[3.7.4.] Putnam’s Anti-Realism.

 

Hilary Putnam

·         Born in Chicago (1926).

·         Ph.D. in philosophy from UCLA, 1951.

·         student of Hans Reichenbach, one of the more important logical positivists

·         taught at a number of schools, including MIT, before settling at Harvard in 1976; retired in 2000

·         before becoming renowned as a philosopher, Putnam made significant contributions within mathematics.”^[1]

·         in the 1950s and 60s, developed a position within the philosophy of mind called functionalism. (Famously, Putnam later argued against this view—he is well-known as a philosopher who is not afraid to change his mind.)

·         in the 1970s, developed a position within the philosophy of language called semantic externalism

·         was enmeshed in the analytic tradition until the mid 1970s, when he began to develop a theory of truth and reality influenced by the American Pragmatists: internalism

 

In his 1981 book Reason, Truth and History, Putnam extended Quine’s doctrine about reference in order to argue for a form of anti-Realism.

 

 

[3.7.4.1.] The Radical Inscrutability of Reference.

 

Putnam will argue that there are no statements that correspond with the world in the way that Realism requires:

 

... the central picture at work in traditional Realism—that of two sets of objects, one consisting exclusively of linguistic items and the other involving mind-independent items, standing in a determinate referential relation—is incoherent... (273)

 

Putnam’s view is more radical than Quine’s. Even on Quine’s view, reference is not completely indeterminate... “gavagai” when spoken by a speaker of the newly discovered language and “rabbit” when spoken by an English speaker both refer to something in the rabbit family (and not to, e.g. the Eiffel Tower, or Barack Obama, or a host of other things...)

 

But Putnam thinks reference is even more inscrutable than does Quine: “Accounts of the referential force of the term ‘rabbit,’ for example, are not limited to items comprising ... the ‘Rabbit’ family. Far more devious mappings are possible.” (274)

 

 

 

 

[3.7.4.2.] The  Possible World Account of Sentence Meaning.

 

A key idea in Putnam’s anti-Realist argument is that the meaning of a sentence is completely specified by an assignment of a truth value (“true” or “false”) to that sentence for every possible situation—or, as Putnam (and many other philosophers) would say, for every “possible world.”^[2]

 

According to this view...

 

In general, the meaning of “p” is specified by saying...	“p” is true in every possible situation (world) in which p and is false in every possible situation (world) in which it is not the case that p. This is a complete account of the meaning of “p”.
For example, the meaning of “The cat is on the mat” is specified by saying:	“The cat is on the mat” is true in every possible world in which some cat is on some mat, and false in every possible world in which no cat is on any mat. This is a complete account of the meaning of “The cat is on the mat.”

 

But according to Putnam, even knowing the truth value of the sentence “The cat is on the mat” in every possible world will not tell us what the words “cat” and “mat” refer to. The reference of the word “cat” and the reference of the word “mat” are undetermined by the assignment of truth values described above. In other words, even if we know for every possible situation whether “The cat is on the mat” is true in that situation, we still will not know what “cat” refers to and what “mat” refers to.

 

On the assumption that a truth value assignment across all worlds gives us a complete account of a sentence’s meaning, Putnam’s claim that such an assignment does not fix the reference of a sentence’s terms implies that there is no fact of the matter about what a sentence’s terms refer to.

 

In fact, the reference of “cat” and “mat” is so undetermined that an interpretation on which “cat” refer to cherries and “mat” refers to trees is just as good as one on which they refer, respectively, to cats and mats. As Putnam puts it:

 

[The sentence “A cat is on a mat”] can be reinterpreted so that in the actual world ‘cat’ refers to cherries and ‘mat’ refers to trees without affecting the truth-value of [that sentence] in any possible world.^[3]

 

 

 

 

[3.7.4.3.] Cats* and Mats*.

 

Putnam’s argumentative strategy^[4] is to consider these two sentences --

 

(8) A cat is on a mat.

