Output

Recent work:

Lidz, J., & Howitt, Katherine. (2026). LLMs are not children: They have to earn our love. Behavioral and Brain Sciences, 49, e210. doi:10.1017/S0140525X26104695 link

Howitt, Katherine, Scontras, G., and M. Polinsky. to appear. English restrictive relative clauses are subject to crossover violations. Linguistic Inquiry. link

Howitt, Katherine, Phillips, C., and Lidz, J. (2025). Strong Crossover, weak evidence: 4 year old’s knowledge of the strong crossover constraint. Boston University Conference on Language Development (BUCLD) 50. abstract.

Howitt, Katherine, Nair, S., Dods, A., & Hopkins, R. (2024). Generalizations across filler-gap dependencies in neural language models. CoNLL 2024 link pdf

Nair, S., Howitt, Katherine, Dods, A., & Hopkins, R. (2024). LMs are not good proxies for human language learners. Boston University Conference on Language Development (BUCLD) 49. slides.

Howitt, Katherine, Phillips, C., and Lidz, J. (2024). Probing 4 year old children’s knowledge of the strong crossover constraint. Experiments in Linguistic Meaning (ELM) 3. poster.

Lee, E., Howitt, Katherine, Dixon, L., Ness, T., Nakamura, M., Muller, H., & Phillips, C. (2023) Alignment between adult and child predictive processing dynamics: Evidence from a gamified cloze study in a museum. (poster) The 36th Annual Conference on Human Sentence Processing, University of Pittsburgh, Pennsylvania, USA. poster.

Howitt, Katherine, Dey, S., Sakas, W. G. (2021) Gradual syntactic triggering: the gradient parameter hypothesis. Language Acquisition, 28:1, 65-96, DOI: 10.1080/10489223.2020.1803329 pdf

You can find my CV here.