SciELO - Scientific Electronic Library Online

vol.29 número2Procedural meaning and definite descriptionsIntroduction: General term rigidity and Devitt's rigid appliers índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados



  • No hay articulos citadosCitado por SciELO

Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google


Análisis filosófico

versión On-line ISSN 1851-9636

Anal. filos. vol.29 no.2 Ciudad Autónoma de Buenos Aires nov. 2009



Buenos Aires Symposium On Definite Descriptions:

Michael Devitt

The Graduate Center, The City University of New York


The present article contains a defense of the thesis that definite descriptions can have referential meanings that include a descriptive component from the following objections contained in the preceding articles: (i) the idea that the thesis at stake cannot adequately account for cases of misdescriptions (Díaz Legaspe), (ii) the claim that referential descriptions should be considered to be purely referential, with no descriptive meaning component whatsoever (Skerk), and (iii) the alleged viability of a pragmatic approach according to which definite descriptions do not have referential meanings but can only be referentially used (Caso). As far as (i) is concerned, it is argued that misdescriptions can be clearly accounted for in pragmatic terms. Against (ii), it is pointed out that sentences of the form 'The/an F is not F' have a contradictory status -they are taken to be contradictory even when the descriptions involved have referential meanings. Finally, regarding (iii), it is argued that, in defending a very audience-directed stance, Caso confuses the epistemic and constitutive dimensions of reference determination, and as a consequence he is committed to an extreme form of linguistic pragmatism according to which the meaning of an expression is constitutively determined by the way in which hearers interpret utterances.

KEY WORDS: Referential meaning for descriptions; Singular proposition; Semantic convention; Pragmatic regularity; The semantics-pragmatics distinction.


El presente artículo contiene una defensa de la tesis según la cual las descripciones definidas pueden tener significados referenciales que incluyen un componente descriptivo frente a las siguientes objeciones contenidas en los artículos precedentes: (i) la idea según la cual la tesis en cuestión no explica adecuadamente los casos de descripciones fallidas (Díaz Legaspe), (ii) la tesis de que no hay razones para creer que los significados referenciales de las descripciones incluyan un componente descriptivo (Skerk), y (iii) la supuesta viabilidad de una perspectiva pragmática, desde la cual se considera que las descripciones no tienen significados referenciales sino que pueden ser usadas referencialmente (Caso). En relación con (i), se sostiene que las descripciones fallidas pueden ser claramente explicadas en términos pragmáticos. Contra (ii), se destaca que las oraciones de la forma 'El/un F no es F' tienen un carácter contradictorio -se considera que son contradictorias aun en los casos en los que las descripciones involucradas tienen significados referenciales. Finalmente, contra (iii), se argumenta que, al defender un acercamiento muy orientado hacia la audiencia, Caso confunde la dimensión epistémica de la determinación de la referencia con su dimensión constitutiva y, en consecuencia, resulta comprometido con una forma extrema de pragmatismo lingüístico según la cual el significado de una expresión se encuentra constitutivamente determinado por la manera en que los hablantes interpretan las emisiones.

PALABRAS CLAVE: Significado referencial de las descripciones; Proposición singular; Convención semántica; Regularidad pragmática; La distinción entre semántica y pragmática.

Earlier versions of the papers in this symposium were delivered at a very enjoyable workshop in Buenos Aires in March 2008. I thank Eleonora Orlando for organizing that workshop and for her helpful introduction. And I thank Justina Díaz Legaspe, Laura Skerk, and Ramiro Caso for their interesting papers. I will respond to them in order.

1. Introduction

I urge the thesis, RD, that the definite description 'the F' not only has a referential use but also a referential meaning (1981, 2004, 2007a,b). Consider a referentially used token of this description. The core of its meaning is provided by its causal-perceptual link to the object that the speaker has in mind. But, I argue, the nominal 'F' also contributes to the meaning. How does it contribute?

