Orthographic relatedness and transposed-word effects in the grammatical decision task

Jonathan Mirault,Jonathan Grainger,Charlotte Leflaëc

doi:10.3758/s13414-021-02421-y

Abstract

In two on-line experiments (N = 386) we asked participants to make speeded grammatical decisions to a mixture of syntactically correct sentences and ungrammatical sequences of words. In Experiment 1, the ungrammatical sequences were formed by transposing two inner words in a correct sentence (e.g., the brave daunt the wind / the daunt brave the wind), and we manipulated the orthographic relatedness of the two transposed words (e.g., the brave brace the wind / the brace brave the wind). We found inhibitory effects of orthographic relatedness in decisions to both the correct sentences and the ungrammatical transposed-word sequences. In Experiment 2, we further investigated the impact of orthographic relatedness on transposed-word effects by including control ungrammatical sequences that were matched to the transposed-word sequences. We replicated the inhibitory effects of orthographic relatedness on both grammatical and ungrammatical decisions and found that transposed-word effects were not influenced by this factor. We conclude that orthographic relatedness across adjacent words impacts on processes involved in parallel word identification for sentence comprehension, but not on the association of word identities to positions in a sequence.

Highlights

Effects of orthographic relatedness among words have been much investigated in single word recognition studies, where the orthographically related words are not physically present (e.g., Andrews, 1989; Grainger et al, 1989; see Grainger, 2018, for a review), but much less so in the context of sentence reading
The remaining dataset was composed of 34,402 observations, which largely exceeds the recommendation of Brysbaert and Stevens (2018)
We estimated power for Experiment 2 a priori using the results of Experiment 1 with SIMR (Green & Macleod, 2016)

Summary

Introduction

Effects of orthographic relatedness among words have been much investigated in single word recognition studies, where the orthographically related words are not physically present (e.g., Andrews, 1989; Grainger et al, 1989; see Grainger, 2018, for a review), but much less so in the context of sentence reading. The same pattern of findings has been obtained using a parafoveal-on-foveal manipulation, such that when fixating word N, the word immediately to the right (N + 1) can be orthographically related to word N or not, and as readers’ gaze moves to position N + 1 the word at that location is changed to become a regular continuation of the sentence Orthographic relatedness has been found to facilitate processing of the target word (i.e., shorter gaze durations) when the parafoveal stimulus (N + 1) is both a word and a nonword (Angele et al, 2013; Dare & Shillcock, 2013; Inhoff et al, 2000; Mirault & Grainger, 2020; Snell et al, 2017). Orthographically related flankers were found to facilitate target word processing independently of whether they were words or nonwords (Snell et al, 2017)

Methods

Results

Conclusion