Our hearts may indeed “skip a beat” when we hear grammatical errors in speech. In a new study published in the Journal of Neurolinguistics, researchers discovered that our heart rate variability, a measure of the heartbeat’s rhythm, changes in response to grammatical errors in speech. This finding suggests there is a deep and measurable connection between our physiological responses and our implicit understanding of language.
Scientists have long been fascinated by how we process language – an ability that often operates below our conscious radar. Understanding this process is crucial, not just for academic curiosity, but for practical applications like designing effective learning tools or diagnosing language-related cognitive issues. Until now, researchers largely relied on direct questions or tasks to gauge language understanding. However, these explicit methods can be skewed by biases or the individual’s ability to articulate their knowledge.
Recognizing the limitations of explicit measures, the authors of the new study turned their attention to implicit measures, which infer knowledge and cognitive processes from indirect, performance-based tasks.
“On the one hand, there was the challenge of developing a reliable method that can be used with a wide range of populations, to tap into aspects of cognition that are not directly observable, or where there is a need to circumvent the requirement to provide an explicit evaluative judgment, e.g., in work with populations that cannot yet or can no longer express their opinion, due to (young/old) age or ill health (be it physical or cognitive/mental),” explained study author Dagmar Divja (@dagmardivjak), a professor at the University of Birmingham and author of “Frequency in Language.”
“On the other hand, there was my hunch that some people (including myself!) get annoyed by grammar errors more than others, and I started to wonder whether this would have a measurable physical manifestation and whether we could capture this. We knew that our pupils react to ‘stress’: when something is scary or difficult, our pupils become larger. Heart rate is regulated by the same autonomic nervous system, so it could be expected that we’d find an effect in heart rate too.”
For their study, the researchers recruited 41 native British English speakers, aged 18 to 44. These individuals were carefully selected to ensure a range of educational backgrounds and familiarity with foreign accents, but without any known learning difficulties or cardiac rhythm issues. Participants listened to 40 short audio samples of English speech. Half of these samples contained grammatical errors specifically related to the use of articles, such as “a”, “an”, and “the”. To add complexity and realism, the speech samples were presented in different accents (British and Polish) and spoken by both male and female voices.
To create these speech samples, the researchers used transcripts from interviews on common topics. They then deliberately introduced errors in half of these transcripts. For example, in some instances, articles were incorrectly added or omitted. The error density in these samples varied from 18% to 56%, meaning some samples had more grammatical mistakes than others.
While the participants listened to these samples, their cardiovascular activity, particularly their heart rate variability, was closely monitored. This measure reflects the variation in time intervals between heartbeats and is influenced by the autonomic nervous system, the part of our nervous system that controls bodily functions like heart rate and digestion.
Divjak and her colleagues found that when participants listened to speech containing grammatical errors, their heart rate variability changed noticeably. There was a significant decrease in heart rate variability in response to speech containing grammatical errors. This decrease was more pronounced when the error density in the speech was between 20% and 40%.
Interestingly, the decrease in heart rate variability was greater when the grammatical errors were made by native English speakers rather than non-native speakers. This suggests that errors made by native speakers were less expected and thus had a greater impact on the listeners’ physiological responses.
“Our pupils react to ‘stress’: when something is scary or difficult, our pupils become larger,” Divjak said. “Heart rate is regulated by the same autonomic nervous system, so it could be expected that we’d find an effect in heart rate too. But showing that in practice was quite another matter!”
“It was fascinating to see how many dimensions of our experience we really capture, even though we are not aware of them: in this case people really showed sensitivity to ‘small’ language errors and also knew that these errors were more unusual for native speakers that for foreign speakers.”
The study bridges the gap between physiological responses and cognitive functions. By showing that our heart rate can be linked to linguistic processing, it supports the idea that cognitive activities, even those as complex as language processing, have physical manifestations in the body.
As Divja told PsyPost: “Our findings bring into focus two things: 1.) a new dimension of the intricate relationship between physiology and cognition, suggesting that cognitive effort reverberates through the physiological system in more ways than previously thought, and 2.) a better idea of how many dimensions of our experience we really capture, how specific our expectations are, even though we are not aware of them. Accurately assessing an individual’s linguistic abilities, regardless of their age and physical or cognitive abilities, is important for many questions pertaining to core areas of life relating to cognition, including brain health.”
But the study, like all research, includes some limitations. First, the study focused on a specific aspect of language — grammatical errors in spoken language — and involved a relatively small and homogenous group of participants. Future studies could expand the scope to include a more diverse range of linguistic aspects and a broader participant demographic. To fully understand the universality of these findings, it would be beneficial to replicate the study across different languages and cultural backgrounds.
“As is always, the case with ‘first’ discoveries, it would be good to see more work using this technique, in a variety of settings involving language,” Divja said. “We also hope that our method will be taken up by researchers and clinicians who are interested in language cognition and work with populations that are difficult to reach or test to facilitate accurate assessment and ensure better outcomes.”
Despite these limitations, the study opens up exciting new avenues for research. It suggests that cardiovascular measures, particularly heart rate variability, could be valuable tools for assessing linguistic knowledge implicitly.
“From an applied perspective, we are now working on establishing whether the reactions we recorded are in any way related to personality and would need to be taken into account in high-stakes real-life situations,” Divjak explained. “For example, if we find that conscientious people react more strongly to language errors, would that mean they are more negatively inclined towards people who make errors, and could that mean, for example, that they would be less likely to hire a foreign applicant if they are interviewing candidates for positions?”
“The research was funded by the Leverhulme Trust via a Research Leadership Award to Dagmar Divjak and was carried out at the University of Birmingham,” she added.
The study, “Physiological responses and cognitive behaviours: Measures of heart rate variability index language knowledge“, was authored by Dagmar Divjak, Hui Sun, and Petar Milin.