I have always favored the view that 'F' plays a role in determining the reference of the referential 'the/an F', as also in determining the reference of the demonstrative 'that F'. So, on my view of singular thoughts...'the/an/that F' would designate an object that 'F' applies to and that 'the/an/that F' is causally grounded in by perception. Other possibilities suggested for 'that F' take 'F' to contribute independently of 'that'. 'That F is G' is treated as equivalent either to'That is F and G' '[The x: x = that and x is an F](x is G)'. The same possibilities are available for the description 'the/an F' insofar as we can treat it as implicitly containing something like the simple demonstrative 'that'. I shall remain neutral between these possibilities. (2004, p. 292)

My neutrality goes a bit further: I resist forcing an account of the referential 'the F' into the theoretical framework of "direct reference" (pp. 292-293). Both Díaz Legaspe and Skerk show a fondness for this framework but I am dubious of it (1996, pp. 179-186, 240-244). Still, whatever account we come up with, a referential token of 'the F is G'will be literally true only if the object in mind is an F.
Díaz Legaspe and Skerk accept RD but object to my handling of the nominal 'F'. In particular, they think that my view has unacceptable consequences in cases of misdescriptions, cases where the object in mind is not 'F'. Caso rejects RD altogether.

2. Justina Díaz Legaspe

Díaz Legaspe rightly draws attention to the conflicting intuitions that we have when confronted with a case of misdescription. Consider her example. She says, with a particular individual in mind,

(1) The man in the corner drinking a martini looks like my uncle

The man in question does indeed look like her uncle but he is actually drinking water. So, is her utterance true or false? We seem drawn to both alternatives. On my view, of course, (1) is literally false because the man she has in mind is not drinking a martini: the convention in using a description to express a thought about a particular object in mind is to use one with a nominal that applies to the object. So that is how I explain the intuition of falsity. But how about the intuition of truth? Díaz Legaspe succeeds in conveying a true message with (1), despite its literal falsity, because she has the water-drinking man in mind in virtue of a causalperceptual link and her audience picks up clues to this link in the usual way for a referential description (or demonstrative, for that matter):

A speaker expressing a singular thought about a certain object participates in the referential convention and thus exploits the causalperceptual link to that object; a hearer participates in the referential convention and thus takes account of clues to what has been thus exploited. (2007a, p. 22; see also 2007b, pp. 51-52)

We might say that the Díaz Legaspe "speaker refers" to the water-drinking man and her audience detects this.

Díaz Legaspe objects to my thus resorting to pragmatics to explain the apparent truth of (1). She notes that in defending RD I reject pragmatic explanations of referential uses. But I, like everyone else, think that pragmatic explanations of linguistic phenomena are sometimes right; see, for example, my discussion (2004, pp. 282-283) of Stephen Neale's example of Jones' saying that "everyone taking my seminar turned up"meaning thereby that only Smith turned up. There can be no sweeping dismissal of pragmatic explanations: each case must be judged on its merits. Clearly I think that the merits of my pragmatic explanation of (1) are good.
Díaz Legaspe proposes a different explanation of the intuitions. The basis of it is her view that speaker S's referential utterance of 'the F is G' with a in mind is literally true if and only if S believes that Fa and it is the case that Ga. She is led to this view by her focus on what is required for successful communication:

in a situation where both speaker and audience believe that an object is F, although it is not, the sentence 'The F is G' will successfully communicate a thought about that particular object...there would be no need for the intended object to fall under F, but instead it would be mandatory that the speaker or the audience believes that it does".

Although the explanation of successful communication typically appeals to conventional meanings as well as pragmatic elements, it is a very different matter from the explanation of those conventional meanings (see section 4 below for more on this). And, as we have seen, successful communication in these cases of misdescription can be easily explained without resorting to anything so curious as the idea that the literal truth of S's utterance about a depends in part on what S believes about a.

