RT list: Fresh perspectives needed on my RT-analysis of emphatic "do"

Dear fellow Relevance enthusiasts,

I was overjoyed to discover this email list. I’ve often thought how amazing
it would be to have a “direct line” to Dan Sperber or Deirdre Wilson. Turns
out the fantasy is true!

Let me start by telling everyone a little bit about myself. (To put my
cards on the table, this email is going to be a request for help. I will
take anything I can get.) I received my B.A. in Philosophy from UC Berkeley
in 2008, where I wrote a thesis on Merleau-Ponty’s philosophy of language
under the supervision of Hubert Dreyfus. Since then, I’ve enrolled in a
Master’s program in Linguistics at San Diego State University, where I’ve
been studying Relevance theory independently for the past 2 semesters and
am now writing a thesis. After wandering aimlessly through the terrain (I
wrote one expository paper on RT, and another paper responding to critiques
that RT is “asocial”), I finally settled on my thesis topic- I’m doing a
Relevance-theoretic analysis of the emphatic “do”. (My professor suggested
that doing an analysis of one specific construction would be the wisest
thing to submit to Linguistics departments for Ph.D. admissions).

Needless to say, I’ve been reading up on a ton of prior work. I thought I
had it all figured out (with respect to emphatic “so”), but then I looked
closer at my analysis and the walls came tumbling down. Now I’m trying to
save what I can without making any major changes (but if major changes are
in order, I’m prepared to do what it takes.)

Let me quickly go through the stage-setting and then I’ll present the
problem (Don’t worry too much about the preliminaries- just skim through
this part.). First, I specify what sense of the emphatic “do” I’m
interested in and I distinguish it from some neighboring senses. I’m
specifically interested in those cases of “implicitly contrastive” emphatic
“do”-- so this rules out “explicitly contrastive” cases, and
“non-contrastive” cases like do-imperatives and the use of emphatic “do”
for politeness. Explicitly contrastive cases are just like they sound: the
verb-complement introduced by the “do” contrasts in some obvious way with
an earlier proposition (or “assumption”, in RT terms) in the discourse.
Some (fairly simple) examples from the British National Corpus:
(SP:PS0WS)<http://corpus.byu.edu/bnc/x4s.asp?s=(SP:PS0WS)>I don't
fight with the girls.
(SP:PS0WN) <http://corpus.byu.edu/bnc/x4s.asp?s=(SP:PS0WN)> You *do*
*fight*with the girls!; .
(SP:KPGPSUNK) <http://corpus.byu.edu/bnc/x4s.asp?s=(SP:KPGPSUNK)> Jo! Don't
even bother joining in.
(SP:PS555)<http://corpus.byu.edu/bnc/x4s.asp?s=(SP:PS555)>Yes, I
*do* *bother*!

Implicitly contrastive cases, on the other hand, are where the the contrast
is between the verb-complement introduced by the “do” and one or more
*inferences* arising from earlier utterances in the discourse. So, for
example,

   1. housing committee's considered this matter and the vast majority of
   people found it fairly unacceptable but we'll ignore that, we'll ignore
   that, fine, well I s'pose if you want to ignore it that's fine but
   (pause) I think (pause) w-- we *do* *face* here a a key issue
   2. (SP:PS645) <http://corpus.byu.edu/bnc/x4s.asp?s=(SP:PS645)> Criminals
   aren't too bothered about the er problems they cause victims in the
   commission for their crime, but er as a general rule er they *do*
*draw*a line somewhere,
   3. (SP:PS1VS) <http://corpus.byu.edu/bnc/x4s.asp?s=(SP:PS1VS)> Now of
   course these sort of situations *do* *cause* a few problems,

I take it as a given that everyone has intuitions about these implicitly
contrastive cases. I formulate a working definition and I illustrate it
with some examples from the BNC. Following this, I consider some semantic
analyses of emphatic “do”. Emphatic “do” is problematic for
truth-conditional semantics (I take it for granted) because it doesn’t seem
to contribute to the truth-conditions of the utterance yet it contributes
meaning of a sort. I show that all the proposed semantic accounts (like
Verum focus or Karttunen and Peters’ well-known analysis where they couch
conventional implicatures in Montague grammar) are insufficient because
they all fail to take account of the speaker’s belief about the hearer’s
belief (or context).

Next, I offer a hypothetical Gricean analysis for comparison with the RT
analysis. Grice would say that emphatic “do” introduces a conventional
implicature with the form “p contrasts in some way with q” (where p is the
proposition following the “do” and q is an inference arising from r, an
earlier proposition in the discourse-- also I realize that the material
after “do” is a verb-complement, not a proposition, but I’ll make
provisions for this elsewhere in the paper). I offer reasons to support
this claim (against the claim that it introduces a conversational
implicature). Among these are the fact that the contrast is part of the
meaning of emphatic “do”.

