Sibley, Kello, Plaut, and Elman (2008) proposed the sequence encoder as a model that learns fixed‐width distributed representations of variable‐length sequences. In doing so, the sequence encoder overcomes problems that have restricted models of word reading and recognition to processing only monosyllabic words. Bowers and Davis (2009) recently claimed that the sequence encoder does not actually overcome the relevant problems, and hence it is not a useful component of large‐scale word‐reading models. In this reply, it is noted that the sequence (...) encoder has facilitated the creation of large‐scale word‐reading models. The reasons for this success are explained and stand as counterarguments to claims made by Bowers and Davis. (shrink)

The forms of words as they appear in text and speech are central to theories and models of lexical processing. Nonetheless, current methods for simulating their learning and representation fail to approach the scale and heterogeneity of real wordform lexicons. A connectionist architecture termed thesequence encoderis used to learn nearly 75,000 wordform representations through exposure to strings of stress‐marked phonemes or letters. First, the mechanisms and efficacy of the sequence encoder are demonstrated and shown to overcome problems with traditional slot‐based (...) codes. Then, two large‐scale simulations are reported that learned to represent lexicons of either phonological or orthographic wordforms. In doing so, the models learned the statistics of their lexicons as shown by better processing of well‐formed pseudowords as opposed to ill‐formed (scrambled) pseudowords, and by accounting for variance in well‐formedness ratings. It is discussed how the sequence encoder may be integrated into broader models of lexical processing. (shrink)

The forms of words as they appear in text and speech are central to theories and models of lexical processing. Nonetheless, current methods for simulating their learning and representation fail to approach the scale and heterogeneity of real wordform lexicons. A connectionist architecture termed thesequence encoderis used to learn nearly 75,000 wordform representations through exposure to strings of stress‐marked phonemes or letters. First, the mechanisms and efficacy of the sequence encoder are demonstrated and shown to overcome problems with traditional slot‐based (...) codes. Then, two large‐scale simulations are reported that learned to represent lexicons of either phonological or orthographic wordforms. In doing so, the models learned the statistics of their lexicons as shown by better processing of well‐formed pseudowords as opposed to ill‐formed (scrambled) pseudowords, and by accounting for variance in well‐formedness ratings. It is discussed how the sequence encoder may be integrated into broader models of lexical processing. (shrink)

The self‐teaching hypothesis describes how children progress toward skilled sight‐word reading. It proposes that children do this via phonological recoding with assistance from contextual cues, to identify the target pronunciation for a novel letter string, and in so doing create an opportunity to self‐teach new orthographic knowledge. We present a new computational implementation of self‐teaching within the dual‐route cascaded (DRC) model of reading aloud, and we explore how decoding and contextual cues can work together to enable accurate self‐teaching under a (...) variety of circumstances. The new model (ST‐DRC) uses DRC’s sublexical route and the interactivity between the lexical and sublexical routes to simulate phonological recoding. Known spoken words are activated in response to novel printed words, triggering an opportunity for orthographic learning, which is the basis for skilled sight‐word reading. ST‐DRC also includes new computational mechanisms for simulating how contextual information aids word identification, and it demonstrates how partial decoding and ambiguous context interact to achieve irregular‐word learning. Beyond modeling orthographic learning and self‐teaching, ST‐DRC’s performance suggests new avenues for empirical research on how difficult word classes such as homographs and potentiophones are learned. (shrink)

It is often assumed that graphemes are a crucial level of orthographic representation above letters. Current connectionist models of reading, however, do not address how the mapping from letters to graphemes is learned. One major challenge for computational modeling is therefore developing a model that learns this mapping and can assign the graphemes to linguistically meaningful categories such as the onset, vowel, and coda of a syllable. Here, we present a model that learns to do this in English for strings (...) of any letter length and any number of syllables. The model is evaluated on error rates and further validated on the results of a behavioral experiment designed to examine ambiguities in the processing of graphemes. The results show that the model (a) chooses graphemes from letter strings with a high level of accuracy, even when trained on only a small portion of the English lexicon; (b) chooses a similar set of graphemes as people do in situations where different graphemes can potentially be selected; (c) predicts orthographic effects on segmentation which are found in human data; and (d) can be readily integrated into a full-blown model of multi-syllabic reading aloud such as CDP++ (Perry, Ziegler, & Zorzi, 2010). Altogether, these results suggest that the model provides a plausible hypothesis for the kind of computations that underlie the use of graphemes in skilled reading. (shrink)

