Listener age and perceived ethnicity in speech adaptation: Evidence from child- and foreigner-directed registers

This study examines how speakers adapt their speech based on listener age and perceived ethnicity, focusing on two socially modulated registers: Child-Directed Speech (CDS) and Foreigner-Directed Speech (FDS). Twelve native English speakers completed read and spontaneous speech tasks directed toward four listener types, represented by images: Asian Adult, Asian Child, White Adult, and White Child. Drawing on models of social cognition, audience design, and Lindblom’s Hyper- and Hypoarticulation framework, the study analyzed phonetic and lexical adaptations across conditions. Results revealed robust age-driven modifications: speakers produced slower speech, expanded vowel spaces, and reduced lexical diversity when addressing children—hallmarks of CDS. In contrast, perceived ethnicity had minimal effects on speech patterns, suggesting that visual cues alone may not strongly elicit FDS features. However, a significant interaction in lexical diversity emerged: White child listeners elicited higher type-token ratios than Asian child listeners. These findings suggest that listener age exerts a stronger and more consistent influence on speech adaptation than perceived ethnicity. They further support the view that speakers construct listener models based on intersecting social cues. The results have implications for theories of register, adaptive speech production, and listener-sensitive communication in both human and machine contexts.