By Daniel A. Kaufman
On tap this week is likely the single most influential paper in the philosophy of language, Gottlob Frege’s “On Sense and Reference” (Über Sinn und Bedeutung), originally published in 1892. So many and wide-ranging are the paper’s implications that philosophers are still talking about it today. While the tide certainly has turned against the Fregean account of meaning, there remain prominent contemporary Fregeans and neo-Fregeans – the late Jerrold J. Katz being, perhaps, the most notable example – and one would be hard pressed to find anyone working in the philosophy of language, who has not spent a good amount of time thinking about this particular paper.
“On Sense and Reference” does a number of things, but the two most significant include: (a) rejecting the idea that meanings are psychological objects, in favor of the notion that they are publicly graspable, abstract objects; and (b) rejecting the idea that the meaning of an expression is exhausted by its referent, the latter view which was articulated by John Stuart Mill, in A System of Logic (1843).
Frege’s innovation was to introduce a new semantic value, “sense” (sinn), which he characterized as a “mode of presentation” – by which he meant, essentially, a description – that uniquely picks out an expression’s referent. Thus, the sense of ‘Superman’ might be “the guy from the planet Krypton, in the blue and red uniform, who can leap tall buildings in a single bound” or “the superhero who had a torrid love affair with Lois Lane” or some other such thing. (Bertrand Russell similarly maintained that a name is essentially a disguised definite description, and this Frege/Russell “Description Theory” of names was dominant in analytic philosophy, until the publication of Saul Kripke’s 1970 Princeton lectures on the topic, under the title Naming and Necessity, in 1980.) The sense of a sentence is a proposition, which consists of an ordered n-tuple, made up of the senses of the component parts of the sentence. Thus, the sense of “Superman flies” is an ordered pair consisting of <SenseSuperman, Senseflies>. Somewhat obscurely, the referent of a sentence, on Frege’s view, is its truth-value, which means that every true sentence has the same referent (as does every false one).
As far as ontological commitments go, senses were a very good buy, in that they could be deployed in solving a number of significant logical and semantic puzzles. Frege, himself, was concerned with two in particular.
While the terms ‘morning star’ and ‘evening star’ refer to the same object (Venus), they clearly have what Frege calls different “cognitive values.” One need know nothing substantial, in order to know that the morning star is the morning star, but to discover that the morning star is the evening star might be a significant empirical discovery, and while it certainly is a necessary truth that the evening star is the evening star, one can easily imagine a possible world, in which the evening star is not the morning star.
These obvious facts are quite mysterious – perhaps, even inexplicable – if the meaning of a term consists in nothing but its referent. With the introduction of sense, however, the mystery disappears and these facts become easy to explain.
Ascriptions of Belief and other Propositional Attitudes
A bedrock principle of logic — one that derives from Leibniz’s law of the indiscernibility of identicals — is the substitutability of co-referential terms, salva veritate; that is without a change of truth value. Thus, if the sentence ‘Superman flies’ is true, then the sentence ‘Clark Kent flies’ must also be true, insofar as Superman is identical with Clark Kent.
And yet, there are linguistic contexts in which this bedrock principle appears to break. Consider, for example, the following sentence:
(i) Lois Lane believes that Superman flies.
Ms. Lane, however, does not know that Superman and Clark Kent are the same person, so the following would be false:
(ii) Lois Lane believes that Clark Kent flies.
Philosophers are quite loathe to think of bedrock principles of logic as being breakable, and certainly, if another explanation can be found, it is preferable. Senses provided Frege with the answer. In an “opaque” linguistic context – like the one created by ‘believes that’ – terms do not have their customary reference, according to Frege, but rather their indirect reference, which just is their customary sense. The referent of ‘Superman’ in (i), then, is “the guy from the planet Krypton, in the blue and red uniform, who can leap tall buildings in a single bound,” and the referent of “Clark Kent” in (ii) is something like “the skinny geek, with glasses, who works at the Daily Planet.” In the opaque context, then, these two terms – ‘Superman’ and ‘Clark Kent’ – do not refer to the same object, and this is what explains their non-substitutability. Bedrock logical principle saved.
Senses proved that they could do a lot more work than this, however, and were deployed in any number of directions. They provided semantic content for non-referring names – on a purely referential account of meaning, how can ‘Santa Claus’ mean anything, given that there is no referent? – and they made it possible to individuate meanings in a more fine-grained fashion, such that one could distinguish synonymy from mere co-referentiality – ‘creature with a heart’ and ‘creature with a kidney’ are co-referential expressions, in that they refer to precisely the same set of creatures, but it would seem crazy to suggest that they are synonymous.
It would take the better part of a century for philosophers to work through all the implications of the ideas expressed in “On Sense and Reference” and just as much time to begin finding the problems with it. And it would take the best of the best in the analytic tradition to really challenge it – the Wittgensteins, Quines, Kripkes, and Putnams of the world.
In short, a ground-breaking paper, in every possible “sense” of the term.