This site © 2007 by the Faculty of English, University of Cambridge, CB3 9DP
Cynthia Allen
School of Language Studies, The Australian National University
Email: Cindy dot Allen at anu dot edu dot au
This paper examines a little-studied construction in which a possessive is followed by a determiner, e.g. on þinre þære sweteste lufa. This Poss Det construction is common in OE manuscripts and is found in some twelfth century manuscripts. My investigation shows that none of the twelfth century examples are from texts which were composed after 1100, suggesting that the later examples were the result of copying and do not reflect a construction which was productive anymore. Two of the post-Conquest examples also lack an adjective, a feature found in all 250 examples in the manuscripts from before the middle of the eleventh century. Comparative evidence from modern Scandinavian varieties where a Poss Det construction is found provide evidence that the lack of adjectives in the OE texts reflects a genuine syntactic fact, rather than a data gap. If we analyse possessives as determiners, rather than adjectives, the OE ‘adjective constraint’ can be explained by assuming that adjective phrases as well as noun phrases had a slot for a determiner, a relic of the origin of weak adjectives as nominalisations. The most plausible explanation for the fact that the adjective is missing only in late examples is that scribes were attempting to reproduce an archaic construction which was no longer part of their productive grammar. Since numerous morphological and syntactic changes took place within the noun phrase in Early ME, a number of hypotheses about why the Poss Det construction died out could be suggested. Syntactic typology can help us to eliminate some hypotheses, since some initially plausible hypotheses do not explain why Poss Det has survived in some Scandinavian varieties but not in English. In fact it does not seem possible to isolate any morphological or syntactic change that could have served as a ‘trigger’ for the loss of Poss Det. A more plausible explanation is that Poss Det is typologically a marked construction, and marked constructions require robust evidence to be incorporated into grammars by language learners. An optional stylistic device, Poss Det was not used in OE by all writers and there is some evidence that it was on the decline in late OE. When its use became too infrequent, language learners simply failed to learn that adjective phrases had a slot for the determiner.