Adjectives and usage-patterns in the [X enough to VERB]-construction

Recent constructionist research into adjectival gradability suggests that scalar adjectival constructions may feaure implied relations of causality (e.g. Bergen & Binsted 2004, Fortuin 2013) The [X enough to VERB]-construction appears to be one such construction:

(1)A couple of them were just puppies, but old enough to know it was good-bye. (COCA 2012 FIC AntiocRev)
(2)But every time I try to get close enough to touch one, the mother shows her yellow teeth and growls. (COCA 2012 FIC Bk:IntoFreeNovel)
(3)With its 18-inch barrel and collapsible stock, it is quite maneuverabe and is light enough to carry with a scope mounted. (COCA 2012 MAG OutdoorLife)

The above examples suggest that the construction involves an implied pragmatic relation of force-dynamics between [X enough] and the infinitive clause, based on Johnson's (1987) force-dynamic image schema of enablement (Fortuin 2013 calls this function sufficiency). In this paper, we will explore the construction in the perspective of usage-based construction grammar with a view to gaining insight into its functionality and mapping some of its usage-patterns. Making use of data from the 2012-section of Davies' (2012) COCA, a number of corpus methods are applied in our anaysis of the construction. For instance, a covarying collexeme analysis (Stefanowitsch & Gries 2004) indicates that the construction does indeed set up a relation of enablement, with semantically coherent pairs such as cool-handle, small-fit, heavy-sink, smart-know, and sensitive-detect appearing among the most strongly coattracted lexeme pairs in the construction in the corpus. While the pairs mentioned above, and the examples in (1-3), feature adjectives in the X-position, we may also find nouns and adverbs in that position:

(4)I hope you guys will be gentlement enough to come and find me in heaven and tell me you're sorry. (COCA 2012 SPOK NBC_Dateline)
(5)Not even light travels fast enough to escape it. (COCA 2012 NEWS CSMonitor)

A simple frequency analysis of the distribution of the three word classes in the X-position suggests that adjectives (n = 720 76.7%) are the most likely candidate to occur, followed by adverbs (n = 202, 21.5%), and then nouns (n = 17, 1.8%). Given that the propositional act function of both adjectives and adverbs is modification while that of nouns is reference (Croft 2003: 184-188), it might be tempting to set up a dichotomy between [ADJ/ADV enough to V] and [N enough to V]. However, if we take into consideration syntactic functions, we find that [ADJ enough to V] and [N enough to V] behave alike, appearing primarily in predicative functions, while [ADV enough to V] appears in adverbial functions. This suggests that [X enough to V] covers two constructions: an adverbial one and an adjectival one. In the latter, gradable adjectives prototypically appear in X-position, and nouns appear to adopt the scalar adjectival features of the construction, as seen in (4). It is further interesting to note that, with a preference for predicative syntactic functions, this construction falls under Croft's (2003: 185) of constructions which are marked codings of the propositional act function of predication.