What pressures lead languages to be similar to each other, and what pressures lead languages to differ? How do these pressures affect learning, and what is the structure of the linguistic knowledge children will ultimately acquire? I address these questions in the following three lines of research:
Fillmore influentially proposed that all languages represent event participants in terms of the same primitive case roles, such as Agent, Patient and Instrument. But are these categories actually universal, and how are they represented in the mind? I have found that deaf children encode the category of Agent in their gestures even when they are not learning an established language. Nonetheless, my work on English shows that there is no single coherent category Instrument encoded by the instrumental markers with and use. But there may still be a prototype concept of tool use that guides semantic organization across languages. I am conducting a typological survey as well as experimental study of Germanic languages to find out.
Relevant papers: Rissman & Majid (2019), Rissman & Goldin-Meadow (2017), Rissman & Rawlins (2017), Rissman (2013), Rissman (2011)
Flexible construal of events
Imagine a causative event such as a cat knocking over a glass of water. We can use language to describe who did what to whom, but we can also convey alternate construals of the event, e.g.: the cat knocked over the water vs. the water was knocked over by the cat. My work on emerging sign languages indicates that when people create new languages, devices for who-did-what-to-whom emerge before devices for flexible construal. I am investigating why this asymmetry emerges, and whether typical language learning shows the same pattern. For example, does flexible construal require more complex conceptual knowledge, such as perspective taking? I have also found that people are more likely to construe an event from the perspective of a patient when they cannot see the agent's face.
Verbs are commonly analyzed as encoding a small number of "arguments," with a wider range of event participants encoded as "adjuncts." My work shows that instrumental verbs such as slice and chop encode an instrumental participant (e.g., a knife) as neither a prototypical argument nor a prototypical adjunct. I propose this behavior emerges because verbs introduce participants into event structure through two representations: the argument structure and the fine-grained semantics of the verb. I am also using crosslinguistic data and data from child homesign to ask whether shared conceptual knowledge shapes the structure of these verbs.