Davidson on Metaphor
In discussions of metaphor, Davidson is (in)famous for claiming that metaphorical
utterances lack any distinctive, nonliteral meaning. But there is much less agreement
about just what he means by this. In this article, I explicate this claim as it occurs in
“What Metaphors Mean” (1978) and relate it to his reflections on language in “A Nice
Derangement of Epitaphs” (1986), leaving the project of locating both papers within
his broader views on interpretation for another time. In Section 1, I argue that despite
some puzzling inconsistencies, the overall thrust of “What Metaphors Mean” is a radical
form of noncogitivism. In Section 2, I argue that in “Nice Derangement,” Davidson
applies several of the arguments offered against metaphorical meanings in “What
Metaphors Mean” to linguistic meaning more generally; but his criteria for what counts
as “meaning” have shifted to include context-local word meaning alongside Gricean
speaker’s meaning. With respect to metaphor, he appears to have abandoned his previous
noncognitivism for an analysis in terms of speaker’s meaning, but it is not clear
that this new view is justified by his new model of meaning. In Section 3, I articulate
and evaluate a neo-Davidsonian view of metaphor, which retains as much as possible
from both papers.
1. “What Metaphors Mean”
The central claim of “What Metaphors Mean” is clear: “Metaphors mean what the
words, in their most literal interpretation, mean, and nothing more” (WWM 32).
However, Davidson never makes explicit just what he means by “mean,” and the term
is notoriously ambiguous in both ordinary and philosophical discourse. In this section,
I work my way through a series of increasingly strong interpretations of what he might
A Companion to Donald Davidson, First Edition. Edited by Ernie Lepore and Kirk Ludwig.
© 2013 John Wiley & Sons, Inc. Published 2013 by John Wiley & Sons, Inc.
21
Metaphor and Varieties of Meaning
ELISABETH CAMP
elisabeth camp
362
mean, ultimately arguing that the most radical view does the most justice to the largest
proportion of what he says.
The weakest reading of the central thesis is simply the denial of the view that words
themselves change meanings when used metaphorically. Depending on what one
means by “semantics” and “pragmatics,” this weak reading is naturally captured in the
relatively uncontroversial slogan that metaphor is pragmatic rather than semantic.1
And much of “What Metaphors Mean” does attack the idea of metaphorical word
meanings by invoking the distinction between meaning and use:
I depend on the distinction between what words mean and what they are used to do. I think
metaphor belongs exclusively to the domain of use. . . . It is no help in explaining how
words work in metaphor to posit metaphorical or figurative meanings, or special kinds of
poetic or metaphorical truth. These ideas don’t explain metaphor, metaphor explains
them. Once we understand a metaphor we can call what we grasp “the metaphorical
truth” and (up to a point) say what the “metaphorical meaning” is. But simply to lodge
this meaning in the metaphor is like explaining why a pill puts you to sleep by saying it
has a dormative power. Literal sentences and literal truth conditions can be assigned to
words and sentences apart from particular contexts of use. This is why adverting to them
has genuine explanatory power. (WWM 33)
Similarly, he says:
The point of the concept of linguistic meaning is to explain what can be done with words.
But the supposed figurative meaning of a simile [or a metaphor] explains nothing; it is
not a feature of the word that the word has prior to and independent of the context of
use, and it rests upon no linguistic customs except those that govern ordinary meaning.
(WWM 40)
Because it is so uncontroversial to place metaphor in the domain of use, these passages
are notable more because they articulate criteria for linguistic meaning – as something
a word has “prior to and independent of” specific conversational contexts, in virtue of
“linguistic customs” – that will become Davidson’s target in “Nice Derangement.” The
weak interpretation of the central thesis is weak in the sense that it holds that this
criterion for word or sentence meaning is compatible with allowing that metaphor
exemplifies some other, more context-bound species of speaker’s meaning.
Davidson does appear to embrace metaphorical speaker’s meaning at several points
in “What Metaphors Mean.” For instance, he grants that metaphor “is effective in praise
and abuse, prayer and promotion, description and prescription” (WWM 33). He also
claims that metaphors can serve as assertions, lies, and promises:
What makes the difference between a lie and a metaphor is not a difference in the words
used or what they mean (in any strict sense of meaning) but in how the words are used.
Using a sentence to tell a lie and using it to make a metaphor are, of course, totally different
uses, so different that they do not interfere with one another, as say, acting and lying
do. . . . It can be an insult, and so an assertion, to say to a man “You are a pig”. . . . What
distinguishes metaphor is not meaning but use – in this it is like assertion, hinting, lying,
promising, or criticizing. (WWM 43)
metaphor
363
On the weakest reading of the central thesis, the speaker of a metaphor exploits her
words’ standard meanings in order to say or assert something else. This is a view that
various theorists, including myself endorse.2
However, it cannot be Davidson’s view,
because he repeatedly commits himself to a significantly stronger interpretation of the
thesis. Perhaps most obviously, he continues the passage above by adding: “And
the special use to which we put language in metaphor is not – cannot be – to ‘say
something’ special, no matter how indirectly. For a metaphor says only what it shows on
its face” (WWM 43, emphasis added). Likewise, he writes that “[a] metaphor doesn’t
say anything beyond its literal meaning (nor does its maker say anything, in using the
metaphor, beyond the literal)” (WWM 32, emphasis added).3
Despite the talk about metaphors underwriting description and assertion, then, Davidson
is committed to a strong version of the central thesis, one which denies that
metaphors express cognitive contents in any way. To get clear on just how strong his
noncognitivism is, and why he endorses it, we need to distinguish two intertwined but
distinct arguments he offers in its favor. The first argument is that metaphors lack the
right kind of effects to count as meaning. Thus, he writes, “we must give up the idea
that a metaphor carries a message, that it has a content or meaning (except, of course,
its literal meaning)” (WWM 45).4
Similarly, he says, “metaphor” is not “a form of communication
alongside ordinary communication [which] conveys truths or falsehoods
about the world much as plainer language does”; it is not “a vehicle for conveying ideas”
(WWM 32). Rather than telling us that things are a certain way, Davidson suggests,
metaphors make us see things in a certain way:
If what the metaphor makes us notice were finite in scope and propositional in nature, this
would not in itself make trouble; we would simply project the content the metaphor
brought to mind onto the metaphor. But in fact there is no limit to what a metaphor calls
to our attention, and much of what we are caused to notice is not propositional in character.
