Jennifer Hinnell (SFU)
Sunday, May 20
In the study presented here I take a usage-based, construction grammar (CG) view of meaning (Fillmore, Kaye & O’Connor 1988;; Goldberg 1995, 2006; Kemmer and Barlow 2000), which assumes that linguistic meaning is a property of both lexical items and grammatical constructions. This study investigates the widely accepted pattern of Canadian English be done+NP (eg. I am done my homework), and argues that a construction is required to account for the semi-idiomatic aspects of the syntax, semantics, and dialectal variation of the pattern. I motivate properties of the be done+NP construction, providing empirical evidence to support the theory that constructions are learned form-function pairings based on schematic mental representation. In a CG approach, the existence and common usage of two variants suggests two separate form-function pairings, and this is borne out in the study presented here.
The pattern: Both American and Canadian English usage contains the pattern be done –ing (examples 1-3) and be done+PP(with) (4-5). However, in Canadian English the pattern also includes be done+NP (6-10), which are highly unacceptable to all American English speakers except for small pockets in northeastern Vermont and Pennsylvania (Yerastov 2008). The construction analysis offered in this paper focuses on the syntactic and semantic differences of the Canadian variant be done+NP in contrast with the variant that is acceptable in both major dialects, be done+PP(with). Data was gathered with WebCorp, a linguistic search engine for accessing the web as corpus (http://www.webcorp.org.uk). The Canadian data was retrieved from Canadian blogs and websites (restricting the domain to .ca), and the American data was collected from American newspapers, blogs and Wikipedia entries using WebCorp as well.
Findings: The data supports previous research that both variants are robust and common in Canadian usage, whereas American usage is restricted to be done+PP(with). In addition it was found that the two Canadian variants are not interchangeable, but rather express different aspectual information. Lastly, the be done+NP variant of Canadian English appears to have a low degree of schematicity (grammatical productivity) and a high degree of lexical specificity, where the NP is most likely to be a lexical item in an educational program frame, or related to household activities such as cleaning or eating. I take this as a starting point for explorations of how constructions evoke frames, and, alternately, how frames constrain constructions. I suggest that the be done+NP construction evokes multiple frames (the domestic/educational ones just mentioned, as well as a ‘completion’ frame) and thus represents a blend. Lastly I show that the construction also contains elements of viewpoint which are not present in related constructions; these facts related to viewpoint are captured in a frames and construction approach.
Conclusions and future research: This study shows that be done+NP encodes different types of semantic and syntactic information than be done+PP(with). Furthermore, the corpus-based, empirical evidence supports a constructionist view of mental representation. As American English lacks the be done+NP form of the construction, future research will investigate how American English usage expresses the semantics that Canadians use that construction to convey.
(See the attached PDF for a charts and references)