To finish in German and Mainland Scandinavian: telicity and incrementality Alexandra Anna Spalek and Kjell Johan Sæbø University of Oslo Abstract Among the words that describe initial or final parts of events, words describing finish- ing stand out in a number of ways: in a language like English, there is a transitive verb which is singularly flexible regarding the type of event retrievable from the context; in a language like German, there is no verb but there is a verbal particle; in either case, there is a requirement of telicity and there is a requirement of theme incrementality. The present paper documents these facts and offers an analysis of the verbal particle. 1 Introduction Aspectual verbs like begin or finish , which take verbal complements, as in (1a) or (2a), but can also take nominal complements, as in (1b) or (2b), have been a subject of attention in formal lexical semantics over some twenty years. 1 (1) a. Before you begin making the cake , heat your oven to 350 degrees and grease and flour a 9 inch round cake pan. b. Now begin the cake by sifting the flour, salt and spices into a large mixing bowl, lifting the sieve up high to give the flour a good airing. (2) a. If your child likes to turn pages before you finish reading the page , that is okay. b. By the time I had deciphered a sentence, my classmates had finished the page . The key observation is that the two underlined phrases convey the same meaning (modulo mood and tense), so somehow, the nominal complement is interpreted as though it were to (in an informal sense) incorporate a verb. The general tendency since Pustejovsky (1995) has been to take these verbs to instantiate logical metonymy and to motivate methods of lexical coercion, and particularly work by Asher (2011) is influential, if not uncontroversial: Egg (2003) and Pi˜ nango and Deo (2016) advance alternative approaches. A question which has not been at the center of attention is whether there are significant differences in how freely aspectual verbs can take nominal arguments; another is whether the pattern seen in (1) and (2) is cross-linguistically stable. 1 Unless otherwise indicated, all examples are authentic or modulations of attested cases. URL source references are omitted for parsimony. 1