Learning to reason about speakers' alternatives in sentence comprehension: A computational accountHendriks, P., van Rijn, H. & Valkenier, B., Nov-2007, In : Lingua. 117, 11, p. 1879-1896 18 p.
Research output: Contribution to journal › Article › Academic › peer-review
We present a computational simulation study of the acquisition of pronouns and reflexives. The computational simulation is based on an Optimality Theory analysis, and is shown to account for the well-known observation that in English and many other languages the correct comprehension of pronouns lags behind that of reflexives (the so-called Delay of Principle B Effect). Comprehension is modelled as a two-step process involving optimization from a given form to its corresponding meaning followed by optimization from this meaning to its corresponding form. This model is implemented using plausible assumptions with respect to the cognitive architecture. The computational simulation shows that lack of processing speed causes the model to produce an output before both steps of the comprehension process have been completed. Because, according to the Optimality Theory analysis, the adult interpretation of pronouns is dependent on reasoning about alternative forms and hence on completion of both steps of the comprehension process, whereas the interpretation of reflexives is not, this results in comprehension errors with pronouns but not with reflexives. We conclude that speed of processing may be an essential factor in explaining the Delay of Principle B Effect and other comprehension delays in language acquisition. (C) 2007 Elsevier B.V. All rights reserved.
|Number of pages||18|
|Publication status||Published - Nov-2007|
|Event||Conference on Languaage Acquisition between Sentence and Discourse - , Netherlands|
Duration: 1-May-2005 → …
Conference on Languaage Acquisition between Sentence and Discourse
01/05/2005 → …Netherlands
- Delay of Principle B Effect, language acquisition, Optimality Theory, bi-directional optimization, cognitive modelling, computational simulation, SEMANTICS, KNOWLEDGE, BINDING