Re: RT list: Cognitive effects and processing effort

Dear Jan,

I believe that your interpretation of Bach's criticism comes pretty
close to what Bach must have in mind. From where I stand, however,
there seems to be a misunderstanding of the theory itself in it. At
least from the relevance-theoretic perspective the focus lies on
interpretation rather than production and for an interesting debate I
would divert you to Carston's response to Horn's criticism about
speaker considerations of relevance in Intercultural Pragmatics (2006,
vol. 2, issue 3, available also on Robyn's homepage). The central point
there is that "considerations of speaker effort do play a role,
although nothing like as central a role as Horn (and I would think
Griceans generally) favour". However, I think and have repetedly
stressed that RT can have significant implications on the study of
linguistic production as well, and research towards that area is not
only fruitful but also necessary to establish the validity of the
notion of relevance Sperber and Wilson put forward. Nevertheless, what
is central in this I believe is not the second but the first principle
of relevance (which was not explicitly formulated until 1995,
considerably after Bach and Harnish 1987 criticism). So, I believe that
if the relevance mechanism exists, it is the cognitive tendency of
humans to MAXIMAL relevance that should shed light to production
issues. As I argue in my forthcoming thesis, if the allocation of our
mental recourses are actually relevance-oriented, then it is the
selected context of the competent communicator that motivates him to
produce an OPTIMALLY relevant utterance, as this context should include
information about the hearer. In this sense, RT is capable of
describing speaker tendencies as well as hearer's ones. Of course, this
is only a suggestive line, but I think that the distinction between
maximal and optimal relevance does play a very big role in the theory
and it is on the grounds of the first that the second is based (a point
of confusion for many scholars and Bach as I interpret him as well).
Either way, what I find particularly appealing is that if there is a
single cognitive tendency it should affect equally production and
interpretation without posing a different (naive) RT for the first. The
mechanisms of production surely differ but the first cognitive
generalisation of Sperber and Wilson should not be treated as specific
to the interpretation process. Then further ideas stemming from it with
respect to the characterisation of optimal relevance during utterance
interpretation (i.e. the path of least effort) might provide equally
appealing arguments of how we access mental content during planning our
utterances, if of course they are manipulated appropriately. Even so,
it would be good to hear what hardcore relevance theorists might have
to say on this issue.

Best wishes,
Stavros

Quoting Jan Straßheim <strassheim@gmx.de>:

> - probably the reason why Bach thinks precise assessment of relevance
> is "the most obvious problem" to RT as a theory of communication is
> that, according to RT, competent communicators make very precise
> predictions indeed about what would be most relevant to individual
> hearers and in what way, and often with fair success (i.e. similarity
> of thought sufficient for the speaker's purposes). While of course
> this doesn't require any quantitative measurement, local comparisons
> of actually competing information or sensitivity to joules and spread
> of activation wouldn't do either, since the assessment is about
> someone else's mind. So prima facie, people seem, independently of
> science and experimentation, to use some naive relevance theory which
> appears to be more fine-grained (or global? or behaviour-oriented?)
> than RT. Anyway, this was roughly Bach's question in the commentary
> to the Précis of 1987, and I'd love to learn what the current RT
> answer is...
>
> Best,
> Jan
>

-- 
Stavros Assimakopoulos
PhD candidate
Department of Linguistics and English Language
University of Edinburgh