| Functional magnetic resonance imaging (fMRI) has provided a wealth of information about the localization of language processing in the intact brain. However, such localization studies can be complicated by many factors including stimuli selection, experimental design, and the use of explicit tasks. In this dissertation, I investigated the neural substrates of visual word processing, and identified and differentiated those brain regions involved in lexical and semantic processing. I focused upon neural regions that are automatically activated by the presence of a word, regardless of task demands. This required experimental designs that limited the influence of extraneous cognitive processes on patterns of activation. In the first two experiments, I used fMRI to establish which brain regions are involved in lexical and semantic processes by presenting visible words and non-words in isolation. I then conducted a series of studies that limited potential top-down influences upon word processing by utilizing subliminal presentations of words. The third experiment was a behavioral semantic masked priming experiment, conducted to establish the utility of the masking procedure. Finally, in the fourth and fifth experiments, I used fMRI to establish the brain regions involved in lexical and semantic processes while simultaneously limiting the contribution from conscious processes by presenting masked words and nonwords.; Activations to lexical aspects (i.e., all word categories) were found bilaterally in angular gyrus and inferior frontal gyri. A role for the angular gyrus in lexical processes is consistent with previous literature. Although inferior frontal gyrus activation has previously been found to semantic manipulations, the present results indicated that eliciting such effects may require controlled rather than incidental semantic processes.; Activations in response to semantic aspects were found in bilateral temporal gyri. These effects were mediated by the semantic aspect of concreteness. A comparison of visible and masked word processing indicated that the strongest effects for the masked words were found in the left hemisphere, particularly in the temporal lobe, suggesting the primacy of the left hemisphere in automatic language processes. |
