Permalink : http://hdl.handle.net/2115/37063

This paper presents the results of a statistical survey of the ways in which let's imperative sentences function in (American) English conversation. Unlike ordinary imperatives, imperatives with let's suggest a shared action by speaker(s) and addressee(s) and normally involve a verbal response indicative of agreement or refusal. However, notable exceptional cases have been noted (Biber et al. 1999: 1117, Huddleston and Pullum 2002: 936 fn., Collins 2004: 302). The present paper focuses on the following questions that remain unanswered: (i) How frequent is the "joint action" use of let's-imperatives in conversation? (ii) How frequent is the case of "action by just one" use, and whose action is more common, the speaker's or addressee(s)'s? (iii) How frequent is the discourse-organizational use of let's as realized in the form let's say or let's see? (iv) How frequently are let's-imperatives actually followed by a verbal response? While let's-imperatives have been extensively analyzed in both descriptive and theoretical terms, relatively little quantitative work has been conducted, although important exceptions include Aarts 1994, and, most notably, Collins 2004. On the basis of 133 tokens of let'simperatives obtained from dialogues in four mystery novels, this paper reports the following findings:(i) the joint action use accounts for nearly 80% of the data; (ii) when let's-imperatives suggest an action by just one party, the case of the addressee's action is a great deal more frequent (14 out of 16) than the speaker's; (iii) the discourse organizational use with the form let's say or let's see accounts for 9.9% of the data (13 out of 133 tokens); and (iv) in only 31% of the data were let's-imperatives followed by explicit verbal response. The last finding suggests that, contra Huddleston and Pullum (2002: 936), the case of response-less let's-imperatives is hardly the exception but rather the norm in actual conversation.