(WWM 46)
And again:
[The problem with a paraphrase is] not only that we can’t provide an exhaustive catalogue
of what has been attended to when we are led to see something in a new light; the difficulty
is more fundamental. . . . Seeing as is not seeing that. (WWM 47)
The “wrong kind” argument thus relies on the assumption, common among philosophers,
that meaning is an exclusively propositional matter. One might well question this
assumption – indeed, in his seminal article “Meaning,” Paul Grice only requires for an
utterance-event to have nonnatural meaning that “some effect” be produced in the
hearer in virtue of his recognition of the speaker’s intention to produce that effect
(Grice 1957: 385): he does not specify what form that effect must take, and even
explicitly allows that “of course, it may not always be possible to get a straight answer
involving a ‘that’ clause, for example, ‘a belief that . . .’” (Grice 1957: 385). More
importantly, Davidson’s positive point here – that metaphors (typically) produce a kind
of nonpropositional insight through seeing-as – is compatible with the possibility that
they might also, at least sometimes, express propositional contents which would qualify
as meaning. (I’ll argue in Section 3 that this is indeed the case.)
elisabeth camp
364
Davidson’s second argument cuts against the possibility of treating metaphors’
effects, propositional or not, as instances of meaning, by arguing that those effects are
not connected in the right way to the uttered sentence:
The various theories we have been considering mistake their goal. Where they think they
provide a method for deciphering an encoded content, they actually tell us (or try to tell
us) something about the effects metaphors have on us. The common error is to fasten on
the contents of the thoughts a metaphor provokes and to read these contents into the
metaphor itself. No doubt metaphors often make us notice aspects of things we did not
notice before. . . . The issue does not lie here but in the question of how the metaphor is
related to what it makes us see. (WWM 45, emphasis added).
To put the point in Gricean terms, it is generally assumed that in order to count as nonnatural
meaning, an utterance’s effects must be connected to the utterance in a rational
way: that is, as Grice says, the production of the intended effect “must be something
which in some sense is within the control of the audience, or [such] that in some sense
of ‘reason’ the recognition of the intention behind x is for the audience a reason and
not merely a cause” (Grice 1957: 385). But with metaphor, Davidson seems to claim,
the requisite rational structure is absent. Thus, at various points, he speaks of metaphors
(along with similes and other analogical devices) as “inspiring” or “prompting”
recognition of some fact or similarity (WWM 47), or as “inviting” or “bullying” hearers
into drawing a comparison (WWM 40–41) – all verbs that suggest a causal rather than
a rational mechanism of production. Similarly, he says that “Joke or dream or metaphor
can, like a picture or a bump on the head, make us appreciate some fact – but not by
standing for, or expressing, the fact” (WWM 46).5
In all of these cases, Davidson thinks, the metaphor, or perhaps the speaker, does
something more like showing or pointing the hearer toward a situation – something,
in Grice’s terms, more like “deliberately and openly letting someone know” or “getting
someone to think” than like “telling” or claiming (Grice 1957: 382). And the problem
with treating showing as a case of nonnatural meaning, according to Grice, is that the
hearer’s recognition of the speaker’s intention fails to play an essential role in producing
the effect: the speaker’s utterance at most prompts or causes him to notice
something that he could have noticed on his own.6
Similarly, Davidson argues, with
metaphor: any cognitive content a metaphor produces is something the hearer generates
for himself, with the speaker merely exercising more or less force in aiding its
production. This insistence on the hearer’s autonomous interpretive role is reflected in
the very first sentences of “What Metaphors Mean,” where Davidson writes, “Metaphor
is the dreamwork of language and, like all dreamwork, its interpretation reflects as
much on the interpreter as on the originator . . . understanding a metaphor is as much
a creative endeavor as making a metaphor, and as little guided by rules” (WWM 31).
Thus, while a metaphor’s speaker might harbor certain hopes or expectations about
the effects her utterance will have, these effects are not systematically predictable on
the basis of the sentence uttered, because they depend too heavily on the particular
hearer and the particular conversational situation. More importantly, on this view,
those effects are not constrained by the speaker’s expectations: the hearer is free to make
of the metaphor what he will, by noticing whatever similarities or facts happen to be
most interesting and notable to him.7
metaphor
365
In effect, the “wrong kind” objection invokes at the level of speaker’s meaning an
analogous criterion to that for linguistic meaning cited earlier. For any species of
meaning, Davidson assumes, we should only posit meanings that play a substantive
explanatory role, and purported meanings that are neither systematically predictable
nor rationally justifiable fail to meet this requirement. If this is right, then the only
theoretically useful notion of “meaning” associated with metaphor is that of literal
sentence meaning. At least we know that sentences have these meanings, as a result of
a compositional interpretive theory; further, we also know that we need to appeal to
these meanings to explain metaphors’ effects, since they, and not the uttered sentences’
mere sounds or shapes, play a crucial role in causing those effects.
In Section 3, I’ll challenge both Davidson’s “wrong kind” and “wrong way” arguments.
For now, by way of exegetical summary, I note that it doesn’t appear possible to
render all of Davidson’s claims about metaphors and meaning in “What Metaphors
Mean” consistent: in particular, the claim that metaphors can be assertions (WWM 43)
appears to conflict directly with the central negative thesis. The most plausible overall
reading, though, is a radically noncognitivist one, on which metaphorical utterances
fail to have or otherwise express any (nonliteral) propositional, cognitive contents.
Specifically, Davidson holds this because he assumes first, that speaker’s meaning must
be propositional, and must involve a rational, intentional relation between utterance
and effect; second, that the primary, distinctive work of metaphors is to produce a state
of seeing-as; and third, that seeing-as violates both these criteria for meaning.
2. “A Nice Derangement of Metaphors”
Turning now to Davidson’s later paper, “A Nice Derangement of Epitaphs,” we can
frame the primary exegetical challenge much as in Section 1. Here again, it is clear that
Davidson advocates a fairly radical negative conclusion, now extended from metaphor
to linguistic meaning across the board: “I conclude,” he says, “that there is no such
thing as a language, not if a language is anything like what many philosophers and
linguists have supposed” (NDE 265). Again, though, this central claim might mean
various things, depending on what Davidson means by “language” and “meaning.”
Further, while his focus is not on metaphor per se, the discussion has direct, significant
implications for the analysis of metaphor. As with “What Metaphors Mean,” I will work
my way through a sequence of increasingly strong interpretations of the central negative
thesis.
The weakest interpretation is that Davidson merely intends to claim that conventional
meaning is not sufficient for communication, because all communication involves
spontaneous, local, non-rule-governed coordination between speaker and hearer. In
effect, on this interpretation, Davidson thinks there is no such thing as a language just
because he takes Gricean interpretive processes to be so pervasive throughout the
course of linguistic interpretation that we have “erased the boundary between knowing
a language and knowing our way around in the world generally” (NDE 265).
