a nice treethe second doors to social sciencesanother nice tree


One of the best ways to make the predictions that a linguistic model makes about language explicit is to computationally implement it. The process of implementation can yield significant insights about the types of computation the model assumes language learners and users to be capable of. Implementation often clarifies exactly what types of patterns the model predicts to occur in human language. Furthermore, building these models necessitates the formalization of many details that may otherwise be left vague. It is these seemingly small details that often have the most significant ramifications for the predictions that linguistic models make.

The primary goals of the Chicago Language Modeling Lab are as follows:
  1. to computationally implement linguistic models;
  2. to discover and implement learning algorithms for those models;
  3. to test the predictions of the models by simulating language learning, use, and change among populations of interacting language-using agents.


U of C seal