Document Type



Sociolinguists generally agree that a diglossic situation is one in which a single speech community employs two or more varieties of language, a H(igh) variety and a L(ow) variety, for different communicative purposes. Ferguson’s (1959) classic definition also includes a structural component: the two forms of language are varieties of the same language, and hence related, but “highly divergent” from one another, more so than a dialect in relation to its standard language. However there is little agreement on this point, and different researchers give different characterizations of how divergent H and L must be. The theoretical status of intermediate varieties of language in a diglossia is crucial to sorting out what sort of relationship exists between H and L. If an intermediate variety were to constitute a distinct canonical variety, a “third pole” distinct from H and L (Ferguson 1991), then the essential “two-ness” of Ferguson’s (1959, 1991) characterization would break down. Likewise, if H and L were extremes on a uniform continuum of varieties, and were not clearly distinct, the “two-ness” characterization would again break down. A peculiar fact about research on diglossia is that while its two-ness is a central theoretical claim, and while this claim is not universally accepted, no one appears to have ever developed an empirical procedure for testing this claim. Research on diglossia typically employs the casual observations of the sociolinguist, who impressionistically makes judgments about the relatedness or divergence of linguistic systems. Claims about one, two, three, or uncountable numbers of canonical varieties are simply asserted for others to accept, without further justification. The two-ness of diglossia is therefore something which needs to be empirically established. In this paper, I examine empirical evidence for two-ness in Sinhala diglossia based on a study of 52 Sinhala text fragments. I conclude that a two-ness does exist in Sinhala diglossia, but that a modified view is necessary. The view I adopt is one where H and L are seen as systems of correspondence among the varieties of language in a diglossia; I adapt Britto’s (1986, 1991) notion of H and L “diasystems” for the purpose of describing these correspondences.


Linguistics | Social and Behavioral Sciences

Publication Date




Included in

Linguistics Commons