On the Analysis of Echo Questions
Echo questions are first and foremost questions, i.e. the speaker tries to get the hearer to close an informational gap. But unlike normal questions they convey the impression that the speaker reopens a gap that both speaker and hearer know has already been closed before. This ‘echo’ effect correlates with likewise distinctive formal features: Thus, in (German) echo wh-questions, on which I concentrate, the wh-phrase is positionally free, always bears the nuclear accent, the final contour is typically a rise, and these formal features may combine with all clause types to yield bona fide echo wh-questions. So far, the standard mode of analyzing echo (wh-) questions has been discourse- based, i.e. their particular form and the echo effect are taken to be derivable from their being ‘quotes’ of a previous utterance, with a wh-phrase and/or prosodic features added to mark the re-questioned aspect. The main aim of this paper is to show that this approach is mistaken, and that an analysis is feasible whereby the pragmatic properties of echo (wh) questions are derived from their formal properties, notably their focusing properties, in a non-ad hoc way.