2017 Volume 9 Pages 27-34
Having recorded event-related potentials (ERPs) while four native speakers of Hebrew silently read the eight stimuli in Table 2, the present writer proposed in his preceding paper (Ikeda et al. 2015) that the difference between N170 and P250 reflects a physical quantity (with or without vowel points) and a cognitive quantity (the number of units) of a given visual stimulus. The present paper aims to revisit this experiment, and to propose a different interpretation of the same results.
We noticed that the stimuli with vowel points produced significantly larger amplitude at N170 (see Bentin et al. 1999) than those without vowel points as already reported by Bar-Kochva (2011), while the influence of the number of linguistic units was limited at this time course (Figs. 3 and 4). On the other hand, the stimuli with a single linguistic unit produced significantly larger positive amplitude around 250 msec (labeled tentatively as P250 here) than those with two linguistic units, while the influence of the vowel points was limited at this time course (Figs. 5 and 6).
These results would rather suggest that visual information and linguistic information in Hebrew are processed separately as early as 170 through 250 msec and that the visual processing precedes the linguistic processing within this time window.