Grice's Desideratum of Conversational Candour

Speranza

Some analytic philosophers (even of the Oxford school!) hold that it is possible to lie without intending to deceive, can you believe it?

A conceptual analysis may be provided, thus, that involves the warranting
context condition, in single condition of believing that one is in a warranting
context.

More specifically

***************************************************************

(L) U lies iff

CONDITION (i) 

U is not the victim of some linguistic or 'expressive' error/malapropism, or using metaphor, hyperbole, or irony (i.e. attempting at a conversational implicature, which may mislead) 

CONDITION (ii) 

U says that p. (This requires an analysis of 'saying' or 'explicating').

CONDITION (iii)

(iii) U believes, however, p to be false (i.e. U does not avow by the desideratum of conversational candoour -- although 'misleading' is NOT necessarily saying something which is false). (This is of course different from self-deceiving or from Moore's paradox, "It is raining, but I don't believe it.")

CONDITION (iv) 

U takes himself to be in what we may dub a "warranting" context.

*******************************************************************************

According to (L), it is not possible to lie if one does not believe that
one is in a warranting context. 

Consider the case of a putative lie told in a totalitarian state, since this happens to be the case of some utterances acctually demanded by a totalitarian state. 

These utterances of sentences supporting the state are made by people (utterers) who don not believe them, to people (addressees) who do not believe them.

Everyone knows that false things are being said (or explicated), and that they are only being said (or explicated) only because they are required by the State in question. 

It seems somewhat reasonable to suggest that, since everyone is forced to make these false utterances, and everyone knows they are false (it is 'common ground that they are false, as Grice would have it, and use the 'square-quote' device to mark this), they cease to be genuine lies, however gross.

People (utterers and addressees) living in a totalitarian state, making pro-state utterances, are a trickier case (as they should be). 

Whether or not the utterances is made in a contexts where "a warrant of truth" is, however, present is not at precisely crystal clear!

If utterer U is making an untruthful statement to addressee A, and everyone (i.e. utterer and addresse)
knows that false things are being said (or explicated), i.e., U knows that the
addressee A knows that the utterer U is being untruthful, the utterer U does
not believe that he is in "a warranting context."

According to (L), however, it is not the case that U lies -- so it seems (L) is too weak to capture the subtleties of the Oxford English ordinary expression of 'lie.'

However, it might be argued that in other cases, utterer and addressee know that false things are being said or explicated, and hence, that the
utterer does not believe that he is in a warranting context.

If this is so, according to (L), neither is lying.

On top of this, some Oxford ordinary language philosophers have argued that lying does not require an "intention" (in the Griceian use of this word -- vide his "Intention and Uncertainty") to deceive (but cfr. his original formulation of the desideratum of conversational conversational as involving a prong (b), as a limiting factor to prong (a) -- "make your stronger conversational contribution" -- provided you don't MISLEAD -- , and that
there can conceptually be non-deceptive bald-faced lies and knowledge-lies. 

However, other Oxford ordinary language philosophers have rejected (L), ont he grounds that (L) seems to entail, as it stands, that one cannot lie when the falsity
of what one is stating is common knowledge (or 'common ground,' as Grice and I would prefer).

This conceptual analysis of lying does
not relieve the narrowness, alas. 

The concept of "warrant," unfortuantely, is not broad enough to
explain how an utterer U can lie in the face of common knowledge (or 'common ground', as Grice prefers -- he symbolises this as "[p]" -- cfr. [There is a king of France and] the king of France is bald."

An utterer can warrant "p" only if "p" *might* be the case (As opposed to: "I saw a square circle on the way to St. John's.") 

When the falsehood of "p" is common knowledge (or 'common ground' or strictly, 'has been assigned 'common ground status'), no party to the common knowledge (or common ground) can warrant "p," simply because "p" happens to be epistemically impossible ("I noticed that on the way to St. John's, I saw an infant who was an adult." -- vide Grice/Strawson, "In defense of a dogma.")

According to some Oxford ordinary language philosopher, a negotiator who tells a falsehood that will lead to
better co-ordination between buyer and seller is telling a bald-faced lie.

Some Oxford ordinary language philosophers, however, provide a conceptual analysis of lying as being tantamount to just asserting what the utterer does not believe. (A violation of Grice's desideratum of conversational candour.)

