Analytic-synthetic distinction

The analytic–synthetic distinction is a distinction used in philosophy to divide an ontology into two parts: an analytic part consisting of terms related by synonymy, and a synthetic part concerning connections between such terms and 'real' objects.

Analyticity and empiricism

In several books and papers, Quine challenged the analytic–synthetic distinction proposed by Carnap. Quine argued that although there are trivial situations in which analyticity prevails (those circumstances which simply replace some elaborate sequence of terms by a tautological equivalent), all interesting propositions in an ontology (all that do not involve simple use of definitions) are synthetic in nature, that is, they inevitably bring forward some empirical fact.

According to Putnam, Quine's position on analyticity is:

‘	“A statement is analytic if it can be turned into a truth of formal logic by substituting synonyms for synonyms.”	’
—Hilary Putnam, ‘Two Dogmas’ revisited

The issue then turns upon whether 'synonym' has a meaning beyond simple tautology. An example involving simple tautology is that All bachelors are unmarried, which holds true simply because, by stipulation, someone is a bachelor if and only if they are unmarried. Quine argued there were no other kinds of analytic statements, and any attempt to extend 'synonymy' beyond such kinds of examples was doomed to failure. Analyticity is possible only by stipulation.

The target here is a resolution of the question whether synonymy is not just a logical matter, but a matter of usage. For example, can stipulation be extended to certain sense perceptions or observations? Such a view seems to propose the matter of 'analyticity' is not one strictly of logic, but possibly also involving confirmation.

‘	“ I shall go on to apply my principle to show the following classes of judgments are exempt from logical criticism: ... One of these consists of perceptual judgments. For example, when I say "The sky is blue", I am not speaking of any external reality but mean only that when I look up I have a sensation of blueness. It is conceivable that this judgment, being an entirely different sort of mental product from a sensation, should misrepresent the sensation. But if we cannot help making that judgment, and up to date there is not the slightest ground for a suspicion that we ever can make it otherwise than we do, it is utter nonsense to inquire whether it is made right or wrong.”	’
—C.S. Peirce, On the presuppositions of logic

Is the analytic-synthetic distinction blurry?

Interpretation

Carnap introduced two types of 'truth' which he called L-truth and F-truth. L-truth was truth by virtue of semantical rules of an adopted language alone and is his definition of 'analytic' truth (L for Logic). On the other hand, F-truth is what Carnap calls synthetic truth and is also described by him as 'contingent' truth, requiring the 'observation of facts' (F for Fact). He points out explicitly that for a statement phrased within a language to be useful, its analytic truth is necessary but has no bearing upon its synthetic truth. The truth of statements, according to Carnap, was not an undeniable consequence of analyticity within their formal structure, that is L-truth alone, but also a matter of F-truth.

Carnap was close to Lewis in thinking that analytic statements were hypothetical and, while tautologically true within their formal structure, could be considered to have some empirical validity only by comparison with experiments. In this respect, Carnap and Lewis agree with Hawking/Mlodinow's proposal of model-dependent realism.

Quine's argument over synonymy can be viewed as stating that the distinction between L-truth and F-truth is untenable, that there is an ineluctable interplay between observation and construction of a language. Or, Quine's position can be taken as in agreement with Carnap over L-truth, and to be an argument over how F-truth is to be established.

‘	Quine's pragmatic challenge to the analytic-synthetic distinction has won many sympathizers, including Nelson Goodman, Morton White, and Hilary Putnam. Critics, however, maintain that some notion of analyticity is indispensable to any coherent account of either formal systems or our everyday use of language. H.P. Grice and P.F. Strawson, for example, contend that analyticity, necessity, and cognitive synonymy constitute an internally coherent family of terms, which Quine takes out of context in appealing to extensional criteria such as semantic reference. Others follow Wittgenstein in restricting analyticity to pure logical tautology, noting that Quine himself does not contest logical truths or stipulations. Even such counterproposals, however, concede a severe curtailment of the distinction, and in the wake of challenges from pragmatism it is hard to envision the analytic and the synthetic ever again enthroned in resplendent isolation.	’
—Frank X. Ryan, "Analytic: Analytic/Synthetic" in American Philosophy: An Encyclopedia'

Either view is in accord with Quine's statements concerning the verification of concepts, namely, that they cannot be established individually, but only entire theories can be assessed, and they stand or fall as a unit.

