Remark on J. L. Speranza's post "English Composition -- and Grice" of Feb 28, 2011

February 28, 2011

Speranza writes the following:

"But," Dale notes, "I don't know whether before Davidson the idea had ever been explicitly suggested that meaning might be explained - or explained away - by stating a relation of some sort between speakers and compositional-semantic theories."

Good point. I wonder about that, too. But to have that 1963 early reference is very good.

I have always been fascinated with the incredibly clear statement of what a compositional semantic theory should do given by Carnap in the Aufbau (and I quote this in my dissertation, too, on p. 57 of the PDF version at russelldale.com and on pp. 81 - 82 of the original version, which can be downloaded free at scribd.com/doc/49598324/The-Theory-of-Meaning-by-Russell-Dale). Here is what Carnap says:

Let us consider the designation relation as it holds between written words and their meanings. Since natural languages do not have general rules which allow us to deduce the meaning of a word from its form, there is no way of indicating the extension of this relation except by enumeration of all its member pairs. If a basic language is already known, then this is done through a dictionary; otherwise, the answer takes on the form, for example, of a botanical garden, that is, a collection of objects, each of which has its name written on it. If the meanings of the words are known, then the answer to the correlation problem of the designation relation for sentences can be solved through a general function, which however, is usually very complicated. It is the syntax of the language in question cast in the form of a meaning rule. A meaning rule may (in an elementary case) have the following form: if a sentence consists of three words, a noun in the nominative case; a verb in the third person singular, present tense, active mood; and a noun in the accusative case, then it designates the state of affairs that the object of which the first word is the sign stands to the object of which the third word is the sign in the relation of which the verb is the sign.
From the correlation problem, we distinguish the essence problem. Here we do not simply ask between what objects the relation obtains, but what it is between the correlated objects, by virtue of which they are connected. The question does not ask for the constitution of the related object, but asks for the essence of the relation itself. Later on,...we shall indicate the difference between science and metaphysics..., and we shall see that the essence problems belong to metaphysics.... [Carnap, Rudolf. (1928). The Logical Structure of the World. Translated by Rolf A. George. Berkeley; Los Angeles: University of California Press, 1967. Pp. 34-35.]

First thing I like to think about here is that this predates by 5 years Tarski's "The Concept of Truth in Formalized Languages" which is often understood as the earliest statement of what a semantic theory would look like. I'm not interested in raising a "Who did it first?" question here. I just find it interesting that few people said things of this sort this early. Did ANYBODY say anything this clearly before? Did Frege or Russell? Can someone point it out to me? I'd be really eager to see this, but I have never found such clarity in either Frege and Russell or anyone else before this time.

Now, what Carnap identifies above as the correlation problem is clearly just the task of providing a formal compositional meaning theory for a natural language, that is, a finitely statable mapping (function) from sentences to propositions, based on meanings assigned to lexical simples (words) and sentence syntax. In 1928, the idea that the syntax of a natural language could have been seriously systematized must have seemed quite a distant dream, perhaps completely unattainable. Indeed, Carnap says in this passage that the correlation he is identifying would be "very complicated" for just that reason. With Chomsky (1957), Syntactic Structures, there certainly suddenly seemed real hope that a syntax for natural language was, ultimately, achievable.

And, Chomsky (1957), of course, paved the way for Katz and Fodor (1963), "Structure of a Semantic Theory". But, Katz and Fodor (1963) lost sight of the fact that they were supposed to give the meaning (or at least the truth conditions) for natural language expressions, and instead ended up giving us a formalism with no interpretation, that is, what came to be called "markerese". Katz and Fodor (1963) mapped natural language sentences onto representations that they called "markers". The provided rules for saying when two expressions meant the same thing (when they mapped onto the same marker) and when they meant different things (when they mapped onto different markers), but they never said how to determine from the markers what the actual meanings of sentences were. That is, the markers were uninterpreted.

Within four years of Katz and Fodor (1963) this basic criticism of Katz and Fodor started to be published and accepted widely. I believe Bruce Vermazen was the first to publish this sort of criticism explicitly in 1967 (Bruce Vermazen. (1967). Review of Katz and Postal's An Integrated Theory of Linguistic Descriptions, and Katz's The Philosophy of Language. Synthse, 17 (1967), 350-365.).

But, I believe that Davidson comments on the Katz and Fodor (1963) approach (if not explicitly) in "Truth and Meaning" (in Synthese, September 1967, pp. 304 - 323; this is the same issue of Synthese in which the Vermazen review above appeared) when he says:

Ask, for example, for the meaning of 'Theaetetus flies'. A Fregean answer might go something like this: given the meaning of 'Theaetetus' as argument, the meaning of 'flies' yields the meaning of 'Theaetetus flies' as value. The vacuity of this answer is obvious. We wanted to know what the meaning of 'Theaetetus flies' is; it is no progress to be told that it is the meaning of 'Theaetetus flies'. This much we knew before any theory was in sight. In the bogus account just given, talk of the structure of the sentence and of the meanings of words was idle, for it played no role in producing the given description of the meaning of the sentence. [Davidson (1967), pp. 30.]

And, Davidson is continuing the commentary on the Katz and Fodor (1963) approach a paragraph later with the often quoted:

Paradoxically, the one thing meanings do not seem to do is oil the wheels of a theory of meaning - at least as long as we require of such a theory that it non-trivially give the meaning of every sentence in the language. [Davidson (1967), p. 307.]

I will leave it there for now. I would love to hear what others have found about the history of these points and matters.

--Russell Dale