of the social conditions that affect the maintenance and transmission of that language. Thus, the fact that a language is diglossic is actually a feature of the linguistic culture[*] of the area where that language is used, rather than of the language per se. To speak of a particular language as diglossic or not is at best imprecise, since a language (e.g. English) as spoken in one part of the world may exhibit little or no diglossia, while the same language (again using English as an example) as used in a Caribbean creole community would have to be considered diglossic. Speakers of a particular language can not be characterized as diglossic; only their behavior, or the behavior of the speech community can be considered diglossic. Thus, beliefs and attitudes about the language condition the maintenance of diglossia as a fact of linguistic culture. In the case of the Tamils, for example, it is the set of beliefs about the antiquity and purity of Tamil that unites all members of the linguistic culture in its resistance to any change in the corpus or status of Tamil, by which of course is meant H-variety Tamil. (Schiffman 1974: 127).

Diglossia and Literacy.

In a society where literacy is not universal, not all speakers control the use of the school-imparted H-variety. This does not mean that illiterates have the option of using the L variety in H-variety domains; rather, the expectation is that they will remain silent[*] rather than exhibit inappropriate linguistic norms. Their linguistic behavior is in fact restricted to the L domains, and use of H domains is de facto the monopoly of the educated few.

Shifting domains and Diglossia.

While diglossia as a fact of linguistic culture may be stable, the distribution of domains reserved for one variety or other can vary; the dominance of a particular domain by a particular variety can shift, with one variety encroaching on domains previously restricted to another. In Tamil, for example, the political speech was once restricted to the domain of the H variety, but nowadays political speeches only begin and end in H; in between, L variety predominates (probably as a mark of solidarity). In journalism, especially in political cartoons, etc. one also sees a shift from H to L in many linguistic cultures. In Alemannic Switzerland and some other linguistic cultures, the development of television has opened up a domain that has become almost exclusively that of the L variety, especially in `live' interviews, talk shows, game shows, sports reporting, etc. where use of H would seem stilted and unnatural.[*][*]

On the other hand, social forces within a particular linguistic culture can move to eliminate diglossia, as was the case when medieval Latin was displaced during the Renaissance by various European vernacular languages; diglossia is giving way in present-day Greece, where it had held sway until a government decree ordained the shift from H (katharevousa) to L ( demotiki) in many domains[*] Diglossia was more extreme in pre-modern Bengali and Telugu than it is today, as a result of movements led by prestigious writers (Tagore for Bengali) to democratize access to literacy and education, and modernize their languages. Latin held on in German linguistic culture until the early 18th century in a number of restricted domains (scholarly writing, university lectures). When and if diglossia is more or less eliminated, or made less extreme, by the choice of a more modern colloquial norm[*] we would have to, by rights, speak of a kind of language shift. To ignore shift when it takes place within a diglossic continuum would be to perpetuate the notion that diglossia is in effect irrelevant.

Diglossia and Linguistic Areas.

If diglossia is an aspect of linguistic culture, it may result from and be maintained by the existence of a linguistic area (Emeneau 1956) in which diglossia is an areal feature as well as a feature of a particular linguistic culture within the area. In South Asia, and in those Southeast Asian linguistic cultures that use Indic writing systems, diglossia seems to be a well-nigh inherent characteristic of the linguistic cultures,[*] since there is a tendency to develop diglossia even in languages that originally may have not exhibited a great degree of it. When Hindustani was chosen as the national language of independent India, supposedly because of its wide use as a lingua franca in the area, steps were immediately taken to develop an H variety, highly Sanskritized in vocabulary, since the vernaculars of Hindi then in existence seemed to be too `Low' for many citizens of the country. Of course diglossicization as a value may vary from sub-culture to sub-culture in the region, but it cannot be denied that the overall view in South Asia and peninsular Southeast Asia is pro-diglossic.

Partial vs. Total Diglossia

Researchers have noted the situation where some speakers control H but others have L as a mother tongue, and learn H as a second system. Thus in some linguistic cultures, all speakers exhibit diglossic behavior (i.e. use both H and L varieties in complementary distribution), while in others, only some members of the society do.[*] This could be illustrated either by a society where everyone controls L, but only some actively control H, or the opposite case where everyone speaks and writes H, but some also control an L variety. We can refer to this dichotomy as total diglossia vs. partial diglossia. This factor is distinct from the issue of whether diglossia is homogeneous or heterogeneous in the area (see below ).

Homogeneous and heterogeneous diglossia.

Even if diglossia is total and universal, we must determine whether the L norm is in fact one variety or more than one, i.e. is it homogeneous or heterogeneous. That is, is there an L variety that can be used for communication throughout the linguistic culture and with all segments of the speech community, such that no one is forced to resort to the H variety (written formal/spoken) or some other language, as a lingua franca.[*] In Switzerland, no one L-variety is recognized as standard; speakers must learn to accommodate their variety to those of others, since the use of H Schriftdeutsch is not considered appropriate between Swiss citizens (Schiffman 1991).

Diglossia and Power and Solidarity.

Brown and Gilman (1960) introduced the notion that the use of certain pronouns (epitomized as T and V) can be an expression of power and/or solidarity. Rubin (1972) extended the analogy of T and V pronouns to the use of L and H varieties in Paraguay, a supposedly `bilingual' linguistic culture in which the two languages, Spanish and Guaranʬ, are in an extended diglossic relationship. In many of the linguistic cultures discussed here, the use (or misuse) of L and H varieties also can raise some of these same issues. Certainly the use of L where H is expected (or vice versa), constitutes a violation of communicative competence rules. If an outsider speaks Hochdeutsch in Alemannic Switzerland, addresses a hotel clerk in Hindi in Madras, or begins a conversation with a well-dressed stranger in Asunciʫn in Guaranʬ, these are violations of social norms that stem from an inadequate understanding of the linguistic culture.

Brown and Gilman (1960) established the notion that use of T pronouns (the familiar, non-respect form) can have several social meanings. Reciprocal use of T by equals expresses solidarity, but between non-equals the giver of T is putting him/herself in a position of power, and the receiver is expected to respond with V. Similarly, reciprocal V usage implies mutual respect and social distance; any non-reciprocal use of these pronouns is an expression of a differential of power.

As Rubin demonstrated, in diglossic situations the use of H or L varieties in a given social exchange (as distinguished from societal patterned usage as a whole) may be seen as the same kind of T/V situation. The use of L may be an expression of solidarity and may not be offered to speakers whose social position is superior or distant. Similarly H may be the only variety appropriate in a given situation because the use of L would imply a solidarity that is only reserved for members of a particular in-group. The use of Black English by white speakers of American English in conversations with African-Americans would probably be considered insulting unless individual allowances had already been negotiated. The use of L-variety Tamil by non-Indians is considered inappropriate by many educated Tamilians, who may respond in H-variety Tamil or in English unless the use of L-variety has already been negotiated (with explanations about the goals of the speaker and disclaimers about intended slurs and put-downs.) The use of H-variety German in Alemannic Switzerland conversely may be seen as a power-trip designed to put the Swiss speaker at a disadvantage. The fact that the Hochdeutsch speaker may have no alternative L to use may be irrelevant; it certainly explains the desire to switch to `neutral' English or French. In Luxembourg, however, L-variety and its use are expressions of Lëtzebuergesch nationality and ethnic solidarity, so while Luxembourg nationals expect L from all Luxembourgers, they switch readily to French or Hochdeutsch or English with foreigners, with no expectation that they will or should be able to speak L.