Rajesh Bhatt's coauthor, former syntax guru Roumi Pancheva, will be presenting their paper at GIST, Generative Initiatives in Syntactic Theory, held March 22-23 at the University of Gent. The title and abstract of their talk follows.
Two Superlative Puzzles
We discuss two cases where superlatives interact with relative clauses with puzzling consequences. The first case concerns the distribution of relative readings of superlatives inside relative clauses (e.g., “the boy who Mary likes the most”). We demonstrate that the facts require that the head be interpreted inside the relative clause. The second case concerns the licensing of subject infinitival relatives by superlatives (“the first man to walk on the moon”). Here too we suggest that the head is interpreted inside the infinitival clause but the superlative is external.
Schueler (2005) observes that relative clauses with relative readings of superlatives are non- intersective. Thus, “John is the boy who runs the fastest” does not entail that John is the fastest person; that could be Sally, with John being the fastest only among the boys. Schueler’s account utilizes the head external analysis of relative clauses with an additional mechanism where the comparison class of the superlative is restricted by the external head of the relative clause. This mechanism ensures that the comparison class is only composed of boys, delivering the desired result. However, we show that the problem identified by Schueler also arises in cases where the relative reading of the superlative is dependent on an element other than the relative pronoun (e.g., “John is the boy who MARY likes the most”), yet his treatment doesn’t generalize to these cases. Adopting an analysis where the head is interpreted internal to the relative clause allows for a uniform treatment of all cases of relative readings of superlatives inside relative clauses.
Bhatt (2006) offers arguments that the licensing of non-modal subject infinitival relatives by superlatives requires reconstruction of the superlative to a relative-clause internal position. Inspired by the discussion in Sharvit (2010), concerning the cross-linguistic distribution of such relatives and their temporal interpretation, we reanalyze these cases as infinitival complements of the superlative operator: [the [first [man to walk on the moon]]], i.e., the superlative does not originate inside the infinitival clause but the head does. One of the arguments in support of the first point, namely the clause-internal origin of the superlative, is that relative readings which should be available go missing (compare “MARY likes the tallest man”, which permits a relative reading where of all the relevant people who like men, the man who Mary likes is the tallest vs. “MARY likes the tallest man to walk on the moon”, which lacks such a reading). We derive this result from the fact that the infinitival clause provides the comparison class. Whenever the comparison class is explicitly given, as in “Mary likes the tallest man of the three”, relative readings are absent. Concerning the second point, namely the origin of the head, we show that the temporal dependence between the head and the infinitival clause, is naturally accommodated if the head is interpreted inside the infinitival clause.
To conclude, superlatives provide two different arguments for the head-internal interpretation of relative clauses.