5. Quine’s Appeal to Use and the Genealogy of Indeterminacy
The envisioning of language that has long marked the analytic tradition involved, at first, only a relatively vague and inexplicit conception of language’s “use,” “application,” or intersubjective “practice. ” Even this vague and inexplicit conception was, as we have seen, already enough to suggest some of the fundamental ambiguities that arise from placing an appeal to language at the center of the methods of philosophy. But it was left to the second generation of analytic philosophers, those who also played the largest role in consolidating and spreading the tradition as a unity, to develop more explicitly the more problematic implications of its methods. One of the most significant and enduring of these expressions is W. V. O. Quine’s model of “radical translation” and the notorious thesis ofindeterminacy of translation
to which it led.
Over a period of twenty-five years, from the period of his first published writings to his seminalWord and Object
, Quine moved by stages away from the “logical syntax” project of his mentor Carnap, and toward the “radical translation” or “radical interpretation” model of linguistic understanding. The model seeks to reconstruct the facts about the meaning and interpretation of a language in terms of the publicly accessible knowledge available, in principle, to a field linguist initially innocent of the language under interpretation. It thus captures, probably as completely as is possible, the thought that to understand a language is to understand astructure
of signs that are offered and consumed in a public, social context. But the most significant implication of the radical translation model is not its formulation of a structuralist picture of language, but rather the way its result undermines this picture from within. For almost as soon as Quine had fully conceived the radical translation model, he also saw its radical implication: that the meaning of ordinary sentences, though entirely grounded in the publicly accessible facts of language-use, is also systematicallyindeterminate
with respect to the totality of those facts.
The indeterminacy result was first articulated inWord and Object
(1960), but it had developed gradually, in Quine’s own thinking, over the twenty-five years of his dialogue with Carnap. Over the period from 1934 to 1950, Quine came by stages to question and then entirely to reject the traditional distinction between analytic and synthetic statements, and with it also the intuitive notions of logical necessity, synonymy, meaning and intention that Carnap and others had used it to explicate. The publication, in 1951, of Quine’s influential “Two Dogmas of Empiricism” marked a watershed moment in this development; in the article, Quine made explicit his rejection of the analytic/synthetic distinction and began to articulate his own, alternative picture of epistemology. But years before this watershed, the seed of both Quine’s divergence from Carnap and his elaboration of the radical translation scenario had already been planted with a subtle but unmistakable appeal that already appears in some of Quine’s first published writings.
What I shall call Quine’sappeal to use
appears already in 1934, in Quine’s first published reactions to Carnap’sLogical Syntax
. There it already marks, as I shall argue, the essential difference of emphasis that would eventually grow into Quine’s rejection of Carnap’s entire picture. For from the time of these first philosophical writings, Quine held that it is impossible to understand thestructure
of language in complete independence of an understanding of the intersubjectivepractice
of its speakers. In this, Quine already diverged from Carnap, whose vision inThe Logical Syntax of Language
called for languages to be treated as arbitrary, rule-based calculi, uninterpreted in themselves. By understanding the significance of this difference for the development of Quine’s thought, we can gain insight into both the underlying reasons for his divergence from Carnap and the larger significance of the indeterminacy result itself. For we can see how it formulates Quine’s far-ranging internal critique of the structuralist picture of language that can otherwise seem, as it did for Carnap, natural and unavoidable, and that continues to determines both ordinary and philosophical thinking about language and its analysis.
I
We can begin to understand the development of Quine’s understanding of language and meaning by considering its origins in his initial reaction to the work that was the basis of his first philosophical writings, Carnap’sLogical Syntax
. Conceived and written over a period of three years, and appearing in 1934,Logical Syntax
made the bold claim that the problems of philosophy and the logic of science could be treated purelysyntactically
: that is, in terms simply of formal rules governing the interrelation and combination of symbols, without reference to their meanings.
Logicians had previously recognized the syntactical nature of the grammaticalformation rules
governing the possibilities of combining symbols into meaningful sentences, given a perspicuous sorting of symbols into grammatical types. In addition to this, Carnap argued,transformation rules
governing inference or derivation of one symbol-sequence from another could also be treated as purely syntactical ones, concerning only the interrelations of symbols.
In this way, the logical analysis of language becomes the purely descriptive “mathematics and physics of language,” the theory of the rules actually governing the inscription and manipulation of signs in a particular language, natural or artificial.
The important notions of analyticity, deducibility, and logical contradiction can then be formulated, Carnap argues, in terms of the syntactical rules for a given language. Their formal properties, moreover, can be investigated in abstraction from any pre-existing interpretation of thesignificance
of those rules.
Indeed, as Carnap urged, the syntactical conception of logic had the substantial merit of exposing the arbitrariness of the logical rules constitutive of anyparticular
language. For any particular language, logical syntax displays the rules constitutive of meaning and logic inthat
language; but we can always imagine, and formulate, alternative sets of rules to suit our particular needs. This shows, Carnap suggests, that the logical analysis of language need not be an investigation of the “single” logic or the “true” logic, as philosophers had formerly supposed.
