Hörberg T. and Jaeger TF. (2021) A Rational Model of Incremental Argument Interpretation: The Comprehension of Swedish Transitive Clauses. Front. Psychol. 12:674202. doi: 10.3389/fpsyg.2021.674202

"The model presented in this paper predicts processing difficulty in terms of Bayesian surprise associated with changes in expectations of possible argument interpretations (SVO vs OVS). It accommodates 36 different lexical-semantic (e.g., animacy) and morphosyntactic (e.g., case marking) features of a wide variety of transitive sentences and is trained on a corpus of 16 552 sentences. The predictions of the model are tested against reading times of a self-paced reading experiment and compared against a linguistic model, which predicts reading times directly from the sentence features. We find that Bayesian surprise as derived from our model – with only a single degree of freedom – predicts reading times equally well as the linguistic model, which is based on a number of features and their interactions. 
This suggests that the effects of lexical-semantic and morphosyntactic cues to argument interpretation are mediated through expectation-based processing.
Importantly, our model brings together two types of models of language comprehension: expectation-based models and constraint-based models, by being concerned with changes in expectations on the basis of linguistic features, rather than on the basis of words or syntactic categories."
Thomas Hörberg, researcher in general linguistics

Abstract

A central component of sentence understanding is verb-argument interpretation, determining how the referents in the sentence are related to the events or states expressed by the verb. Previous work has found that comprehenders change their argument interpretations incrementally as the sentence unfolds, based on morphosyntactic (e.g., case, agreement), lexico-semantic (e.g., animacy, verb-argument fit), and discourse cues (e.g., givenness). However, it is still unknown whether these cues have a privileged role in language processing, or whether their effects on argument interpretation originate in implicit expectations based on the joint distribution of these cues with argument assignments experienced in previous language input. We compare the former, linguistic account against the latter, expectation-based account, using data from production and comprehension of transitive clauses in Swedish. Based on a large corpus of Swedish, we develop a rational (Bayesian) model of incremental argument interpretation. This model predicts the processing difficulty experienced at different points in the sentence as a function of the Bayesian surprise associated with changes in expectations over possible argument interpretations. We then test the model against reading times from a self-paced reading experiment on Swedish. We find Bayesian surprise to be a significant predictor of reading times, complementing effects of word surprisal. Bayesian surprise also captures the qualitative effects of morpho-syntactic and lexico-semantic cues. Additional model comparisons find that it—with a single degree of freedom—captures much, if not all, of the effects associated with these cues. This suggests that the effects of form- and meaning-based cues to argument interpretation are mediated through expectation-based processing.

Read the full article in Frontiers in Psychology:
A Rational Model of Incremental Argument Interpretation: The Comprehension of Swedish Transitive Clauses >>