ArticlePDF Available

Models of Reading Aloud: Dual-Route and Parallel-Distributed-Processing Approaches

Authors:
Max Coltheart
Max Coltheart
Max Coltheart
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Brent Curtis
Brent Curtis
Brent Curtis
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Paul W. B. Atkins at ProSocial World
Paul W. B. Atkins
  • ProSocial World
Michael Haller
Michael Haller
Michael Haller
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Download full-text PDF
Read full-text
Download citation

Abstract

It has often been argued that various facts about skilled reading aloud cannot be explained by any model unless that model possesses a dual-route architecture (lexical and nonlexical routes from print to speech). This broad claim has been challenged by M. S. Seidenberg and J. L. McClelland (1989, 1990). Their model has but a single route from print to speech, yet, they contend, it can account for major facts about reading that have hitherto been claimed to require a dual-route architecture. The authors identify 6 of these major facts about reading. The 1-route model proposed by Seidenberg and McClelland can account for the 1st of these but not the remaining 5. Because models with dual-route architectures can explain all 6 of these basic facts about reading, the authors suggest that this remains the viable architecture for any tenable model of skilled reading and learning to read. The dual-route cascaded model, a computational version of the dual-route model, is described. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Coltheart, Curtis, Atkins and Haller (199
3).pdf
Content uploaded by Paul W. B. Atkins
Author content
All content in this area was uploaded by Paul W. B. Atkins on Oct 10, 2016
Content may be subject to copyright.
Coltheart, Curtis, Atkins and Haller (199
3).pdf
Content uploaded by Paul W. B. Atkins
Author content
All content in this area was uploaded by Paul W. B. Atkins on Oct 10, 2016
Content may be subject to copyright.
A preview of this full-text is provided by American Psychological Association.
Learn more
Content available from Psychological Review
This content is subject to copyright. Terms and conditions apply.
... A complicating factor for this approach is the presence of multi-letter graphemes. At least for monosyllabic words, grapheme parsing can be, in principle, automated (Coltheart et al., 1993;Pritchard et al., 2016); however, ambiguities remain, for example, when it comes to silent letters. Here, we rely on a semiautomated procedure (see Supplementary Materials 2 on https://osf.io/fmxy8/?view_only=b9b6b0c4e1fa48b0b3af4bca31e0f4af for more details). ...
... Second, there is no unambiguous way to determine the GPC rules. Even in the implemented versions of the DRC in English (Coltheart et al., 2001), German , French (Ziegler et al., 2003), Italian (Mulatti & Job, 2003), and Dutch (Schmalz et al., 2015), the GPC rules have been hand-picked by the modelers (but see Coltheart et al., 1993;Pritchard et al., 2016). This involves arbitrary decisions. ...
How we should measure Orthographic Depth: Or should we?
Preprint
Full-text available
  • Sep 2024
Cross-linguistic reading research often focuses on the effect of orthographic depth - the closeness of the relationship between print and speech. To understand its effect on reading, we need to be able to objectively quantify the level of orthographic depth of a given orthography. Previous work has suggested that different dimensions underlie orthographic depth, and it is not always clear if and how existing quantifications map onto these underlying dimensions. Here, we first examine how different measures relate conceptually to underlying theoretical dimensions. Then, we quantify the relative depth of 8 European orthographies. We use both 11 existing methods and two new approaches which have not been previously used to quantify orthographic depth: Distance-based measures relying on the closeness of the phonology of orthographically similar words, and mutual information, as a theory-neutral approach. The relationship between the different measures suggests that they map on two separate dimensions: the size of the orthographic units that map onto phonology and the systematicity of the mapping, in line with previous theoretical work which drew a distinction between complexity and unpredictability. The measures derived based on different theoretical assumptions largely show agreement. From a theoretical perspective, this prevents us from making differential predictions based on different approaches. From a practical perspective, this suggests that different measures may yield comparable results, as long as they tap into the same underlying dimension of orthographic depth.
View
... Devemos ressaltar que o modelo descrito bem como outros modelos do mesmo período ainda se limitavam à descrição do processamento de palavras monossílabas, na medida em que requeriam apenas a relação entre grafema e fonema (Marslen-Wilson;Tyler, 1980;Rumelhart, 1981;Seidenberg;McClelland, 1989;Rastle, 1994;Coltheart, 2000). Propõe-se, posteriormente, uma modelagem computacional que visa à incorporação de palavras dissílabas (a expandir para palavras polissílabas) do inglês, modelo esse que passa a ser denominado de Modelo de Dupla Rota em Cascata (DRC) (Coltheart et al., 1993;Coltheart, 2000). ...
