When a Man Says He Is Pregnant: Event-related Potential Evidence for a Rational Account of Speaker-contextualized Language Comprehension

Spoken language is often, if not always, understood in a context formed by the identity of the speaker. For example, we can easily make sense of an utterance such as “I’m going to have a manicure this weekend” or “The first time I got pregnant I had a hard time” when spoken by a woman, but it would be harder to understand when it is spoken by a man. Previous ERP studies have shown mixed results regarding the neurophysiological responses to such speaker-content mismatches, with some reporting an N400 effect and others a P600 effect. In an EEG experiment involving 64 participants, we used social and biological mismatches as test cases to demonstrate how these distinct ERP patterns reflect different aspects of rational inference. We showed that when the mismatch involves social stereotypes (e.g., men getting a manicure), listeners can arrive at a “literal” interpretation by integrating the content with their social knowledge, though this integration requires additional effort due to stereotype violations-resulting in an N400 effect. In contrast, when the mismatch involves biological knowledge (e.g., men getting pregnant), a “literal” interpretation becomes highly implausible or impossible, leading listeners to treat the input as potentially containing errors and engage in correction processes-resulting in a P600 effect. Supporting this rational inference framework, we found that the social N400 effect decreased as a function of the listener’s personality trait of openness (as more open-minded individuals maintain more flexible social expectations), while the biological P600 effect remained robust (as biological constraints are recognized regardless of individual personalities). Our findings help to reconcile empirical inconsistencies and reveal how rational inference shapes speaker-contextualized language comprehension.

💡 Research Summary

The paper investigates how listeners incorporate speaker identity into real‑time language comprehension and why previous ERP studies have reported either an N400 or a P600 response to speaker‑content mismatches. The authors propose a rational inference framework: listeners first attempt a literal interpretation of an utterance; if integration with speaker knowledge is merely effortful, an N400 is elicited, whereas if a literal interpretation is implausible, listeners treat the input as potentially erroneous and engage a correction process reflected in a P600.

To test this, 64 participants heard sentences that were either socially stereotypical violations (e.g., a man saying “I’m going to have a manicure this weekend”) or biologically impossible statements (e.g., a man saying “The first time I got pregnant I had a hard time”). Speaker identity was conveyed via a photo and voice, ensuring that the same linguistic content could be paired with a male or female speaker. EEG was recorded, focusing on the 300‑500 ms N400 window and the 600‑900 ms P600 window across frontal and central electrodes.

Results showed a robust N400 for the social mismatch condition. Importantly, the amplitude of this N400 decreased as a function of participants’ Openness (a personality trait from the Big Five), indicating that more open‑minded listeners have more flexible social expectations and thus experience less integration difficulty. In contrast, the biological mismatch condition produced a strong P600, with no modulation by Openness, suggesting that biological constraints are treated as absolute and trigger an error‑correction mechanism regardless of individual differences.

The authors argue that these findings reconcile earlier contradictory ERP reports. Prior studies that observed N400s likely involved mismatches that could be reconciled through flexible social knowledge, whereas studies reporting P600s involved violations that render a literal interpretation impossible, prompting a revision process. The rational inference account thus positions the N400 as an index of probabilistic semantic integration and the P600 as an index of post‑lexical monitoring and repair when the input is judged noisy or erroneous.

The discussion highlights theoretical implications for models of language processing that emphasize context integration, such as the “noisy channel” view of comprehension. It also underscores the role of individual differences: personality traits can shape how readily social expectations are updated, but not how biological facts are processed. Limitations include the cultural specificity of the stimuli, the focus on only two types of mismatches, and the absence of other personality dimensions. Future work is suggested to test the framework across diverse languages, additional speaker attributes (age, socioeconomic status), and to combine ERP with behavioral measures of correction.

In sum, the study provides compelling neurophysiological evidence that speaker identity is an integral part of the mental model listeners use during comprehension, and that the brain flexibly employs either an N400‑driven integration effort or a P600‑driven correction process depending on whether the mismatch is socially negotiable or biologically untenable. This advances our understanding of how contextual information and rational inference shape real‑time language processing.