Older adults adapt their lexical choices based on their beliefs about their interlocutor’s age.

Effective communication often requires speakers to adapt their lexical choices by reusing (entraining to) their conversational partner’s lexical choices, even if those words are not usually favoured (e.g. using “spectacles” over “glasses” in the UK). Lexical adaptation may also include maintaining use of these disfavoured words with future partners. While our preliminary research has shown that lexical entrainment and maintenance of entrained terms increase with age, little is known about the cognitive mechanisms driving these differences. The present (OSF-preregistered) research investigates age-related effects on lexical entrainment and maintenance of entrained terms during dialogue, specifically looking at how these phenomena are influenced by speakers’ age and their beliefs about their partners’ age.
Experiment 1 used a 2(Participant’s age: old vs. young) X 2(Partner 1’s age: old vs. young) X 2(Partner 2’s age: same vs. different age to Partner 1) design. 160 older (60-79 years old) and 160 younger (18-39 years old) adults completed two sessions of an online picture-matching-and-naming task with different (actually computer-simulated) partners in each session. Participants learned about both partners’ ages at the start of the task. Then in each session, they took turns with the partner to name objects that had a disfavoured and a favoured name. Entrainment was measured in session 1 as participants’ tendency to reuse disfavoured terms used by Partner 1; maintenance was measured in session 2 as participants’ tendency to reuse entrained terms (from Session 1) with Partner 2.
Older speakers were significantly more likely to entrain and maintain disfavoured terms than younger speakers (entrainment: 57%[30%] vs. 28%[26%] - see figure 1; maintenance: 46%[32%] vs. 18%[22%] - see figure 2). Crucially, while younger speakers’ entrainment and maintenance was not affected by (their beliefs about) either partner’s age, older speakers’ entrainment was significantly influenced by both partners’ ages: they were more likely to entrain to an older than a younger Partner 1 (64%[30%] vs. 50%[30%]; z=-2.11, p=.03). They also entrained more often to Partner 1 when anticipating later interaction with a Partner 2 from the same age-group than different age-group as Partner 1 (63%[30%] vs. 52%[30%]; z=-2.66, p=.008). Experiment 2 replicated the effect of Partner 1’s age on older adults’ entrainment with another group of 160 older adults. We found null effects of Partner 2’s age on entrainment when this information was undisclosed during session 1, confirming that our Exp1’s finding indeed reflects audience design effects (and was not due to a measurement error).
These results suggest that older (though not younger) speakers adapt their lexical choices based on their partner’s age, both in present and anticipated future interactions. This supports audience design effects in older adults’ entrainment. We also discuss other potential explanations, including age-related differences in executive functions.