 

(8') A cat* is on a mat*.

 

and then argue the following: even if we know everything there is no know about the meaning of those two sentences, we still will not know exactly what “cat” and “mat” refer to (as they occur in sentence (8)) and still now know what exactly what “cat*” and “mat*” refer to (as they occur in sentence (8')).

 

To define the terms “cat*” and “mat*”, we first need to define three different types of situation:

 

case (a)	case (b)	case (c)
some cat is on some mat and some cherry is on some tree	some cat is on some mat but no cherry is on any tree	no cat is on any mat (whether any cherry is on any tree does not matter)

 

cat* (df.): x is a cat* if and only if either case (a) holds and x is a cherry or case (b) holds and x is a cat or case (c) holds and x is a cherry.

mat* (df.): x is a mat* just in case either case (a) holds and x is a tree or case (b) holds and x is a mat or case (c) holds and x is a quark.

 

So whether or not an object is a cat* or a mat* depends on what situation obtains:

 

case (a)	case (b)	case (c)
some cat is on some mat and some cherry is on some tree   cats* are cherries mats* are trees	some cat is on some mat but no cherry is on any tree   cats* are cats mats* are mats	no cat is on any mat       cats* are cherries mats* are quarks

 

The sentence

 

(8') A cat* is on a mat*.

 

is true in every possible world in which

 

(8) A cat is on a mat.

 

is true, and (8') is false in every possible world in which (8) is false.

 

We can tell this by checking three types of possible world:

 

case (a)	case (b)	case (c)
some cat is on some mat and some cherry is on some tree   cats* are cherries mats* are trees   (8) “A cat is on a mat” T (8') “A cat* is on a mat*” T	some cat is on some mat but no cherry is on any tree   cats* are cats mats* are mats   (8) “A cat is on a mat” T (8') “A cat* is on a mat*” T	no cat is on any mat       cats* are cherries mats* are quarks   (8) “A cat is on a mat” F (8') “A cat* is on a mat*” F

 

Knowing the meaning of (8) (i.e., knowing what truth value (8) takes in every possible situation / world) will not tell us whether “cat” and “mat,” as they occur in that sentence, refer to cats and mats, or whether they refer to cats* and mats*. We can interpret “cat” and “mat” as referring, not to cats and mats, but to cats* and mats*, without changing the truth value “The cat is on the mat” has in any possible world, and therefore without changing its meaning.

 

Putnam’s conclusion: “...if nothing prevents us from taking (8) to be about cats* and mats*, then since the actual world is an A-type world and in an A-type world, cats* are cherries and mats* are trees, nothing prevents us from taking (8) to be the claim that some cherry is on some tree.” (276)

 

And there’s nothing special about cats and mats and cherries and trees in this example. We could run this same argument with different terms to show that “The cat is on the mat” (or any other statement) can be interpreted to be about anything at all.

 

 

[3.7.4.4.] Away with Correspondence.

 

The lesson that Putnam draws from all this is that the sort of correspondence required by Realism Claim #2 (the correspondence theory of truth) simply does not obtain. There are no mind-independent relations correlating words and objects (“cat” and cats, for example), and so the account of truth and meaning assumed by Realism must be inaccurate.

 

... Realism ... takes the truth of our assertoric or statemental sentences to consist in their matching, fitting, mirroring, or corresponding to the states of affairs obtaining in a mind-independent world; but they can succeed in this only if there are determinate referential relations that provide one-to-one correlations between the words making up those sentences and the objects entering into the states of affairs the sentences are supposed to match. But now we find that it is not merely indeterminate whether a term like ‘cat’ picks out three-dimensional enduring substances or their temporal parts; it is not even determinate whether what the term picks out is an animal or a piece of fruit. (276-77)

 

 

Stopping point for Wednesday March 25. For next time, finish reading the chapter from Loux (pp.279-90).