Agentivity and intentionality -correction

[dan everett]

     My statement on RRG actually overstates the case a bit. RRG
     distinguishes between intentional vs. nonintentional actors in
     principle, but says that most languages will not avail themselves of
     this distinction in most constructions. The claim is that such
     distinctions are normally 'neutralized'. However, RRG allows for the
     flexibility I appealed to in my last posting, by recognizing that in
     some constructions in a one language or in all constructions in some
     other language, this neutralization may fail to hold, leading to
     distinctions in the relevance of intentionality (and whether the roles
     involved are agent, effector, etc.).

     The point is that it is misguided to try to impose a 'universal rule'
     on how this teeny-weeny little bit of intentionality gets
     'linguistified'/grammaticized.

     Dan Everett