Davidson illustrates the insufficiency of conventional meaning for successful com-
municationbyappealingtomalapropisms,appropriations,andneologisms:forinstance,
Goodman Ace’s sustained barrages of puns and twists, which Mark Singer describes as
elisabeth camp
366
a“baffleof wits,”orArchieBunker’sinvitationto“Leadthewayandwe’llprecede”(NDE
251). In all of these cases, speaker and hearer converge on a common, nonconventional
interpretation of the speaker’s words, even when one or both parties is ignorant of, or
ignores, the relevant expression’s conventional meaning. As Davidson says:
What is interesting is the fact that in all these cases the hearer has no trouble understanding
the speaker in the way the speaker intends. It is easy enough to explain this feat on the
hearer’s part: the hearer realizes that the “standard” interpretation cannot be the intended
interpretation; through ignorance, inadvertence, or design the speaker has used a word
similar in sound to the word that would have “correctly” expressed his meaning. . . . It
seems unimportant, so far as understanding is concerned, who makes a mistake, or
whether there is one. (NDE 252)
In this passage, we might say, Davidson invokes broadly Gricean reasoning to explain
successful communication of nonconventional meaning, generalizing Grice’s account
to encompass, first, cases in which the speaker’s departure from conventional meaning
is unintentional, and second, cases where the departure involves the interpretation of
specific uttered expressions rather than the generation of unsaid implicatures. The
interpretive strategy in both cases is fundamentally the same as for classic Gricean
implicature: the hearer assumes that the speaker is a basically rational agent who wants
to produce some specific effect, and he puts what he knows about the utterance the
speaker actually produced together with what he knows about the sort of person the
speaker is, what she believes, and what her purposes might be, to identify her intended
effect (Grice 1989).8
If the central negative claim of “Nice Derangement” was just that conventional
meaning does not suffice for the successful assignment of values to a speaker’s words
because it must be embedded within a more general interpretive machinery, it would
not be particularly radical. But Davidson claims, further, that conventional meaning is
also not necessary for successful communication, as illustrated by cases where the
speaker invents an entirely new expression rather than “converting” an old word to a
new use: “Sheer invention is equally possible, and we can be as good at interpreting it
(say in Joyce or Lewis Carroll) as we are at interpreting the errors or twists of substitution”
(NDE 259). Given this, Davidson concludes that conventions are altogether
explanatorily otiose in a theory of communication:
what interpreter and speaker share, to the extent that communication succeeds, is not
learned and so is not a language governed by rules or conventions known to speaker and
interpreter in advance; but what the speaker and interpreter know in advance is not (necessarily)
shared, and so is not a language governed by shared rules or conventions. (NDE
264)
Indeed, it seems we might use some of the arguments Davidson deployed in “What
Metaphors Mean” against metaphorical meaning to go further: to undermine the very
cases he invokes here as being instances of meaning at all. In “What Metaphors Mean,”
Davidson argued that metaphorical meanings lacked “genuine explanatory power”
because they could not “be assigned . . . apart from particular contexts of use” (WWM
33), and did not explain anything further that could be done with words. But if con-
metaphor
367
ventional semantics is now being ruled out as otiose, then, the only viable species of
linguistic or word meaning that remains would seem to be context-local speaker’s
meaning.9
Moreover, the way in which hearers arrive at these interpretations, Davidson
now says, is nothing more than a “strategy,” a “mysterious” and highly contextdependent
process (NDE 264): “there are no rules for arriving at passing theories, no
rules in any strict sense, as opposed to rough maxims and methodological generalities”
(NDE 265). But if genuine meaning must be predictive and rule governed, as assumed
in “What Metaphors Mean,” then this criterion threatens to undermine all varieties of
speaker’s meaning – both intentional and unwitting, and both implicatures and word
meanings.10
We thus appear to have a truly radical meaning skepticism on our hands.
This is emphatically not the conclusion of “Nice Derangement,” however. On the
contrary, Davidson says he “take[s] for granted . . . that nothing should be allowed to
obliterate or even blur the distinction between speaker’s meaning and literal meaning”
(NDE 252). Indeed, he thinks we are driven to give up the very thing he once treated
as explanatory bedrock – conventional, context-independent word meaning – precisely
in order to preserve this distinction (NDE 252). So why, and how, does he stop short
of wholesale skepticism? The answer, I think, is that where in “What Metaphors Mean”
Davidson took conventional, context-independent word meaning as fundamental,
and looked with suspicion on purported cases of meaning that failed to fit that model,
by the time of “Nice Derangement,” he has come to think that merely “strategic,”
broadly Gricean interpretive processes are sufficiently robust to provide an alternative
theoretical groundwork. While the central negative thesis of “Nice Derangement” is
that the pervasiveness of Gricean interpretive processes renders conventional meaning
explanatorily otiose, this is complemented by a positive project of showing that a
revised notion of compositional word meaning (of the sort Davidson himself spent
much of his philosophical career developing) still plays a crucial role within the
Gricean framework.
The crucial question, then, is how to understand this alternative, anticonventionalist
notion of compositional word meaning. Davidson rejects talk of “literal meaning”
as “too incrusted with philosophical and other extras to do much work,” and offers
“first meaning” in its place. “The concept,” he says,
applies to words and sentences as uttered by a particular speaker on a particular occasion.
But if the occasion, the speaker, and the audience are “normal” or “standard” (in a sense
not to be further explained here), then the first meaning of an utterance will be what
should be found by consulting a dictionary based on actual usage (such as Webster’s Third).
