Dietmar Zaefferer - Are German control nouns with infinitival complements inherently definite?

Fifty years ago the second part of Gunnar Bech’s influential ‘Studies on the German Verbum Infinitum’ (Bech 1955/1957) appeared. One of the less discussed proposals of this text is the distinction between explicative (1) and implicative (2) constructions among the governed infinitivals of German:

(1)    Die Fähigkeit, zaubern zu können
        The ability do-magic to be-able
(2)    Die Fähigkeit, zu zaubern
         The ability to do-magic

In (1) the ability consists in being able to do magic, in (2) it relates to doing magic. In the first case Bech assumes that the infinitival ‘explicates’ the definite article in that the latter cataphorically refers to the former, in the second case there is no such relation between article and attribute. Although the proposal appears somewhat obscure, it inspired Fabricius-Hansen and von Stechow to attempt an explication of Bech’s intuition in their analysis of extended nominals in German (1989), which analyses the infinitival in (1) as an adjunct and in (2) as a complement. It is argued that Restle (2006) is right in rejecting this analysis and proposing a purely semantic approach instead which treats along the lines of genetives which also can have defining/explicating (the vice of alcoholism) and other interpretations. Based on this approach a more fine-grained ontology is proposed that predicts the definiteness behavior of German control nouns with infinitival complements by assuming two things. First, the explicative-implicative distinction is reconceptualized in terms of a direct versus indirect specification, where the latter means creation of a subconcept via saturation of a participant role. Second, it is assumed that control nouns code functional concepts which yield unique concepts when applied to concepts coded by infinitivals: There are many abstract abilities, but there is only one abstract ability to do magic. And there are many instantiations of specified abstract abilities in concrete persons, by only one in each person. This explains Bech’s intuition that the definite article is required in both cases. Finally, an explanation is given for the fact that indefinite nominals of the given kind are negative polarity items:

   (3)    Von einem Recht, so etwas zu tun, kann keine Rede sein.
            Of a right such a-thing to do can no mention be.

References

Bech, Gunnar (1955/1957): Studien über das deutsche verbum infinitum. Historisk-filologiskemeddelelser/udg. af Det Kongelige Danske Videnskabernes Selskab 35:2 (Bd. 1)und 36:6 (Bd. 2). 2. unveränderte Auflage 1983. Tübingen:Niemeyer.

Fabricius-Hansen, Cathrine und Arnim von Stechow (1989). Explikative und implikative Nominalerweiterungen im Deutschen. Zeitschrift für Sprachwissenschaft 8,173-205.

Restle, David (2006). Kontrollnomina. Eine Untersuchung zum Verhalten attributiver Infinitivkonstruktionen im Deutschen (Habilitationsschrift Universität München)