3. Laura Skerk

In uttering (1) Díaz Legaspe misdescribes the object she has in mind because she is wrong about it. Skerk offers a nice example where the speaker deliberately misdescribes the object. A teacher says,

(2) The star that is first seen in the evening is actually a planet and it's called 'Venus',

intending thereby to correct the mistaken belief of her students that the object seen in the evening is a star. On my view, the teacher is following what is in fact a common communicative practice, conveying a true message by deliberately saying what is false. The explanation of the teacher's success with (2) is mostly similar to that of Díaz Legaspe's success with (1): in uttering (2) the teacher is exploiting her causalperceptual link to that object seen in the evening and her students pick up clues to that link in the usual way for a referential description. However, in this case, she provides a further clue to the students by trading on their mistaken belief that the object is a star. And she thereby achieves a nice rhetorical effect.
Skerk responds to (2) with the proposal that the nominal in "a referential description can be regarded as secondary or instrumental...just a guide to the identification of the object" and not part of the meaning. But this view that the nominal is of only pragmatic significance has been subjected to powerful objections. I illustrate these as follows:

the following argument forms, which are obviously valid when the description is attributive, seem so also when it is referential: 'All Fs are G; so, if the/an F exists it is G'; 'The/an F is G; so, some F is G'; 'The/an F is G; so, something is F and G'. And statements of the following form seem contradictory: 'The/an F is not F'. It is hard to see how this could be so if 'F' were not making a semantic contribution to the referential 'the/an F'. (2004, p. 291)

Skerk does not respond to such objections.1

4. Ramiro Caso

A consequence of RD is that definite descriptions are ambiguous, sometimes being referential, sometimes attributive. Influenced by recent, very stimulating, pragmatics literature, Caso proposes instead "a univocality account" "that posits a linguistically encoded meaning for descriptions that is neither referential nor attributive". This meaning has a "conceptual" part consisting of the implication of uniqueness together with the meaning of the nominal. The meaning also has a "procedural" part: "look for an object x such that jx and satisfies the expected relevance". Caso gives more details of this procedural meaning:

Interpret d referentially; for every i in F = {x : x Dc (fx)}, taken according to a preference order P on F, construct the i-dependent proposition pi that corresponds to (Sd)(i/d); if optimal relevance is achieved, then retrieve pi and stop; else, interpret d attributively; construct the corresponding object-independent proposition q; if optimal relevance is achieved, then retrieve q and stop; else stop

I have no problem with Caso's conceptual meaning but I have three with his procedural meaning.

A. This proposal is reminiscent of one by François Recanati (1989) that was subjected to a seemingly devastating criticism by Neale (1990, pp. 110-112). Caso notes the criticism but thinks that it does not count against his proposal. It is hard to see why it doesn't. As Neale points out, according to Recanati, 'the F is G' can be used to express "two utterly different types of proposition", sometimes [used referentially] an object-dependent one and sometimes [used attributively] an object-independent one. Ditto, according to Caso. Neale goes on: Recanati's claim that 'the F' is

unambiguous within his theory is close to being on a par with the claim that the noun 'bank' is unambiguous: on some occasions it is used for ground alongside a river and on others it is used for a type of financial institution. (p. 112, n. 36)

This criticism is just as devastating against Caso's proposal.

B. According to Caso, the meaning of a definite description instructs the hearer to first seek a referential interpretation and only if this fails seek an attributive one. It is not plausible that hearers do or should always follow this order of procedure. Wherever it is unlikely that the speaker has any particular object in mind, the hearer surely immediately and rightly settles on an attributive reading without considering any possible referential readings. Consider the following, for example: "The winner of tomorrow's lottery will probably squander the proceeds"; "The worst hurricane this year will strike Florida". Caso defends his position by claiming that "obviously, one has to go first; otherwise, it wouldn't be that much of a procedure." But, of course, the idea that "one has to go first" arises only from his pragmatics-inspired commitment to a definite description having a procedural meaning at all.

C. This brings me to my deepest disagreement with Caso. It is with his commitment, influenced by Relevance Theory (Sperber and Wilson 1995), to a fairly extreme form of pragmatism. This is not the place to argue the vexed pragmatics-versus-semantics issue. I shall have to content myself with a few brief bold statements.

