There have been many recent studies that explain language use and structure from the point of view of communicative efficiency. The linguistic explananda include correlations between frequency (or surprisal) and word length, phonetic reduction of predictable units, minimization of dependency distances in word order, the use and omission of complementizers, relativizers, case markers and other grammatical units. Communicative efficiency means that language users act rationally when choosing linguistic expressions, minimizing the costs of communication without jeopardizing the benefits. Information theory provides a convenient framework for measuring these costs and benefits.

In general, this approach is fruitful, but it also has some problems, which I will address in my talk. In particular, there is evidence that language users often fail to take the perspective of the interlocutor when encoding their message. In addition, efficient trade-offs, often visualized as Pareto frontiers, are not unproblematic in the presence of more than two types of communicative costs. Moreover, communicative context and encyclopaedic knowledge are often excluded from these accounts. I will also argue that correlational studies are not sufficient and propose an experimental paradigm that allows us to test the efficiency-related explanations in vitro.

Natalia Levshina is a linguist working at the Max Planck Institute for Psycholinguistics in Nijmegen. Her main research interests are cognitive and functional linguistics, pragmatics, typology, corpora and data science. She obtained her PhD at KU Leuven in 2011 and got her habilitation qualification at Leipzig University in 2019. In addition to papers on causatives, differential case marking, politeness, word order variation and other linguistic topics, Natalia is the author of a best-selling statistical manual “How to Do Linguistics with R” (Benjamins, 2015).