Instead, in logical investigations, a “principle of tolerance” reigns, allowing the logician to stipulatearbitrary
rule-determined languages to suit particular needs. Logical investigations can henceforth be liberated from any assumption or question of correctness or incorrectness, and alternative logics and languages freely pursued. Carnap suggests that this will lead to the solution of many troubling philosophical problems, including problems in the foundations of mathematics. These disputes can henceforth be seen simply as involving alternativeproposals
for the form of a language, rather than the substantive disagreements about the nature or forms of objects or entities that they might otherwise appear to be.
The syntactical conception of language thereby gave Carnap a powerful new suggestion for resolving philosophical disagreements by treating them as resulting from disagreements about conventional language forms.
At the same time, though, the conception of logic as syntax also makes possible an account of theorigin
of philosophical and metaphysical error and confusion that would prove decisive for Carnap’s ongoing critique of metaphysics. According to Carnap inSyntax
, most metaphysical sentences in fact arise from the confusion of two ways of speaking, what Carnap calls theformal
and thematerial
modes. The sentences of logical syntax, sentences about symbols and the rules that govern them, are expressed in the formal mode. According to Carnap, all philosophical and logical claims can be written in this mode, since all logical claims in fact characterize the syntax of language. In ordinary usage, though, these formal, syntactic claims are often mistaken for claims in the material mode, or claims about objects and entities rather than about symbols. This becomes particularly problematic when such claims appear to license general ontological or metaphysical conclusions. Thus, for instance, we might be tempted to assert in the course of metaphysical theorizing that “5 is not a thing, but a number” or that “Friendship is a relation.”
But the appearance of substantial theory vanishes when we transform these material-mode sentences into their formal-mode correlates, the syntactical propositions “ ‘5’ is not a thing-word, but a number word” and “ ‘Friendship’ is a relation-word.”
By transforming the material-mode philosophical claims into the formal mode, we reveal their hidden root in the conventional form of the language.
With this revealed, it becomes possible to see what might otherwise seem to be substantial philosophical claims as in fact resting on nothing more than the conventionally determined rules of a particular language. Even claims about meaning, Carnap argues, can be treated as propositions of syntax mistakenly formulated in the material mode. Rightly understood, the claim that one sentencemeans
the same as another is simply the syntactical claim that the two sentences are intersubstitutable, according to the syntactical rules of the language, without altering grammatical or derivational relations to other sentences.
The body ofLogical Syntax
develops these suggestions by developing two specific artificial languages. The rules of Carnap’s “Language I” allow for the formation of meaningful terms and predicates, relations of logical inference between sentences, and a syntactic property of analyticity. The syntactical rules for Language I are themselves, as Carnap demonstrates using a method akin to Gödel’s method of arithmetizing syntax, formulable in Language I itself. Thus the formulation of logical syntax does not require any problematic hierarchy of meta-languages, since each language of a certain degree of complexity has the resources to describe its own syntax.
The second formal language, Language II, is an expansion of Language I, produced by adding to it unlimited quantifiers that allow its sentences to refer to an infinite range of objects. In the context of the logical syntax project as a whole, the two specialized artificial languages have the role of simplified models. Carnap compares their introduction to the physicist’s use of abstractive constructions such as the simple pendulum to help establish the underlying principles of the much more complicated natural world. Just as reflection on these abstractions can illuminate the basic principles of more complicated natural situations, Carnap suggests, the construction of simplified artificial languages like Languages I and II will illuminate the principles and rules underlying the “vastly more complicated” natural languages.
For Carnap, it was thus essential to the possibility of logical syntax that languages, both the artificial ones he developed in the book and the actually spoken natural languages, could be treated asformal calculi
. Such calculi are pure rule-based systems for the combination and transformation of symbols, themselves conceived as lacking any determinate individual meaning.
Examples include not only natural and artificial linguistic systems, but even rule-based systems that include nothing recognizable as symbols; for instance, the game of chess, considered as an uninterpreted system of positions and rules for the transformation of positions, is such a calculus. The procedure of considering calculi without reference to the intended meaning of their symbols, according to Carnap, ensures that what we discuss as the “meaning” of sentences can be treated “exactly,” as emerging from the explicit and definite rules of syntax, rather than defined inexactly and ambiguously, as it would have to be if it depended on the introduction of specific meanings for words:
Up to now, in constructing a language, the procedure has usually been, first to assign a meaning to the fundamental mathematical-logical symbols, and then to consider what sentences and inferences are seen to be logically correct in accordance with this meaning. Since the assignment of meaning is expressed in words and is, in consequence, inexact, no conclusion arrived at in this way can very well be other than inexact and ambiguous. The connection will only become clear when approached from the opposite direction: let any postulates and any rules of inference be chosen arbitrarily; then this choice, whatever it may be, will determine what meaning is to be assigned to the fundamental logical symbols.
