A pronoun like he or she might refer to dozens of different people, just in the course of reading a few news articles. How do our brains keep all of this straight? Keep reading to find out.
If I say, “My sister loves Taylor Swift. She knows all her songs and all the outfits she’s worn on tour”, it sounds pretty innocuous, but linguistically-speaking something remarkable happens. In just one sentence, she goes from being my sister to being Taylor Swift.
This is the magic of pronoun resolution, when the interpretation of a pronoun gets linked to a particular referent. As discussed in Pronouns in Bio, Pt. 1, part of what helps this process happen is that the pronoun and referent share a few features in common, like number and gender.
Figure 1. Taylor Swift still likes country music.
Who still likes country music?
When scientists record electrical activity from the brain using electroencephalography (EEG), they see a couple of characteristic responses when pronoun resolution gets difficult. Much like how the knee reflex is a characteristic response to tapping right below the kneecap, these responses are part of healthy nervous system function and not under conscious control. For studying pronouns with EEG, the pronoun is like the tap below the kneecap, and the characteristic response is a complex set of changes in the brain’s electrical activity that scientists break down into components.
Figure 2. Nref and P600 responses can be seen by comparing anomalous sentences to a baseline sentence that contains no ambiguity or feature mismatches (black). The plot shows the brain’s response to each bolded pronoun, recorded from one of the many electrodes on the participant’s scalp. The teal line shows the response recorded to the teal, ambiguous sentence, and the orange line shows the response recorded to the orange sentence containing a feature mismatch. The shaded areas highlight the difference that means an Nref response (teal) or P600 response (orange) is happening.
For pronoun resolution, one of these components is the Nref, a negative shift in the electrical signal recorded over the front of the head when participants hear a pronoun with an ambiguous referent (van Berkum et al., 1999). The she in a sentence like “When we last talked about Taylor Swift, my sister said that she still likes country music” would likely elicit an Nref response. This is because it could refer to either Taylor Swift or my sister, even if you think it’s more likely to refer to my sister.
Another component of the characteristic response is the P600, a positive shift in the electrical signal that occurs when the features of a pronoun don’t seem to match those of any referent we’ve heard so far (Osterhout & Mobley 1995). This kind of response would likely occur in a sentence like, “When we last talked about Taylor Swift, my sister said that he still likes country music”. In this case, the pronoun’s gender and number features don’t seem to match any referent we’ve heard.
If you’ve read Pronouns in Bio, Pt. 2, you may also wonder what happens in someone’s brain when a new pronoun is introduced in a language, like the Swedish gender neutral pronoun hen. There aren’t any EEG studies on hen yet, but there has been one on a similar pronoun in English.
What’s up with them?
In English, the pronouns they, them, and theirs most typically refer to plural referents, as in “Three people left their coats behind”, but they can also refer to singular referents. The use of they to talk about a single person of unknown or unimportant gender has a long history in English, going back hundreds of years (Balhorn 2004). This is the kind of they in a sentence like “Someone left their coat behind”. In contrast, the use of they to refer to a specific, named person, sometimes called ‘non-binary they’, as in “Sam left their coat behind”, is more recent.
When people are asked to rate how natural different uses of they sound, uses of they like “Sam left their coat behind”, tend to be rated lower than uses of they to refer to plural referents, such as “the dentists”, or unnamed people, such as “my friend” (Camilliere et al. 2021). However, these judgments vary by participants’ age and familiarity with non-binary people. Younger people accept singular they in a wider variety of contexts, as do people who report that they have more experience with and positive feelings towards non-binary people (Camilliere et al. 2021).
Interestingly though, when electrical activity from the brain is recorded, non-binary they elicits both an Nref response and a P600 response relative to plural they, regardless of whether the participant rated the naturalness of non-binary they highly (Chen et al. 2021). This suggests that processing non-binary they involves some neural processing that plural they does not. In particular, the elicitation of an Nref response suggests that the response to non-binary they has something in common with the response to an ambiguous pronoun, and the P600 response suggests that difficulty with feature matching may play a role.
What does it all mean?
There are a couple of possible reasons why participants who accept sentences with non-binary they still show neural signs of extra processing effort when they encounter it.
One is the inherent ambiguity of they: the referent of they can be plural or singular, definite or indefinite, depending on the kind of they being used. This increases the likelihood of referential ambiguities that trigger the neural processes responsible for an Nref response. The ambiguity also increases the likelihood that a participants’ first interpretation of the sentence results in a feature mismatch that would trigger the neural processes responsible for a P600 response.
Another factor is familiarity. Participants probably have less experience with non-binary they compared to more common pronouns. This would mean that even though some participants find it highly acceptable, non-binary they could still require more effort for the brain to process than other pronouns. But since familiarity changes with time, this raises an interesting question: Do these neural responses change too?
To learn is to change
Research on learning new languages shows that learners can develop a P600 response to grammatical feature mismatches in a new language as they learn the language’s grammar (White et al. 2012, Gabriele et al. 2021). Some researchers hypothesize that the converse should be true too: The P600 response should wane if something that used to be a feature mismatch becomes a match. If this prediction is correct, an individual’s P600 response to non-binary they should diminish over time, as they gain more experience with it and it becomes more grammatical for them.
Chen et al. (2021) provide some evidence to support this. All participants in their study were American undergraduate students (18~22 years old) at a school where “every student is introduced to preferred pronouns, taught about nonbinary gender identities, and encouraged to provide their preferred pronouns as part of orientation.” Students in the study who were older, who had been in that environment longer, had smaller P600 responses to non-binary they.
This suggests that exposure to and practice with non-binary they in a supportive environment contributes to changes in brain activity similar to those that occur when learning a second language. In other words, when we learn new grammar, whether in our own language or a new one, the brain is engaging similar processes.
Research on this topic is still developing, and more data will be necessary to fully understand how pronouns get their meaning, old and new alike. Data from languages like Swedish, whose single gender-neutral pronoun hen is distinct from other pronouns in the language, could help with this. Does hen elicit Nref and P600 responses? We can’t wait to find out.
References
Balhorn, M. (2004). The rise of epicene they. Journal of English Linguistics, 32(2), 79-104.
Camilliere, S., Izes, A., Leventhal, O., & Grodner, D. (2021). They is changing: Pragmatic and grammatical factors that license singular they. In Proceedings of the Annual Meeting of the Cognitive Science Society (Vol. 43, No. 43).
Chen, P., Leventhal, O., Camilliere, S., Izes, A. and Grodner, D. (2021). Singular they in transition: ERP evidence and individual differences. In 34th Annual CUNY Conference on Human Sentence Processing.
Gabriele, A., Alemán Bañón, J., Hoffman, L., Covey, L., Rossomondo, A., & Fiorentino, R. (2021). Examining variability in the processing of agreement in novice learners: Evidence from event-related potentials. Journal of Experimental Psychology: Learning, Memory, and Cognition, 47(7), 1106.
Osterhout, L., & Mobley, L. A. (1995). Event-related brain potentials elicited by failure to agree. Journal of Memory and language, 34(6), 739-773.
Van Berkum, J. J., Brown, C. M., & Hagoort, P. (1999). Early referential context effects in sentence processing: Evidence from event-related brain potentials. Journal of memory and language, 41(2), 147-182.
White, E. J., Genesee, F., & Steinhauer, K. (2012). Brain responses before and after intensive second language learning: Proficiency based changes and first language background effects in adult learners. PloS one, 7(12), e52318.