It is a condition on telling a lie that the utterer U makes an *assertion* (For Grice, asserting correlates to "judging", and ordering to "willing" -- and he further assumes that judging REDUCES to willing -- Method in Philosophical Psychology).  

Some Griceians differentiate between an "assertion" and a "non-assertion" according to 'narrow plausibility.' 

To qualify as an "assertion," a lie must have narrow plausibility. 

Thus, an utterer who only had access to the assertion might believe it.

This is the grain of truth behind the assumption that lying requires the Griceian "intention to deceive" if not to (weaker) mislead.

A bald-faced lie shows that an assertion does not need to meet a requirement of "wide" (but rather 'narrow') plausibility, that is, credibility relative to the utterer's total evidence (Grice draws this concept from Davidson -- vide Grice and Baker, "Davidson on weakness of the will')

Some Griceians provide, as examples of an assertion, and hence, lies, 

-- the servant of a maestro telling an unwanted female caller that the sounds she hears over the phone are not the maestro and that the servant is merely dusting the
piano keys, and 

-- a doctor in an Iraqi hospital, during the Iraq war, telling a
journalist who can see patients in the ward in uniforms that 

“I see no
uniforms." 

The claim that these each of these utterances count as an "assertion," however, and therefore
a "lie," is somewhat controversial by Griceian standards (i.e. by his high standards of linguistic botanising)

A statement like the two examples above (the alleged 'assertion') neither expresses the utterer's
belief, nor aim to affect the belief of the addressee in any way, since
their falsehood is common knowledge (or has 'common ground status,' as Grice prefers)

Some Griceians, alas, do not offer a definition
of "asserting" a proposition (with necessary and sufficient conditions). Grice tried, but failed.

To
the extent that some Griceians do not fully analyze the concept of "assertion" (Schiffer tried in his D Phil Oxon under P F Strawson but, for Grice, Schiffer failed, too!) their conceptual analysis of lying is far from crystal-clear.

It may be argued against this that an utterance such as the above in question is not an assertion, and hence, on this account, fail to be a Kantian lie.

But then, some Griceians hold that it is possible to lie without intending to deceive (never mind 'mislead,' to use the formulation of Grice's desideratum of conversational candour).

But some Griceians have been adamant in defending the assertion condition for lying. 

Utterer U lies when
U asserts something that U believes to be false. 

(Cfr. Grice's "Desideratum of conversational candour," that only mentions that the utterer should try NOT TO MISLEAD). 

Some Griceians hold that an utterer U 
assert something when U makes a statement and U believes that U is
in a situation in which the Griceian "conversation maxim," (a joke he uses 'echoing Kant,') or desideratum of conversational candour, ‘Do not say what you believe to be false,’ is in
effect. 

This leads to an akin conceptual analysis of lying:

U lies if and only if

CONDITON (i)

U *states* that p.

(ii) Addressee A believes that U makes this statement "p" in a context where a Griceian conversational maxim (Grice's joke in 'echoing Kant') is in effect, "Do not make statements that you believe to be
false" (Grice's desideratum of conversational candour)

(iii) U believes that "p" is false. (Cfr. Grice on Strawson's ditto theory of truth in "Logic and Conversation, ii -- a variant, Grice thinks, of Ramsey's redundance theory, yet allowing for Tarski's counter-examples)

Alleged counter-examples to this revised conceptual analysis of lying have prompted a
revision of it in order to accommodate them, almost alla Popper:

(C')
U lies if and only if

CONDITION i

U says that p.

CONDITION ii

U believes that p is false (or at least that p will be
   false if U succeeds in communicating or 'meaning' that p), and

CONDITION iii

U intends to violate Grice's conversational maxim (his joke 'echoing Kant,' fr. the desideratum of conversational candour) against communicating
   something false by communicating that p.

And there is a further variant:

   (C'')

   U lies if and only if 

CONDITION i.

U say that p.

CONDITION ii.

U believes that p is false (or
   at least that p will be false if U succeed in communicating or meaning that p -- "getting his meaning across")), and

CONDITION iii.

 U intends to communicate something false by communicating that p.

Both (C') and (C'') are able to accommodate a counterexample
to the earlier attempt at a conceptual analysis.