Carnap’s method of securing meanings by treating languages as calculi hearkens back to the Fregean idea that the meaning of a sentence can be determined purely by the logical rules that govern its relations of inference and derivation (see chapter 2). It combines this inferentialist conception of meaning with a formalist conception, akin to Hilbert’s, of the nature of a symbolic system. The synthesis makes it clear that the meaning of a sentence, at least insofar as it is relevant to logic, has nothing to do with the ideas, intuitions, or psychological associations that might be connected, in any person’s consciousness, with the particular words that make it up. Rather, meaning is, from the outset, explicitly public, since the syntactical rules definitive of it are shared ones, introduced as a matter of stipulation or public agreement. The philosophical logician’s task is, then, simply to consider the variety of linguistic systems, both actual and possible, and to compare the systems underlying actually existing languages with the simplified and artificial ones he may readily create.
But in requiring that syntactical rules be bothcompletely arbitrary
and wholly constitutive of the sentential meaning that will emerge from the linguistic practice using them, Carnap’s view invites a certain significant tension regarding the institution, stipulation, or adoption of these rules themselves. The tension is almost evident in the first words of the Foreword ofLogical Syntax
:
For nearly a century mathematicians and logicians have been striving hard to make logic an exact science. To a certain extent, their efforts have been crowned with success, inasmuch as the science of logistics has taught people how to manipulate with precision symbols and formulae which are similar in their nature to those used in mathematics. But a book on logic must contain, in addition to the formulae, an expository context which, with the assistance of the words of ordinary language, explains the formulae and the relations between them; and this context often leaves much to be desired in the matter of clarity and exactitude. In recent years, logicians representing widely different tendencies of thought have developed more and more the point of view that in this context is contained the essential part of logic; and that the important thing is to develop an exact method for the construction of these sentences about sentences. The purpose of the present work is to give a systematic exposition of such a method, namely, of the method of “logical syntax”
In the course of the actual practice of constructing artificial languages, the explicit introduction of specialized symbolism will always depend onauxiliary
explanations and interpretations. These will specify theintended
significance and implications of the new symbolism in a convenient, already existing language. As Carnap notes, it is typical to regard such explanatory auxiliaries, as they might occur in the introduction of special symbolism in a textbook, as strictly inessential to the symbolism thereby introduced. The explanatory auxiliariesmust
, in fact, be strictly inessential to the language itself, if it can be considered to be apure
logical calculus, arbitrarily chosen from among all such possible systems. But carrying out the project of logical syntax itself requires that the explanatory introduction of syntactical rulesnot
be inessential in this way. For the actual stipulation or formulation of rules is not simply descriptive of, but actuallyconstitutive
of, the specialized languages created by the syntactician. And it is difficult to imagine that, as a matter of theoretical practice, the syntactical rules constitutive of a language can in fact generally be formulated without any specific intended meaning in mind.
Carnap, in other words, problematically construes the discursive explanations that accomplish the exposition of the system of syntax as bothexternal to
andnecessary for
our understanding of that system itself. For Carnap’s requirement of arbitrariness to be satisfied, it is essential that the significance of the auxiliary explanations and interpretations be extrinsic to the significance of the rules themselves. But even where this specification takes place in the object language, it relies, in practice, onsome
existing understanding of the intended significance of the rules laid down. The particular rules Carnap introduces inSyntax
for Languages I and II, for instance, are introduced with a variety of such devices and auxiliary formulations. Even the introduction of the most basic rules for the sentential connectives, ‘>’, ‘~’, etc. depends on the reader’s antecedent understanding of the ordinary usage of the words “or”, “not,” etc.
This difficulty about the role of interpretation in the formulation of syntactic rules is compounded further in the case of the study of already existing natural languages. Here, the theoretician’s explicit introduction of syntactic rules that purport to represent the actual syntax of the language in question can only be motivated by some antecedent sense, even if only a vague one, of the significance of these rules in terms of the actual practice of the language’s speakers.
The theoretician seeking to describe this practice syntactically can legitimately abstract from most of the vast variety of causal and inferential linkages, evident in the actual use of a language, between individual words and their ordinary referents. But his introduction of rules meant to capture theactual
logic of inference in the language can hardly portray them as completely arbitrary. The introduction of any rule that purports to re-describe the underlying logic of an already-existing language will inevitably rely on discursive explanations that express that rule in antecedently familiar terms, and so will make backhanded reference to forms of speech already familiar to the language’s speakers. Given Carnap’s description of the analytical procedure of logical syntax, it seems impossible to avoid this reference. But given that it must occur, it is extremely difficult to preserve Carnap’s commitment to the genuine arbitrariness and conventionality of all of our language systems.
0%
Author: Paul M. Livingston
