I'm a postdoctoral researcher at MIT Brain and Cognitive Sciences, currently working with Kim Scott and Josh Tenenbaum on what infants know about physics, using the online platform Lookit to collect longitudinal data from infants over the web. I am also actively collaborating with Ev Fedorenko and Jesse Snedeker (Harvard Psychology) on projects to do with how argument structure (the connections between how sentences are built and what they mean) shapes our linguistic abilities at the cognitive and neural levels.
I enjoy verbs, events, early conceptual representations, replicability, open science, and metadata.
I'm broadly interested in questions about how early development informs how we understand adult cognition and language. First, how are pre- or non-linguistic concepts like causation, agency and physical space mapped to language? Second, how do early cognitive capacities like social cognition and awareness of information structure impact language learning and use? More specifically, I am interested in how children (and adults) use syntactic structures to make inferences about what sentences mean, and to choose the right things to say to get their own meanings across. I study how children use their conceptual and linguistic representations of causation and motion to make inferences about particular events in the world, and about the meanings of syntactic structures such as the transitive (Jane broke the lamp) and periphrastic causative (Jane made the lamp break).
Beyond my individual and within-lab scientific work, I am passionate about improving our scientific practices as social scientists, including promoting replication, data sharing, and large collaborations to improve the reliability and usability of what we learn about the minds and brains of young children. I am on the governing board of ManyBabies, which is preparing to release its first dataset, a study of more than 1,500 babies from over 50 labs aiming to replicate a 'best practices' version of the Infant-Directed Speech preference.
Ask me about Conference Codes of Conduct!
Printable CV last updated June 2018
Some slightly tongue-in-cheek advice about giving academic talks.
Most other recent things I've written/adapted can be found on Gitub or OSF. Stay tuned for some highlight panels here. Feel free to reuse and adapt any of those experimental scripts or the ones here, with citation.
Here is a link to an older page of resources, including an implementation of Talk2Tobii and the Willow python package, and a little R script called CLANtoR that very slowly gives you CHILDES corpora in a data frame. Update 10/6/17: don't use CLANtoR, use childes-db!!