Max Coltheart vs Mark Seidenberg

The role of semantic information in reading aloud
Chair: Jeffrey Binder

Thursday, November 7, 5:45 – 7:15 pm, Crystal Ballroom

Max Coltheart was born in Frankston, near Melbourne, and grew up in various small towns in Australia. He studied psychology and philosophy at the University of Sydney and went on to complete a PhD in psychology there. He then held academic positions at Monash University, the University of Waterloo, the University of Reading and the University of London before taking a position as Professor of Psychology at Macquarie University in 1987, and then moving from that position to be Director of the Macquarie Centre for Cognitive Science (an Australian Research Council Special Research Centre) from 2000 until 2009. Macquarie University awarded him a DSc in 2001 and an Honorary LLD in 2010, and he was appointed a Member of the Order of Australia in the 2010 Australia Day honours for his services to children with learning difficulties. He is a Fellow of the British Academy, the Australian Academy of Science, and the Academy of the Social Sciences in Australia. He is currently a member of the Centre for Cognition and its Disorders at Macquarie, working on the cognitive neuropsychology and computational modelling of reading, and also on several topics in cognitive neuropsychiatry, especially delusional belief.

Abstract: I will consider evidence from cognitive neuropsychology, computational modelling and experimental psychology which I take to support the view that there are distinct lexical and nonlexical routes from print to speech that subserve reading aloud, and that within the lexical reading route one can distinguish a lexical but nonsemantic processing route (direct communication from visual word recognition to spoken word production) and a lexical-semantic processing route (communication from visual word recognition to the semantic system followed by communication from the semantic system to spoken word production). According to this framework, any word can be read aloud without any contribution from the lexical-semantic processing route, so the question of the role that semantic information actually plays in reading aloud is an empirical one; I will discuss evidence relevant to this open question.


Mark SeidenbergMark S. Seidenberg is the Hilldale Professor and Donald O. Hebb Professor in the department of psychology at the University of Wisconsin. He is a cognitive neuroscientist who has  studied reading and language since the disco era.  Seidenberg received a Ph.D. and several other degrees from Columbia University, where his ideas about language were greatly influenced by Nim, the chimpanzee who was being taught sign language and mostly talked about food, a shared interest. His reading research addresses the nature of skilled reading, how children learn to read, developmental reading impairments, and the brain bases of reading, using the tools of modern cognitive neuroscience: behavioral experiments, computational models, and neuroimaging.  Current research focuses on the “achievement gap,” the chronically low reading performance of poor and minority children in the US.  His reading research is tied to parallel issues concerning language, particularly the statistical structure of language and its role in determining the character of linguistic information (e.g., morphology), in language acquisition, and in age-related changes in language learning abilities.

Abstract: Reading involves learning to compute the meanings of words from print; being able to read aloud is just a by-product. Characteristics of reading aloud are therefore determined by how people solve the reading problem, as well as by characteristics of the orthography-phonology mapping, which vary across writing systems, and individual differences, which may be constitutional or experiential in origin.  These factors determine the “division of labor” between different components of the lexical system relevant to tasks such as reading aloud,  giving rise to a variety of effects, including semantic influences on reading aloud.  I’ll consider relevant empirical evidence and related issues concerning the adequacy of competing computational models of word naming and reading.