(NDE 252)
So far, this just says that first meaning is a lot like conventional meaning, except in those
cases when it’s not. A more helpful understanding, he suggests, comes from thinking
of first meaning as “first in the order of interpretation,” as determined by the speaker’s
communicative intentions: “The intentions with which an act is performed are usually
unambiguously ordered by the relation of means to ends (where this relation may or
may not be causal)” (NDE 253). As Davidson notes, “first meaning” is thus a species of
Gricean nonnatural meaning. However, he also observes, as we noted in Section 1, that
nonnatural meaning need not be distinctively linguistic:
elisabeth camp
368
Nothing said so far limits first meaning to language; what has been characterized is
(roughly) Grice’s non-natural meaning, which applies to any sign or signal with an
intended interpretation. What should be added if we want to restrict first meaning to
linguistic meaning? (NDE 254)11
Davidson suggests two additional conditions to achieve linguistic meaning: first, the
method for determining the intended interpretation must be systematic. That is, this
particular utterance meaning must be related in a rule-governed way to the meanings
of other potential utterances, such that the interpreter’s assignments of values to this
utterance’s parts and their mode of combination forms part of an overall theory that
enables him to interpret a wide range of other utterances (NDE 254). Second, the operative
theory must be shared between the interlocutors: that is, they must converge upon
a common understanding of the speaker’s utterance using a sufficiently similar method
(NDE 256). Davidson cautions, however, that this second requirement does not entail
that the hearer would use the same theory either to produce his own utterances or to
understand a different speaker in another context: all that “must be shared” for successful
communication “is the interpreter’s and the speaker’s understanding of the
speaker’s words” (NDE 257). Nor, of course, is it necessary that speaker and hearer
share this interpretive theory of the speaker’s words prior to the conversation; this is
the point of malapropisms and neologisms, and the central negative thesis. Perhaps we
might sum up Davidson’s view here in the slogan “All semantics (and syntax) is local.”12
How does this context-local notion of “first meaning” avoid “obliterating” the
distinction between word meaning and speaker’s meaning?The answer should be intuitively
obvious: given, for instance, that Archie Bunker’s interpretive theory assigns as
the semantic value for “precede” what I (and presumably you) would assign for
“proceed,” and given that we realize this fact about him, Archie can successfully use
the sentence “Lead the way and we’ll precede” to implicate something more to us – say,
that it’s time to go to the party. Further, we need to appeal to this context-local “first
meaning” in order to explain Archie’s overall speech act, including both the additional
content that the hearer recognizes Archie intended for him to entertain, and the way
in which Archie intended him to identify it.
While the basic distinction between first/word meaning and speaker’s meaning and
its explanatory utility are intuitively obvious, spelling out the distinction in a principled
way, and especially making sense of Davidson’s specific cases, is rather more difficult.
In particular, returning to our central topic, the status of metaphor becomes
particularly vexed on this view. Davidson only mentions metaphor once in “Nice
Derangement,” and only in passing, when he says that metaphor is an instance of
something that is done “all the time,” where a speaker “has said something true by
using a sentence that is false” (NDE 258). The first point to make here is obvious: this
represents a dramatic shift from the staunch noncognitivism of “What Metaphors
Mean.” Aside from the question of intertextual consistency, though, the remark raises
a puzzle internal to “Nice Derangement”: what conception of “first meaning” entitles
Davidson to treat metaphor in this way? In particular, why does metaphor not count
as a form of first meaning?
To see the puzzle, we need to compare two cases, both originally from Keith Donnellan,
that together illustrate how Davidson is drawing the distinction between word and
metaphor
369
speaker’s meaning.The first is Donnellan’s (1966) classic example motivating a distinction
between referential and attributive uses of definite descriptions: Jones utters
“Smith’s murderer is insane” in order to talk about someone (say, George) who Jones
believes falsely to have murdered Smith (perhaps George is on trial for the murder).
Davidson here agrees with Donnellan that the speaker, Jones, successfully refers to
George; but he insists, seemingly against Donnellan, that Jones’s words refer to the
actual murderer:
. . . the reference is none the less achieved by way of the normal meanings of the words.
The words therefore must have their usual reference. All that is needed, if we are to accept
this way of describing the situation, is a firm sense of the difference between what words
mean or refer to and what speakers mean or refer to. . . . Jones has said something true by
using a sentence that is false. This is done intentionally all the time, for example in irony
or metaphor. A coherent theory could not allow that under the circumstances Jones’ sentence
was true; nor would Jones think so if he knew the facts. Jones’ belief about who
murdered Smith cannot change the truth of the sentence he uses (and for the same reason
cannot change the reference of the words in the sentence). (NDE 258)
Davidson’s reasoning here is just an instance of the standard Gricean machinery, generalized
as we saw earlier to include cases where the hearer realizes that the speaker
doesn’t realize something, either about her words (as in the case of Mrs. Malaprop) or
about the world (as here).
The second case is one in which Donnellan (1968) responded to an accusation by
Alfred McKay that his description of the previous case showed that Donnellan “shared
Humpty Dumpty’s theory of meaning: ‘When I use a word’, Humpty Dumpty said, . . .
‘it means just what I choose it to mean’” (NDE 257). (Earlier in his conversation with
Alice, Humpty Dumpty had used the word “glory” to mean “a nice knockdown argument.”)
In response to McKay’s accusation, Donnellan says,
If I were to end this reply to MacKay with the sentence “There’s glory for you” I would be
guilty of arrogance and, no doubt, of overestimating the strength of what I have said, but
given the background I do not think I could be accused of saying something unintelligible.
I would be understood, and would I not have meant by “glory” “a nice knockdown argument”?
(Donnellan 1968: 213)
Donnellan’s main point here is that while Humpty Dumpty could not succeed in
meaning “a nice knockdown argument” by “There’s glory for you” just out of the blue,
Donnellan can successfully mean that in his current context, because the right sort of
Gricean background is in place: specifically, Donnellan knows that McKay knows (that
Donnellan knows, etc.) about Humpty Dumpty’s failed attempt at meaning.
For Donnellan, then, the two examples, of Jones referring to George and of Donnellan
saying he has a nice knockdown argument, are on a par: in both cases, the speaker
succeeds in meaning or saying something even though he departs from his words’
standard, conventional meanings. But Davidson treats the two cases very differently:
in the case of “Smith’s murderer,” he thinks the words retain their standard meaning
although the speaker uses them to refer to something else;13
while with “glory,” he
thinks, the words themselves take on new meaning:
elisabeth camp
370
When Donnellan ends his reply to MacKay by saying “There’s glory for you,” not only he,
but his words, are correctly interpreted as meaning “There’s a nice knock-down argument
for you.” That’s how he intends us to interpret his words, and we know this, since we have,
and he knows we have, and we know he knows we have (etc.), the background needed to
provide the interpretation. (NDE 260)
Given the structural similarity between Donnellan’s use of “glory” and the intentional
malaprops deployed by Goodman Ace, we can see why Davidson would want to treat
Donnellan’s “glory” as a case of context-local word meaning. However, recall that “first
meaning” was also supposed to be literal meaning: it is the meaning the words have in
that context. And this seems like an odd thing to say about “glory” here. In particular,
with cases of intentional twists and appropriations like Goodman Ace’s and Donnellan’s,
we would miss out on an important part of what’s funny – and thus an important
part of the effect the speaker intends to produce – if we simply took the speaker’s ultimately
intended word meaning as the words’ “first” meaning.
