On the etymology of the Japanese plural suffix and its possible connection to Korean

This paper presents an etymological analysis of the Japanese plural suffix tachi, Old Japanese tati. I propose that tati originates from a grammaticalization of an earlier Pre-Old Japanese phonological form *totwi, the non-bound reflex of which is the Old Japanese quasi-collective marker dwoti ‘fellow (person), everyone, together’. The reconstruction of a Pre-Old Japanese stem *totwi (Pre-Proto-Japanese /*tətəj/) with quasi-collective and plural function clarifies the possible connection of the Japanese plural suffix to the Korean plural suffix tul (Middle Korean tólh), which Whitman (1985, p. 217) proposed to be cognates but which has since been criticized on phonological and distributional grounds. I show that reconstructing the earliest form of the Japanese plural suffix as /*tətəj/ resolves each of the three phonological issues with the Japano-Koreanic comparison, creates a better morphosyntactic match between the two languages, and rules out a loanword relationship of the Japanese and Korean forms.