(i) Caso rightly thinks of definite descriptions "as providing clues that help in the identification of referents". How do they provide clues. My simple explanation is that they do so because they have referents. As noted in section 1, the convention for referential descriptions is to refer in virtue of a certain sort of causal-perceptual link to objects. The hearer of a particular description participates in this convention and so identifies the referent by looking for clues to the causal-perceptual link exploited by the speaker. That is the interpretive procedure the hearer follows. So, it is the proper subject of study by pragmatics, taken as the theory of communication. But that procedure is not the proper subject of semantics because the procedure is not part of the meaning. Rather, the procedure is pragmatically appropriate given that the meaning is as just described. In general, the theory of linguistic communication must start from some semantic assumptions about the language. Relevance Theory is an impressive theory of communication but is largely irrelevant to semantics.
(ii) The pragmatists' focus on communication leads them, and leads Caso, to what he aptly describes as "a very audience-oriented stance" toward meaning. This frequently results in the fundamental mistake of thinking that meanings are somehow constituted by the way hearers interpret utterances.2 Consider the "meaning-properties" of an utterance in as broad a sense as you like, covering not only what is said but also what is meant or implicated. What constitutes one of those properties is one thing, how the hearer discovers the property, another. The property is constituted by what the speaker does, by the conventions she participates in, the objects she has in mind, or messages she intends to convey. None of these properties are constituted in any way at all by what the hearer does in trying to interpret what is said or meant. Paul Grice made very clear that something like his "Cooperative Principle" must play a role in the hearer's decision about what the speaker implicated but did not say. Later pragmatists have demonstrated that something like that principle -perhaps the "Principle of Relevance" (Sperber and Wilson 1995)- must play a role also in the hearer's interpretive decision about what is said. Some such principle will guide her in figuring out what conventions the speaker is using (including what language or dialect the speaker is using),3 what objects the speaker has in mind, and so on. The processes that the hearer uses to interpret an utterance may indeed provide evidence about a meaning-property but they do not constitute it. The hearer might do everything right, acting in accord with all appropriate communicative principles, and still get the wrong interpretation. Caso shows some sensitivity to the danger of making the fundamental mistake but I think he still does make it.


1 She finds support for her view in the fact that, in extensional contexts, replacing a definite description with a coreferential term in a sentence yields some sort of equivalence. But this equivalence is in truth values not meaning whilst our concern with the description is with its meaning. So no support for her view is to be found there.

2 Stephen Levinson (2000, pp. 186-187) provides a striking example. His mistake is facilitated by equivocation over 'determine', taking it to refer both to what constitutes what is said and to how the hearer tells what is said.

3 Thus, to interpret 'shallot' one needs to determine whether a person is speaking English-English, Australian-English, or American-English.


1. Devitt, M. (1981), "Donnellan's Distinction", in French, P. A., Uehling Jr., T. E. and Wettstein, H. K. (eds.), Midwest Studies in Philosophy,Volume VI: The Foundations of Analytic Philosophy, Minneapolis,   University of Minnesota Press, pp. 511-524.        [ Links ]

2. Devitt, M. (1996), Coming to Our Senses, Cambridge, Cambridge University Press.        [ Links ]

3. Devitt, M. (2004), "The Case for Referential Descriptions", in Reimer, M. And Bezuidenhout, A. (eds.), Descriptions and Beyond, Oxford, Clarendon Press, pp. 280-305.        [ Links ]

4. Devitt, M. (2007a), "Referential Descriptions and Conversational Implicatures", European Journal of Analytic Philosophy, 3, pp. 7-32.        [ Links ]

5. Devitt, M. (2007b), "Referential Descriptions: A Note on Bach", European Journal of Analytic Philosophy, 3, pp. 49-53.        [ Links ]

6. Grice, H. P. (1989), Studies in the Way of Words, Cambridge, MA, Harvard University Press.        [ Links ]

7. Levinson, S. (2000), Presumptive Meaning: The Theory of Generalized Conversational Implicature, Cambridge, MA, The MIT Press.        [ Links ]

8. Neale, S. (1990), Descriptions, Cambridge, MA, The MIT Press.        [ Links ]

9. Recanati, F. (1989), "Referential/Attributive: A Contextualist Proposal", Philosophical Studies, 56, pp. 217-249.        [ Links ]

10. Sperber, D. and Wilson, D. (1995), Relevance: Communication and Cognition, 2nd edn. (1st edn. 1986), Oxford, Blackwell Publishing.         [ Links ]