Putting aside brute intuitions about whether “glory” itself or only Donnellan as
speaker means “a nice knockdown argument,” the more important question from a
theoretical perspective is what differentiates “glory,” which Davidson does want to treat
as a case of first, literal, albeit context-local word meaning, from metaphor, which he
does not. For it is at least as plausible that the words take on a context-specific meaning
in the case of metaphor, as it is with intentional twists like Donnellan’s use of “glory.”14
In particular, first, with metaphor as with “glory,” it is typically15
possible to identify a
specific word or phrase as metaphorical while the other terms in the uttered sentence
are used literally. Moreover, unlike with “glory,” it is typically possible to extend that
metaphorical interpretation to related expressions, for instance through tense, pluralization,
adjectivalization, colligation, and so on. Second, metaphor, even more clearly
than an intentional twist like “glory,” permits echoic speech and attitude reports, anaphoric
reference and (dis)agreement (e.g. “That’s true” or “She’s not that way at all”),
and embedding within complex constructions like conditionals and attitude reports.
And third, with both intentional twists and metaphor, the speaker’s contextually
intended interpretation of her words can serve as input for implicature: for instance,
by saying “She’s the Taj Mahal,” a speaker might implicate that he would like to date
the woman in question (Tsohatzidis 1994; cited in Bezuidenhout 2001).
But if metaphorical assignments of values to particular expressions contribute
to the compositional determination of sentence meanings which in turn serve as
inputs to implicature, then it seems fairly clear that metaphor falls squarely within the
domain of what Davidson calls “first meaning” – even as he also equates this with literal
meaning:
Every deviation from ordinary usage, as long as it is agreed on for the moment (knowingly
deviant, or not, on one, or both, sides), is in the passing theory as a feature of what the
words mean on that occasion. Such meanings, transient though they may be, are literal;
they are what I have called first meanings. (NDE 261; see also 252)
Thus, unless he can identify a principled basis for distinguishing metaphor from intentional
twists like “glory,” it appears that Davidson has not merely “blurred” but “obliterated”
the distinction between literal and speaker’s meaning after all. And in turn, it
metaphor
371
appears that he also lacks the resources to claim that in speaking metaphorically, a
speaker “says something true by using a sentence that is false”: rather, it seems he
should hold that when all goes well, the sentence uttered is itself true, on the contextually
operative interpretation. This would take us far from the view in “What Metaphors
Mean,” indeed.
Although I think this is a real problem for Davidson’s view here, I also think the best
basis for distinguishing the two cases relies on a feature of metaphor that Davidson
himself cited in “What Metaphors Mean.”There, he argued that in a metaphorical utterance,
the words’ literal meanings play a crucial role in producing the metaphor’s effects
in two related senses. First, unless a hearer appreciates the change from literal to metaphorical
use, “most of what is thought to be interesting about metaphor is lost” (WWM
37). Second, in the case of metaphor, “our attention is directed . . . to what language is
about” (WWM 37), rather than to the language itself: we notice new features of Juliet,
or Louis XIV, or Richard III, by thinking about them in comparison to the sun, and not
just as falling within the extension of “the sun.” That is, it is the normal semantic value
of the metaphorical expression, and not just the expression itself, which plays a central
role in achieving the metaphors’ effects. Both of these features are missing from a
genuine malapropism: precisely because the speaker is confused about the expression’s
normal semantic value, she cannot intend for the hearer to play with that normal value
in order to achieve some effect. In the case of Donnellan’s “glory,” the first feature is
present – the utterance is supposed to be interesting because we know Donnellan knows
the normal meaning of “glory.” But the second feature is still absent: there is no interesting
connection between the two phenomena of glory and a nice knockdown argument
in themselves; there is only the arbitrary connection that Humpty Dumpty attempted to
create between the two expressions for them. Other cases, like Goodman Ace’s intentional
twists and many puns, fall somewhere in between “glory” and metaphor here.
What this shows, I think, is that if Davidson wants to retain “first meaning” as a
form of meaning that is both context-local and literal, he needs to appeal to more varieties
of meaning than simply “prior” and “passing” word meanings and propositional
“speaker” meaning. At a minimum, he needs to acknowledge that distinct theoretically
relevant roles are played first, by each interlocutor’s “prior” lexical entry for a given
expression (conventional or not); second, by the context-specific “passing” “first” word
meaning, which both parties must converge on for successful communication; third,
by any “secondary” context-specific expression-values which are parasitic on this initial
passing meaning, such as metaphor and some sarcasm (and perhaps other phenomena
like deferred reference); and finally, by classic Gricean implicatures. (Of course, in many
cases two or even all of the first three species of meaning may coincide.) Only by taking
seriously all of these varieties of word meaning can Davidson make good on his positive
project of showing that an anticonventionalist theory can still reliably distinguish word
from speaker’s meaning.
3. Evaluating a Hybrid Davidsonian View of Metaphor
At this point, it is time to step back from the details of the two papers, to ask what form
a broadly Davidsonian view of metaphor, one that preserves the most important aspects
of both papers, should take. I think we can identify four central and independently
elisabeth camp
372
plausible theses from each paper. From “What Metaphors Mean,” we get first, the claim
that literal meaning plays a crucial role in the production of metaphor’s effects:
Whether or not metaphor depends on new or extended meanings, it certainly depends in
some way on the original meanings; an adequate account of metaphor must allow that
the primary or original meanings of words remain active in their metaphorical setting.
(WWM 34)
Second, there are no clear rules for getting from literal sentence meaning to a metaphor’s
effects: in this sense, metaphorical meaning is not strictly “calculable,” as Grice
claimed implicatures to be. Third, meaning is propositional: it is a form of cognitive
content, which is in turn analyzed in terms of conditions for truth or satisfaction.
Finally, seeing-as constitutes an essential, inextricable aspect of what metaphors do,
and is not itself reducible to seeing-that.
From “Nice Derangement,” we have the following claims. First, conventional
meaning is neither necessary nor sufficient for successful communication. Second, the
input to compositional construction (“word meaning”) is often context-local rather
than conventional. Third, the process by which interlocutors develop “passing” interpretive
theories from “prior” ones is highly context specific and strategic rather than
strictly rule governed. Finally, accepting these three theses does not undermine the
distinction between word and speaker’s meaning.
We can incorporate all eight of these theses to produce the following broadly
Davidsonian view of metaphor, one which is both noncognitivist and anticonventionalist.
Literal word meaning, which may itself be context-local serves as the input to
a compositional interpretive theory and delivers a “passing” meaning for the uttered
sentence. This meaning then serves as a springboard, via mysterious, non-rule-governed
“strategies,” for the hearer’s seeing one thing in terms of another – in the simplest
case, for seeing the object or situation denoted by the subject expression (e.g., Juliet) in
the light of the object, situation, or quality denoted by the predicate (e.g., being the
sun). This cognitive state is not itself a candidate for meaning because it is not propositional.
