On Descriptions
Singular terms
Singular terms are expressions which purport to denote or designate particular individual people, places, or other objects.
Four Puzzles about singular terms (Russell):
The Problem of Apparent Reference to Nonexistents
(1) The present King of France is bald.
K1 (1) is meaningful
K2 (1) is a subject-predicate sentence
K3 A meaningful subject-predicate sentence is meaningful only in virtue of its picking out some individual thing and ascribing some property to that thing. (Frege rejected this, by positing abstract he called “senses” and arguing that a singular term is meaningful in virtue of having one of those over and above its referent — or in the case or a nonreferring singular term, instead of a referent.
K4 (1)’s subject term fails to pick out or denote anything that exists.
K5 Thus, either (1) is not meaningful or (1) picks out a thing that does not exist, but:
K6 there is no such thing as a “nonexistent thing.”
The Problem of Negative Existentials
(2) The present King of France does not exist.
(2) seems to be true and seems to be about the present King of France. But if (2) is true, (2) cannot be about the present King of France, for there is no such King for it to be about. Likewise, if (2) is about the present King of France, then it is false, for the King must then in some sense exist.
Frege’s Puzzle about Identity
(3) Elizabeth Windsor = the present Queen of England
(3) contains two singular terms, both of which (if the statement is true) pick out or denote the same person or thing. If so, then the statement is trivial. But it is not. Frege, again, held that the two singular terms have different “senses,” that is why (3) is informative/non-trivial.
The Problem of Substitutivity
(4) Albert believes that the author of Nothing and Beingness is a profound thinker.
Basically same problem as Frege described in “On Sense and Nominatum,” such as in the sentence “S believes Superman is strong,” we cannot substitute “Superman” with “Clark” without changing the truth-value of the sentence (if S does not know Clark is Superman) though, in reality, Superman is Clark.
Russell’s Theory of Descriptions
(5) The author of Waverley was Scotch
“The” abbreviates a more complex construction involving what logicians and linguists call quantifiers, words that quantify general terms (KR: quantifiers QUANTIFIES general terms; they do not denote/refer any specific things). Russell thinks that (5) as a whole abbreviates a conjunction of three quantified general statements, none of which makes reference to Scott in particular:
(a) At least one person authored Waverly, and
(b) at most one person authored Waverly, and
(c) whoever authored Waverly was Scotch.
Russell maintained that the apparent singular term in (5), “The author of Waverley,” is not really a singular term at all at the level of logical form, but a convenient abbreviation of the more complicated quantificational structure displayed in (a)-(c). As he puts it, the apparent singular term “disappears on analysis.” Our puzzles have arisen in fact from applying principles about singular reference to expressions that are not really singular terms (hence no nominatum existed) at all but only masquerade as such.
Russell’s solution on the paradoxes
Apparent Reference to Nonexistents
Paraphrase (1) according to Russell’s method:
At least one person is presently King of France, and
at most one person is presently King of France, and
whoever is presently King of France is bald.
Russell therefore denies K2, since he denies that “The present King of France” is “really” a singular term. Hence this is not a subject-predicate sentence.
Negative Existentials
Two ways to paraphrase (2)：
One (primary position):
At least one person is presently King of France, and
at most one person is presently King of France, and
whoever is presently King of France does not exist.
This paraphrase does not solve the problem since the first conjunct asserts the existence of a present King, while the third denies it.
Two (secondary position):
Not: (At least one person is presently King of France, and
at most one person is presently King of France, and
whoever is presently King of France exists).
Frege’s Puzzle
Russell’s paraphrase:
At least one person is presently Queen of England, and
at most one person is presently Queen of England, and
whoever is presently Queen of England is one and the same as Elizabeth Windsor
Now we see why the original identity-statement is nontrivial. We learn something knew with this paraphrase. And of course the identity-statement is contingent, since someone else might have been Queen.
Substitutivity
Russell’s paraphrase:
Albert believes the following: (At least one person authored Nothing and Beingness, and at most one person authored Nothing and Beingness, and whoever authored Nothing and Beingness is a profound thinker).
You cannot substitute “the author of Sizzling Veterinarians” into (4) because the corresponding paraphrase would be:
Albert believes the following: (At least one person authored Sizzling Veterinarians, and at most one person authored Sizzling Veterinarians, and whoever authored Sizzling Veterinarians is a profound thinker).
It is not what Albert believes.
The four puzzles made it clear that definite descriptions do not hook into the world by directly naming and nothing else. Notice that even though definite descriptions are not assigned referents in the way that names are, and even though they are not really singular terms at all, they still purport to have singular individuals that answer to them; when a description does in fact have the corresponding individual that it purports to have, those unique so-and-so would be called the description’s semantic denotatum or semantic referent. But the “hook” between a definite description and its semantic referent is far less direct than is the hook between a simple name and its bearer.
Objections to Russell’s theory
Strawson: Strawson thought of referring not as an abstract relation between an expression and a thing but as an act done by a person at a time on an occasion. Hence, Strawson holds that expressions do not refer at all; people refer, using expressions for that purpose.
