How the brain encodes the order of letters in a printed word: The SERIOL model and selective literature review

Abstract

This paper describes a novel theoretical framework of how the position of a letter within a string is encoded, the SERIOL model (sequential encoding regulated by inputs to oscillations within letter units). Letter order is represented by a temporal activation pattern across letter units, as is consistent with current theories of information coding based on the precise timing of neural spikes. The framework specifies how this pattern is invoked via an activation gradient that interacts with subthreshold oscillations and how it is decoded via contextual units that activate word units. Using mathematical modeling, this theoretical framework is shown to account for the experimental data from a wide variety of string-processing studies, including hemispheric asymmetries, the optimal viewing position, and positional priming effects.

Parafoveal preprocessing of word initial trigrams during reading in adults and children.

“…Our data showed a robust TLE in Position 12 for single fixation and gaze durations, and in Position 23 for all dependent variables, but there was no TLE in Position 13. First, this pattern of effects seemed consistent with use of bigrams as the unit for lexical access, thus supporting contextual coding models such as the SERIOL (Whitney, 2001) and Open Bigram models (Grainger & van Heuven, 2003; see also . These models differ, however, in their proposed mechanism-the Open Bigram model implements a contextual mechanism for bigram encoding, while the SERIOL model assumes a spatial coding mechanism.…”

“…Again, this indicates that letter position information was extracted flexibly from the parafovea through bigrams. Again, by extension, we might argue that this result supports contextual coding models such as the Open Bigram and SERIOL (Whitney, 2001) models. The three-way interaction was not significant in Model 2 for any of the dependent measures, indicating that TLEs on the target word were comparable in children and adults.…”

“…Alternatively, a transposed letter effect for letters 1 and 2 but not 1 and 3 would suggest that children encode relative letter position information (e.g., bigrams), supporting contextual coding models such as the Open Bigram and SERIOL (Whitney, 2001) models. This latter possibility would also support Grainger and Ziegler's model (2011), indicating that children in Year 4 are able to use the coarse-grained orthographic route.…”

“…The presence of this TLE indicated that letter position information was extracted from the parafovea independent of letter identity information, and provides evidence for parafoveal flexible letter position coding. On the assumption that it is reasonable to generalize, and assume that models of isolated word recognition might be used to generate predictions about how word identification might proceed (at least to some degree) during normal reading, then we might argue that these results provide evidence in support of models such as the Open Bigram , SOLAR (Davis, 1999(Davis, , 2010, and SERIOL (Whitney, 2001) models (see also the Overlap model by Gómez et al, 2008). Of course, we note that generalization of these findings to these models requires that they be considered in relation to processing that is distributed (spatially and temporally) across fixations.…”

Linking memory and language: Evidence for a serial-order learning impairment in dyslexia

“…Several recent models of reading stress the Serial-order learning in dyslexia importance of the temporal alignment of the serial orthographic representations (i.e., letter position and identity) and phonological representations in reading acquisition (e.g., the SERIOL model, Whitney, 2001; the overlap model, Gomez, Ratcliff, & Perea, 2008).…”