It may, however, lead the hearer to notice that the two entities share interesting
resemblances, and that the subject possesses heretofore unnoticed properties. Finally,
a Davidsonian theory might perhaps allow that a speaker can exploit the anticipatedto-be-induced
state of seeing-as to implicate (but not to say or assert) that the subject
possesses these properties.
What should we make of this view? I think it has three important theoretical merits.
First, its emphasis on the continued contextual relevance of “first,” literal meaning
serves as an important corrective to the view of many contextualists,16
who argue that
in metaphorical utterances, words take on ad hoc meanings, which are the only relevant,
operative meanings of those words in that context. As Davidson says on such a
view “there is no difference between metaphor and the introduction of a new term into
our vocabulary: to make a metaphor is to murder it” (WWM 34). Second, at the same
time, it pushes us to acknowledge that the literal meaning which serves as input for
generating metaphorical effects need not itself be conventional. Thus, if I say “George
is a primate,” intending to suggest that he is territorial, prone to violence, and not too
smart, then the relevant literal meaning of “primate” is one which has already been
metaphor
373
narrowed to exclude human beings as relevant instances (Camp 2008). Finally, Davidson’s
emphasis on the nonpropositional effects of metaphor, and specifically on the
production of a rich, open-ended, often imagistic state of seeing-as, is a crucial corrective
to the narrower, exclusively propositional focus of many philosophical discussions
of metaphor and related uses of language.
However, what I think the contextualists get right, and what this Davidsonian
view misses, is that metaphorical interpretation typically, perhaps always, also generates
a speech act in which a proposition other than the uttered sentence’s “passing”
meaning is presented with assertoric or other illocutionary force. In this sense, I take
metaphor to be importantly akin to “primary” propositional meaning (Recanati
2001), and distinct from implicatures. The speaker of metaphors like “She’s the Taj
Mahal,” “Her hair is burnished copper,” or even “Thou art a boil in my corrupted
blood” has not merely intimated or suggested or pointed toward something: she has
committed herself outright to some content, which can be false as well as true, and
which can genuinely inform the hearer and not merely bring him to notice something
he could have ascertained independently (Camp 2006b). Further, this content
can itself in turn serve as the input for implicatures, or for sarcasm (Camp 2012).
Finally, I take it that this content is identified by way of cultivating a state of seeing-as
(Camp 2003, 2008).
Accepting these points would obviously involve abandoning Davidsonian noncognitivism
– though as we saw, Davidson himself abandons that view in “Nice
Derangement.” To adjudicate between the two positions, we need to evaluate Davidson’s
two strongest arguments for noncognitivism: that the effects produced by metaphor
are of the wrong kind, and are connected in the wrong way, to count as meaning;
I take the “wrong way” argument first.
The claim that neither seeing-as nor any propositions arrived at through it are
governed by strict “rules” for metaphorical interpretation is certainly correct: as Davidson
emphasizes in “What Metaphors Mean,” metaphorical interpretation requires
imagination and intuition – though as he also points out in both papers, this is also
true of all communication to some degree. Nor, at least by the time of “Nice Derangement,”
does Davidson take a role for interpretive imagination to disqualify an effect as
one of meaning. And this seems right: as many theorists have noted, the Gricean
requirement of calculability needs to be construed quite loosely if it is to cover even
canonical cases of implicature.
More importantly, in most if not all interpretive contexts, metaphorical interpretation
is normatively constrained by the speaker’s intentions in a manner characteristic
of non-natural speaker’s meaning. So, for instance, suppose a hearer interprets Romeo’s
utterance of “Juliet is the sun” in a way that employs the same perspective, and so
attributes the same features, that he would have employed if he heard Achilles or Louis
XIV described as the sun, or even as he would if he heard Romeo’s friend Benvolio or
his romantic rival Paris utter the very same sentence. I think it is intuitively clear that
such a hearer would have misinterpreted Romeo’s metaphor, in virtue of failing to take
Romeo’s interpretive intentions into account (Camp 2008). The view that metaphors
are self-sufficient objects of contemplation, which simply draw the hearer’s attention
to features that he could have noticed on his own, is dubious enough in the context of
lyric poetry, but clearly false in the case of conversation.
elisabeth camp
374
The objection that seeing-as, and any further cognitive effects it produces, are the
wrong kind of entity to count as meaning is, I think, considerably more plausible. As I
noted in Section 1, the assumption that speaker’s meaning is inherently propositional
is widespread, even if it is not part of Grice’s influential definition. I myself am tempted
by the thought that a nonpropositional perspective can be part of what a speaker means
to communicate by her utterance, but I recognize that this is a controversial view.
However, even if we reject seeing-as as a candidate for content, this does not establish
that there are no other propositional candidates for illocutionary content in the offing.
And it seems simply incredible to deny that speakers ever, or even often, commit themselves
to contents by their metaphorical utterances. So, for instance, it seems clear that
in the context of a conversation about who should be the next department chair, a
speaker might utter “Sam is a bulldozer” and thereby claim, roughly, that Sam is forceful,
determined, and not particularly considerate of others’ feelings (Bezuidenhout
2001).
The natural response to this is to object, as Davidson does (echoing Max Black and
Cleanth Brooks before him), that the effects produced by metaphor are still too indeterminate
and open-ended to count as meaning: as he says, “When we try to say what a
metaphor ‘means’ we soon realize there is no end to what we want to mention” (WMM,
46). Here again, I think Davidson’s emphasis on indeterminacy and open-endedness is
salutary. (Although it is also important to note that a speaker’s metaphorical illocutionary
content may be quite specific: for instance, whenTed Hughes writes in “Pike,” about
a fish lying on shore, “The gills were kneading quietly,”17
the metaphorical verb describes
a quite specific action; likewise, the description of Sam earlier is fairly limited in its
intended effect.
Once again, though, I do not think indeterminacy or even open-endedness should
disqualify propositions as being meant, so long as the requisite Gricean intentions
are in place: that is, so long as the speaker intends for her hearer to take her to be
committed to this rough body of propositions, and to other related ones, and for this
commitment to be achieved by way of the hearer’s recognition of this very intention.
Indeed, as Grice (1989: 58) notes, indeterminacy is typical of pragmatic communication;
among other factors, it is rhetorically useful to have communicative methods
that preserve the indeterminacy of our corresponding mental states (Camp 2006a).