Objection 1
(1) is not false or true. For Strawson, the speaker has produced an only ostensibly referring expression that has misfired; the speaker has simply failed to refer to anything and so has failed to make a complete statement. In other words, the sentence is not incorrect but abortive; it does not even get a chance to be false. Strawson therefore solves the Problem of Apparent Reference to Nonexistents by denying K3: (1) is meaningful, in that it has a legitimate use in the language and could be used to say true or false things if the world were more cooperative, but not because it succeeds in picking out any individual things.
Objection 2 (misplaced)
Strawson criticizes the claim that part of what a speaker would be asserting in uttering (1) would be that there at present existed one and only one King of France. This claim is implausible for although the speaker presupposes that there is one and only one king, that is not part or what the speaker asserts.
This objection is a misunderstanding, for Russell said nothing at all about acts of asserting. Perhaps Strawson is assuming on Russell’s behalf that whatever is logically implied by a sentence is necessarily asserted by a speaker who utters that sentence.
Objection 3
Strawson points out that many descriptions are context-bound. For example:
(6) The table is covered with books.
With normal paraphrase, (6) would logically implicate that there is at most one table, in the entire universe. Two solutions:
(a) Ellipsis hypotheses: Russell may say that there is ellipsis here, that in the context, “The table” is short for a more elaborate description that is uniquely satisfied. Hence, the sentences really do have the logical forms he posits. However, the question would become “what material is ellipsized?” And the answer matters because (6) will turn out to say completely different thing depending upon which candidate you pick. If we say that “The table” means the table in this room, then we have introduced the concept “room” and construed (6) as being literally about a room, indeed as having the predicate “room” hidden in its underlying logical structure.
(b) Restricted quantification: domains over which quantifiers range need not be universal, but are often particular classes roughly presupposed in the context.
The appeal to (b) differs from (a) in that it does not require that explicit conceptual material (like “room”) be clandestinely mentioned in (6). The quantifier restriction is more like a silent demonstrative pronoun: “At least one table of that sort…,” where the context fixes the reference of “that.”
Objection by Keith Donnellan
Donnellan called attention to what he called the “referential use,” as opposed to the “attributive use,” of a definite description, such as “The Holy Roman Empire.” (Basically titles)
(7) Smith’s murderer is insane.
There are two possible uses of “Smith’s murderer”:
Attributive: (7) means that whoever committed this terrible crime is insane. “Smith’s murderer” is hence only a placeholder that the attribute “insane” would attach to.
Referential: suppose Jones has been arrested and charged and being prosecuted and we all presume Jones is guilty; also, he is rolling his eyes and drooling in a homicidal manner. Here, if we say (7), we are using the description “Smith’s murderer” to refer to the person we are looking at, the defendant, regardless of what attributes he has (he may not be the murderer or not insane after all).
A further characterization by Donnellan: in the attributive use of “The A is Y,” if nothing is the A then nothing has been said to be Y, while in the referential case the fact that nothing is the A does not have this consequence.
Speaker-reference: The speaker or utterer’s referent of a description on an occasion of its use is the object, if any, to which the speaker who used the description intended to call to the attention of her audience. Speaker-reference and semantic-reference can be, and usually are, different. Such distinction matters to the truth value of sentences that embed descriptions within clauses of certain kinds. For example:
(8) I know that’s right because I heard it from the town doctor.
(9) I wish her husband weren’t her husband.
In (8), the truth value may depend on whether “the town doctor” is used attributively or referentially. In (9), the truth value may depend on which way the descriptions (husband and…husband) are taken.
There might be a third kind of referent, conceptually distinct from the other two: actual referent. What is meant seems to be that the actual referent is the object about which the speaker actually succeeded in making a statement, it being left open whether this tracks the literal semantic interpretation of the sentence uttered.
Anaphora
In general, an anaphoric expression inherits its meaning from another expression, its antecedent, usually though not always occurring earlier in the sentence or in a previous sentence. For example:
(10) The man who lived around the corner was eccentric. He used to snack on turtle heads.
Geach called such a term a “pronoun of laziness” and suggested that it merely abbreviates a boilerplate repetition of the antecedent phrase. Under such understanding, then (10) poses no problem for the Theory of Description.
However, see
(11) Just one turtle came down the street. It was running as if it were being pursued by a maniac.
(12) A rabbit appeared in our yard after dinner. It seemed unconcerned.
(11)’s second clause is not equivalent to “just one turtle was running as if…” and (12)’s second clause is not equivalent to “a rabbit seemed unconcerned” for that paraphrase misses the fact that the original “it” referred to the particular rabbit that appeared in the yard.
Russell may rejoin that neither (11) nor (12) contains a definite description, but see
(13) Just one turtle came down the street. The turtle was running as if it were being pursued by a maniac.
(14) A rabbit appeared in our yard after dinner. The rabbit seemed unconcerned.
For (13) it is plausible to take “the turtle” as abbreviating “the turtle that came down the street” in which case it does not threaten Russell’s analysis, but this will not do for (14). If we take “the rabbit” abbreviates “the rabbit that appeared in our yard after dinner” then by the usual uniqueness clause, (14) would entail that at most one rabbit appeared in the yard, which (14) itself does not entail.