When Marc'Antonio says to a Roman addressee


"Bruto is an honorable man."


this Roman addressee knows 

i. that Marc'antonio does not believe that Bruto was
an honorable man.

ii. that Marc'antonio is subject to a Griceian conversational maxim (Grice's joke 'echoing Kant,' cfr. his desideratum of conversational candour), against saying
things that Marc'antonio believes to be false, and 

iii. that Marc'antonio is a
cooperative (or 'helpful' as Grice prefers in his earlier Oxford lectures on logic and conversation -- "cooperation is pretentious") participant in the conversation so far. 

Thus, the Roman addressee is led to
conclude that Marc'antonio is FLOUTING the 'conversational maxim' (Grice's joke, 'echoing Kant,' cfr. the desideratum of conversational candour) in order to
communicate something other than what he explicated -- i.e. Marc'antononio is relying that the Roman will 'catch' his implicature.  

In fact, the best
explanation of Marc'antonio statement is that Marc'antoinio wills to communicate the exact
opposite of what he explicates: 

"It is not the case that Bruto is an
honorable man." 

Since Marc'antonio does not intend to violate this
conversational maxim against communicating something that he believes to be
false (that Brutus is an honorable man) by saying


“Brutus is an honorable man,”


or, more simply, since Marc'antonio does not intend to communicate something
false with his untruthful statement, Marc'antonio is not lying.


However, in the case of a guilty witness, Tony, against whom there is
overwhelming evidence, and who says


“I did not do it."


without the intention that any utterance believe him, the utterer does intend to violate
the conversational maxim (Grice's joke, echoing Kant, cfr. his desideratum of conversational candour) against communicating something that the utterer believes
to be false (that he did not do it) by saying


“I did not do it,”


or, more simply, the utterer does intend to communicate something believed-false
with his untruthful statement, even though he does not intend that any addressee
believe this. 

Some Griceians have contended that a "non-deceptive liar" does not intend
to communicate anything believed-false with his untruthful statement,
and, indeed, may even intend to communicate something believed-TRUE with
his untruthful statements. 

But some other Griceians (yes, they comes in varieties) reject the claim that a "non-deceptive liar" does not intend to communicate anything believed-false, even if U intends
to communicate something believed-TRUE. 

A bald-faced liar might want to
communicate something TRUE. 

For instance, Tony may be trying to communicate
to his addressee that that he will never be convicted.

But that does not mean
that Tony does not also intend to communicate something FALSE in violation of
the the 'conversational maxim' (Grice's joke, echoing Kant, cfr. his desideratum of conversational candour).

Tony wants what he actually said or explicated to be understood and accepted for
purposes of the conversation. 

It is not as if


“I did not do it”


is simply a euphemism for


“You shall never take me alive, coppers!”  

However, in the case of polite
untruths, such as

“Madam is not at home,”


the untruthful statement is simply, broadly speaking a euphemism (and the addressee A must take it as such).

The expression:


"Madam is not at home."


delivered by a servant or a relative at the door, has become a mere
euphemism for indisposition or disinclination. 

Cfr. Grice's example to Schffer,

"We must meet for lunch sometime."

("In Oxford, this EXPRESSION, never mind the implicatures, MEANS "Get lost.")

In the case of polite
untruths, it seems, there is no intention to communicate anything
believed-FALSE

In the case of the servant who tells the female caller,


“I’m dusting the piano keys."


or the Iraqi doctor who tells the journalist


“I see no uniforms."


or the negotiator who tells the other negotiator


“That is the highest I can go."


or the person living in the totalitarian state who makes the pro-state
utterance, it is arguable that there is no intention to communicate
anything believed-FALSE>


If this is true, there is some support for the claim that a
"non-deceptive liar" does not m-intend (alla Grice) to communicate anything believed-FALSE
with their untruthful statements, and hence, that he is not lying
according to (C') and (C'').

Some Griceians hold that it is possible to lie without
intending to deceive. 

Some have also defended the assertion condition for
lying.

You lie when you assert something you believe to be false. 

To assert
that p is to say that p and thereby propose that p become common ground. 

A
proposition, p, acquires, to use Grice's parlance "common ground status" in
a group if all members accept (for the purpose of the conversation) that p,
and all believe that all believe that all accept that p, etc. 

Lying gets thus analysed as follows

(C'') U lies 

iff

CONDITION i.

