Rissman & Majid (2019). Thematic roles: Core knowledge or linguistic construct? Psychonomic Bulletin & Review.
A range of psycholinguistic, typological and developmental evidence indicates that the role Agent is part of core knowledge. For the roles Goal, Recipient and Instrument, however, evidence for cognitive universals is much more tenuous.
Rissman, Woodward & Goldin-Meadow (2019). Occluding the face diminishes the conceptual accessibility of an animate agent. Language, Cognition and Neuroscience.
English speakers are more likely to describe events with passive voice when they cannot see the face of the agent. This suggests that visual prominence has a strong effect on how we think about events.
Rissman, Rawlins & Landau (2019). Event participants and verbal semantics: non-discrete structure in English, Spanish and Mandarin. In A.K. Goel, C.M. Seifert, & C. Freksa (Eds.), Proceedings of the 41st Annual Conference of the Cognitive Science Society. (pp. 960-966). Montreal, QB: Cognitive Science Society.
Verbs in Spanish and Mandarin such as cortar ('cut') and cì ('stab') encode an instrument through a non-discrete semantic structure rather than through argument structure. This result is similar to the English findings in Rissman et al. (2015).
Rissman & Majid (2019). Agency drives category structure in instrumental events. In A.K. Goel, C.M. Seifert, & C. Freksa (Eds.), Proceedings of the 41st Annual Conference of the Cognitive Science Society. (pp. 2661-2667). Montreal, QB: Cognitive Science Society.
The most prototypical instrumental events in English involve an intentional agent, where the instrument is a direct extension of the agent.
Rissman, Goldin-Meadow & Woodward (2018). From visual prominence to event construal: influences (and non-influences) of eyegaze. In T.T. Rogers, M. Rau, X. Zhu & C.W. Kalish (Eds.), Proceedings of the 40th Annual Conference of the Cognitive Science Society (pp. 2334-2339). Austin, TX: Cognitive Science Society.
When people watch videos of an agent and a patient, they are more likely to direct their early eyegaze to the agent when the agent's face is visible than when it is occluded.
English with & use (e.g. Vivienne sliced the pumpkin with a knife) encode different features of instrumentality, inconsistent with the theory that Instrument is a primitive semantic category.
Rissman & Goldin-Meadow (2017). The development of causal structure without a language model. Language Learning and Development
In a longitudinal case study of a child homesigner, we show that this child uses handshape morphology to encode causation, but only after a lengthy developmental period.
Cartmill, Rissman, Novack & Goldin-Meadow (2017). The emergence of iconic features in co-speech gesture and homesign. Language, Interaction and Acquisition
The gestures of hearing children and the signs of a deaf homesigning child appear to be structured based on the same fundamental building blocks.
Rissman, Rawlins & Landau (2015). Using instruments to understand argument structure: evidence for gradient representation. Cognition
English verbs such as slice and poke encode an instrument, but not through a discrete argument structure. When we consider how verbs encode relations between event participants, there are possibilities beyond the familiar one-participant, two-participant and three-participant categories.
Rissman (2015). Cinderella broke and broke: object deletion and manner-result complementarity. CLS 51
Syntactic judgments and corpus data show that instrumental verbs like slice share properties of both manner verbs and result verbs, supporting the theory that verbal roots can be linked to more than one part of event structure, inconsistent with manner-result complementarity.
Rissman, Legendre, Landau (2013). Abstract morphosyntax in two and three-year-old children: evidence from priming. Language Learning and Development
A production priming experiment showed that children represent auxiliary is (e.g. the pig is jumping) and auxiliary are (e.g. the birds are singing) as part of the same abstract category.
Instrumental use (e.g. Vivian used a knife to slice the pumpkin) encodes quantification over worlds compatible with the goals of an agent.
Goal-based interpretations of instrumental with sentences are generated pragmatically.
Semantic and syntactic diagnostics indicate that locata (e.g., Bill stuffed the cabbage with rice) pattern like arguments, whereas instruments (e.g. Bill stuffed the cabbage with a spoon) pattern like adjuncts. At the same time, argumenthood is more gradient rather than categorical.