RT list: Fwd: Non-member submission from Marina Terkourafi <mt217@cam.ac.uk>

From: Nicholas Allott (n.allott@ucl.ac.uk)
Date: Mon Oct 11 2004 - 16:03:57 GMT

  • Next message: Richard Breheny: "RT list: CfP: Experimental Pragmatics Conference"

    > In response to Hanno Beck’s question regarding when and how social
    > information is taken into account during utterance interpretation:
    > Both examples cited,
    > (1) "The Melvin Hall dormitory is on fire."
    > (2) "The dean wishes to speak with you."
    > amount to requests for the addressee to do something (e.g. “Do
    > something
    > about putting the fire out” in case (1), “Go and see the dean” in case
    > (2)). As I understand the problem raised by Hanno, the question is not
    > so
    > much whether the hearer will arrive at these intended implicatures (the
    > request reading of these utterances), but whether s/he will decide to
    > act
    > on them, which to me is a question not about the implicatures
    > conveyed, but
    > about the perlocutionary effects of these utterances. I.e. in both
    > cases,
    > there is no question about the sincerity of the speaker’s intention to
    > get
    > the addressee to do something (notwithstanding the possibility that the
    > speaker knows the addressee thinks the speaker is unreliable, joking
    > etc.).
    > The question is, once the addressee has understood that the speaker
    > wants
    > to get him/her to do something, on what grounds does s/he go on and do
    > it?
    > Nevertheless, we may begin one step earlier, and wonder if there is any
    > chance the addressee may not arrive at the intended request readings.
    > This
    > seems to me to depend on the extent to which the linguistic expression
    > used
    > each time is conventionalised/ stereotypical for expressing the
    > intended
    > propositional content. E.g. saying, instead of (1),
    > (1’) “The Melvin Hall dormitory is in the process of being eliminated
    > by
    > fire.”
    > or, instead of (2)
    > (2’) “The dean desires to converse with you”
    > which express more or less the same propositional content, sound less
    > stereotypical than (1) and (2) respectively. [This may be a matter of
    > the
    > frequencies of the individual lexical items used (which, in this case,
    > correlates with matters of register-specificity, semantic generality/
    > versatility etc.) or of the expressions as a whole. It is also a
    > matter of
    > individual speaker style—e.g. whether the addressee knows the speaker
    > to
    > have a preference for contrived expression, again, a matter of
    > frequency of
    > particular items/constructions, but this time in the individual
    > speaker’s
    > speech. In all cases, this intuition has something to do with previous
    > experience, and should in principle be empirically verifiable, e.g.
    > using
    > corpus data etc.]
    > The addressee may not arrive at the intended request readings if (1’)
    > or
    > (2’) are used, and in any case, if s/he arrives at these readings, it
    > will
    > be after greater processing effort, since s/he will have to figure out
    > why
    > a longer/unusual expression was used. It seems to me that the immediate
    > association of (1) and (2) with the request readings can be nicely
    > captured
    > via Levinson’s (2000) I-principle (‘What is simply described is
    > stereotypically and specifically exemplified’) and the not-so-immediate
    > association of (1’) and (2’) with the same readings via its
    > complementary
    > M-principle (‘Marked descriptions warn “marked situation”’). I have
    > proposed (Terkourafi 2001, 2003) that the intuition of
    > stereotypicality can
    > be propositionally represented as a belief (e.g. “when person
    > such-and-such
    > (this type of speaker, or this particular (token) speaker) says
    > such-and-such s/he is trying to get this type of result) that can
    > participate in the inferential process and determine its outcome.
    > Now, to the problem of whether to act on the request readings. Previous
    > experience is of course important also in this case, as is the notion
    > of
    > stereotypicality. If the addressee has no previous experience of this
    > particular type of speaker or this individual (token) speaker lying,
    > then
    > s/he may well believe him/her, i.e. form the belief that “The Melvin
    > Hall
    > dormitory is on fire”, or “The speaker wishes to speak to me” without
    > further ado. [NB: forming this belief is the perlocutionary effect
    > sought
    > by the speaker, but totally up to the addressee]. However, previous
    > experience of the speaker (this particular token speaker, or the type
    > of
    > speaker s/he represents) lying acts as kind of pointer to a marked
    > situation. The addressee may then well expend more effort deciding
    > whether
    > to believe the speaker or not (this also depends on the addressee’s
    > personal interest, e.g. are all of his/her belongings at Melvin Hall?)
    > In both cases, the issue seems to turn more on how to capture the
    > effect of
    > previous experience on the interpretation process—the points at which
    > such
    > experience will enter the process, if at all, and in what way. The
    > proposal
    > made here is that, for theory’s sake, we can represent such experience
    > propositionally and consider that it is drawn upon based on
    > similarities of
    > previously experienced situations (which may be more or less abstractly
    > propositionally represented) and the perceived situation (again
    > propositionally represented). [Of course, storing need not be
    > propositional; it may also be, e.g. visual, and the discovery of
    > similarities may rely on a comparison of visual—perceived and
    > recalled—stimuli].
    > This proposal is compatible with RT, if the notion of perceived
    > markedness
    > is understood as a measure of relevance (cf. Dessalles’s work on
    > newness
    > vs. expectability of information as a measure of relevance; e.g.
    > Dessalles
    > 1998). This proposal amounts to an add-on to both the Neo-Gricean and
    > RT
    > views, inasmuch as a new theoretical construct, representations of
    > experienced contexts to various degrees of abstraction (which can be
    > represented as frames, cf. Escandell-Vidal 1996, Terkourafi 2001) is
    > required, on which the proposed principles (heuristics or the
    > principle of
    > relevance) operate.
    > I hope some of this helps!
    > With best regards,
    > Marina Terkourafi
    > References
    > Dessalles, J.-L. (1998) Altruism, status, and the origins of
    > language.
    > In: Hurford, J. et al. (eds.) Approaches to the evolution of language:
    > social and cognitive bases. Cambridge: Cambridge University Press.
    > 130-147.
    > Escandell-Vidal, V. (1996) Towards a cognitive approach to
    > politeness.
    > In: Jaszczolt, K. & Turner, K. (eds.) Contrastive semantics and
    > pragmatics.
    > Oxford: Pergamon. 629-50.
    > Terkourafi, M. (2001) Politeness in Cypriot Greek: A frame-based
    > approach. Ph.D. thesis. University of Cambridge.
    > Terkourafi, M. (2003) Generalised and particularised implicatures of
    > politeness. In: Kühnlein, Peter, Hannes Rieser & Henk Zeevat (eds.)
    > Perspectives on Dialogue in the New Millennium. Pragmatics & Beyond New
    > Series 114. Amsterdam: John Benjamins. 151-166.
    > --
    > Marina Terkourafi
    > A.G. Leventis Fellow
    > The British School at Athens Department of Linguistics
    > 52, Odhos Souidhias University of Cambridge
    > GR-10676 Athens Sidgwick Avenue
    > Greece Cambridge CB39DA
    > U.K.
    > http://www.cus.cam.ac.uk/~mt217

    This archive was generated by hypermail 2b29 : Mon Oct 11 2004 - 16:04:23 GMT