Philosophy 4331, Lecture 5: Russell on Descriptions

David Boylan

What are Descriptions?

Names are not the only expressions that appear to refer to objects. Descriptions
do too:

 Consider:

(1) I met a dog in the park. “A dog" is an indefinite description.

(2) The President of the US lives in the White House. “The President" is a definite description.

 Both of these seem to talk about particular objects, just like names do!

∗ You might think that “a dog" refers to some particular animal. If you read “On Sense and Reference" closely
you’ll see this was basically Frege’s view.
And that “The President of the US" refers to the President.

 But the different with descriptions is that they are obviously complex ex-
pressions that take other meaningful expressions as parts.

In “On Denoting" Russell argues against this natural view. He instead claims
that definite descriptions are a kind of quantifier.

What are Quantifiers?

 Consider:

(3) Two students got As.

(4) Every dog has an owner.

 Neither of “two students" or “every dog" are names, even though they appear Subject position is where the subject of the
in subject position. sentence goes.

 Rather they say something about a class of objects, that some proportion of
the members of the class have a particular property.
For this reason, we call them quantifiers.

Initially, the view that descriptions refer is probably the most appealing.
Why is this, do you think?
 2

Problems for a Theory of Descriptions

Russell’s argument for his theory of descriptions is by best explanation: he

thinks his theory gives the best explanation of some puzzles about descriptions.
Puzzle 1. Non-existence:

 There is no King of France. So, if descriptions refer, then “the present King
of France" doesn’t refer to anything.

 But if the meaning of a description is just what it refers to, what is the mean- Just like we did when we talked about empty
ing of: names, we can give a more thorough compo-
sitionality argument here.

(5) The present King of France is bald.

Wouldn’t this just be meaningless? Exercise for you: spell this argument out in
more detail.
 Russell thinks that it is actually false. Do you agree?

Puzzle 2. Informativity:

 As we saw, sentences like the following are known a priori:

(6) JK Rowling is JK Rowling.

 If descriptions are semantically just like names, then the following should
mean exactly the same thing:

(7) JK Rowling is the author of the Harry Potter books.

After all, since she is the author, that’s what “the author of the Harry Potter
books" should refer to.

 But this looks false! We know (6) a priori; but (7) could be a legitimate
discovery.

 Another way to put it: the first can be true while the second is false:

(8) Alice knows that JK Rowling is JK Rowling.

(9) Alice knows that JK Rowling is the author of the Harry Potter books.

Puzzle 3. Logic:

 You might think that, for any sentence p, it’s a logical truth that:

LEM. Either p or not p.

 3

In logic, this is called the law of excluded middle.

 But again, return to the King of France. LEM means one of these must be
true:

(10) The King of France is bald.

(11) The King of France is not bald.

How could either of these be true if the King of France is like a name that
doesn’t refer?

What do you think Frege would say to these arguments?

(Remember, some of Frege’s own arguments were very similar!)

Russell’s View

Let’s think a bit more about quantifiers and what they mean:

 It’s hard to say what a quantifier means by saying what it stands for.
(As we saw, quantifiers don’t really appear to refer to anything.)

 But we can get a better sense of what they mean by paying closer attention
to the contribution they make to sentences as a whole:

∗ “Some Fs are G" in general says that at least one thing exists which is F
and which is also G.
∗ “Two Fs are G" in general says that at least two things exist which are F
and which are also G.
∗ “Most Fs are G" in general says that at least half of the things which are
F are also G.

With that in mind, how might descriptions work as quantifiers?

 What in general does this say?

(12) The F is G.

 Russell’s answer:

∗ “The F is G" says that there is exactly one thing that is F and that thing is
also G.

Why is this the same as Russell’s formulation of the view?

 4

 According to Russell, “The F is G" says that there is some object o which is
F is also G and that anything which is F is identical to o.

 Why does this require that there be exactly one thing which is F?

∗ Consider:

(13) The car outside is red.

According to Russell, this says that:

1. there is some object o which is a car outside;
2. o is red;
3. anything which is a car outside is identical to o.
∗ Suppose there are two cars outside, o and o 0 .
∗ Then condition three is false: o 0 is something which is a car outside but
is not identical to o.

Russell makes the odd claim in a number of places that descriptions are
(in some sense meaningless):

 Relevant passage:

This is the principle of the theory of denoting I wish to advocate: denot-

ing phrases never have any meaning in themselves, but that every propo-
sition in whose verbal expression they occur has a meaning.

Why do you think he says this?

Do you think it’s essential to his argument here?

Scope

Before solving the puzzles, Russell introduces an important idea, the notion of
primary/secondary occurrence.

 Consider:

(14) Alice knows some student at TTU is Tom Cruise’s daughter.

 This can mean at least two things;

∗ Alice just knows there is some such student; but she has no particular
person in mind.
∗ Alice has a particular person in mind who she knows is Tom Cruise’s
daughter; and that person also in actuality happens to be a TTU student.
(Alice may not herself know that this person is a TTU student.)
 5

The second way of understanding the sentence could be expressed this way:

(15) There is some TTU student who Alice knows is Tom Cruise’s daugh-
ter.

 In more modern terminology, this gets called a scope ambiguity.

∗ On the first way of understanding the sentence, we say that “some TTU
student" takes narrow scope wrt “Alice knows".
∗ On the second, we say that “some TTU student" takes wide scope wrt
“Alice knows".

Russell notes that descriptions also give rise to scope ambiguities:

 Consider:

(16) John thinks the professor is an idiot.

Imagine consider two different situations:

∗ John reads the first email from the professor; there are numerous typos,
the professor has gotten the lecture times wrong, he’s spelled his own
name wrong, etc.
∗ John sees some guy in the cafeteria do something stupid; he has no idea
that this guy is the professor.

 Like with the quantifier “a thief", descriptions can take narrow-scope or

wide-scope wrt to other expressions like “Alice thinks".

Solving the Puzzles

Straightforward to see how we solve the first puzzle:

 Recall:

(17) The King of France is bald.

 According to Russell, part of what this says is that there is a unique King of More strictly speaking, this is an entailment.
France.

 Since this part is false, the whole sentence is false.

Solution to the second problem: descriptions and names mean different things.

 Recall:

(6) Alice knows JK Rowling is JK Rowling.

 6

(7) Alice knows that JK Rowling is the author of the Harry Potter books.

These say different things.

 Here’s how to spell out the second on Russell’s account:

(18) Alice knows that JK Rowling is the unique person such that they
wrote the Harry Potter books.

This clearly says something quite to (6). For instance, it does not say that It isn’t a logical truth that Rowling wrote the
Alice knows a logical truth. Harry Potter books!

 As Russell points out, you cannot in general replace a description with a

name, even if the description picks out the same thing as that name.

The third puzzle is where scope becomes particularly relevant:

 Recall:

(19) The King of France is not bald.

On Russell’s theory, there are two ways to interpret a sentence like this, de-
pending on whether the description take narrow or wide scope.

 This is the narrow scope meaning:

(20) It is not the case that: there is a object which is the unique King of
France and it is bald.

This is true if there is no King of France.

 This is the wide-scope meaning:

(21) There is an object which is the unique King of France and it is not
bald.

This is false because there is no King of France.

 Either way, there is no need to say that the Law of Excluded Middle fails.

Names and Descriptions

I started by contrasting names and descriptions.

 I said that Russell thinks that, surprisingly, Russell doesn’t think that de-
scriptions have the same kind of meanings as names.

 In fact, this is exactly how we explained the solution to Puzzle 2 as well.

 7

Is there any reason you think to make a distinction here?

Or might there be reasons to think that names work the same way as de-
scriptions?