In lieu of an abstract, here is a brief excerpt of the content:

American Speech 20.1 (2002) 70-99

[Access article in PDF]

Past be in Tristan da Cunha:
The Rise and Fall of Categoricality in Language Change

Daniel Schreier
University of Canterbury, Christchurch


THIS ARTICLE EXAMINES linguistic regularization in a contact scenario that involves several transplanted varieties of English. It documents the extent and directionality of past be leveling in the colonial setting of the island of Tristan da Cunha in the South Atlantic Ocean. The aim is to show that a combination of language-external factors—such as new-colony formation, extreme geographic isolation, reduced in-migration after an initial founding period, dense social networks, and limited contact with a prestigious standard variety—may have an accelerating effect on linguistic regularization. I suggest that the special sociolinguistic scenario that gave rise to Tristan da Cunha English (TdCE) resulted in rapid homogenization of language-inherent changes and led to unprecedented regularization of past tense be with was as a pivot form. Increasing geographical mobility and off-island education, on the other hand, result in a significant increase of standard were forms.


The loss and reduction of marked or minority variants (so-called LEVELING or REGULARIZATION) 1 has been subject to extensive scrutiny in contact scenarios (Trudgill 1986; Siegel 1987, 1997) and is one of the most widespread manifestations of language change which operate during new-dialect formation. Perhaps one of the most amply documented leveling processes in English involves the past tense morphemes of the verb be. The verbal paradigm of past tense be is notorious for its irregularity as, with the two distinct allomorphs was and were, it is the only English verb that has preserved person-number concord in the past tense paradigm. As Wolfram, Hazen, and Schilling-Estes (1999, 75) note, "The irregular status of be is without parallel in the current configuration of subject-verb concord." In English, this irregularity is explained by the historical development of [End Page 70] the verb, which can be traced to no fewer than three separate Indo-European verbs (see discussion in Hazen 1994). In view of general regularization processes operating mostly in nonstandard dialects of English (Chambers 1995), be is a prime candidate for analogical change. Ample literature scrutinizes its evolution in varieties of English around the world, for instance in the British Isles (Cheshire 1982; Britain 1991, forthcoming; Tagliamonte 1998), the United States (Labov 1972; Wolfram and Christian 1976; Feagin 1979), Australia (Eisikovits 1991), Canada (Meechan and Foley 1994), and the Caribbean (Tagliamonte and Smith 2000).

A cross-dialectal comparison reveals that nonstandard dialects differ both quantitatively and qualitatively. Past be regularization has not advanced at the same rate in the varieties where it has been documented, and there is a significant degree of variability in its directionality. The qualitative differentiation is most plausibly explained by the competition between the different morphemes was and were (and weren't; see below), which may function as pivot forms in a more regular—or more precisely, in a less irregular—past tense paradigm. Indeed, past be regularization has a long-standing historical continuity in English, and Quirk and Wrenn (1960) speculate that some alternation among distinct patterns existed already in the Old English period. By the same token, leveling is well recorded in Middle English, and Visser (1970) and Curme (1977) show that from the fourteenth century on there was a marked tendency for was to occur in contexts of were. Visser (1970, 3: 72) cites the following example from the epic poem "Richard Coeur de Lion," written around 1300:

1. Thrytty knyghtes . . . forsothe WAS in that companye

The historical literature suggests that, whereas was predominantly co-occurred with the first and third person singular and were with the plural persons, the second person singular was subject to considerable regional variation. Forsström (1948), for instance, finds a sharp division between the south of England, where were was predominantly used, and the northern varieties, which historically used was with the second person singular. 2 He also points out that the paradigm has undergone extensive modifications inasmuch as "preterit...


Additional Information

Print ISSN
pp. 70-99
Launched on MUSE
Open Access
Archive Status
Archived 2005
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.