Then I move on to the central part of the paper-- a Relevance-theoretic
analysis of the emphatic “do”. So far, this is what I have: I want to argue
that emphatic “do” has two roles in the utterance interpretation process.
First, and less importantly, it generates (or “introduces”, if you are
being nit-picky) a higher-level explicature on the model of “believes” or
“regrets”, that the Speaker “insists” that p. The higher-level explicature
will come into play later.

Second, I claim that emphatic “do” is a type of procedurally encoded
information (as innovated by Blakemore). Rather than contributing to the
conceptual content of the utterance, it gives the hearer instructions to
look for (and make) inferences arising from earlier propositions in the
discourse that contrast with the proposition following “do”. So it acts as
a constraint on the hearer’s context, which reduces his processing effort
to help him arrive at the intended interpretation (which is the
explicatures of p plus the contrast).

Having called up the right inference(s), the hearer is made to notice the
contrast. As Blakemore points out, the hearer does not necessarily abandon
the earlier assumption at this point (since there is a contradiction, this
is an instance of the 3rd way that an assumption can be relevant-- by
weakening an existing assumption), but might choose to do so after
evaluating the evidence. (This account has obvious similarities to
Blakemore’s analysis of “but” as a denial of expectation.)

The problem with this analysis is that the contrast is not a conceptual
representation (i.e. an implicature). But doesn’t the contrast in the case
of the emphatic “do” seem to be something the speaker is deliberately and
consciously intending to convey? I don’t think we can say *both* that
emphatic “do” encodes procedural information leading to a recognition of
the contrast (and weakening existing assumptions) *and* that it introduces
an implicature. The procedural-encoded move was supposed to explain the
implicature.

I have some ideas to help solve this, or lead to a solution.

One idea: The present analysis is fine. There doesn’t have to be a
conceptual representation to account for the sense of deliberate contrast.
The mere process of calling up the inference, noticing that the contradicts
the assumptions made manifest by the utterance, and deliberating whether or
not to abandon it, could suffice to produce the conceptual (or “conscious”)
effect.

Second idea. The key is to focus on the process of deliberation. When the
hearer evaluates the evidence to help decide whether to abandon the
conflicting assumption, he takes into account the higher-level explicature
of the speaker’s “insistence” (part 1 of the analysis). This might be the
locus of the felt contrast-- the weighing of the evidence, along with the
speaker’s insistence of the truth of p. Then the problem might hinge on the
hearer’s level of trust.

Three. The contrast is a higher-level explicature of some kind. This is
questionable because all the constructions that have been analyzed in this
way (as far as I know) have been speech-act and propositional attitudes
(e.g. in Wilson and Sperber’s Linguistic Form and Relevance). I am not
well-versed enough in semantics/pragmatics to come up with convincing
arguments to support this (but I can try).

Four. Consider this quote from SW’s Relevance: Communication and Cognition:
“Contextual effects [presumably including assumption
weakening/cancellation] and processing effort are “non*-*representational
dimensions of mental processes. They exist whether or not the individual is
consciously assessing them, whether or not they are conceptually
represented. When they are represented [presumably in the case of
contextual implicatures], we claim that they are represented in the form of
comparative judgements. These judgements are intuitive; they have their
basis in the monitoring of physico-chemical parameters.” It’s easy to see
how contextual weakening can be realized as a physico-chemical change in
the system. If a contextual implicature is similarly a physiochemical
effect, then it would be easy to claim that emphatic “do” simply actuates a
physiochemical effect. The explanation in the same in either case (there is
no fundamental difference between contextual implicatures and contextual
weakenings/cancellation.

I’ll be glad to hear anyone’s opinion on this, but it would be
super-exciting to hear from Diane Blakemore, Dan Sperber, Deirdre Wilson,
or Robyn Carston. I know there is nothing in it for you to write back (as I
am a lowly Master’s student at a mid-tier college), but it would help me
TREMENDOUSLY. I’d like to hear your impressions, instincts, or whether you
feel one of the proposed solutions is on the right track. Even better would
be a definite view on the matter. There may not be a right answer, but I
want to make a good argument. And by all means, direct me to relevant
readings (I have just discovered Yus Ramos’ Relevant theory reference list,
this late in the game).

The thesis is due November 2nd, so I’d prefer to hear back before then. If
no one responds, I’m glad to have had the chance to write out my thoughts
and get clearer to myself where I stand on this.

Thanks for taking the time to read this. Keep up the good work everyone!

Regards,

Salpi Vartivarian
Received on Tue Oct 9 07:36:00 2012