Editor for this issue: Marie Klopfenstein <marie
linguistlist.org>
I thank the following for their interesting responses to my query of last week on unaccusative imperatives: Werner Abraham (hello Werner!) Julian Bradfield Esmail Hosseini Michael Johnstone Tom Joyce Jeff Lilly Margaret Lubbers Andrew McIntyre Peter A. Michalove Bill Morris Bert Peeters (hello Bert!) And Rosta (hello And!) Michael Swan David Wilmsen There was a high degree of agreement among respondents, though some introduced ideas or arguments that no one else did. The general conclusion is that (a) the suggested constraint does not exist and (b) to the extent that it does it is 'pragmatic' rather than 'structural'. In support of (a), many counterexamples were produced including the following: (2) Open Sesame. (3) Get well soon. (4) Wake up. (5) Freeze! (6) Die, you (dirty) dog! (7) Fall off a cliff, you sleazoid scumbag! (the _you X_ phrase appears a great facilitator of acceptability, for reasons that remain unclear!) Further, some respondents noted that (one argument) verbs of motion are generally considered unaccusative, and that they form imperatives freely. Levin (like some others) does not count the verbs of motion among the change of state verbs: (8) Come here. (9) Go away. (10)Arrive at noon. In support of (b), it was pointed out that the behaviour of the motion verbs rules out a purely structural account (at least for believers in syntactic unaccusativity). The fact that subjects of unaccusative verbs are projected from internal arguments was held to rule out their referring to agents. However the semantic properties of unaccusative subjects are determined, they are in conflict with those of the 'understood subjects' of imperative verbs, which typically refer to (volitional) agents. Note that of the examples in (2-7) all but one are addressed to humans and the remaining one (2) is addressed to a magically autonomous door (I was also referred to Star Trek style voice-activated doors). Clearly, subjects of motion verbs can be construed as being in (volitional, agentive) control of the motion, which explains why the constraint does not apply to them. It was suggested by a couple of respondents that examples like (3) and my original (1) (_Boil you blighter!_) represent a marginal (non-prototypical) class of imperatives which refer to wishes rather than instructions. I received one reference (relevant to the question whether internal arguments can be agents): Abraham, W. "The aspect-case typology correlation: perfectivity and Burzio's generalisation." Eric Reuland (ed). _Arguments and Case_. Amsterdam: John Benjamins, 131-194. Thanks again to all, Jasper.Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue