Position representation: General principles or domain-specificity?

Speaking, spelling, and serial recall require a sequence of items to be produced one at a time in the correct order. The research in this dissertation addresses whether similar principles underlie serial order processing in different cognitive domains. Sequence representations, at a minimum, contain information about the identity and position of the items in the sequence. Although, various hypotheses---or position representation schemes---have been proposed for how position is represented, little empirical work has directly contrasted them. In this dissertation, I investigate item position representation for sequences in different domains: letters in reading and spelling, phonemes in spoken production, and words, spatial locations and object orientations in working memory. For each domain, a large set of position representation schemes are examined to determine which scheme best explains the patterns of errors produced in these tasks. This work builds on previous research that investigated the representation of letter position in spelling (Fischer-Baum, McCloskey & Rapp, 2010).

In spelling, letter perseveration errors produced by individuals with acquired dysgraphia were analyzed. For example, "edge" produced as ERGE immediately after producing the response FRENCE. Letter perseveration errors (such as the R in this example) appeared in the same position of the error and source responses when position is defined relative to the beginning or end of the sequence---a scheme referred to as a "both-edges scheme." None of the other representational schemes accounted for the data as well. Similar methods are used to investigate position representation in the other domains; intrusion errors with clear sources in nearby responses are analyzed to determine whether they appear in the same position in the error and source responses according to different representation schemes. For most domains, the both-edges scheme best accounted for the data. On this basis, I conclude that representing element position relative to both edges may be a basic principle of sequence representation. However, for other domains---specifically the representation of position for a sequence of spatial locations---a different representation scheme is identified. In the end, I discuss a theoretical framework that can account for these results.