An Opinion piece in the January issue of Trends in Cognitive Sciences by Jonas Obleser and Frank Eisner highlights the necessity to derive abstract representations in spoken language comprehension. Saying this (i.e. the necessity of abstraction) often is worthwhile, I think, because there continues to be controversy. Here my $0.02: claiming that there exist abstract representations is not tantamout to denying the existence of episodic/indexical effects. It seems to me that on that topic there is controversy -- but no issue. I think the evidence for indexical effects is strong, and I think the evidence for abstract representations is overwhelming. Both aspects of speech recognition need to be accomodated in any successful theory.
A second aspect of Jonas' and Frank's paper that is helpful is that they point to the recent developments in the analysis of imaging data (well exemplified by Elia Formisano's recent work) that raise the bar in terms of eludicating spatial results.
Now ... a dfifferent issue is whether even such granular spatial analyses get us closer to computational theories of perception a la neural coding. But that's just my thing ...
No comments:
Post a Comment