Do people routinely pre-activate the meaning and even the phonological form of upcoming words? The most acclaimed evidence for phonological prediction comes from a 2005 Nature Neuroscience publication by DeLong, Urbach and Kutas, who observed a graded modulation of electrical brain potentials (N400) to nouns and preceding articles by the probability that people use a word to continue the sentence fragment (‘cloze’). In our direct replication study spanning 9 laboratories (N=334), pre-registered replication-analyses and exploratory Bayes factor analyses successfully replicated the noun-results but, crucially, not the article-results. Pre-registered single-trial analyses also yielded a statistically significant effect for the nouns but not the articles. Exploratory Bayesian single-trial analyses showed that the article-effect may be non-zero but is likely far smaller than originally reported and too small to observe without very large sample sizes. Our results do not support the view that readers routinely pre-activate the phonological form of predictable words.

Composing sentence meaning is easier for predictable words than for unpredictable words. Are predictable words genuinely predicted, or simply more plausible and therefore easier to integrate with sentence context? We addressed this persistent and fundamental question using data from a recent, large-scale ( n = 334) replication study, by investigating the effects of word predictability and sentence plausibility on the N400, the brain's electrophysiological index of semantic processing. A spatio-temporally fine-grained mixed-effect multiple regression analysis revealed overlapping effects of predictability and plausibility on the N400, albeit with distinct spatio-temporal profiles. Our results challenge the view that the predictability-dependent N400 reflects the effects of either prediction or integration, and suggest that semantic facilitation of predictable words arises from a cascade of processes that activate and integrate word meaning with context into a sentence-level meaning. This article is part of the theme issue ‘Towards mechanistic models of meaning composition’.

In current theories of language comprehension, people routinely and implicitly predict upcoming words by pre-activating their meaning, morpho-syntactic features and even their specific phonological form. To date the strongest evidence for this latter form of linguistic prediction comes from a 2005 Nature Neuroscience landmark publication by DeLong, Urbach and Kutas, who observed a graded modulation of article- and noun-elicited electrical brain potentials (N400) by the pre-determined probability that people continue a sentence fragment with that word ('cloze'). In a direct replication study spanning 9 laboratories (N=334), we failed to replicate the crucial article-elicited N400 modulation by cloze, while we successfully replicated the commonly-reported noun-elicited N400 modulation. This pattern of failure and success was observed in a pre-registered replication analysis, a pre-registered single-trial analysis, and in exploratory Bayesian analyses. Our findings do not support a strong prediction view in which people routinely pre-activate the phonological form of upcoming words, and suggest a more limited role for prediction during language comprehension.

Composing sentence meaning is easier for predictable words than for unpredictable words. Are predictable words genuinely predicted, or simply more plausible and therefore easier to integrate with sentence context? We addressed this persistent and fundamental question using data from a recent, large-scale (N = 334) replication study, by investigating the effects of word predictability and sentence plausibility on the N400, the brain's electrophysiological index of semantic processing. A spatiotemporally fine-grained mixedeffects multiple regression analysis revealed overlapping effects of predictability and plausibility on the N400, albeit with distinct spatiotemporal profiles. Our results challenge the view that the predictability-dependent N400 reflects the effects of either prediction or integration, and suggest that semantic facilitation of predictable words arises from a cascade of processes that activate and integrate word meaning with context into a sentence-level meaning. Keywords Predictability; Plausibility; Semantic Similarity; N400Composing sentence meaning is easier with predictable words than with unpredictable words: for example, 'bicycle' is easier to process than 'elephant' in "You never forget how to ride a bicycle/an elephant once you've learned." This effect of predictability can be observed in behavioural measures of comprehension such as reading times [1] and on amplitude modulation of the N400 ERP component [2]. The N400 is a negative-going and centroparietally distributed component of the ERP, which occurs approximately 200-500 ms after word onset and that is strongly associated with lexico-semantic processing [2-3]. Predictable words elicit smaller N400 amplitude than unpredictable words, suggesting facilitation during semantic processing. However, to what extent is such an effect of predictability driven by other relationships between a word and its context, such as the plausibility of the described event? We addressed this issue by re-analysing data from a large-scale (N=334) replication study [4]. In a temporally fine-grained analysis, we explored whether predictability, plausibility, and semantic similarity have dissociable effects on word-elicited ERP activity and how these effects unfold over time.Word predictability is commonly operationalized as 'cloze probability', the probability of being used in a non-speeded, offline sentence completion test. The correlation between word predictability and N400 amplitude is well-established [5], with some studies reporting correlations as high as or even higher than r = .8 [2,[6][7]. Such results are often considered to demonstrate effects of prediction: people pre-activate a word fully before it appears (as when a specific lexical form can be strongly anticipated in a highly constraining context) or partially (as when some semantic features are activated due to passive spreading of information). Prediction facilitates the semantic activation processes that are initiated when the word appears and are reflected in N400 activity [5].

Research on reading comprehension assessments suggests that they measure overlapping but not identical cognitive skills. In this paper, we examined the potential of eye-tracking as a tool for assessing reading comprehension. We administered three widely-used reading comprehension tests with varying task demands to 79 typical adult readers while monitoring their eye movements. In the York Assessment for Reading Comprehension (YARC), participants were given passages of text to read silently, followed by comprehension questions. In the Gray Oral Reading Test (GORT-5), participants were given passages of text to read aloud, followed by comprehension questions. In the sentence comprehension subtest of the Wide Range Achievement Test (WRAT-4), participants were given sentences with a missing word to read silently, and had to provide the missing word (i.e., a cloze task). Results from linear models predicting comprehension scores from eye-tracking measures yielded different patterns of results between the three tests. Models with eye-tracking measures always explained significantly more variance compared to baseline models with only reading speed, with R-squared 4 times higher for the YARC, 3 times for the GORT, and 1.3 times for the WRAT. Importantly, despite some similarities between the tests, no common good predictor of comprehension could be identified across the tests. Overall, the results suggest that reading comprehension tests do not measure the same cognitive skills to the same extent, and that participants adapted their reading strategies to the tests’ varying task demands. Finally, this study suggests that eye-tracking may provide a useful alternative for measuring reading comprehension.