Names and Descriptions

Names and Descriptions

It has been commonly supposed that names and definite descriptions have an affinity, a connection. Names of people and places, in particular, are associated with widely known attributes: for example, the name “Ringo Starr” is associated with the description “the drummer for the Beatles” and “London” is associated with “the capital of England”. This association has led to the Description Theory of Names—the theory that names meandescriptions, express them, are synonymous with them. And that has been welcomed as solving Frege’s problem of the informativeness of identity statements. We thus have a neat logical package: from the name-description link, to the sense of a name, and hence to the solution of a puzzle. What could be wrong with that? However, it has also been felt that a name is not equivalent to a description for a quiver of reasons: the statement “Ringo Starr was the drummer for the Beatles” is not analytic; speakers don’t always make the association; names are rigid designators but descriptions in general are not; it is not an a priori truth that Ringo drummed for the Beatles; names don’t accept scope distinctions but descriptions do. These objections are powerful—they seem to refute the “famous deeds” description theory and hence the proposed solution to Frege’s puzzle. Maybe names are actually just “directly referential” and have their meaning fixed by their bearer, in which case we will be faced by the informativeness problem again. This is pretty much where we stand today on the marriage between names and descriptions: they can’t live together but life apart is a problem. What is a good marriage counsellor to do? How can she engineer a rapprochement?

She might begin with an anodyne observation: just because we can’t analyze names by descriptions doesn’t mean they are not naturally connected; they may not be legally married but they are certainly intimately linked. It isn’t just that descriptions apply to things named as a matter of fact; names connote descriptions—descriptions are suggested by names, habitually tied to them. Not in the philosopher’s sense of “connote”, which requires synonymy, but in the vernacular sense, i.e., that there is a common and predictable association.[1] This may be individual or communal: the descriptive connotation may be peculiar to an individual speaker, depending on his knowledge of the name’s bearer, or it may be widely shared by speakers (as in “famous deeds”). The name reminds the hearer of something, sets up a mental link, taps into memory. It isn’t that the name means the connoted attribute; it merely evokes it in people’s minds. In particular, it doesn’t generate a priori analytic truths about the bearer of the name; the truths in question are a posteriori and synthetic. Still, they are commonly recognized, reliably correlated. This is quite compatible with accepting either of two theories of the strict meaning of names: the direct reference theory (the sense is the object named), or some other description theory of the name’s meaning (not famous deeds or generally known empirical facts). A leading contender for the latter type of theory is the metalinguistic theory: a name is self-referential, quoting itself, as in “The name ‘Ringo Starr’ means “the person named ‘Ringo Starr’”. For the sake of argument, I will endorse this theory here (it is actually a pretty decent theory): it tells us exactly what the name means in and of itself. In other terminology, it denotes the sense of the metalinguistic description. Accordingly, it is analytic and a priori that Ringo Starr is the person named “Ringo Starr”; and everyone who is familiar with the name is aware of this. Thus, we can say that the bearer of the name is the man Ringo Starr, the denotation (meaning, sense) of the name is given by “the person named ‘Ringo Starr’”, and the connotation of the name (for most people) is given by “the drummer for the Beatles”. That sounds like an attractive package, respecting the facts, not overstating things. We could call the connotation the “cultural meaning” of the name and not step on anyone’s toes, with the metalinguistic synonym as its “semantic meaning”. We could even call the bearer of the name its “referential meaning” and put it in the same box as cultural meaning, i.e., not part of strict and literal meaning but an aspect of the overall significance of the name. The name “connotes” that object as it connotes the property of being the Beatle’s drummer. Meanwhile the metalinguistic description tells us what the name means at its semantic heart.

