This dataset contains elicitation data collected through a questionnaire on superlative strategy choice in English (X-est vs. most X). Native speakers (n = 675) were asked to indicate their preferred superlative variants of the adjectives used in n = 120 systematically manipulated prompt sentences. The data enable us to quantify the effects of seven contextual constraints and four speaker variables on the superlative alternation.
[publication abstract, “The superlative alternation in British and American English: Questionnaire-based insights."]
Drawing on data gleaned from a large-scale questionnaire study, this paper investigates the role of contextual constraints in the alternation of synthetic (-est) and analytic (most) superlative forms. The multifactorial analysis suggests that variables that have been interpreted along the lines of processing complexity in previous studies increase the odds of the analytic superlative, albeit to varying degrees. These findings substantiate the hypothesis that the cognitive support strategy of more-support, as identified by Mondorf (2009b) for the comparative, also extends to the superlative. Despite this empirical support, my analysis exposes the limitations of a complexity-based account of variation in gradation strategy choice. It is further observed that speakers of the two main reference varieties British English and American English do not systematically differ in their sensitivity to the cognitive mechanisms that underlie this alternation. This observation adds to the body of evidence that suggests a high degree of cross-varietal homogeneity in morphosyntactic alternations. From a methodological perspective, it is argued that elicitation studies of this kind constitute a valuable complement to the extant corpus studies conducted on this alternation in that they allow for the operationalisation of linguistic and extralinguistic variables in a controlled setting.
LimeSurvey, 3.5.1 (http://www.limesurvey.org)
R: A language and environment for statistical computing, 3.6.3 (https://www.r-project.org)
