Research

My research focuses on expanding our understanding of learning-theoretic elements of natural language by 1) designing learning algorithms for linguistically relevant formal language classes which can learn from small data, 2) creating efficient software implementations of these algorithms which can be run on naturalistic datasets, and 3) investigating the impact of representational choices and abductive principles to ultimately refine these algorithms and hone in on learning approaches which resemble human behavior.

Logan Swanson, Sarah Payne, Sam van der Poel, Adil Soubki, and Jeffrey Heinz (2026). The Long and Short of It: An experimental investigation into the role of shorter strings in learning regular languages. Proceedings of The 17th Annual Conference on Grammatical Inference. To appear.

Logan Swanson (2026). Syntactic Learning over Tree Tiers. Logics, 4(2), 5.

Logan Swanson, Jonathan Rawski, and Jeffrey Heinz (2026). Phonotactic Learning with Structure, not Statistics. Linguistic Inquiry, 1-24.

Logan Swanson (2025). Learning Multi Tier-Based Strictly 2-Local Languages. Proceedings of Mathematics of Language 2025

Logan Swanson, Kenneth Hanson, and Thomas Graf (2025). Are Syntactic Categories ISL-2 Inferrable? A Corpus Study. Proceedings of Society for Computation in Linguistics 2025

Logan Swanson (2024). Syntactic Learning over Tree Tiers. Proceedings of 35th European Summer School in Logic, Language, and Information Student Session

Logan Swanson and Jeffrey Heinz (2026). The Long and Short of It: An experimental investigation into the role of shorter strings in learning regular languages. Talk at The 17th Annual Conference on Grammatical Inference.

Logan Swanson (2025). Learning over Multiple Tiers. Talk at The 2025 Rutgers Subregular Phonology Workshop.

Logan Swanson (2025). Learning Multi Tier-Based Strictly 2-Local Languages. Talk at Mathematics of Language 2025.

Logan Swanson (2025). Are syntactic categories ISL-2 inferrable? A corpus study. Talk at Society for Computation in Linguistics 2025.

Logan Swanson, Jonathan Rawski, and Jeffrey Heinz (2024). Phonotactic Learning and Constraint Selection Without Statistics. Poster at Annual Meeting on Phonology 2024.

Logan Swanson (2024). The Bottom-Up Factor Inference Algorithm and Applications. Talk at The 2024 Rutgers Subregular Phonology Workshop.

Logan Swanson (2024). Phonotactic Learning with Abductive Principles. Invited talk at MIT’s Phonology Circle group.

Logan Swanson (2024). Syntactic Learning over Tree Tiers. Talk at 35th European Summer School in Logic, Language, and Information student session.

Logan Swanson (2024). Expression or Identity?: Exploring Online Discussion of Pronouns. Poster at Lavender Languages and Linguistics #30.

John David Storment, Logan Swanson, and Daniel Greeson (2023). It's 🏕: Queer Identity and Emoji Use Online. Talk at Lavender Languages and Linguistics #29.