Meilin Zhan Abstract:
Natural language is highly versatile and complex at many different levels. Still it is the most natural way of communication between us as humans, and everyone is considered more or less an expert in at least one language. Can the seeming complexity in language inform our understanding of language processing and acquisition, indicating that language is an optimal code for communication and learning? For the purpose of communication, I will illustrate this question from perspectives of comprehender-oriented optimization and speaker-oriented optimization. For the purpose of learning, I will look at the relationship between direct experience, productivity, and systematicity in language. This leads to more questions: What are the productive components in the language system? How systematic are those components? How do learners acquire the knowledge needed to be able to generalize new examples? We propose that using numeral classifiers, a grammatical category in some of the world’s languages offers a promising way to gain new insights to these questions.
Morteza Sarafyazd Abstract:
When an action results in an error, we attempt to identify the cause so that decision policies can be adjusted to increase the probability of future success. Often the culprit is noise in the sensorimotor pathway. However, inherent nonstationarities in the environment can also lead to errors. Humans reason cognitively about such errors and update their decision policies accordingly. I designed a behavioral task to ask whether monkeys, like humans, use such cognitive reasoning. Here, I provide evidence that monkeys rely on their confidence to infer the source of errors and adjust their decision policies rapidly and rationally. This finding indicates that humans are not alone in this high level cognitive capacity, and provides an opportunity for investigating the underlying computational principles at the level of individual neurons.
Note: In order to fit multiple talks within the hour-long slot, the first talk will start promptly at 12:05, and lunch has been ordered to arrive at 11:50. Please try to come early, collect your lunch, and be seated by 12:05.
UPCOMING COG LUNCH TALKS:
10/24/17 - Maddie Pelz (Schulz Lab), Matthias Hofer (Levy Lab), and Andres Campero (Tenenbaum Lab)
10/31/17 - Mika Braginksy (Levy Lab), Jenelle Feather (McDermott Lab), and Andrew Francl (McDermott Lab)
11/07/17 - Richard McWalter, Ph.D. (McDermott Lab)
11/14/17 - Kevin Ellis (Tenenbaum Lab)
11/21/17 - Dian Yu (Rosenholtz Lab)
11/28/17 - Yang Wu (Schulz Lab)
12/05/17 - Melissa Kline, Ph.D.