17.05.2022 Lena Jäger

Theories of human sentence processing generally assume that the cognitive mechanisms involved in language processing are qualitatively identical not only across speakers of a language but even cross-linguistically. However, within the history of cognitive science, this assumption of an idealized human cognition has not remained uncontested and several researchers have pointed out that individual differences should play a more prominent role in theories explaining human cognition (e.g., Estes 1956; Levinson, 2012). Indeed, over the past decade, evidence has accumulated indicating that individual differences in a comprehender’s cognitive capacities play an important role in sentence processing (e.g., Vuong & Martin, 2014; Nicenboim et al., 2015; Farmer et al., 2017). Although these studies demonstrate the impact of individual differences on sentence processing, it is still not well understood whether individual differences in psycholinguistic effects reflect quantitative or qualitative differences in the way individuals process language, and whether the observed individual differences are inherent characteristics of an individual and are hence stable over time, or whether they rather reflect temporal cognitive states or processing strategies at the moment of testing.

From a methodological point of view, the first step for a principled investigation of individual differences in sentence processing is to establish the test-retest measurement reliability of individual differences in theoretically relevant psycholinguistic effects (Parsons et al., 2019). We can’t take this measurement reliability as a given because of the so-called reliability paradox (Hedge, 2017). It states that test-retest measurement reliability at the individual level is necessarily lower for manipulations with high between-subjects reliability (i.e., replicability at the group level). However, it is likely that precisely effects with high replicability at the group level constitute the set of well-established psycholinguistic phenomena which build the foundation of sentence processing theories.

In this talk, I will present ongoing work in which we investigate the test-retest measurement reliability of individual differences in cognitive capacities on a range of theoretically relevant phenomena of sentence processing. I will present a large-scale data collection in which we collect eye-tracking and self-paced reading data on naturalistic texts from a total of 200 participants, each of them participating in four recording sessions. Besides the behavioral measurements, we administer a comprehensive battery of psychometric tests assessing the participants’ cognitive capacities (including verbal and non-verbal working memory capacity and cognitive control, IQ, and reading skills) as well as a vigilance task to measure their momentary level of wakefulness. I will present preliminary results from a Bayesian data analysis investigating the temporal as well as the cross-methodological measurement reliability of predictability, syntactic locality and integration effects and their interaction with an individual’s cognitive capacities.