Graded strength of comparative illusions is explained by Bayesian inference

2025-11-18

AI Summary

Comparative illusions (CIs)—sentences like “More students have been to Russia than I have”—are systematically judged acceptable despite being logically ill-formed; yet the mechanisms underlying their perceived acceptability and the gradient nature of illusion strength remain poorly understood. Method: We develop a Bayesian noise-channel model integrating statistical language modeling with human acceptability judgments to simulate probabilistic inference during sentence comprehension. Contribution/Results: This is the first computational model to quantitatively predict CI strength gradients. It reveals that pronominal subjects robustly enhance illusion strength, whereas full noun phrases attenuate it—a novel modulating factor not previously identified. The model precisely replicates established experimental effects while uncovering this new syntactic-semantic interface effect. These findings provide strong empirical support for the applicability of noise-channel theory to phenomena at the syntax–semantics interface, advancing our understanding of how probabilistic inference shapes grammaticality judgments.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Natural Language Processing: Lexical Semantics and Morphology
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Like visual processing, language processing is susceptible to illusions in which people systematically misperceive stimuli. In one such case--the comparative illusion (CI), e.g., More students have been to Russia than I have--comprehenders tend to judge the sentence as acceptable despite its underlying nonsensical comparison. Prior research has argued that this phenomenon can be explained as Bayesian inference over a noisy channel: the posterior probability of an interpretation of a sentence is proportional to both the prior probability of that interpretation and the likelihood of corruption into the observed (CI) sentence. Initial behavioral work has supported this claim by evaluating a narrow set of alternative interpretations of CI sentences and showing that comprehenders favor interpretations that are more likely to have been corrupted into the illusory sentence. In this study, we replicate and go substantially beyond this earlier work by directly predicting the strength of illusion with a quantitative model of the posterior probability of plausible interpretations, which we derive through a novel synthesis of statistical language models with human behavioral data. Our model explains not only the fine gradations in the strength of CI effects, but also a previously unexplained effect caused by pronominal vs. full noun phrase than-clause subjects. These findings support a noisy-channel theory of sentence comprehension by demonstrating that the theory makes novel predictions about the comparative illusion that bear out empirically. This outcome joins related evidence of noisy channel processing in both illusory and non-illusory contexts to support noisy channel inference as a unified computational-level theory of diverse language processing phenomena.

Problem

• Explaining graded strength of comparative illusions through Bayesian inference
• Modeling posterior probability of interpretations using language models
• Testing noisy-channel theory predictions for sentence comprehension illusions

Innovation

• Bayesian inference models sentence comprehension as noisy channel
• Statistical language models predict comparative illusion strength quantitatively
• Model explains gradations using human behavioral data synthesis