Speaking while thinking

Abstract Solitude speech has been viewed as dialogic, especially by Western scholars ( Bakhtin [1929] 1984 ; Vygotsky [1934] 1986 ). This study investigates the prevailing belief by focusing on a ‘self-regulatory’ type of Japanese solitude speech, based on data collected in a sequential multiple-task experiment. It was revealed that Japanese self-regulatory solitude speech shows dialogic traits in terms of lexico-grammatical features and sequential structures. The results contrasted with those of our earlier research on an ‘expressive’ type of solitude speech ( Izutsu et al. 2022 ), which were found to be monologic for Japanese response cries but comparatively more dialogic in the case of their American English counterparts. This shows that expressive solitude speech is conceived as either monologic or dialogic, depending on cultural beliefs of the speakers, while self-regulatory solitude speech prefers a dialogic conception with less cultural diversity. It is also suggested that such a dialogic perspective is likely to arise from “the demands of thinking” ( Nishida [1911] 1990 ).