Reader Comments

Post a new comment on this article

Distributed processing

Posted by Potsdam_EM_Group on 28 Sep 2007 at 22:17 GMT

We agree that there are many similarities between RSVP and normal reading. There are, however, also reliable influences on reading fixations from upcoming words that, by definition, are absent in RSVP. Obviously, these so-called preview benefits and parafoveal-on-foveal effects are absent in RSVP reading (see Kliegl, Nuthmann, & Engbert, 2006; Rayner, 1998, for reviews). Moreover, even in RSVP it is not clear that word processing is strictly serial. Possibly, an earlier word is still being processed while the next word is presented. Indeed, given the high-speed experimenter-paced presentation, this seems a very plausible scenario. It was unclear to us how the present experiment could distinguish between serial and parallel processing of words, given that the performance also depended on the retrieval of information after a string of six words. Retrieval itself may additionally bias the processing towards seriality.

Kliegl, R., Nuthmann, A., & Engbert, R. (2006). Tracking the mind during reading: The influence of past, present, and future words on fixation durations. Journal of Experimental Psychology: General, 135, 12–35.
Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research. Psychological Bulletin, 124, 372–422.

RE: Distributed processing

DenisPelli replied to Potsdam_EM_Group on 26 Oct 2007 at 16:51 GMT

"It was unclear ..."
We know that the words were presented serially and that the observers' responses (spoken words) were serial. Physiological measures show that the response of relevant brain areas to a word have much longer duration than the word-to-word interval of fast RSVP reading (or the fixation rate of page reading), so it is very likely, as you say, that there is temporal overlap.

Denis Pelli & Katharine Tillman