"Serial" effects in parallel models of reading

Simultaneous simulations of pure, surface and phonological acquired dyslexia within a full computational model of the primary systems hypothesis

According to the primary systems hypothesis, reading requires interactions of visual-orthographic, phonological and semantic systems. Damage to each primary system generates very different types of acquired dyslexia. Variants of the connectionist 'triangle' models of reading have been developed to investigate individual acquired dyslexia. However, only a few studies have investigated multiple acquired alexia within one framework. Importantly, there are no studies that simultaneously simulate both central dyslexia (e.g. surface and phonological dyslexia) and peripheral dyslexia (e.g. pure alexia). That is largely due to the lack of a visual component in the traditional reading models. To verify the predictions made by the primary systems hypothesis, we developed a connectionist 'deep' multi-layer triangle model of reading including visual, orthographic, phonological and semantic processing layers. We investigated whether damage to the model could produce the general behavioural patterns of impaired performance observed in patients with the corresponding reading deficits. Crucially, damage to the visual-orthographic, phonological or semantic components of the model resulted in the expected reading impairments associated with pure alexia, phonological dyslexia and surface dyslexia, respectively. The simulation results demonstrated for the first time that neurologically-impaired reading including both central and peripheral dyslexia could be addressed within a single triangle model of reading. The findings are consistent with the predictions made by the primary systems hypothesis.

Differential Phonological and Semantic Modulation of Neurophysiological Responses to Visual Word Recognition

BACKGROUND

Reading words for meaning relies on orthographic, phonological and semantic processing. The triangle model implicates a direct orthography-to-semantics pathway and a phonologically mediated orthography-to-semantics pathway, which interact with each other. The temporal evolution of processing in these routes is not well understood, although theoretical evidence predicts early phonological processing followed by interactive phonological and semantic processing.

METHOD

This study used electroencephalography-event-related potential (ERP) analysis and magnetoencephalography (MEG) source localisation to identify temporal markers and the corresponding neural generators of these processes in early (∼200 ms) and late (∼400 ms) neurophysiological responses to visual words, pseudowords and consonant strings.

RESULTS

ERP showed an effect of phonology but not semantics in both time windows, although at ∼400 ms there was an effect of stimulus familiarity. Phonological processing at ~200 ms was localised to the left occipitotemporal cortex and the inferior frontal gyrus. At 400 ms, there was continued phonological processing in the inferior frontal gyrus and additional semantic processing in the anterior temporal cortex. There was also an area in the left temporoparietal junction which was implicated in both phonological and semantic processing. In ERP, the semantic response at ∼400 ms appeared to be masked by concurrent processes relating to familiarity, while MEG successfully differentiated these processes.

DISCUSSION

The results support the prediction of early phonological processing followed by an interaction of phonological and semantic processing during word recognition. Neuroanatomical loci of these processes are consistent with previous neuropsychological and functional magnetic resonance imaging studies. The results also have implications for the classical interpretation of N400-like responses as markers for semantic processing.

An Extension of a Parallel-Distributed Processing Framework of Reading Aloud in Japanese: Human Nonword Reading Accuracy Does Not Require a Sequential Mechanism

Humans can pronounce a nonword (e.g., rint). Some researchers have interpreted this behavior as requiring a sequential mechanism by which a grapheme-phoneme correspondence rule is applied to each grapheme in turn. However, several parallel-distributed processing (PDP) models in English have simulated human nonword reading accuracy without a sequential mechanism. Interestingly, the Japanese psycholinguistic literature went partly in the same direction, but it has since concluded that a sequential parsing mechanism is required to reproduce human nonword reading accuracy. In this study, by manipulating the list composition (i.e., pure word/nonword list vs. mixed list), we demonstrated that past psycholinguistic studies in Japanese have overestimated human nonword reading accuracy. When the more fairly reevaluated human performance was targeted, a newly implemented Japanese PDP model simulated the target accuracy as well as the error patterns. These findings suggest that PDP models are a more parsimonious way of explaining reading across various languages.

Visual word learning in skilled readers of English

Three experiments are reported analysing the processes by which adult readers of English learn new written words. Visual word learning was simulated by presenting short (four-letter) and longer (seven-letter) nonwords repeatedly and observing the reduction in naming latencies and the convergence in reaction times (RTs) to shorter and longer items that are the hallmarks of visual word learning. Experiment 1 presented nonwords in ten consecutive blocks. Naming latencies reduced over the first four or five presentations. The effect of length on naming RTs was large in block 1 but non-significant after four or five presentations. Experiment 2 demonstrated some reduction in RTs to untrained nonwords following practice on a trained set, but the reduction was less than for the trained items and RTs to shorter and longer nonwords did not converge. Experiment 3 included a retest after seven days which showed some slowing of RTs compared with the end of the first session but also considerable retention of learning. We conclude that four to six exposures to novel words (nonwords) are sufficient to establish durable lexical representations that permit parallel processing of newly-learned words. The results are discussed in terms of theoretical models of reading and word learning.

What lies beneath: a comparison of reading aloud in pure alexia and semantic dementia

Exaggerated effects of word length upon reading-aloud performance define pure alexia, but have also been observed in semantic dementia. Some researchers have proposed a reading-specific account, whereby performance in these two disorders reflects the same cause: impaired orthographic processing. In contrast, according to the primary systems view of acquired reading disorders, pure alexia results from a basic visual processing deficit, whereas degraded semantic knowledge undermines reading performance in semantic dementia. To explore the source of reading deficits in these two disorders, we compared the reading performance of 10 pure alexic and 10 semantic dementia patients, matched in terms of overall severity of reading deficit. The results revealed comparable frequency effects on reading accuracy, but weaker effects of regularity in pure alexia than in semantic dementia. Analysis of error types revealed a higher rate of letter-based errors and a lower rate of regularization responses in pure alexia than in semantic dementia. Error responses were most often words in pure alexia but most often nonwords in semantic dementia. Although all patients made some letter substitution errors, these were characterized by visual similarity in pure alexia and phonological similarity in semantic dementia. Overall, the data indicate that the reading deficits in pure alexia and semantic dementia arise from impairments of visual processing and knowledge of word meaning, respectively. The locus and mechanisms of these impairments are placed within the context of current connectionist models of reading.

Connectionist neuropsychology: uncovering ultimate causes of acquired dyslexia

Acquired dyslexia offers a unique window on to the nature of the cognitive and neural architecture supporting skilled reading. This paper provides an integrative overview of recent empirical and computational work on acquired dyslexia within the context of the primary systems framework as implemented in connectionist neuropsychological models. This view proposes that damage to general visual, phonological or semantic processing abilities are the root causes of different forms of acquired dyslexia. Recent case-series behavioural evidence concerning pure alexia, phonological dyslexia and surface dyslexia that supports this perspective is presented. Lesion simulations of these findings within connectionist models of reading demonstrate the viability of this approach. The commitment of such models to learnt representations allows them to capture key aspects of performance in each type of acquired dyslexia, particularly the associated non-reading deficits, the role of relearning and the influence of individual differences in the premorbid state of the reading system. Identification of these factors not only advances our understanding of acquired dyslexia and the mechanisms of normal reading but they are also relevant to the complex interactions underpinning developmental reading disorders.