Elicitation and experimentation: implications for English sociolinguistics

In a recent discussion on the replication crisis and its consequences for linguistics, Grieve (2021) convincingly argues that the complexity of language as a social phenomenon inherently limits the capacity for linguists to investigate variables such as language variety or dialect using experimental approaches as opposed to observational approaches. Grieve points out a sociological propensity for some linguists to aim for experimentation as the optimal and superior empirical approach to studying language, while underestimating the value of observational approaches, an undesirable fallacy for the future of the field. In this paper, we build upon his remarks and discuss some of their implications for future studies in English sociolinguistics and dialectology, especially those focusing on grammatical variation with elicitation-based methodologies. We argue that understanding elicitation as a concept distinct from the dichotomy of observation versus experimentation enables us to avoid the pitfall of equating structured data elicitation with experimentation. We also argue that research in these strands of language variation should continue to embrace the ultimately observational nature of its results and use experimental terminology with great care, as part of a scientific approach to uncovering relationships that matter in the real world such as variation in the social practice of language, rather than describing causal links directly. We illustrate our points with two case studies in English: an exploratory survey of double modal acceptability in Southeast England, and a sociolinguistic questionnaire on the use of necessity modals in Singapore English.