Singh, S. & Mishra, R. K.
Center of Behavioral and Cognitive Sciences, University of Allahabad, India
This study explores the role of context in lexical ambiguity resolution in a lesser studied and typologically different language Hindi, on the basis of views put forward by selective access model (Kellas; 1994) and reordered access model (Duffy, Rayner et al; 2001), which claim the context selective activation and simultaneous activation of lexical items respectively. Some studies show the primacy of context over lexical input while other claim that lexical items are activated independent of context. This study explores the issues put forward by Huettig and Altmann(2005), which support the simultaneous meaning activation of homophones independent of context, by making assumptions based on Embodied cognition. In a set of two eye tracking studies employing visual-world paradigm, where auditory input included the neutral sentences and sentences biased towards the dominant meaning of homophones, the visual images included homophone referent (shape and semantic competitor of subordinate meaning) along with three distracters.
Experiment 1 included the presentation of a shape competitor of the subordinate meaning of the homophone in both neutral and biased conditions, while we presented semantic competitors of the subordinate meaning of the homophone in the second experiment. The eye movement data obtained in both the experiment provide strong evidence that participants can quickly activate the subordinate meanings and also their competitors on several dimensions both in neutral as well as biased sentential contexts. Overall these data from Hindi provide further support to those models of lexical access that assume a parallel activation of both the meanings of an ambiguous word. We further argue for a model of lexical access where sentence context plays little role in lexical access during online spoken sentence processing.