   U says that p to A, and

CONDITION ii.

   U proposes that p become common ground, and

CONDITION iii.

   U believes that p is false.

In the case of an utterer U making an IRONIC untruthful statement, the
utterer obviously does not propose that the believed-false proposition (e.g., Marc'antonio's “Brutus
is an honorable man”) become common ground

However, in the case of a non-deceptive liar, the utterer does propose
that the believed-false proposition (e.g., “I did not cheat”) become common
ground.

The fact that in the case of a non-deceptive lie it is common knowledge
that what the utterer U is saying is (believed to be) false does not alter
the fact that the utterer U is proposing that the believed-falsehood become
common ground.

Indeed, even if the (believed) truth is initially common ground, before
the utterer U proposes that the believed-falsehood become common ground, it
is still the case that the non-deceptive liar is proposing to 'update' the status
common ground with her utterance. 

E.g., the case of a student and the
dean,


“The student wants himself and the Dean to mutually accept that he did
not plagiarize."

It is possible to argue that this account of assertion, and hence (C''') is
faced with a dilemma when it comes to a non-deceptive lie. 

Either, in the
case of a non-deceptive lie, the utterer does propose that the
believed-FALSE proposition become common ground, but becoming common ground
is too weak to count as asserting, or becoming common ground is strong
enough to count as asserting, but, in the case of a non-deceptive lie, the
utterer does not propose that the believed-false proposition become common
ground. 

Consider Donnellan's alleged counter-example of a guest at a party saying
to another guest,


"The man drinking a martini is a philosopher."

(Grice's treatment of Donnellan is to be found in Grice's "Vacuous Names." Donnellan remained a Griceian: "Putting Humpty Dumpty together again" and essays by T. Patton).


-- and of the two guests proceeding to talk about the philosopher, when it
is common knowledge (or has assigned common ground) that the drink in question is NOT a martini. 

Perhaps it
is mutually recognized that it is not a martini, but mutually recognized
that both parties are accepting that it is a martini.

(what Grice calls a 'code': his telling a English girl, the French for "Napoleon was a bad man.") intending the girl to believe Grice means, "Help yourself with a piece of cake.")


The pretense will be rational if accepting the false presupposition is an
efficient way to communicate something TRUE.

However, if proposing that a
believed-FALSE proposition become "common ground" can mean engaging in and
sustaining a “pretence,” possibly in order to communicate truths, then it
is not clear that this counts as making an assertion. 

(Grice learns from Rogers Albritton that to the 'ironic' is to 'pretend,' etymologically).

Hence, a
non-deceptive liar may be proposing that hs believed-FALSE proposition
become common ground without this being an act of making an assertion. 

But
this means that he is not lying, according to (C''')

 Alternatively, if proposing that a believed-FALSE proposition become
common ground means something more than this, such that the utterer intends
or wants herself and his addressee “to mutually accept” her believed-FALSE
proposition, it is not clear that a non-deceptive liar intends or
wants this.


If this is correct, a non-deceptive lie fails to be a lie according to (C''').

And we shall have to leave, alas, (C'''') beyond for a longer day!

The desideratum of conversational candour is indeed double.

Grice refines it in terms of two throngs:

(a) the U should try to make his strongest conversational move.
(b) as a limiting factor to (a), U should try not to mislead.

Grice combines this desideratum with a desideratum of conversational clarity. 

And he adds two principles, the principle of conversational self-love and the principle of conversational benevolence.

By the time he was writing his "Valediction," he had changed his mind (or soul -- vide "Method in philosophical psychology") a few times!

But this is mostly what his Oxford tutees at that small room at St. John's got! (When he was not too late -- he was often referred to as "Godot.")

REFERENCES

Grice, H. P. (1961). The causal theory of perception. The Proceedings of the Aristotelian Society, repr. in G. J. Warnock, "The philosophy of perception," Oxford readings in philosophy.

Grice, H. P. The desideratum of conversational candour.

Grice, H. P. (1989) WoW -- Way of Words. Harvard. 

Grice, H. P. The conversational maxim, "Try to make your conversational

contribution one that is true."

Grice, H. P. (1988). Valediction -- repr. in WoW.

Grice, H. P. On Kant on lying -- Aspects of reason. 

Speranza, J. L. Join the Grice Circle!