Electrophysiological differences in older and younger adults’ anaphoric but not cataphoric pronoun processing in the absence of age-related behavioural slowdown

Abstract

This study reports on an event-related potentials experiment to uncover whether per-millisecond electrophysiological brain activity and analogous behavioural responses are age-sensitive when comprehending anaphoric (referent-first) and cataphoric (pronoun-first) pronouns. Two groups of French speakers were recruited (young n = 18; aged 19–35 and older adults n = 15; aged 57–88) to read sentences where the anaphoric/cataphoric pronouns and their potential referents either matched or mismatched in gender. Our findings indicate that (1) the older adults were not less accurate or slower in their behavioural responses to the mismatches than the younger adults, (2) both anaphoric and cataphoric conditions evoked a central/parietally distributed P600 component with similar timing and amplitude in both the groups. Importantly, mean amplitudes of the P600 effect were modulated by verbal short-term memory span in the older adults but not in the younger adults, (3) nevertheless, the older but not the younger adults displayed an additional anterior negativity emerging on the frontal regions in response to the anaphoric mismatches. These results suggest that pronoun processing is resilient in healthy ageing individuals, but that functional recruitment of additional brain regions, evidenced with the anterior negativity, compensates for increased processing demands in the older adults’ anaphora processing.

Age-related differences in understanding pronominal reference in sentence comprehension: An electrophysiological investigation.

“…This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. pronouns, as previously reported by Arslan et al (2020). Arslan et al (2020) proposed this anterior negativity reflects additional resources recruited by older adults to compensate for increased processing demands for feature-mismatched pronouns.…”

“…Importantly, these Nref responders did not show a P600 mismatch effect, whereas those who did not show a Nref response to ambiguous pronouns continued to show a P600 to mismatched pronouns (Supplemental Figure S3). Our data thus suggest that the P600 and the anterior negativity in response to mismatched pronouns observed in Arslan et al (2020) most likely came from different subsets of older participants.…”

“…Alternatively, the anterior negativity to mismatched pronouns observed in older adults by Arslan et al (2020) may reflect an inferential process similar to the Nref mismatch effect previously seen in young adults (Nieuwland, 2014), which treats mismatched pronouns like ambiguous ones when a straightforward coreference is lacking. In view of findings from young adults suggesting that high-span individuals are more likely than low-span ones to show a Nref to mismatched pronouns (e.g., Nieuwland, 2014; Experiment 1 and 2), this negativity is likely elicited only by a subset of older adults as an alternative response to mismatched pronouns rather than a compensatory response elicited by all older adults to aid the P600 mismatch effect.…”

“…Using sentences containing pronoun-antecedent mismatches, Arslan et al (2020) and Kemmer et al (2004) found robust young-like P600 mismatch effects in older adults. Intriguingly, Arslan et al (2020) noted an additional negativity to mismatched pronouns in older adults over a subset of anterior electrode sites in the same time window as the Nref. Arslan et al (2020) proposed that this effect reflects the recruitment of additional brain activity by older adults to compensate for their lower working memory capacity.…”

Lexico-semantic access and audiovisual integration in the aging brain: Insights from mixed-effects regression analyses of event-related potentials

“…This approach is endorsed by Kulme, Vo, and Drashkow (2021) who recently developed a simulation-based method for estimating sample sizes in a statistical design such as ours, for cases where data from a sufficiently similar prior study may not be available to run a convincing data-driven estimation procedure. Our experiment's sample size compares favourably to other published work in this regard (e.g., Arslan, Palasis, & Meunier, 2020,;18 young, 15 old, 52 items per condition;Federmeier, Kutas, & Schul, 2010, , Expt 1, 16 young adults, 20 older adults, this experiment included several manipulations). Their category manipulation involved 120 cues paired with three types of exemplars, for 40 items per condition.Taken together, we therefore expected that our experiment would be sufficiently powered to compare younger and older adults on the key comparisons of interest.…”