The primary difference in the case of metaphor is that there is no determinate, literal,
conventionally determined “core” which is asserted. As the later Davidson insists,
however, it is not clear that such a core is always present even for cases of literal
meaning. Thus, I conclude that neither the wrong way nor the wrong kind arguments
for noncognitivism succeed; we should allow that in speaking metaphorically,
speakers do indeed “say something [potentially] true by using a sentence that is
[usually] false” (NDE 258).
Before closing, I want to step back from the specific topic of metaphor to the broader
question of conventional meaning. I think we should grant that conventional meaning
is neither necessary nor sufficient for successful communication. But I believe we should
reject Davidson’s inference to the conclusion that conventions have no theoretical role
toplayinanaccountof meaninganddiscourse.Davidsonclaimsthat“errorormistake,”
in the sense of deviating from “what a good dictionary would say, or what would be
metaphor
375
found by polling a pod of experts whose taste or training I trust,” “is not philosophically
interesting” (NDE 252). I disagree: in addition to accounting for cases where hearers
are sufficiently cognitively agile and charitably motivated to figure out what nonstandard
meaning or effects the speaker must have intended to achieve with her utterance,
a theory of meaning should also explain cases of communicative failure: why and how
interlocutors respond as they do when hearers are too uninformed, lazy, or stupid to go
the extra mile required to reconstruct the speaker’s communicative intentions, or when
hearers do recognize what the speaker is trying to do, but refuse to let her “get away
with it” (NDE 258). Speakers who diverge from conventional meaning do so “‘under
license’ from other participants,” as Grice (1989: 45) says, and while hearers are typically
willing to cooperate by “accommodating” a manifestly intended meaning (Lewis
1979: 340), this sort of charity goes above and beyond the minimal conditions on
cooperative conversation.18
What Davidson rightly emphasized in “What Metaphors
Mean” is that literal, conventional meaning provides a stable, cross-contextual “least
common denominator” for communication: a basis for constructing meaning that both
speakers and hearers are entitled to assume their interlocutor acknowledges. This
common basis rarely coincides with utterances’ ultimate meanings, and speakers and
hearers can indeed get along without it. And in any case, it needs to be embedded within
a larger machinery of Gricean intention recognition. But if we ignore literal, conventional
meaning altogether, we deprive ourselves of the resources to understand how
speakers intentionally manipulate conventional meaning to nonconventional ends, for
the sake of obfuscation, insight, or entertainment, and how hearers sometimes fight
back against such manipulations.
Acknowledgment
The author would like to give thanks for discussion to Josh Armstrong, Endre Begby,
Louise Daoust, Carlos Santana, and participants at the NEH Summer Seminar on Quine
and Davidson.
Notes
1 For recent dissension from pragmatic models, see Hills (1997) and Stern (2000).
2 Bergman (1982) and Camp (2008).
3 We could render Davidson’s claims here consistent by assuming a distinction between what
a speaker says and what she asserts. Again, this is a position I myself endorse (Camp 2007a).
If so, however, one would expect him to note this explicitly. More importantly, it is incompatible
with the arguments for noncognitivism offered below. Finally, we cannot take him to
endorse an interpretation of the central thesis on which the sentence the speaker utters, as
well as what if anything she says and asserts, all have only their ordinary literal meanings,
but she still puts forward a nonliteral meaning in some other way, such as implicating,
hinting, or insinuating. It is true that Davidson says he has “no quarrel” with a view
of metaphor summarized by Heraclitus’s statement about the Delphic oracle: “It does
not say and it does not hide; it intimates” (WWM 46). And we might plausibly think
of intimation as a form of Gricean implicature, as with Grice’s (1989: 33) classic letter of
elisabeth camp
376
recommendation. However, Davidson explicitly rejects intimation as a form of meaning: in
discussing T.S. Eliot’s poem “The Hippopotamus,” which consists of series of contrasts
between the hippopotamus and the Church, he says,
Here we are neither told that the Church resembles a hippopotamus (as in simile) nor bullied
into making this comparison (as in metaphor), but there can be no doubt the words are being
used to direct our attention to similarities between the two. . . . The poem does, of course, intimate
much that goes beyond the literal meaning of the words. But intimation is not meaning.
(WWM 41)
4 Specifically, he resists any attempt to “shrug [ his arguments] off as no more than an insistence
on restraint in using the word ‘meaning.’This would be wrong.The central error about
metaphor is most easily attacked when it takes the form of a theory of metaphorical
meaning, but behind that theory, and statable independently, is the thesis that associated
with a metaphor is a cognitive content that its author wishes to convey and that the interpreter
must grasp if he is to get the message. This theory is false, whether or not we call the
purported cognitive content a meaning” (WWM 46).
5 As Richard Rorty (1987: 290) says in elaborating Davidson, metaphors “do not (literally)
tell us anything, but they do make us notice things. . . . They do not have cognitive content,
but they are responsible for a lot of cognitions.”
6 Note, however, that Grice actually makes this condition fairly weak. He rejects the case of
Herod showing Salome the head of St. John the Baptist on a charger as one of nonnatural
meaning because it is Salome’s appreciation of the situation itself – the natural connection
between a severed head and the death of the person whose head it is – which produces the
effect in her; Herod’s role is merely one of staging. However, in cashing out this requirement,
Grice says only that “the [hearer’s] recognition [of the speaker’s intention to produce a
certain belief in him] is intended by [the speaker] to play its part in inducing the belief, and
if it does not do so something will have gone wrong with the fulfillment of A’s intentions.
Moreover, A’s intending that the recognition should play this part implies, I think, that he
assumes that there is some chance that it will in fact play this part, that he does not regard it as
a foregone conclusion that the belief will be induced in the audience whether or not the intention
behind the utterance is recognized” (Grice 1957: 384, emphasis added). As I argue in
Section 3, it is plausible that a speaker’s communicative intentions do satisfy this condition
in the case of seeing-as produced by metaphor.
7 For a more recent defense of this view, see Lepore and Stone (2010). In effect, Davidson
thinks of a metaphor as an object that is created by a speaker but then enters the world on
its own. Davidson’s stance here is strikingly similar to the New Critical view of poetry
articulated by Wimsatt and Beardsley (1954): “The poem is not the critic’s own and not the
author’s (it is detached from the author at birth and goes about the world beyond his power
to intend about it or control it).” Along similar lines, Davidson says,
The critic is, so to speak, in benign competition with the metaphor maker.The critic tries to make
his own art easier or more transparent in some respects than the original, but at the same time
he tries to reproduce in others some of the same effects the original had on him. In doing this
the critic also . . . calls attention to the beauty or aptness, the hidden power, of the metaphor
itself. (WWM 47)
8 Davidson might bolster the generality of the point that all communication involves broadly
Gricean reasoning made earlier by noting, as Bach and Harnish (1979: 10) do, that even
when an utterance’s intended interpretation does fully coincide with its conventionally
metaphor
377
determined one, identifying this fact itself requires the hearer to employ the same nuanced,
context-sensitive interpretive abilities as conversion and invention do.
