Poster C17, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

ERPs reveal early orthographic and phonological selectivity during single word reading.

Laurie S. Glezer1, Katherine J. Midgely1, Karen Emmorey1, Phillip J. Holcomb1;1San Diego State University

Previous fMRI research provides evidence for a hierarchical organization in the ventral visual pathway for the visual word form, which has led to the proposal that running posterior to anterior neurons are tuned to increasingly complex word features. There is also evidence to suggest that there is a region at the top of this processing hierarchy that contains neurons that are selectively tuned to the written form of whole real words. However, due to the poor temporal resolution of fMRI, it is unclear at what time point this selectivity occurs. The Bimodal Interactive Activation Model (BIAM, Grainger and Holcomb, 2009), which is based on findings from behavioral and ERP research using priming paradigms, supports the idea that single word reading is accomplished in a mostly hierarchical, feedforward fashion and that orthographic and phonological whole word recognition is achieved by 325 ms. In this model it is proposed that an earlier component (N250) reflects mapping of sublexical features onto lexical representations. However, the evidence to support the sublexical nature of the N250 component is largely based on nonword to real word priming paradigms, which may not fully probe lexical processing. We conducted two experiments using real words in both the prime and target positions and systematically altered the orthographic and phonological similarity to examine precisely when selectivity to orthography occurs. For Experiment 1 there were three conditions of interest: (1) “same” (S), in which the same stimulus was presented twice (as prime and target; e.g., coat-coat); (2) “one letter difference” (1L), in which the prime and target differed by only one letter (e.g., boat-coat); and (3) “different” (DIFF), in which the target shared no letters with the prime (e.g., fish-coat). In Experiment 2, there were three conditions of interest: 1) S, (e.g., hair-hair); 2), a homophone of the target (H), (e.g., hare-hair); and 3) a different word from the target (Control), (e.g., hear-hair). For Experiment 1, in posterior left hemisphere electrodes we find that both 1L and DIFF have similar negative going waves in comparison to the S condition. In addition, we find the same thing for Experiment 2 with H and Control exhibiting similar negative going waves in comparison to the S condition. This was not the case in these electrodes during the N400 window. These results suggest that no priming was present in left posterior electrodes during the N250 window for real words that had overlapping orthography and phonology. That is, even though these items shared sublexical features, there was no attenuation of the signal in the N250 window, suggesting that the full lexical representation was being encoded during this time frame. Our results provide evidence that when using a short SOA and real word stimuli, a whole word lexical response can be identified within the N250 window.

Topic Area: Perception: Orthographic and Other Visual Processes