Recent research has found that the omission of accent marks in Spanish does not produce slower word identification times in go/no-go lexical decision and semantic categorization tasks [e.g., cárcel = carcel], thus suggesting that vowels like á and a are represented by the same orthographic units during word recognition and reading. However, there is a discrepant finding with the yes/no lexical decision task, where the words with the omitted accent mark produced longer response times than the words with the accent (...) mark. In Experiment 1, we examined this discrepant finding by running a yes/no lexical decision experiment comparing the effects for words and non-words. Results showed slower response times for the words with omitted accent mark than for those with the accent mark present, thus suggesting a bias toward a “word” response for accented items in the yes/no lexical decision task. To test this interpretation, Experiment 2 used the same stimuli with a blocked design and a go/no-go lexical decision task. Results showed similar response times to words regardless of whether the accent mark was omitted. This pattern strongly suggests that the longer response times to words with an omitted accent mark in yes/no lexical decision experiments are a task-dependent effect rather than a genuine reading cost. (shrink)

It has been suggested that there is a close relationship between visual attention span (VAS) and fluent reading. This relation may be modulated by participants’ age, and exhibits various patterns in different reading modes (i.e. oral v.s. silent reading) and different reading levels (e.g. sentence v.s. character/word levels). Moreover, the modulation effects from the above factors might be more remarkable in the framework of languages with a deep orthography. Therefore, the present study investigated the developmental pattern of the relationship between (...) VAS skills and reading fluency in Chinese, a language with particularly deep orthography, by recruiting 292 participants from primary schools, middle schools and universities. Two tests were utilized to assess fluent reading skills at the single-character and sentence levels with oral and silent reading modes. A visual 1-back task was adopted to reflect VAS capacity with non-verbal stimuli and no verbal response. Results showed that the VAS capacity of low-grade primary school students could significantly account for the variance in single-character reading fluency in the oral mode and that it was a significant predictor of sentence reading fluency in the oral mode among high-grade primary school students. VAS abilities of middle school students allowed a unique and stable prediction of their silent sentence reading. With increasing reading ability, VAS skills of adults showed significant and similar predictive power for estimating the variations in fluent sentence reading in both silent and oral modes. These results revealed developmental changes in the contribution of VAS to fluent reading in Chinese, and provided evidence unveiling whether the underlying mechanisms of oral and silent reading were shared or different. (shrink)

Cognitive Science, Volume 46, Issue 5, May 2022.

The visual attention (VA) span is defined as the amount of distinct visual elements which can be processed in parallel in a multi-element array. Both recent empirical data and theoretical accounts suggest that a VA span deficit might contribute to developmental dyslexia, independently of a phonological disorder. In this study, this hypothesis was assessed in two large samples of French and British dyslexic children whose performance was compared to that of chronological-age matched control children. Results of the French study show (...) that the VA span capacities account for a substantial amount of unique variance in reading, as do phonological skills. The British study replicates this finding and further reveals that the contribution of the VA span to reading performance remains even after controlling IQ, verbal fluency, vocabulary and single letter identification skills, in addition to phoneme awareness. In both studies, most dyslexic children exhibit a selective phonological or VA span disorder. Overall, these findings support a multi-factorial view of developmental dyslexia. In many cases, developmental reading disorders do not seem to be due to phonological disorders. We propose that a VA span deficit is a likely alternative underlying cognitive deficit in dyslexia. (shrink)