Exhaustivity, Focus (and Incorporation) in Hungarian

The general question: to what extent do we need specific semantic rules to account for differ-ent phenomena such as focus sensitivity, identificational vs. non-identificational focus etc.? For this discussion Hungarian data are specifically interesting, because it has been widely (and very influentially) argued that Hungarian pre-verbal focus has an intrinsic semantic ex-haustive interpretation, while post-verbal focus completely lacks such an interpretation. This is the main motivation for the classic distinction between information and identificational focus (È. Kiss 1998). I would like to show that such a distinction may not be necessary, since both “types of focus” can be modeled assuming the very same semantic mechanism of focus interpretation. 1. Hungarian focus and exhaustivity The major semantic point about Hungarian pre-verbal focus is that it triggers truth conditional differences even without any overt focus sensitive particles, as shown in (1) from Szabolcsi (1981): (1) Nem [Péter]F aludt a padlón, hanem [Péter és Pál]F (aludt a padlón). Not Peter slept the floor-on but Peter and Paul slept the floor-on ‘It isn’t Peter who slept on the floor; it’s Peter and Paul who slept on the floor.’ The standard way of dealing with this peculiarity is to assume some covert exhaustivity op-erator (such as a covert “only ” (formally shown in (2))). This point is common to nearly all approaches to the semantics of Hungarian focus, even if they differ in details: (2) λP λx P(x) ∧ ∀y P(y) → y = x The exhaustivity operator arguably prevents the focused expression of being negated and hence allows for a continuation as in (1). The necessity of the stipulation of such an operator has been discussed widely in the litera