Three factors in explaining scalar diversity

Authors

  • Eszter Ronai
  • Ming Xiang

DOI:

https://doi.org/10.18148/sub/2022.v26i0.1026

Abstract

Rates of scalar inference, whereby the utterance of a weaker term (e.g., some) leads hearers to infer the negation of a stronger term (all), have been found to vary substantially across lexical scales. For instance, the some but not all scalar inference arises much more robustly than good but not excellent. This finding has been termed scalar diversity. In this paper, we first replicate scalar diversity on 60 different pairs of scalar expressions, which represent a better balance across grammatical categories than has been tested in previous work. We then turn to the open question of what can explain scalar diversity, proposing three factors: 1) a language production-based metric of how accessible the stronger alternative (all) is; 2) the distinctness of the two scalar terms (some vs. all), as measured by posterior degree estimates; and 3) the meaning of the negated strong scalar term (not all), again measured by degree estimates. We report on three experiments showing that these factors can indeed explain some of the observed variation in scalar inference rates.

Downloads

Published

2022-12-22

How to Cite

Ronai, E., & Xiang, M. (2022). Three factors in explaining scalar diversity. Proceedings of Sinn Und Bedeutung, 26, 716–733. https://doi.org/10.18148/sub/2022.v26i0.1026