LLF – Bât. ODG – 5e étage – Salle du conseil (533)
Matías Guzman (LLF)
Solving the Paradigm Cell Filling Problem: an analogical account of the Russian noun class system
In this talk I argue that analogical classification is a possible solution to the Paradigm Cell Filling Problem. Using the Russian noun class system as an example, I will show that analogy greatly reduces system complexity, and helps speakers predict the inflection class of Russian nouns. Additionally, I will show that metrics like accuracy are more suited for cross-linguistic comparison than entropy.