‘	“ Or, to change the figure, total science is like a field of force whose boundary conditions are experience. A conflict with experience at the periphery occasions readjustments in the interior of the field. Truth values have to be redistributed over some of our statements. Re-evaluation of some statements entails re-evaluation of others, because of their logical interconnections -- the logical laws being in turn simply certain further statements of the system, certain further elements of the field. Having re-evaluated one statement we must re-evaluate some others, whether they be statements logically connected with the first or whether they be the statements of logical connections themselves.”	’
—Willard Quine, Empiricism without the dogmas, Part VI of 'Two dogmas of empiricism'

Quine took this stance further, to argue for a 'theory of everything', to use modern parlance.

Price proposes that the real argument between Quine and Carnap is not the analytic-synthetic distinction, but is over the issue of multiple languages, "the assumption that there is some sort of principled plurality in language which blocks Quine’s move to homogenize the existential quantifier." This 'homogenization' of Quine's is made possible by Quine's dismissal of the internal-external distinction, the view that these multiple languages are only a plethora of specialized languages, all of which can be subsumed under one general, over-all language. "What is to stop us treating all ontological issues as internal questions within a single grand framework?"

One approach to this question is the work of Wittgenstein, who pointed out the futility of looking for the meaning of a term that subsumes all its particular meanings. In a prolonged discussion of the possible meanings of the word 'game', for example, Wittgenstein introduced the idea of of imposing boundaries or frontiers upon its meaning, or circumscribing its meaning.

‘	“ For how is the concept of a game bounded? Can you give the boundary? No. You can draw one; for none has so far been drawn. ... To repeat, we can draw a boundary — for a special purpose. Does it take that to make the concept usable? Not at all. (Except for that special purpose.) No more than it took the definition: 1 pace=75 cm to make the measure of length 'one pace' usable. And if you want to say "But still, before that it wasn't an exact measure", then I reply: very well, it was an inexact one. — Though you still owe me a definition of exactness.”	’
—Ludwig Wittgenstein, Philosophical Investigations, excerpted from §68 & §69

This idea of introducing circumscribed meanings for special purposes is what Carnap meant by a 'linguistic framework' and what Hawking/Mlodinow mean by a model-dependent realty. Some of the modern ideas of 'length' are discussed in length measurement, illustrating Wittgenstein's comments about the unit 'pace' and the redefinition of what is meant by an 'exact' measure of length as measurement methods evolve and the circumstances shift in which 'length' is used.

Thomasson also supports the fundamental point as the role of multiple languages. She says that the focus upon the analytic-synthetic distinction in the Carnap-Quine debate is misplaced; the important point is the internal-external distinction (which Quine dismissed as trivial): "The real distinction instead is between existence questions asked using a linguistic framework and existence questions that are supposed to be asked somehow without being subject to those rules—asked, as Quine puts it ‘before the adoption of the given language’." Carnap considered this issue as well: "In order to speak about any object language...we need a metalanguage. We shall use as our metalanguage M a suitable part of the English language which contains translations of the sentences and other expressions of our object languages..., names(descriptions) of those expressions, and special semantical terms."

Remarks

So, to go back to the beginning, "Is the analytic-synthetic distinction blurry?" There is no doubt that the question is still under discussion. It appears that the subject is shifting toward a new formulation: suggesting that while "analyticity" may be supported within each particular 'linguistic framework' as suggested by Carnap, the plethora of possible frameworks poses the complicated problems of comparing them. In addition, the invention or conception of a new linguistic framework is a process whose dependence upon observation is unclear, as Quine suggested. As Aristotle posited, we might never have invented the concept of 'circularity' were it not for observations of objects that are seemingly circular.