Inner Voice Experiences in Processing of Direct and Indirect Speech
Previous research shows that we may be more likely to mentally simulate an "inner voice" during silent reading of direct speech quotations (e.g., Mary said: "This dress is lovely") as opposed to indirect speech quotations (e.g., Mary said that the dress was lovely). Such "inner voices" were reflected in increased activations of the temporal voice areas of the auditory cortex and in modulations of reading speed by story contexts. The current research explored the time course of "inner voice" processing using EEG/ERP. We found that "inner voice" simulations in silent reading of direct speech emerged in 225-275 ms and 375-475 ms latency windows, implying a potentially two-staged processing. The early ERPs (~250 ms) indicate a non-semantic nature of "inner voices". Preliminary source analysis results suggest potential involvement of language production in "inner voice" simulations. The research to date suggests that the "inner voice" experience in silent reading of direct (vs. indirect) speech quotations is common among hearing individuals. However, congenitally deaf participants do not distinguish between direct and indirect speech in silent reading and seem to mentally represent them in signs. The implications of our results are discussed in relation to embodied cognition and the implicit prosody hypothesis.