RE: RT list: Non-sentential utterances, logical form, explicatures (e.g. in poetry)

Dear all,

Thank you for the responses on the topic so far. Let me try to specify more
clearly what bothers me. My focus is theory-oriented, esp. about the
concepts of 'logical form' and 'explicature' as they are used in RT.

Here are simpler examples to focus on my point:
Consider the expression "the gardener" - grammatically just an NP.

Now, this expression is used as an ostensive stimulus in the following
contexts:
1) Context: Someone asks who left the gate open. Answer: "The gardener"
2) Context: Someone asks whom they saw in the afternoon. Answer: "The
gardener"

So, what about the logical forms involved in the answers? All the input that
the linguistic decoding system has in each case is the phonological
representation of the phrase 'the gardener', which it probably would parse
as an NP, and the decoding process would lead to a semantic representation
like GARDENER - with the - presumably procedural - information that the
referent is to be assumed as known or something like that (the details don't
matter here). The grammar, qua decoding system, is not given any clue how to
construct a well-formed sentence of English from this phrasal input only.
For example, the grammar cannot tell that in case 1) this phrase is to be
construed as a subject, and in case 2) as an object nor does this input
provide any predicate.

Question: if a verbal expression encodes no more than a concept, can it be
said to encode a logical form? In other words, can a logical form be
identical to just a concept?

Looking in the RT literature, this does not seem to be the original
intention:
"A logical form is a well-formed formula, a structured set of constituents,
which undergoes formal logical operations determined by its structure.”
(Sperber and Wilson 1995:72)

If the output decoded by the grammar component are well-formed sentences
(semantic representations), then an NP alone would not be a well-formed
output string.

A likely fix might seem to be that perhaps the preceding utterance provided
some kind of 'sentence schema' which the grammar could use as part of the
input - e.g. if the preceding utterance had been a question with a structure
like [[X]NP [LEAVE-OPEN [GATE]NP]VP]S, the grammar could have slotted the
encoded NP into the X spot - and it would have been the subject. (Perhaps
this is what Andrea Rocci had in mind in her answer, saying "You have
predicates, arguments, quantifiers ... you have everything you need to have
a logical form.")

There seem to be at least three possible problems here:
a) Is it admissable/likely that the grammar component utilises contextually
available structures as input to its decoding process? This touches, of
course, on the whole issue of encapsulation of the language faculty, a
complex matter I don't want to go into here.
b) Quite apart from that issue, there are clearly quite common situations
where no such grammatical structure is contextually available; for example:
3) Context: Shop assistant asks customer who has just come in, "How can I
help?". Customer's response: "Two cokes".
c) Perhaps more importantly: Is it - pragmatically - likely that the mind
would bother constructing a whole sentence first, outputting as 'logical
form' a well-formed string of English grammar, arbitrarily supplying a
number of unspecified constituents, which would then have to be developed
into a propositional form, resolving all remaining ambiguity and
underdeterminacy, when all that is actually needed is the construction of an
assumption a mental representation, that satisfies the requirements of
optimal relevance? In other words: is it likely or necessary that our mind
takes the 'detour' of constructing an elaborate linguistic representation
from minimal input - just because that input was verbal?

The most efficient - and hence most relevant - route would be to use the
concept decoded from the expression - GARDENER - and use it directly to
construe an assumption that would satisfy the communicative POR. This
construal could be a strictly pragmatic process, using some highly
accessible assumption schema - not going via a sentence schema.

There is a further puzzle to me in the RT account relating to logical form -
and that is that is has been made the distinctive criterion for the
explicature-implicature distinction:
“Any assumption communicated, but not explicitly so, is implicitly
communicated: it is an implicature. By this definition, ostensive stimuli
which do not encode logical forms will, of course, only have implicatures.”
(Sperber and Wilson 1995:182)
Why is this claim a problem? Consider the following variant to the
'gardener' example above:
4) Context: Someone asks who left the gate open. Response (non-verbal): The
person asked points at the gardener.

Now it would seem that the assumption communicated by the response in 4) is
essentially the same as that communicated in 1): 'The gardner left the gate
open'. However, while in 1) it would be an explicature, if an NP is a
'logical form' or if it is expanded by the grammar into an S, in 4) it would
be an implicature. Now, as an implicature, it could be either a contextual
assumption or a contextual implication - nothing else. It does not seem to
qualify as a contextual assumption of the act of communication 4), but
rather its outcome (which might then, of course, function as a contextual
assumption in subsequent acts of communication). Yet, if it is a contextual
implication, then it constitutes a cognitive effect. This means the
assumption communicated by 4) is in itself a cognitive effect, whereas in 1)
the same assumption communicated by a verbal utterance is not a cognitive
effect, but to be relevant at all must lead to cognitive effects. In other
words, ceteris paribus, the non-verbal response is, so to speak, "one
cognitive effect ahead" of the verbal one. To me, at least, this outcome
seems somewhat unexpected.

So, what exactly is a logical form? Can it be the semantic representation
just a concept (encoded by just a noun or noun phrase) or should it have
some predication encoded as well? As Carston has pointed out,
"Unfortunately, different uses of the term 'logical form' abound (Carston
2002:89, note 26). I wonder whether in RT itself perhaps the distinction
between 'logical form' as an abstraction of a mental representation in
general and the Chomskyan output of the LF component has become blurred at
times.

It seems to me that the answers to these questions may have wideranging
ramifications, e.g. with regard to the whole explicature-implicature
distinction, which has alredy come under some dicussion (e.g. Carston 2002:
"The distinction between explicatures and implicatures, the two kinds of
communicated assumptions, is primarily a derivational distinction and may
have no greater import than that" 366 versus Iten 2005: "This means that
there is an important difference between explicatures and implicatures ..."
86).

At the end of the day, the answers to these questions should be empirical.
For example, when a verbal expression consisting only of a noun or noun
phrase is given, either the mind passes it through the grammar, trying to
construct a well-formed sentence, or it does not but uses the decoded
concept directly as input to inferential processes. Perhaps this has already
been investigated - if so I'd be happy to know.

Best wishes,
Ernst-August

References:
Carston, R. 2002. Thoughts and utterances: The pragmatics of explicit
communication. Oxford: Basil Blackwell.
Iten, C. 2005. Linguistic Meaning, Truth Conditions and Relevance. Palgrave
Macmillan.
Sperber, D., and D. Wilson. 1995. Relevance: Communication and cognition.
Oxford: Blackwell.

 
> -----Original Message-----
> From: owner-relevance@linguistics.ucl.ac.uk
> [mailto:owner-relevance@linguistics.ucl.ac.uk] On Behalf Of
> ernst-august_gutt@sil.org
> Sent: venerdì, 11. febbraio 2011 17:26
> To: relevance@linguistics.ucl.ac.uk
> Subject: RT list: Non-sentential utterances, logical form,
> explicatures (e.g. in poetry)
>
> Dear all,
>
> According to the definition of explicature, it is dependent
> on an utterance having a logical form.
> Question: when does a verbal utterance have a logical form?
>
> The first stanza of Oscar Wilde's poem "The dole of the
> king's daughter"
> reads as follows:
>
> Seven stars in the still water,
> And seven in the sky;
> Seven sins on the the King's daughter,
> Deep in her soul to lie.
>
> Does this stanza have any logical forms, explicatures (if
> only weak ones)?
>
> Any insights/comments appreciated.
>
> Ernst-August Gutt
>
Received on Mon Feb 14 13:42:13 2011