R Cartson writes: (as "Speranglished" by yours truly, with apologies):
The Problem of SPED (Significant Processing Effort Difference): what Does RT
Predict? an apparent clash.
"I. Noveck & al's claim is that a metaphorical utterance takes *more
processing effort* & has more cognitive effects than a non-figurative
equivalent'. Building on Gibbs they run some experiments to test the claim
to find the results support it. They see this is
precisely what the RT account of
metaphor predicts.
"Gibbs also takes the *extra processing effort* prediction to follow from
RT. But, unlike Noveck et al, finds it to be *at fault* with his
experimental results:
psychological research shows that addressees
do not devote an extra processing resource to
understanding a metaphor as compared with
a nonmetaphoric utterance. The RT's "metaphor
-as-loose-talk" view may *not* see metaphor as
a violation of communication norms
(A la good ol' H P Grice. "Logic & Conversation", in Studies in the Way of
Words, p.34. JLS)
but still incorrectly assumes that metaphor & other
tropes demand an extra cognitive effort. It is this
psychological evidence that Noveck & al. are calling into question.
An opposite prediction:
"The opposite view, however, of what RT *predicts* about metaphor processing
is to be found. In _Poetic Effects: A RT perspective_ Pilkington discusses
Gibbs's and Gerrig's psycholinguistic exeriments which, Pilkington claims,
show that
an utterance used metaphorically does NOT
necessarily take longer to process than the same
utterances used literally.
[Excursus:
Well, I guess it also has to do, inter alia, with
whether English is your first language or not, or
how language-aware you are. A friend of mine lives in
Boston, USA, but English is not his first language,
and he came to learn
red herring
(before learning what a "herring" was) as meaning
"a distraction from the topic" rather than "a special
kind of fish red-smoked as used in mediaeval times
by the peasants to divert the lord's dogs from the
trail of the game the peasants lived on". JLS]
According to Pilkington 'the RT account of metaphor interpretation is quite
consistent with these psycholinguistic results'.
The Puzzle.
"So what's going on here? On the face of it, it looks as if someone has to
be wrong about what the RT prediction is. Before I check back on what the
founding parents
(and the granddad? H Paul Grice. Please do! JLS)
"have to say, note that talk of extra processing effort raises the question of
more than what?
The Distinction: TRUTH-CONDITION vs. IMPLICATURE.
"What is the metaphorical utterance being compared with? I think there are 2
different comparisons at work, and this may be the source of the apparently
contradictory positions. Sperber/Wilson discussing such highly conventional
metaphors as
1. Robyn is a lion(ess).
which communicates the strong implicature that
2. Robyn is a brave (lad, lass).
say that such a metaphor
must suggest some extra line of thought
if their indirectness and its extra processing
cost is to be justified
and, of course, this holds all the more for more creative metaphors, such as
3. Robyn is a pink pantheress.
The processing effort comparison here is between uttering (1) and uttering
(2), or, if you like, between
4. You are a grice.
5. You are a grubby lad!
(OED. grice: a piglet. Also used metaphorically to refer to a "grubby, dirty
lad" - A "lad" is what ever overmetaphorical Americans would call a "kid",
i.e. the offspring of a goat. :)). It is this sort of comparison that Noveck
et al. have in mind.
Sperber's proposal of a solution: THE PROBLEM QUESTION MAKES NO SENSE (SPED
= 0), or rather, both metaphor and non-metaphor may involve the same
processing effort. It's a context-sensitive thing.
Consider now an example in Sperber:
6. Robyn is a territorial.
"In a particular context C1 the addressee might access an interpretation
along the lines that
7. Robyn is a very dutiful
b loyal
c obedient
d a team-player
e self-sacrificing
but not including that
8. Robyn a is a member of the military
b is trained to kill
c has weapons
"This would be a case of a "loose" or metaphorical use of "territorial", or
'soldier'. In a different context C2, however, the LITERAL
(truth-conditional) properties of being a member of the military, being
trained to kill, etc. might be accessed instead. For Sperber, in BOTH cases
the addressee follows a path of
least effort
in accessing the interpretation and stops when the expectation of maximal
relevance is satisfied. There doesn't seem to be a prediction that
there will be any significant processing effort
difference (SPED) between the loose/metaphorical
use in C1 and the literal (truth-conditional) use
in C2.
It's clear that in a some arbitrary context Cn the loose/metaphorical
reading may be actually
easier to access
than the truth conditional one."
[My example of "red herring" above. Actually, few native speakers know why a
red-herring is called a red herring. I had to search in the OED2 myself and
I'm not native, hey!]
