Früh ( Er ) ‘ Early ( Er ) ’ , Spät ( Er ) ‘ Late ( R ) ’ , and Phase Adverbs

There is a rich literature about the temporal conjunctions before/after, but at the time I gave the talk that underlies this paper I was not aware of any analysis of the temporal comparatives früher/später ‘earlier/later’, which may be used to express similar states of affairs, but are constructed differently. Recently I got acquainted with the del Prete’s thesis about It. prima/dopo, which analyses prima as a comparative and dopo as a preposition. This is the only paper known to me that goes into the same direction as the following proposal. Del Prete’s analysis is very different from mine and I must leave the discussion of his theory to another occasion. The semantics of before/after is notoriously controversial and the semantics of the related adjectives is therefore interesting in itself. A study of the adjectives gains additional interest from the fact that they are entirely differently constructed: they are degree adjectives and have a comparative, an equative and a positive variant. I will study each of them. I will devote particular attention to the positive forms of the adjectives because I have a novel proposal for a positive operator that solves a long standing puzzle in the semantics of degree adjectives, namely the question of how we can give a uniform treatment of the positive form of both poles of antonym pairs such as tall/short. The semantics of my positive operator will shed new light on the analysis of the notoriously difficult phase particles schon ‘already’ and noch/erst ‘still’/‘only’. I will state their presuppositions in terms of ‘early’ and ‘late’. If I were not concerned with these particles, I could have dealt with English alone in this paper. By and large the relevant facts seem to be the same for English and German. But since my proposal for the analysis of schon/erst is different from a recent proposal due to (Condoravdi, 2003), I felt safer with the German examples. As I said, the comparative variants ‘earlier’/’later’ are particularly intriguing because they are temporal comparatives, and time is a dimension that has not been studied in the semantics of degree. Times play a double and perhaps confusing role in the semantics of ‘early/late’. They denote the temporal position of an event and they play the role of degrees of