Monday, June 16, 2014




P.S. To “How Not to Read/Teach  ‘Two Dogmas’"

Rereading the previous, I regret that I failed to say something more positive about Quine’s great achievement in “Two Dogmas of Empiricism”. What Quine did (along with his friend and my undergraduate teacher, Morton White—see “The Analytic and the Synthetic; an Untenable Dualism”) was to raise the stunning question, whether there is any intelligible sense of “analytic” in which, as the positivists claimed, mathematics is analytic, or in which “light travels in straight lines”  is analytic, or, in which any fundamental principle of science is analytic (and hence unrevisable).  Once one sees that there isn't any sense in which some of our fundamental beliefs are analytic—or “synthetic apriori” as Kant thought, either—one views the history of human knowledge (not just science!) in a different way. Quine’s rejection of the notion of “meaning” was, I believe, mistaken; but accepting that there is such a thing as describing the meaning of words does not commit one to accepting the overblown notion of  the analytic that Quine and White attacked. This is what I said long ago in "The Analytic and the Synthetic”, and I should have repeated it.

4 comments:

  1. "What Quine did (along with his friend and my undergraduate teacher, Morton White—see “The Analytic and the Synthetic; an Untenable Dualism”) was to raise the stunning question, [...] in which any fundamental principle of science is analytic (and hence unrevisable)"

    I think the parenthetical remark, at the end, is questionable. I am not completely sure about the other logical positivists, but it is unfair to attribute to Carnap the view that analytic sentences are unrevisable. Carnap was one of the main logical positivists and the main target of "Two Dogmas." But as far back as Logical Syntax one can see Carnap explicitly supporting Duhemian holism, that sentences cannot be tested in isolation, and that all sentences, including analytic ones, are revisable. By the early 50s, what mattered for Carnap, regarding analytic sentences, was just that they followed from the rules of meaning for the language. But those rules of meaning could change.

    This is part of the reason, I think, that some have argued that Quine's discussion of revisability in the final section of Two Dogmas misses the mark on Carnap's philosophy. Carnap already accepted much of those insights. But, nonetheless, what I take the strong Quinean argument against Carnap to be is that: *given* Carnap's commitments to holism, complete revisability, and empiricism, it follows that the distinction between a person holding a sentence true in virtue of meaning, or not, is unjustified. This is because holism and revisability imply that any sentence, analytic or not, may be held true or rejected. But then there seems to be no real behavioural difference between accepting a sentence as analytic, or otherwise. And then, given empiricism, Carnap should accept that the analytic/synthetic distinction is unjustified.

    Or maybe this is better understood as a *challenge* to Carnap, or others, to provide a behavioural criterion for holding a sentence analytic or not. Indeed, Carnap understood it in this way (and some letter correspondence between Carnap and Quine around that time reveal Quine had similar thoughts).There are some, myself included, who believe Carnap can meet the challenge.

    ReplyDelete
  2. I have a different way of looking at analyticity and syntheticity. Instead of looking at analytic and synthetic statements, I say there are analytic and synthetic analyses. Analytic analyses are through mathematics and logic, where relevance relations are followed and new statements are produced or read with reference to other statements of definition or synonymy. Synthetic analyses are done with reference to scientific or existential relations between statements and what makes these statements acceptable. So, for example, "A=A" can be analytically read, while "Evolution is generally accepted as true" can be synthetically read with reference to statistical analyses of scientists.

    Do you think this is a promising or even tenable way of addressing these concepts, or do you feel there are serious problems I am missing?

    ReplyDelete
  3. “The Analytic and the Synthetic; an Untenable Dualism. … Quine’s rejection of the notion of “meaning” was, I believe, mistaken; but accepting that there is such a thing as describing the meaning of words does not commit one to accepting the overblown notion of the analytic that Quine and White attacked.”


    Even this analytic/synthetic distinction is wholly depended on the issue of ‘meaning’ which is 100% hinged on the “Martian Language Thesis (MLT) -- Any human language can always establish a communication with the Martian or Martian-like languages”.


    MLT is based on the solid fact that both human and Martian share the followings:
    One, the same meta-space (encompasses the same physics laws, the same history of this universe, etc.)

    Two, the same ‘meaning-space’ (the meaning of every ‘object’ and ‘event’ of this universe (having an intrinsic existential meaning (IEM) and radiating out some additional info into the environment) is the same for both human and Martian).


    This MLT is absolute even in the multiverse case when other universes have a different meta-space (that is, also having a different meaning-space). When the other universes are totally ‘disconnected’ to this universe, there will be no ‘communication’ between them and this universe, that is, there is no ‘meaning’ issue among them. If a species of the other universes is able to come to this universe, then there must have a ‘connection’ between the other universes and this one, and that connection will unify the different meta-spaces into a new and bigger unified-meta-space, and thus again reestablish a new and bigger meaning-space. So, the rejection of ‘meaning’ can only be done ‘subjectively’.


    The syntax/semantics dualism was intensely debated at Scientia salon (see http://scientiasalon.wordpress.com/2014/06/12/the-turing-test-doesnt-matter/comment-page-1/#comment-3911 ). I think that that analysis is also applicable on this analytic/synthetic duality.

    ReplyDelete