"I think these are the sorts of comparison that Pilkington (and Gerrig) are
making, and they are quite opposite to those by Gibbs and Noveck."
A WAY OUT?
"So is that it? Problem solved? Probably not!
One assumes that not only cases such as (6) (Robyn is a soldier), but also
cases like one (4) - "you're a grice" - have been tested and have given a
result of no "SPED" (a word of confirmation from Gibbs and or Glucksberg
would be welcome!)."
[also, perhaps, to recall Grice PERE. Principle of Economy of Rational
Effort. In R. Grandy & R. Warner, PGRICE, Philosophical Grounds of
Rationality, Intentions, Categories, Ends (p.83) to which Wilson & Sperber
contributed. Grice thinks that implicit ratiocination is just as valid as
the explicit variety). Grice writes:
Where there is a ratiocinative procedure for
arriving rationally at a certain outcome, a
procedure which, because it is ratiocinative,
involves an expenditure of time & energy, then,
if there is a NON-ratiocinative & so more economical
procedure which is likely, FOR THE MOST PART,
to reach the same outcome as the ratiocinative
procedure, then, provided the stakes ARE NOT TOO
HIGH, it will be rational to employ the
CHEAPER tho' somewhat LESS RELIABLE
NON-RATIOCINATIVE proceudre as a substitute
for ratiocination" (op. cit. p.83).
RC: "If so, i.e. if no SPED is observed, this would still seem to be at odds
with the RT account, unless, of course, the measures (reading times &c) used
in the experiment simply can not tap the sort of effort differences at issue?"
A Further Pragmatic Compliction: REFERRING vs. PREDICATING.
"Furthermore, there seems to be a striking difference in the psychological
processing of a metaphorical predication and a metaphorical referential
expression. Whereas the predication (which all the examples so far discussed
are) may take no more effort than literal paraphrases
SPED = 0
a case of metaphorical reference, such as when I say of my cat (9) or my
bedroom (10)
9. The princess is sleeping on a rug!
10. This junkyard my mother shouldn't see.
does take LONGER than literal reference, i.e.
SPED > 0
"This is pointed out by Gibbs & Onishi/Murphy who ease the way for a
metaphorical reference but had yet to admit that there just *IS* a heavier
processing load for a metaphorical refererntial expression than for a
corresponding literal one, such as:
11. My cat is sleeping on a rug
(but is it really "mine"?)
and
12. This mess of a room my mum should not see.
(but is this really "a mess of a something", or is it just a "messy"
something).
"Of course, there are plenty of possible explanations for this
"referring/predicating" difference, some of which are discussed by Gibbs and
O&M."
(And another may be JR Searle's DPhil Oxon under PF Strawson, "Reference &
Predication". Strawson and Grice having worked for years on the
philosophical issues involved therein. JLS).
A Possible Partial Solution-cum-explanation.
"Now, the Noveck et al. experiments in which a metaphorical/literal
processing difference (SPED) *was* found were all cases of REFERRING. Their
results do reinforce Gibbs and O&M's claim (that SPED > 1), but they could
be subject to an explanation for the extra effort
in terms of special properties
of reference
that make metaphor (or non-truth conditional utterances at large) more
difficult when used in this way (i.e. when used to perform the pragmatics of
"referring" rather than of "predicating"). It might be of greater interest
for specification of the RT prediction about non-truth-conditional
utterances if an experiment is made such that it reveals a SPED > 1 when the
non-truth conditional component is used in
predicating
a property of an already truth-conditionally determined referent. I hope
these are planned! If not, should we accept it's all just
contextual-dependent after all and that that is the ultimate solution and
explanation?
Refs:
GERRIG R Empirical constraints on computational theories of metaphor.
Cognitive Science 13
GIBBS R Comprehending figurative referential descriptions.
J Exp Psych 16
GIBBS R The Poetics of Mind. Cambridge.
NOVECK I, M BIANCO & A CASTRY Costs/benefits of metaphor.
Metaphor & Symbol 16
ONISHI K. & G MURPHY
Metaphoric reference:
when metaphors are not understood as easily
as literal expressions.
Memory & Cognition 21
PILKINGTON A Poetic effects: a relevance theory perspective. Benjamins
SPERBER D Metarepresentations in an evolutionary perspective. In
Metarepresentations: a multidisciplinary perspective, Oxford
SPERBER D. & D Wilson Loose talk. Aristotelian Society 86
======
This archive was generated by hypermail 2b29 : Wed Jun 27 2001 - 05:28:34 GMT