Functional Partitioning and Possible Limits on Variability

Abstract

English adjective comparison presents a textbook case for variation analysis, exhibiting robust variation between synthetic (angrier/angriest) and analytic (more/most angry) forms. However, our understanding of this alternation relies largely on evidence from written genres; very little is known about adjective comparison in speech, particularly in the vernacular. Since it is likely that analytic comparison gained traction via analogy with French, the shortage of spoken evidence proves a critical gap in our knowledge. This article examines comparison strategies in a large corpus of spoken New Zealand English (NZE), and compares the results with those from another colonial variety, Canadian English. Consistent with reports elsewhere, synthetic comparison is the preferred overall strategy. However, the analysis reveals a system that is partitioned: comparison is not particularly variable in these materials, either synchronically or diachronically. Individual adjectives tend to pattern either synthetically or analytically, raising questions about an across-the-board functional overlap between comparative forms. This article explores a number of explanations (e.g., variation is register-specific or variety-specific, or may be visible only in extremely large corpora), and ultimately suggests that in contrast to the situation found in written and expository genres more generally, variation is limited in vernacular speech.