Abstract for Ferenc Kiefer's talk

A new look at information structure in Hungarian
Ferenc Kiefer (Budapest)
April 7, 2005 (Thursday)
4:15 PM - ; Room 6417, the CUNY Graduate Center

Hungarian sentence structure has been claimed to be based on 'topic-comment' structure rather than on grammatical subject and grammatical predicate (cf., E.Kiss 2002 for a recent discussion of this approach). 'Topic' has been identified as the logical subject of the sentence; at the same time it has also been claimed that 'topic' represents known information. It will be shown that this approach leads to a number of problems. First of all, non-specific noun phrases, too, can function as topics, though they do not represent known information. Second, quantified noun phrases which are not topics can nevertheless be logical subjects. Third, sentences with an indefinite subject NP can be ambiguous: the indefinite NP can either refer to an individual already mentioned in the discourse but it can also denote a completely unknown entity. This ambiguity is left unexplained by the 'topic-predicate' approach. It will be claimed that 'topic' should be distinguished from 'logical subject' and 'comment' from 'logical predicate'. Sentence structure is determined by 'logical subject - logical predicate' structure, 'topic-comment' structure having to do with information structure proper. The two structures may, but need not coincide. It will be shown that the 'two-level' approach has several advantages over the 'one-level' approach.