Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers
Poster Session B: Wednesday, August 13, 1:00 – 4:00 pm, de Brug & E‑Hall
Neural Substrates of a Symbolic Action Grammar in Primate Frontal Cortex
Lucas Y. Tian1, Kedar U. Garzón2, Daniel Hanuska1, Xiao-Jing Wang3, Joshua B. Tenenbaum4, Winrich Freiwald1; 1Rockefeller University, 2Columbia University, 3New York University, 4Massachusetts Institute of Technology
Presenter: Lucas Y. Tian
At the core of intelligence is the capacity to solve new problems. In turn, problem-solving has been hypothesized to depend on cognitive operations resembling symbolic grammars (Newell & Simon, 1976), with two core components: discrete units (symbols) and rules for recombining symbols into new composite representations (syntax). Whether and how symbolic grammars are implemented in neuronal substrates remains unknown. Here, we establish a research program to elucidate the neural basis of action grammars. In a drawing task, macaque monkeys learn action grammars, which guides how they compositionally generalize to draw new images. Our behavioral analyses indicate that these grammars consist of symbolic action primitives and syntactic rules. In recordings of neuronal activity across motor, premotor, and prefrontal areas, we identified separate populations encoding action grammar components, including motor primitives, action symbols, and syntactic rules. Here, we report the discovery of an action symbol representation in ventral premotor cortex (PMv). Specifically, we found that PMv encodes planned stroke primitives, and does so in a manner exhibiting three symbolic properties: abstraction, categorical structure, and recombination. Thus, we have established a paradigm to study compositional generalization using action grammars, and identified a representation of action symbols in PMv. In ongoing work, we are studying how neural activity, in PMv and interconnected areas, may implement the systematic composition of symbols using syntactic rules.
Topic Area: Language & Communication
Extended Abstract: Full Text PDF