Online representations of non-canonical sentences are more than good-enough

Abstract

Proponents of good-enough processing suggest that readers often (mis)interpret certain sentences using fast-and-frugal heuristics, such that for non-canonical sentences (e.g. The dog was bitten by the man) people confuse the thematic roles of the nouns. We tested this theory by examining the effect of sentence canonicality on the reading of a follow-up sentence. In a self-paced reading study 60 young and 60 older adults read an implausible sentence in either canonical (e.g. It was the peasant that executed the king) or non-canonical form (e.g. It was the king that was executed by the peasant), followed by a sentence that was implausible given a good-enough misinterpretation of the first sentence (e.g. Afterwards, the peasant rode back to the countryside), or a sentence that was implausible given a correct interpretation of the first sentence (e.g. Afterwards, the king rode back to his castle). We hypothesised that if non-canonical sentences are systematically misinterpreted then sentence canonicality would differentially affect the reading of the two different follow-up types. Our data suggested that participants derived the same interpretations for canonical and non-canonical sentences, with no modulating effect of age group. Our findings suggest that readers do not derive an incorrect interpretation of non-canonical sentences during initial parsing, consistent with theories of misinterpretation effects that instead attribute these effects to post-interpretative processes.

Misretrieval but not misrepresentation: A feature misbinding account of post-interpretive effects in number attraction

“…This difference between on- and offline measures is relevant to post-interpretive accounts of the misinterpretation of noncanonical sentences (Bader & Meng, 2018; Cutter et al, 2022; Meng & Bader, 2021). This account is based largely on the finding that agent–patient naming tasks induce a larger working memory burden together with the finding that real-time plausibility judgements diverge from these high error rates.…”

“…Furthermore, the offline findings from Patson and Husband (2016) were corroborated with the current study's stimuli in Experiment 3, suggesting that the systematic rate of erroneous responses is likely due to the nature of the comprehension question task rather than a product of faulty representational processing of the target NP's global number feature. This difference between on-and offline measures is relevant to post-interpretive accounts of the misinterpretation of noncanonical sentences (Bader & Meng, 2018;Cutter et al, 2022;Meng & Bader, 2021). This account is based largely on the finding that agent-patient naming tasks induce a larger working memory burden together with the finding that real-time plausibility judgements diverge from these high error rates.…”

“…This is used to argue against a heuristic processing stream, which together with an algorithmic processing stream makes up the main argument of good-enough processing accounts of misrepresentations (Christianson, 2016;Christianson et al, 2010;Ferreira et al, 2002;Ferreira & Patson, 2007). The findings from the post-interpretive studies (Bader & Meng, 2018;Cutter et al, 2022;Meng & Bader, 2021) suggest that comprehension questions forcing participants to search in memory for feature-noun associations are prone to interference not typically encountered in normal language processing. One reason researchers likely assume that forcing participants to search in memory for features is not antithetical to natural linguistic behaviour is because it is superficially similar to cue-based retrieval processesthey both involve memory retrieval.…”

“…Experiment 1 used a phrase-by-phrase two-sentence discourse self-paced reading paradigm (cf. Cutter et al, 2022; Slattery et al, 2013) to test whether attractor nouns can lead comprehenders to misrepresent number features of the head noun in a complex NP during real-time processing. The goal is to reconcile the discrepant findings from Patson and Husband’s offline number probe from findings in the cue-based retrieval literature showing no attraction at the verb in grammatical sentences (Tanner et al, 2014; Wagers et al, 2009).…”

Processing noncanonical sentences: effects of context on online processing and (mis)interpretation

“…According to the good-enough model of sentence comprehension (henceforth referred to as GE model, Christianson, 2016;Ferreira & Patson, 2007;Karimi & Ferreira, 2016), misinterpretation errors reflect the application of heuristic strategies that are assumed to operate in parallel to algorithmic processing routines, including heuristics favoring the agent-before-patient over patient-before-agent order, and plausible over implausible interpretations (see Townsend & Bever, 2001 for a related proposal). Alternative accounts have ascribed misinterpretation effects to processes involved in maintaining or operating on memory representations for sentences (Bader & Meng, 2018;Cutter et al, 2022;Meng & Bader, 2021;Paolazzi et al, 2019). For example, Bader and Meng (2018) proposed that misinterpretations reflect processes of memory retrieval required by the agent/patient naming task which was used to assess comprehension in Ferreira (2003) and in Experiment 1 of Bader and Meng (2018).…”