[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
On Tuesday 21 January 2003 10:00 am, And Rosta wrote: > John: > #And Rosta scripsit: > #> Agnosticism isn't possible with synonymy. Either you believe two > #> words have the same sense, or you don't. In this instance, they > #> didn't. But I do. Arrant nonsense. I can easily believe that a lexeme doesn't have the same meaning as itself (in different contexts) and I can easily be unaware of the meaning of a word, and unable to tell whether it is a synonym of some other word. Furthermore, when two words have substantial but not complete overlap of meaning, I can answer the question, "Are they synonyms" by saying "It depends". There are many possible theories of synonymy, and I don't know a priori which one is appropriate to a given discussion. This statement is based on a theory of meaning known as pragmatism, which held, in its original form, that many useless arguments can be avoided by noting that people use the same word with different meanings, and that negotiation over definitions is more productive than fiat. You may hold another theory, but you can't tell me that I don't believe in mine. > #I'm not sure whether "you don't" means "you believe they have different > #senses" or merely "you ~ believe they have the same sense". > > I'm saying that they amount to the same thing, if (as I say) synonymy > is a linguistic rule that says "the sense of word X = the sense of word > Y". I have such a rule for 'Wolfram' and 'Tungsten'. If you lack such a > rule, then X and Y aren't synonyms, even though their senses may be > extensionally equivalent. Most Americans don't know the word Wolfram, so they can't know whether or not it is a synonym for Tungsten. Ingorance has nothing to do with belief, or with the meanings of the words. > #> That's right. (it so happens that one of the few things I know about > #> these two words is that they denote the same thing.) You can conclude > #> that for you they are synonymous, unless there is strong evidence > #> from the rest of usage that they aren't generally recognized as > #> such, in which case you may create a superconcept that covers > #> them both, but continue to think of a woodchuck as a M. monax that > #> would chuck wood and of a groundhog as a M. monax that appears on my > #> birthday to foretell the coming weather. > # > #So synonymy is really a 3-place relationship between two terms and a > #believer, x1 is synonymous with x2 in the usage of x3? No. It at least requires two terms, a person, a theory of meaning, and some linguistic information about the terms, and it may well require an explicit ontology. In any case, meanings are not that precise no matter what your theory of meaning or your ontology. I suppose we could add in a theory of fuzzy logic or some other notion of "sufficiently the same". Of course, even with all of that you can't reduce synonymy to an equivalence relation. > Gack -- *please* change "usage" to "language"! But yes, with that > change, then indeed synonymy is a 3-place relationship. > > #> As you recognize, cases like "square of 2" and, if it is understood > #> compositionally, "H2O" are not candidates for synonymy because they > #> have a compositional meaning. And it is clearly possible to believe > #> that 4 is not the square of 2. > # > #How is this fundamentally different from believing that furze is not > gorse, #or that "not" is not a mark of negation? > > If furze and gorse are aren't synonyms (they aren't for me), then it is > not fundamentally different. If they are synonyms, then believing that > furze is not gorse is literally the same thing as believing that furze is > not furze, so "furze is not gorse" must receive a nonliteral > interpretation, under which the two terms are not synonyms. Believing that "furze" does not refer to gorse cannot be correctly changed by substitution into believing that "furze" does not refer to furze. Use vs. mention, and standard rules for substitution of bound variables. > Believing that "not" is not a mark of negation is essentially a lexical > belief about sound--meaning correspondences. In contrast, believing > that 2 + 2 = 5 is a nonlinguistic (mathematical) belief. Except in the case that it is a lexical error, as sometimes happens among those learning English as a foreign language. > #> * Synonymy exists: we can have the knowledge "word X and word Y have > #> the same sense (whatever the sense is)". Under some theories of meaning, but not all. > #But this seems to be true iff we believe it, or more operationally, terms > #are synonymous for me iff I use them interchangeably. > > Yes. No. The phrase "synonymous for me" is not synonymous with "synonymous". > #This makes hash of the distinction between de dicto and de re belief. > > Why? De dicto beliefs are not linguistic beliefs. We can report de dicto > in Lojban the beliefs of someone who does not know Lojban. > Synonymy would be a fact about my knowledge of Lojban, not about > the believer's beliefs. > > --And. -- Edward Cherlin Generalist & activist--Linux, languages, literacy and more "A knot! Oh, do let me help to undo it!" --Alice in Wonderland