Getting on the modality ladder: event-oriented modal
markers between voice and modality

Heiko Narrog
Tohoku University

The evolution of grammatical categories and their mutual relationship in synchronic and diachronic perspective has been a topic of particular interest for the past fifteen years or so. In this presentation I want to put the spotlight on the relationship between two categories that has not received much attention yet, namely modality and voice. This is done so on the background of a model of modality which has two dimensions, namely volitivity on the one hand and event-orientation vs. speaker-orientation on the other hand. There I claim that it is in particular the event-oriented, "low-level" modalities such as participant-internal possibility or necessity which show affinity to voice, or voice-like phenomena, i. e. non-canonical marking of participants.

The point of departure for my investigation is Japanese. In Japanese linguistics, the markers of participant-internal possiblity, -(r)are.ru and - (r)e.ru are not even recognized as part of modality, the reason being that they are derived from a voice construction and still can optionally take non-canonical marking. However, there is incidental evidence from other languages that these are not isolated cases. This led me to conduct a survey of potential and necessity markers in 200+ languages the results of which I will present. I will further argue that there is a conceptual foundation for the fact that it is the event-oriented modalities that show non-canonical marking, and not the speaker-oriented modalities.