COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring. |
University of Cambridge > Talks.cam > RCEAL Tuesday Colloquia > Discrete elements: The essence of language? Notes on the neural side of words and rules
Discrete elements: The essence of language? Notes on the neural side of words and rulesAdd to your list(s) Download to your calendar using vCal
If you have a question about this talk, please contact Chris Cummins. In linguistics, language is described as a discrete system of lexical elements and recursive rules linking them together in sentences and larger chunks of text (Jackendoff, 2002). The notion of discrete representations has, however, been questioned on the basis of neural networks simulating syntactic and semantic processing (Elman et al., 1996). Fully distributed networks, which map probabilities and process sentence structures to a surprising degree, do not seem to develop discrete representations analogous to lexical entries or to combinatorial rules (McClelland & Patterson, 2002). This situation creates a trench between linguistic and neural approaches to language. I will examine whether such a trench is necessary or motivated and find a clear answer: No. We first ask whether the absence of discrete representations is a necessary feature of networks. By using networks that are more brain-like than the standard 2-4 layer networks employed to approach questions of lexical and syntax processing, we demonstrate that both lexical units and syntactic rules emerge if relevant features of brain anatomy and function are mirrored by neuronal networks (Knoblauch & Pulvermuller, 2005; Garagnani et al., 2008; Pulvermuller & Knoblauch, 2009). A digression will highlight how rule mechanisms may interact with neuronal dynamics and feedback regulation to yield a neuronal device equivalent to a pushdown memory (Pulvermuller, 1993; 2003). A second empirical part will ask whether we have evidence for the existence of discrete units at the lexical and syntactic rule level. This will culminate in a brief review of recent results from the Cambridge lab that provide neurophysiological (EEG, MEG ) support for discrete functional units in language processing (Pulvermuller & Assadollahi, 2007; Pulvermuller et al. 2008; Garagnani et al., 2009). In conclusion, computational and neuroscience results show that the brain’s language system includes an assembly of discrete neuronal lexical representations (words and affixes) plus a set of discrete neuronal combinatorial rules. Discrete neuronal words and rules may well be the essence of language. Developing mechanistic models of these entities may be critical for advancing the cognitive neuroscience of language. This talk is part of the RCEAL Tuesday Colloquia series. This talk is included in these lists:Note that ex-directory lists are not shown. |
Other listsSpanish & Portuguese CUED Computer Vision Research Seminars Book LaunchesOther talksAre hospital admissions for people with palliative care needs avoidable and unwanted? Barnum, Bache and Poe: the forging of science in the Antebellum US What can we learn about cancer by modelling the data on it? Cohomology of the moduli space of curves Group covariance functions for Gaussian process metamodels with categorical inputs Dynamics of Phenotypic and Genomic Evolution in a Long-Term Experiment with E. coli Mathematical applications of little string theory Black and British Migration Speculations about homological mirror symmetry for affine hypersurfaces The role of the oculomotor system in visual attention and visual short-term memory A transmissible RNA pathway in honeybees Regulation of progenitor cells in adult lung and in lung cancer |