Time for a Change: Why Psycholinguistics Should Move Beyond the Content-Function Word Distinction.

Abstract

The traditional distinction between content and function words continues to underpin much psycholinguistic research, even though other areas of linguistics have long recognized it as problematic (e.g. Bennis et al., 1983, and Ishkhanyan et al., 2017 within aphasiology; Diessel, 1999: 150-153, and Bisang, 2010: 291 within linguistic typology; Emonds, 1985: 162-164, and Rauh, 1993 within grammatical theory). We argue that relying on this pre-theoretical distinction in psycholinguistics can obscure meaningful variation and potentially lead to misleading results. We illustrate this with Staub (2024), who, in a large-scale eye-tracking study, found no evidence that so-called function words are skipped more often or receive shorter fixations than so-called content words. We argue that this null finding likely stems from Staub's function word group being heterogeneous, containing both lexical and grammatical items. In support of this, we reclassify the words in Staub's function word group as either lexical or grammatical based on a usage-based theory of the lexical-grammatical contrast (Boye & Harder, 2012), and rerun the analysis in his Experiment 1 on this group of words. Our analysis suggests that the grammatical words receive shorter fixations than the lexical words, providing support for the heterogeneity of the function word group. Given this, it is unsurprising that Staub found no differences in reading measures between the function and content word categories. This case demonstrates how relying on the traditional function-content distinction in psycholinguistic research can obscure meaningful patterns, and it highlights the potential of adopting Boye and Harder's (2012) usage-based alternative.