O processamento do acento primário no português brasileiro à luz do Modelo de Dupla Rota em Cascata
Article
Full-text available
  • Nov 2024
Este trabalho contrasta o processamento do acento primário em pseudopalavras do português brasileiro a partir do Modelo de Dupla Rota em Cascata (Coltheart; Rastle, 1994; Rastle; Coltheart, 2000). O Modelo postula a existência de duas rotas de processamento conforme o tipo de palavra: palavras reais familiares são processadas pela rota lexical e palavras reais não familiares e pseudopalavras são processadas pela rota não lexical. Para verificar se o acento em português pode ser analisado sob tal ótica, foram realizados dois estudos experimentais que avaliam se os falantes, ao lerem uma pseudopalavra, aplicam a regra geral (Hipótese Métrica de Bisol, 1994) por meio da ativação da rota lexical; ou se se valem de informações lexicais, como a similaridade fonológica com uma palavra real, informações morfológicas e segmentos finais, por meio da ativação da rota não lexical. O experimento 1 avaliou a similaridade fonológica e o experimento 2 avaliou padrões morfológicos derivacionais ambos em relação à regra geral. Os resultados indicam que as pseudopalavras resgatam informações lexicais, segmentais e morfológicas ao serem processadas, colocando em questionamento o processamento a partir apenas da existência de duas rotas guiadas pelo tipo de palavra. Há evidências de que a ativação do padrão acentual pode ocorrer por diferentes mecanismos.
View
... O modelo de dupla rota de leitura em voz alta postula que existem duas rotas da palavra impressa à fala: uma rota lexical e uma não lexicalessa última também chamada de rota fonológica. Dessa maneira, Coltheart et al. (1993) informam que a leitura pela rota lexical depende do conhecimento da palavra no léxico mental 1 e que a leitura pela rota não lexical ou fonológica não depende desse léxico. Portanto, a leitura oral de um texto escrito pode ocorrer Página3 por meio de um processo visual direto ou por meio de um processo envolvendo mediação fonológica. ...
Avaliação do efeito de extensão na leitura oral de palavras isoladas da Língua Portuguesa
Article
  • Sep 2024
Este estudo objetiva avaliar um possível efeito da extensão na leitura oral, particularmente na variável velocidade de conversão. Para isso, foram avaliados 18 participantes do 2º, 3º e 4º anos do Ensino Fundamental. Com cada um dos escolares, gravaram-se 180 palavras/pseudopalavras isoladas, divididas em três categorias de extensão: quatro grafemas ou menos, cinco a sete grafemas e oito grafemas ou mais. Diferentemente do que se encontra documentado na literatura, os resultados da leitura oral de nossos participantes sugerem que a extensão das palavras não impacta o tempo médio de conversão grafofonêmica – nem na leitura de palavras, nem na de pseudopalavras. Em relação à leitura das palavras, a justificativa para tal achado pode estar no perfil de nossos participantes, os quais, já com avançado nível de aprendizado do sistema de escrita, fazem a leitura mais pela rota lexical do que pela fonológica. Essa última, em geral, é a rota em que mais facilmente se observa o efeito de extensão, de acordo com o que está documentado na literatura. Contudo, como tampouco observamos influência da extensão na conversão de pseudopalavras, recomendamos que novos estudos ampliem o escopo de proficiência dos participantes.
View
... 2.1. Identificación y escritura de palabras: el modelo de doble ruta Los modelos sobre los procesos en la producción ortográfica están estrechamente ligados a los modelos explicativos de la lectura de palabras, siendo el más ampliamente referenciado el modelo llamado de doble ruta (Coltheart, Curtis, Atkins y Haller, 1993;Coltheart, Rastle, Perry, Langdon y Ziegler, 2001). Este modelo incluye una ruta fonológica o indirecta que permite leer y escribir palabras a condición de que resulte adecuada la traducción grafema-fonema (G-F) y fonemagrafema (F-G). ...