9 Indeed, Davidson’s skepticism about preestablished conventional meaning extends beyond
semantics, to encompass syntax: thus, Davidson thinks we have good evidence that “internal
grammars do differ among speakers of ‘the same language,’” for instance in their rules
for pluralization (NDE 263).
10 One of Grice’s key reasons for thinking that the connection between what is said and what
is implicated is rational, and so counts as a case of nonnatural meaning, is that implicatures
are “calculable” on the basis of what is said. In this sense, at least, Gricean implicatures are
also classically assumed to be “rule-governed.”
11 Note that Gricean nonnatural meaning encompasses more than “first” meaning, contrary
to what Davidson here appears to assume.
12 Note that Davidson’s two conditions – or even all three, including conventionality – do not
suffice to distinguish linguistic meaning from other forms of “first” meaning, as he claims.
Many nonlinguistic representational systems, most notably cartographic and diagrammatic
systems, are systematic, shared, and conventional (Camp 2007b, 2009).
13 It is difficult to determine whether Donnellan himself really does think the words change
their meanings in either case; in particular, in his discussion of “glory,” Donnellan only
claims that he meant “a nice knock-down argument” by “glory.” See Kripke (1977) for an
argument that Donnellan must be committed to a semantic analysis of the referential/
attributive distinction, and for a classic defense of the alternative pragmatic interpretation
invoking the distinction between semantic and speaker’s meaning.
14 For advocacy of a word-based approach to metaphorical interpretation, whether semantic
or pragmatic, see inter alia Bezuidenhout(2001), Carston (2002), Hills (1997), Levin
(1977), Recanati (2001), Sperber and Wilson (1985), and Stern (2000). Further, note that
some cases of sarcasm behave in much the same way (Camp 2012).
15 The existence of phrasal and sentential metaphors, in which the entire unit serves as a
vehicle for thinking about some other entity or situation, as in “I see storm clouds gathering
over the hill,” said in the context of an increasingly hostile conversation, constitute an
important, and I think theoretically revealing, exception to this claim.
16 For example, Bezuidenhout (2001), Recanati (2001), and Sperber and Wilson (1985).
17 Ted Hughes, “Pike.”
18 For further discussion of interpretive antagonistic conversations and their implications for
the theory of meaning, see Camp (2013).
References
Bach, K. and Harnish, R. (1979). Linguistic Communication and Speech Acts. Cambridge, MA: MIT
Press.
Bergmann, M. (1982). Metaphorical assertions. The Philosophical Review 91(2):229–245.
Bezuidenhout, A. (2001). Metaphor and what is said: a defense of a direct expression view of
metaphor. Midwest Studies in Philosophy 25:156–186.
Camp, E.. (2003). Saying and seeing-as: the linguistic uses and cognitive effects of metaphor.
Dissertation, University of California, Berkeley.
——— (2006a). Metaphor and that certain “je ne sais quoi.” Philosophical Studies 129(1):
1–25.
———. (2006b). Metaphor in the mind: the cognition of metaphor. Philosophy Compass 1(2):
154–170.
elisabeth camp
378
———. (2007a). Prudent semantics meets wanton speech act pluralism. In Context-Sensitivity
and Semantic Minimalism: New Essays on Semantics and Pragmatics, G. Preyer and G. Peter (eds).
Oxford: Oxford University Press, pp. 194–213.
———. (2007b). Thinking with maps. In Philosophical Perspectives 21:1, Philosophy of Mind,
J. Hawthorne (ed.). Oxford: Wiley-Blackwell, pp. 145–182.
———. (2008). Showing, telling, and seeing: metaphor and “poetic” language. In The Baltic
International Yearbook of Cognition, Logic, and Communication (Volume 3), A Figure of Speech:
Metaphor, E. Camp (ed.). Manhattan, KS: New Prairie Press, pp. 1–24.
———. (2009). A language of baboon thought? In Philosophy of Animal Minds, R. Lurz (ed.).
Cambridge, UK: Cambridge University Press, pp. 108–127.
———. (2012). Sarcasm, pretense, and the semantics/pragmatics distinction. Noûs 46(4):
587–634.
———. (2013). Working to conversational rule: deniability and pedantry in antagonistic conversational
contexts.
Carston, R. (2002). Thoughts and Utterances: The Pragmatics of Explicit Communication. Oxford:
Blackwell.
Donnellan, K. (1966). Reference and definite descriptions. The Philosophical Review 75(3):
281–304.
———. (1968). Putting Humpty Dumpty together again. The Philosophical Review 77(2):
203–215.
Grice, H.P. (1957). Meaning. The Philosophical Review 66(3):377–388.
———. (1989). Logic and conversation. In Studies in the Way of Words. Cambridge MA: Harvard
University Press, pp. 22–40.
Hills, D. (1997). Aptness and truth in verbal metaphor. Philosophical Topics 25(1):117–153.
Kripke, S. (1977). Speaker’s reference and semantic reference. Midwest Studies in Philosophy 2:
255–276.
Lepore, E. and Stone, M. (2010). Against metaphorical meaning. Topoi 29(2):165–180.
Levin, S. (1977). The Semantics of Metaphor. Baltimore, MD: Johns Hopkins University Press.
Lewis, D. (1979). Scorekeeping in a language game. Journal of Philosophical Logic 8:339–359.
Recanati, F. (2001). Literal/nonliteral. Midwest Studies in Philosophy 25:264–274.
Rorty, R. (1987). Unfamiliar noises I: Hesse and Davidson on metaphor. Proceedings of the Aristotelian
Society 61(Supp.):283–296.
Sperber, D. and Wilson, D. (1985). Loose talk. Proceedings of the Aristotelian Society 86:
153–171.
Stern, J. (2000). Metaphor in Context. Cambridge, MA: MIT Press.
Tsohatzidis, S. (1994). Speaker meaning, sentence meaning and metaphor. In Foundations of
Speech Act Theory: Philosophical and Linguistic Perspectives, S.L. Tsohatzidis (ed.). London:
Routledge, pp. 365–373.
Wimsatt, W.K. Jr. and Beardsley, M.C. (1954). The intentional fallacy. In The Verbal Icon: Studies
in the Meaning of Poetry. Lexington: University of Kentucky Press.