Metalinguistic Negation Echoic Negation of

22 - 26 August 2011 Harvard-Australia Workshop on: Language, Learning and Logic

Metalinguistic Negation, Echoic Negation or Pretence of Descriptive
Francesco Gentile – University
of Nottingham
Negation
? – Department of Philosophy –
apxfpg@nottingham.ac.uk
(1) We didn’t see [rhinocerosai]–We saw [rhinoceroses].
(2) A: Auntie’s meal was good.
B: It wasn’t good–it was delicious!
(3) A: Is the king of Italy bald?
B: The KoI isn’t bald–there’s no king of Italy.
Horn (1989: 377): MN “can be glossed 'I object to U’, where U is
crucially a linguistic utterance or utterance type rather than an
abstract proposition.”

(i) How to make sense of these utterances: Do they form a
natural
category, or do they differ in some important
way?

(ii) How shall we explain the speaker’s behaviour, and their
apparent commitment to a contradiction?
(iii) Does negation play the same role in all these examples?

1: Negation is always descriptive, truth-conditional;
2: Negation is generally perceived as metaconceptual
because of the material that falls under it, which is
echoically used;
3: Once this material is conceptually adjusted (i.e.,
enriched) so that a fully propositional form is recovered,
the utterance is fully assessable for truth;
4: Therefore, no need to appeal to a metalinguistic
operator. Contradictions are resolved pragmatically, at the
level of the proposition expressed by the speaker, i.e., its
explicature.
(4) I don’t eat [tomeiDouz]; I eat [toma:touz].

i. All there is to negation is its being
descriptive.
ii. Negation is unambiguous: it doesn’t allow for a

wide vs. narrow scope distinction. Negation is
maximally permissible: it always has a ‘hole’
reading

(iii) Point of view = Portion of the common ground a
speaker accepts, or thinks other speakers accept at a
particular time of the conversation

iv. By uttering a rectifying second clause, speakers
make clear that they do not share a salient part
of the common ground on the basis of some
specific elements, e.g., presuppositions, implicatures,
formal aspects of the utterance, considered as
erroneous or disruptive.

(iv) Asserting or denying a proposition is a way of
accepting a certain take on the CG

Hypothesis:


(4) I don’t eat (what is properly called [tomeiDouz]); I eat
(what is properly called [toma:touz]).

Truth-conditional
operators only
scope over atissue contents

[SEMANTIC LEVEL]

[EXPLICIT CONTENT]

Supporting
evidence:

(2) A. Auntie’s meal was good.
B. It wasn’t good–it was delicious!

Grammar does
not provide rules
for determining a

special behaviour
of not wrt these
other elements of
the CG

(2) Not (auntie’s meal was good). [SEMANTIC LEVEL]
(2) Not (auntie’s meal is [appropriately considered as]
‘good’).
[EXPLICIT LEVEL]

(i) (1)-(3) involve a particular act of denial, i.e.,
metalinguistic negation (MN). MN does not operate on the
proposition expressed by the utterance, but on the utterance
itself.
 
(ii) A speaker who utters a complex sentence of the form
‘Not P*,
Q’ does not assert the denial of ‘P’, but objects to a previous
utterance of P on the basis of some element ‘*’, which makes
the conceptual or formal representation of that utterance

infelicitous.
 
(iii) regular, descriptive, negation is unambiguous; however,
instances of presupposition cancellation or implicature denial
may be pragmatically accounted for by appeal to MN.
1: The status of ‘metalinguistic’ is unclear: if its sense
is ‘I object to U’, then an explanation of how ‘not’ has
acquired this sense should be offered.
 
2: The autonomy of contradiction:
Procedural Priority Thesis. The expression of a contradiction
is what launches the reassessment of utterances like (1)-(3).
 
Logical Immunity Thesis. The contradiction has an
autonomous role, which cannot be accommodated by

(5) A thinks that auntie’s meal is good.
(6) B thinks that auntie’s meal is delicious.
 
(7) B disagrees with A as to whether auntie’s meal is merely

good.
#(8) B disagrees with A as to whether auntie’s meal is
appropriately considered as good.

Carston’s view: contradictions are resolved at the level of the
utterance's explicature
My view: contradictions are essential to the effectiveness
of these linguistic constructions.
We can’t modify the semantic profile of those utterances, but we
can accept the idea that sometimes speakers explore
contradictions or inconsistencies with particular expressive
intents.

(ii) But speakers may use contradictions: for instance,
to criticise the point of view that another speaker
would adopt, or has already adopted, by expressing
the semantic content of the sentences embedded in
the first clauses of these constructions, along with any
further assumption derivable from them.


iii. The particular use of negation in these cases
is pretended: the speaker is only pretending to deny
an at issue content

(4) Not (I eat tomatoes).

1: Echoicity is neither a necessary nor a sufficient condition
for irregular, marked negation.
 
2: The disagreement problem:

(i) Speakers cannot be committed to contradictions

“It wasn’t good.”
Not P (P = The meal was good)
Not Q (Q = The meal was
better than good)
Common Ground1 =
Not P & not Q
“It was delicious!”

Q (Q = the meal was delicious)
Common Ground2 =
(Not P & Not Q) & Q

Upgraded content: Q
(Not Q gets discarded)

METARULE1:
Upgrade your
context with what
is asserted and
what is implicated
(e.g., scalar
implicatures,
presuppositions)
METARULE2:
Upgrade your
context in such a
way as to discard
inconsistent

assumptions
(in our case,
discard either Q or
not Q)
RULE implementing
MR2: Discard or
revise any
assumption based
on whichever
element receives

(v) Neither of these two actions requires actual belief
in the proposition(s) expressed by an utterance
(Stalnaker, Yablo)
A word of caution: Two senses of pretence are
active here:
 
1. Pretence of descriptive negation: a performative
action
2. Pretence of sharing a CG with one’s own

interlocutor: a mental state
You enter in the second pretence just by performing
the first one.

Comparing the Three Views
Meaning

Ambiguit
y

Inference

Horn’s
MetaNegation

Two
Meanings
(Neg & MN)

Yes


Gricean?

Echoic
Negation

One
Meaning
(Neg)

No

Modulation

Pretence

One
Meaning

No

Basic Logical
Reasoning

Burton-Roberts, N. (1989). On Horn’s Dilemma: Presupposition and Negation. Journal of
Linguistics, 25: 95-125.
Carston, R. (2002). Thoughts and Utterances. Oxford: Blackwell.
 
Currie, G. (2006). Why Irony is Pretence. In S. Nichols (ed.), The Architecture of Imagination:
New Essays on Pretence, Possibility, and Fiction, Oxford: OUP.
 
Geurts, B. (1998). The Mechanisms of Denial. Language 74, 2: 274-307.
 
Grice (1967). Studies in the Way of Words. Cambridge, Mass.: Harvard University Press. 
Horn, L. R. (1989). A Natural History of Negation. Stanford: CSLI Publications.
 
Kadmon, N. (2001). Formal Pragmatics. London: Blackwell.
  
Sperber, D. and Wilson, D. (1981). Irony and the Use-Mention Distinction. In P. Cole (ed.),
Radical Pragmatics. Academic Press: 295-318.
 
Stalnaker, R. (1999). Context and Content. Oxford: OUP.
Thanks to Greg Currie, Stefano Predelli, and all my crew: Alba, Felice, Anna, Franca
and Jiyeun, Becci and Chris!