Mecanismos implicados en la escritura de palabras del castellano
Article
Full-text available
  • Dec 2014
El artículo examina los mecanismos cognitivos implicados en la escritura de palabras en lengua castellana. Este tema ha sido ampliamente estudiado en sistemas ortográficos inconsistentes/profundos como el inglés y el francés, pero mucho menos en sistemas consistentes/superficiales como el del castellano. Los datos analizados sugieren que los dos mecanismos de escritura que constituyen la base modelo de la doble ruta: el fonológico y el ortográfico-lexical, intervienen en la escritura de palabras. En comparación con otras ortografías menos consistentes, la adquisición de estos dos mecanismos es más rápida en castellano. Este resultado es conforme a las predicciones del modelo de autoenseñanza. Complementariamente estudiamos otros dos mecanismos implicados en la escritura de palabras: uno ortográfico de carácter subléxico, basado en la frecuencia de regularidades ortográficas, y el otro basado en el conocimiento morfológico. Los datos a este respecto son escasos en castellano. No obstante, los resultados muestran que los aprendices adquieren rápidamente rutinas ortográficas basadas en regularidades ortográficas subléxicas, y que utilizan conocimientos morfológicos para escribir palabras. Esto muestra que ambos factores intervienen en la escritura en todos los sistemas ortográficos alfabéticos, sea cual fuere su grado de consistencia-profundidad. Las diferencias entre sistemas no parecen ser cualitativas sino cuantitativas. Además, todos los mecanismos examinados en castellano se adquieren más rápidamente que en los sistemas menos consistentes.
View
... Seidenberg and McClelland's study (as cited in Coltheart, Curtis, Atkins, & Haller, 1993) provided a glimpse into the world ...
View
... Seidenberg and McClelland's study (as cited in Coltheart, Curtis, Atkins, & Haller, 1993) provided a glimpse into the world ...
View
... For instance, a person can quickly make sense of the letter combination 'Biddle' as an inaccurate word despite it not being in their mental lexicon. In such a case, the strategy used to access word meanings from a written text without phonological representation is the OC [31]. After analyzing the mental lexicon, it indicates that this word does not have a corresponding item in that lexicon. ...
Iraqi EFL university students' word recognition strategies and vocabulary proficiency: A correlational study
Article
  • Mar 2024
Word recognition is a cognitive activity widely acknowledged as a hallmark of fluent L2 readers. Understanding a text can only be accomplished with the ability to decode printed words efficiently. Another factor affecting students' comprehension is vocabulary. Therefore, this research aims to investigate whether there is any significant relationship between word recognition strategies and vocabulary knowledge among Iraqi university students. This study made use of four instruments to collect the data. A comprehensive framework for evaluating students' vocabulary proficiency consists of two tests: one measures the reception and production of vocabulary information and the other assesses orthographic coding and phonological decoding procedures. The results show that the respondents use both strategies to an average degree with over-reliance on the orthographic one in recognizing English words by employing a descriptive correlational research design with 400 respondents. Furthermore, a positive relationship was observed between word recognition strategies and vocabulary proficiency.
View
Brain mechanisms of Chinese word reading
Chapter
  • Jun 2018
This book provides readers with a unique array of scholarly reflections on the writing systems of Chinese, Japanese, and Korean in relation to reading processes and data-driven interpretations of cross-language transfer. Distinctively broad in scope, topics addressed in this volume include word reading with respect to orthographic, phonological, morphological, and semantic processing as well as cross-linguistic influences on reading in English as a second language or a foreign language. Given that the three focal scripts have unique orthographic features not found in other languages – Chinese as logography, Japanese with multi-scripts, and Korean as non-Roman alphasyllabary – chapters expound script-universal and script-specific reading processes. As a means of scaling up the body of knowledge traditionally focused on Anglocentric reading research, the scientific accounts articulated in this volume importantly expand the field’s current theoretical frameworks of word processing to theory building with regard to these three languages.
View
View
View
Surface Dyslexia and Dysgraphia: Dual Routes, Single Lexicon
Article
Full-text available
  • Jun 1992
  • NEUROCASE
The dual route interpretation of surface dyslexia as a deficit in word-specific activation has been challenged recently by computational models that incorporate a unitary print-to-sound mechanism. The most current of these makes no allowance for word-specific nodes, but obtains the pronunciation of regular and exception words by weighted connections between graphemic and phonemic units. Damaging the model in a variety of ways produces a pattern that appears consistent with the performance of many surface dyslexic patients. Exception words are mispronounced more often than regular words, though accuracy deteriorates on both categories. In addition, frequency has no clear-cut effect on the probability of reading an exception word correctly. We describe the existence of a variant of the syndrome that is not fully captured by the above simulations. MP, a surface dyslexic, demonstrates a dissociation between lexical and nonlexical pronunciation of written words. We also show that performance on irregular words varies as a function of their frequency. We provide evidence that the locus of the subject's deficit arises at the level of the representations in a single orthographic lexicon that subserves both reading and writing.
