UC San Diego

Reading is learned in the presence of an already formed auditory language network. However, unlike auditory language reading is a recent cultural invention made possible by an extensive period of learning. Understanding the relationship of visual language with auditory language is key to understanding the novel human construct of reading. Articulatory motor movements are a potential bridge between the existing auditory language network and the developing visual reading network. Children who vocalize while learning to read and who understand the relationships between letters and sounds learn at a faster and more successful

rate. However, in neuroanatomical models of silent reading the precentral gyrus, associated with articulatory motor movements, is largely omitted.

The first section of the dissertation presents evidence that the precentral gyrus is involved in the dorsal reading route, putatively in grapheme-to-phoneme conversion. Chapter 1 presents evidence from a speeded semantic decision task. Word-level linguistic effects in the Precentral Gyrus and significant early phase-locking activity between the Fusiform and Precetral Gyrus were identified. Chapter 2 presents evidence from a Match/Mismatch task between sequentially presented graphemes and phonemes. Again, the precentral gyrus is implicated as a central hub by the combination of letter-specific effects, Mismatch effects, and significant connectivity with the Fusiform Gyrus. Chapter 3 examines the overlap and separation of Visual and Auditory language using a semantic decision task performed in each sensory modality. We find that while the Visual language processing that was present significantly overlaps with Auditory language processing, only a fraction of the Auditory language network is recruited during Visual language processing.

The second section details methodological advances to aid in the study of language using intracranial iEEG. Chapter 4 details the use of carbon-based electrodes to increase the possible spatial resolution that iEEG can measure while retaining high signal-to-noise ratio. Chapter 5 details a multimedia tablet which was created to facilitate increased data collection on patients without increasing the effort necessary from either patients or staff. By increasing the possible spatial resolution and the possible amount of data collected, these two chapters demonstrate how to build upon the work in the first three chapters.