Text difficulty modulates the surprisal effect in self-paced reading
Bibliographic Data
| ID | 22417153 |
|---|---|
| Authors | Lin Chen (0000-0001-5135-3998), Lin Yuan Chen (University of Illinois Urbana-Champaign, corresponding author), Gaisha Oralova (0000-0003-0531-9451, University of Pittsburgh), Xiaoping Fang (0000-0003-3410-277X, Beijing Language and Culture University), Shannon Clark (0000-0002-2600-8145, University of Alberta), Daniela Teodorescu (University of Alberta), Maxwell Helfrich, Maxwell R Helfrich (University of Pittsburgh), Alona Fyshe (0000-0003-4367-0306, University of Alberta), Carrie Demmans Epp (0000-0001-9079-4921, University of Alberta), Charles Perfetti (0000-0002-0211-8518, University of Pittsburgh) |
| Year | 2026 |
| Publication date | 2026-02-11 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Reading and Writing (JOURNAL) |
| Journal identifiers | ISSN: 0922-4777 • E-ISSN: 1573-0905 |
| Publisher | Springer Science and Business Media LLC (PUBLISHER) |
| DOI | 10.1007/s11145-026-10768-7 |
| OpenAlex | W7128521667 |
| Language | EN |
| References cited | 64 |
Surprisal, the statistical likelihood of a word or syntactic structure given its preceding context, is an important metric for understanding the predictive/integrative process in reading. Although individual words and syntactic structures vary in surprisal within any text, regardless of its overall difficulty, it remains unclear whether text difficulty modulates how surprisal influences reading. We hypothesized that higher text difficulty increases readers’ reliance on contextual information during reading, leading to stronger surprisal effects. To test this hypothesis, we examined how text difficulty modulates the strength of word and syntactic surprisal effects and whether this modulation generalizes across computational language models with different architectures. We compared reading times for two sets of texts that differed in difficulty but were matched on average surprisal. Results showed larger word and syntactic surprisal effects in more difficult texts, consistently across models. These results suggest that the predictive/integrative processes are shaped by global text properties.
Cognition · Psycholinguistics · Syntax · Reading and Literacy Development · Text Readability and Simplification · Writing and Handwriting Education
The effect of word predictability on reading time is logarithmic
Lexical complexity and fixation times in reading
Coh-Metrix
Data from eye-tracking corpora as evidence for theories of syntactic processing complexity
Moving beyond Kučera and Francis
Updating a mental model
What do we mean by prediction in language comprehension?
The lexical nature of syntactic ambiguity resolution.
Linguistic complexity
Word Knowledge in a Theory of Reading Comprehension
Derivation of New Readability Formulas (Automated Readability Index, Fog Count and Flesch Reading Ease Formula) for Navy Enlisted Personnel
Reading Ability
The Brain Basis of Language Processing
Expectation-based syntactic comprehension
Probabilistic word pre-activation during language comprehension inferred from electrical brain activity
Random effects structure for confirmatory hypothesis testing
Fitting Linear Mixed-Effects Models Using lme4
LmerTest Package
Eye Movement Traces of Linguistic Knowledge in Native and Non-Native Reading
Prediction in reading
Large-scale benchmark yields no evidence that language model surprisal explains syntactic disambiguation difficulty
Probabilistic Top-Down Parsing and Language Modeling
Younger and Older Adults' "Good-Enough" Interpretations of Garden-Path Sentences
Assessing Readability
Coh-Metrix
The late frontal positivity reflects incremental mental model updating
Information-theoretical Complexity Metrics
Processing Advantages of Lexical Bundles
| Citation velocity | historical |
|---|---|
| Highly cited | No |