View
On the connection between connectionism and data: Are a few words necessary? Psychological Review, 97, 432-446
Article
Full-text available
  • Jul 1990
Parallel distributed processing (PDP) models represent a new and exciting approach to the study of visual word recognition in reading. M. S. Seidenberg and J. L. McClelland's (see record 1990-03520-001) model is examined because the strongest and widest claims for the viability of a connectionist account of visual word recognition have been made on the basis of their model. The current implemented version of their model fails to account for important facts concerning how human subjects read aloud and carry out lexical decisions, despite the fact that these tasks are central to the performance domain that the model purports to explain. The incorporation of multiple routines and an explicit lexical level of representation into the model may help resolve some of the difficulties. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
View
View
View
More words but still no lexicon: Reply to besner et al. (1990)
Article
  • Jan 1990
The major points in the Besner, Twilley, McCann, and Seergobin (1990) critique of the Seidenberg and McClelland (1989) model are addressed. The model's performance differs from that of people in ways that are predictable from an understanding of the limitations of the implementation. The principal limitations are the size of the training corpus and the phonological representation. The issue of pseudohomophone effects is discussed, and Besner et al.'s new data are shown to be consistent with the Seidenberg and McClelland account of lexical decision.
View
Automatic phonetic activation in silent word reading: Evidence from backward masking, Journal of Memory and Language 27: 5970
Article
  • Feb 1988
Visual access to a printed word may be accompanied by a very rapid activation of phonetic properties of the word as well as its constituent letters. We suggest that such automatic activation during word identification, rather than only postlexical recoding, routinely occurs in reading. To demonstrate such activation, we varied the graphemic and phonetic properties shared by a word target and a following pseudoword mask. Graphemic (mard) and homophonic (mayd) masks, equated for number of letters shared with a word target (made), both showed a masking reduction effect relative to a control mask. There was an additional effect of the homophonic mask over the graphemic mask, attributable to phonetic activation. A second experiment verified this pattern of mask reduction effects using conditions that ruled out any explanation of the effect that does not take account of the target-mask relationship. We take the results to suggest that a phonetic activation nonoptionally occurs (prelexically) during lexical access.
View
View
View
Stochastic interactive processes and the effect of context on perception*1
Article
  • Feb 1991
The effects of context on perceptual identification responses given without time pressure are well-described by classical models in which contextual and stimulus information exert independent effects. A recent article by Massaro (1989) raises the possibility that interactive models, such as the TRACE model of speech perception, are inherently incompatible with these classical context effects. The present article shows that this incompatibility hypothesis can be rejected. Mathematical analysis and computer simulation methods are used to show that interactive models can exhibit the classical effects of context, if there is variability in the input to the network or if there is intrinsic variability in the network itself. A variety of interactive models which incorporate variability can all produce the classical context effects, at least under some conditions; the conditions are rather general in the case of one of the variants. The findings suggest that interactive models should not be viewed as alternatives to classical accounts, but as hypotheses about the dynamics of information processing that lead to the global asymptotic behavior that the classical models describe.
View
An Interactive Activation Model of Context Effects in Letter Perception: Part 2. The Contextual Enhancement Effect and Some Tests and Extensions of the Model
Article
  • Jan 1982
The interactive activation model of context effects in letter perception is reviewed, elaborated, and tested. According to the model, context aids the perception of target letters as they are processed in the perceptual system. The implication that the duration and timing of the context in which a letter occurs should greatly influence the perceptibility of the target is confirmed by 9 experiments with 128 undergraduates, demonstrating that early or enhanced presentations of word and pronounceable-pseudoword contexts greatly increased the perceptibility of target letters. According to the model, letters in strings that share several letters with words should be equally perceptible whether they are orthographically regular and pronounceable (SLET) or irregular (SLNT) and should be much more perceptible than letters in contexts that share few letters with any word (XLQJ). The prediction was tested and confirmed. Overall results are accounted for, with some modification of parameters, although there are some discrepancies in detail. Several recent findings that seem to challenge the model are considered, and a number of extensions are proposed. (34 ref)
View