Perturbating factors: affectedness and grammatical aspect

Certain factors seem to interact with the semantic basis for class membership. Mithun (1991:520–3,531–2) describes how, in Central Pomo, Caddo, and Mohawk, states that do not significantly affect their participants end up in the S_A category, despite their lack of agentive properties; for instance, inherent-state predicates, such as being tall, strong, good, big, or alive. States in which the participant is significantly affected, either eventive ones such as falling, dying, or slipping, or patientive, affecting, non-inherent-state ones, such as being cold, afraid, or sleepy, are in the S_O category.

 The other important factor is aspect. In Mohawk, intransitive verbs inflected for perfect aspect take S_O person markers, regardless of other semantic factors (Mithun 1991:532–3). Thus, a S_A intransitive verb that takes S_A person markers in the habitual present and punctual (perfective) forms has to take S_O person markers in the perfect form; speakers have no choice. This can be seen in the following Mohawk example(from Mithun 1991:533; person markers are in bold):

(1)

Neither interfering factor is relevant in Cariban languages. Affectedness concerns mainly states, i.e. potential S_O intransitives. As we have seen, most (if not all) states in Cariban languages are not verbs, regardless of the level of affectedness, e.g. Tiriyó kure nai məe ‘s/he is good/well’, kawə nai məe ‘s/he is tall’ (non affecting states), nari ke nai məe ‘s/he is afraid’, əənGhpo nai məe ‘s/he is sleepy’ membership. Grammatical aspect does not cause any perturbations either: all finite tenses consistently take the same person markers, as illustrated in the Tiriyó paradigms in 2.

(2)