The overall goal of the proposed research is to observe the process of word learning, and track a new word as it is ?rst learned and then stored into semantic memory. In order to accomplish that, the project will record electrophysiological activity from neurosurgical patients while they engage in a verbal learning task. Neurosurgical patients are well suited to participate in this study because, for clinical reasons, they have electrodes implanted beneath their skull for a period of time ranging from 1-3 weeks. During this time, these electrodes will record electrocorticography (ECoG) while patients engage in a learning task, offering a high spatiotemporal resolution glimpse into brain activity during normal verbal word learning functioning. In this study, the task is a verbal word learning task in which patients are presented with pseudo words with no meaning, and gradually learn to associate speci?c meaning to each word. After words are presented, the patient is then given four options and asked to identify the correct semantic meaning. The main innovative concept of this project is that we will use a classi?er to decode whether a word is known or unknown, and then use that classi?er to determine the point at which the unknown words become known during the learning process. We hypothesize that word learning goes through two stages: an episodic stage and a semantic stage. The former is characterized by hippocampal activation and the latter is characterized by increased network connectivity in the lateral temporal cortex. If successful, this proposal will not only provide experimental evidence supporting the complementary learning system view of word learning, but it would also represent a ?rst step toward understanding how words enter the mental lexicon. This knowledge could potentially help rebuild the mental lexicon in cases where it is pathologically altered, such as aphasias.
The relevance of this research to public health is twofold. First, by linking memory theoretical concepts to linguistics, this project will help build better speech processing devices, and also add to basic scienti?c knowledge about the neural circuitry underlying episodic memory. Second, by mapping the process by which words go through the hippocampus and into semantic storage, this research will help identify potential targets for memory deep brain stimulation devices to bolster cognitive function in cases of pathological decline.