What then of informativeness? We have two options: adopt the metalinguistic paraphrase for this purpose, or appeal to the descriptive connotations. The former strategy is familiar and I will not expatiate on it; the latter is fresher and more intriguing. The idea is that when we learn a true identity statement our knowledge is augmented by the connotations that come with the name: for example, “Hesperus” connotes “the evening star” and “Phosphorus” connotes “the morning star”, so that when we learn that Hesperus is identical to Phosphorus, we learn that the evening star and the morning star are one and the same planet. This is not the same proposition as that expressed by “Hesperus is Phosphorus” (that is given by the metalinguistic paraphrase), but is a substantive piece of knowledge associated with the identity statement. It isn’t a tautology but a real cognitive step forward.[2] It might easily be mistaken for the knowledge conveyed by the original statement, but actually it corresponds to a different statement—the one connoted. It is connotational (cultural) knowledge not denotational (semantic) knowledge. If you want to know what the conveyed denotational knowledge consists in, you will have to look at the denotations of the contained names, viz. the metalinguistic descriptions. That, too, is not tautological (it isn’t analytic that the same planet is called both “Hesperus” and “Phosphorus”). So, actually, two pieces of descriptive knowledge are obtained when the identity statement is discovered to be true—metalinguistic (denotational) and empirical (connotational). Lots of knowledge is conveyed in this composite picture, so we are not short of resources with which to answer Frege’s question. Nor does the resulting theory violate any of the principles employed in the standard objections to the description theory of names; we just have to be more careful about the formulation of that theory. We mustn’t be too simplistic and rigid about how the descriptions do their job: connotation can do it as well as denotation (strict intrinsic meaning). The marriage counsellor has advised a looser relationship as the solution to his clients’ problems, though not one without substantial commitments (live apart but always keep each other in mind).

Not all words have connotations, at least not in any marked manner; names stand out in this respect. Connectives and prepositions don’t do much connoting; nor do articles and pronouns. Verbs do some, but nouns do the most. This includes singular and common nouns (e.g., kind terms, natural and artificial). The reason for this is presumably that these are the things of most interest to us—in particular, people and places (just listen to the Beatles song “In My Life”). We have thoughts and feelings about the bearers of names, and these stick in our minds. Names are connotation-rich. It is not surprising that there are quasi-semantic associations here, interpenetrations. Frege and Russell were right to sense an affinity, though they overdid it to the point of asserting identity. The pragmatics of names is clearly description-directed. So, we should expect that descriptions will play a role in the functioning of names, though an auxiliary role. Names thus point towards descriptions without being equivalent to them. But the pointing goes in the other direction too: descriptions also connote names. If you hear the description “the drummer for the Beatles”, you naturally think of Ringo Starr, his face, his hair, his drumming style—the name strikes you in a certain way. This explains why we tend to regard definite descriptions as name-like, contrary to Russell’s theory. We expect descriptions to function like names because they irresistibly suggest names to us (it isn’t just that they occur in subject position); hence Russell’s theory comes as a surprise. The connotations of descriptions instill the idea that they are name-like—just as the connotations of names instill the idea that names are description-like. The mutual connotation encourages semantic assimilation. We mistake affinity for overlap, even identity. This is the psycholinguistics of names and descriptions, their depth psychology (the “semantic unconscious”). Russell’s theory of descriptions could be entirely correct, and known to be so, and yet descriptions persist in having name-involving connotations; just as Mill’s theory of names could be entirely correct and yet names persist in having descriptive connotations. And connotations count: they shape conceptions.[3]

Connotations need not be accurate; denotations have to be. You can get things wrong about the objects you refer to without detriment to your referential success, but you can’t be wrong in what you mean and expect to lock onto the right object. You might think Ringo was the drummer for the Stones but still refer to Ringo with “Ringo” not Charlie, but you can’t refer to Ringo with “Ringo” if you don’t grasp that Ringo is named “Ringo” and not “Charlie”. Denotation is infallible in this sense, but connotation is not: you can’t grasp the meaning of “tree” and not refer to trees with “tree”, but you can grasp the connotation of the word “tree” and this connotation not be an accurate representation of trees (for you “tree” might connote “plant that poisonous snakes hide in”—because you once saw a documentary about poisonous tree snakes in Africa). Connotations are certainly not entailed by denotations (meanings)—they are just stuff you associate with the word. But they can cooperate with denotations in the use and understanding of language—they come with words (hence con-notation: “con” as in “together with”). We should really pay more attention to the phenomenon of connotation in the philosophy of language.[4]

[1] See my “On Denoting and Connoting” for some background to this paper.

[2] One thing is perfectly clear: affirmations of self-identity can never be informative.

[3] Connotations are not the same as conversational implicatures, though they belong to the same general territory. Neither follows from strict literal meaning; both are broadly pragmatic (or penumbral). Connotations are more like psychological associations, cognitive and affective; implicatures have to do with conversational rules governing the purposes of communication. A comparative study of connotation and implicature would be worth undertaking.

[4] Words have reference, sense, force, tone, and connotation. Each does its job in accounting for language use. Consider all the connotations of “I went for a job interview in New Jersey yesterday”, let alone “The Second World War killed millions of people”. Language is heavy with connotation.

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.