Prof. Aniello De Santo at the LSA 2025
Aniello presented a talk at the Annual Meeting of the LSA, as part of a special session
on "Computational Models of Learnability and Acquisition of Morphology and Phonology".
The title of his talk was "Learnability insights from (and into) the TSL Neighborhood",
read the abstract below!
-----
"Learnability insights from (and into) the TSL Neighborhood"
The Tier-based Strictly Local languages (TSL; Heinz et al., 2011) account for a variety
of local and long-distance phonotactic phenomena. In this talk, I review new insights
into the computational primitives needed to account for linguistic dependencies in
phonotactics and beyond (De Santo and Graf, 2019 a.o.). I argue that recent grammatical
inference results for these classes (De Santo and Aks ̈enova 2021; Johnson and De
Santo 2023; 2024) can help with the interpretation and design of experimental investigations
into human learning. Finally, I show how attention to the fine-grained differences
between these classes cast new light on past results from Artificial Grammar Learning
experiments (Lai, 2015; De Santo and Rawski, 2022), help design studies to differentiate
computational complexity constraints from other limitations imposed on the learner,
and probe the connection between complexity biases and general cognitive resources.