Page 89 - 6688
P. 89

89
                  [25]  Him: Oh, Mary, I'm glad you're here.
                     Her: What's up?
                     Him: I can't get my computer to work.
                     Her: Is it broken?
                     Him: I don't think so.
                     Her: What's it doing?
                     Him: I don't know. I'm useless with computers.
                     Her:What kind is it?
                     Him: It's a Mac. Do you use them?
                     Her: Yeah.
                     Him: Do you have a minute?
                     Her: Sure.
                     Him: Oh, great.

              The extended interaction in [25] may be called a 'requesting' speech event without a central
              speech  act  of  request.  Notice  that  there  is  no  actual  request  from  'him'  to  'her'  to  do
              anything. We might characterize the question 'Do you have a minute?' as a 'pre-request',
              allowing the receiver to say that she's busy or that she has to be somewhere else. In this
              context, the response 'Sure' is taken to be an acknowledgement not only of having time
              available, but a willingness to perform the unstated action. The analysis of speech events is
              clearly another way of studying bow more gets communicated than is said.
              The usefulness of speech act analysis is in illustrating the kinds of things we can do with
              words and identifying some of the con-ventional utterance forms we use to perform specific
              actions.  However,  we  do  need  to  look  at  more  extended  interaction  to  understand  how
              those actions are carried out and interpreted within speech events.

                                                       Directives

                In order for directives/requests for action to be heard and interpreted as legitimate, they
              must satisfy certain felicity conditions (Gordon and Lakoff 1971:64);
              1. Speaker wants hearer to do act.
              2. Speaker assumes hearer is able to do act.
              3. Speaker assumer hearer is willing to do act.
              4. Speaker assumer hearer would not do act in the absence of the request.
                     According  to  N.Bonvillain,  “directives  are  particularly  sensitive  to  contexts  of
              speaking and to specific social characteristics of the issue and addressee. Their complexity
              stems  from  the  fact  that  a  speaker  should  phrase  requests  so  as  to  have  the  greatest
              likelihood of positive result, namely compliance; but because a social relationship of some
              sort exists between interlocutors (even if it is one of “stranger”), speakers must be sensitive
              to addressees’ feelings. An issuer of directives needs to navigate between two extremes of
              clarity: He must make his request clear enough so that the addressee comprehends the
              directive intent, yet he must also pay attention to the addressee’s needs not to be imposed
              on by a blunt presumption of the speaker’s power (1997:111).”
                     Largely because of the demand directives place on the addressee, and because of
              the fact that they can be realized by a variety of syntactic forms, the choice of directive type
              can express a great deal about the social context of discourse and the relative status of the
              interlocuters, e.g. their age, sex, occupation, and familiarity (Ervin-Tripp, 1976)
   84   85   86   87   88   89   90   91   92   93   94