Logblog: Richard Zach's Logic Blog
(This is a guest post by my linguistics colleague Darin Flynn)
I was intrigued by your last post—that Carnap (apparently) gave serious consideration to suggestions by Gödel and Behmann that he use “semantics” rather than “syntax” in the title of his 1934 book. The story we’re told in linguistics is that Carnap learned to love semantics (e.g. 1939, 1942, 1947/1956) only after logicians (notably Tarski) showed it to be competitive with logical syntax, but his goal in the early 30s was precisely to develop a formal theory of language (though not of natural language) based on syntax rather than semantics—not just because the latter had a less reputable scientific status at the time, but also because Carnap considered only the first part of the familiar linguistic form/meaning dichotomy to be properly formal:
“A theory, a rule, a definition, or the like is to be called formal when no reference is made in it either to the meaning of the symbols (for example, the words) or to the sense of the expressions (e.g. the sentences), but simply and solely to the kinds and order of the symbols from which the expressions are constructed.” (p. 1)
Some of America’s most important theoretical linguists—notably Leonard Bloomfield, Zellig Harris and Noam Chomsky—believed linguistics to be about form rather than meaning, so were impressed by Carnap’s semantics-free theorizing, and by his first example, “Pirots karulize elatically” (p. 2). The latter demonstrated that a sentence could be analyzed as well-formed phonologically, morphologically and syntactically in the absence of meaning. Carnap also used this pseudo-sentence to illustrate “logical syntax” (e.g. deduction): “A is a Pirot” ∴ “A karulizes elatically”. Crucially, “neither the meaning of the words nor the sense of these three sentences need be known” (p. 3).
Chomsky was suspicious of Carnap’s “syntactically” defined logical deductions (semantics in sheep’s clothing?), but was otherwise (like his former mentor, Harris) a careful and appreciative reader of Carnap, and of Goodman’s and Quine’s takes on Carnap. As Heitner (2005) describes in “An odd couple: Chomsky and Quine on reducing the phoneme”:
“Chomsky’s early philosophical interaction with Nelson Goodman (1906–1998) and Morton White (b. 1917)—two prominent figures within analytic philosophy of language deeply suspicious of an uncritical reliance on “meaning” in philosophy—with whom he took graduate classes with as an undergraduate is no doubt also relevant. After all, it was Goodman who recommended Chomsky for a Junior Fellowship in the Society of Fellows at Harvard where Chomsky would come into close contact with an elite Harvard philosophical circle revolving around Goodman and Quine... Chomsky recalls how “studying at Penn with Zellig Harris and Nelson Goodman was a highly stimulating experience”; a “remarkable opportunity” where he spent “a good deal of time in courses, seminars, discussions, primarily with philosophers at Harvard—Quine, Austin (who was visiting Harvard then), White, and others... a very lively and stimulating period in the Cambridge area for a student with my particular interests” (p. 17). … [I]n addition to these interpersonal relations, others have also detected a significant line of intellectual descent connecting Chomsky to Quine through the influence of Goodman, and through Goodman and Quine, to the philosophical heritage of Rudolph [sic] Carnap (1891–1970)—one of the leading positivist figures of the Vienna Circle of Logical Empiricism. For while Otero (1994, vol. II) reports that among non-American philosophers, it was only Rudolf Carnap whom Chomsky read as a student (p. 3), Tomalin (2002, 2003) has more recently provided valuable research documenting the distinctive Carnapian aspirations and “constructive nominalist” methodology adopted throughout Chomsky’s early formulations of generative grammar as reflected in The Logical Structure of Linguistic Theory (1955/1975) and his first published paper “Systems of Syntactic Analysis” (1953). According to Tomalin, Chomsky’s early commitment to constructing formal systems of syntactic analysis independent of semantic information is traceable to the empiricist methodology of Carnap and the meta-mathematical efforts of Goodman and Quine to devise a formal system for mathematics without appeal to abstract objects (2003, p. 1236). (See Newmeyer (1996, p. 15) for similar considerations.) In fact, in “Logical Syntax and Semantics, Their linguistic relevance” Chomsky (1955a, p. 36) favorably references Quine’s (1953) extremely influential critiques of meaning, and even explicitly defends Quine’s anti-mentalistic philosophy in The Logical Structure of Linguistic Theory.” (p. 2-3)
For instance, compare Carnap’s quotation above with Chomsky’s in The Logical Structure of Linguistic Theory (the title of which may plausibly be construed as a mashup of Carnap’s The Logical Structure of the World & The Logical Syntax of Language): “In the strict sense of the word, an argument, a characterization, a theory, etc. is ‘formal’ if it deals with form as opposed to meaning, that is, if it deals solely with the shape and arrangement of symbols” (1955/1975, p. 83). Compare, too, Chomsky’s “Colorless green ideas sleep furiously” with Carnap’s pseudo-sentence. According to Tomalin (2006), these are examples among many of Carnap’s influence on Chomsky and on his former mentor, Harris. Another notable example: “the transformations developed by Harris and Chomsky were related to the transformation rules presented in (the English translation of) Carnap LST” (p. 168).
Tomalin’s research on the philosophical and historical background of transformational generative grammar is fascinating—particularly his original attention to Carnap, and Goodman’s modifications. For critical reviews, see especially Scholz & Pullum (2007) “Tracking the origins of transformational generative grammar” and Seuren (2009) “Concerning the roots of transformational generative grammar” [sorry, this one's paywalled :( -RZ].