Some scales require cognitive effort: A systematic review on the role of working memory in scalar implicature derivation.

If some inferences require cognitive effort, could that mean, that all of them do? The scalar term "some" has long fascinated academics from various backgrounds, as it can be interpreted either purely semantically, as "some and possibly all", or pragmatically, as "some and not all". The pragmatic reading implies the generation of what is called a scalar implicature. While scientific investigation of such implicatures has given rise to many potential explanations of the "pragmatic enrichment" phenomenon behind them, the debate between the two dominant frameworks-the literal-first and the default accounts-has not convincingly been settled. With the birth of a new interdisciplinary field, appropriately dubbed experimental pragmatics, the last 20 years have led to a substantial amount of new empirical data on scalar implicatures. In this ongoing investigation, the loading and measuring of Working Memory has become an important experimentation tool, as it allows to test the contrasting hypotheses with regard to the cognitive effort of implicature generation, which are made by the two main theoretical accounts. The current systematic review evaluates the relevant literature until March 08, 2022 in an attempt to shed light on the role of Working Memory in implicature derivation. A comprehensive search, and two-step review procedure yielded a sample of 18 studies, containing data of 23 relevant experiments. Findings were bundled in a narrative synthesis and combined through two separate meta-analyses. Our results support the literal-first account, by showing that the derivation of scalar implicatures is a cognitively effortful process that is sensitive to changes in the available Working Memory resources. However, as the reported effects are relatively weak and capricious, we argue that the development of more sophisticated paradigms and eventually, stronger theories within the field, will be crucial in order to both fully understand the current results and set-up fruitful future research.