by JLS
for the GC
From:
books.google.com.ar/books?isbn=0415095557...
R. Willis states:
"It follows that our attempt to translate Hopi cosmological concepts accurately into ordinary English is doomed to failure."
Why?
---
His answer is NOT a simple one!
How does this relate to Grice? Grice was a specialist of metaphysics. His very early 'Metaphysics' (co-authored with Strawson and Pears) appeared in "The nature of metaphysics" (London, Macmillan, 1957) and he was aware of the main issues that would concern an Oxford philosopher of his generation. In later years, he became more self-conscious about the metaphysical nature of his endeavour and became more and more involved in determining a methodology for metaphysical research.
At the core of his scheme is the concept of a 'category' --. If R. Willis says that Hopi cosmology doesn't quite translate to English, the 'explanation' has to be categorial --. Philosophers surely can discuss categories in languages other than English. Or more strictly, a discussion of categories -- as embedded in English, say -- need not be conducted in a language which is merely reduced to its role of an "object-language". After all, Whitehead wrote in more or less ordinary English his 'process metaphysics' -- which is supposed not to fit exactly the Aristotelian category scheme that many postulate at the core of 'ordinary English'. So some liberal approach to these issues seems to be in order!
Subscribe to:
Post Comments (Atom)
Posts
Reminds one of the venerable Sapir-Whorf hypothesis. While I don't exactly believe in the S-W hyp., I find it interesting (tho' the Chomsterites tried to nix it back in the 60s or something). Your latin examples for example indicate the problems of translation, even from related languages--latin to english. The latinate verbs do not map at least exactly to Anglo; really most Ang. native speakers have difficulties with spanglish and franch irregular verbs, not to say masc./fem nouns, adjectives . Same for pinche aleman (German as in Deutsch); the cases and articles rather challenging (..Allemaigne the term since Caesar). German, confound that wretched language as Mark Twain called it--Twain/Clemens did speak/read some french at least (and probably had Latin in his grammar school, back when many Merican schools required it).
ReplyDeleteAt any rate it's possible syntax conditions thought to some degree; language relates to metaphysics, even for the most reductionist of dweebs. Time is not Zeit (or tiempo...actually 'la hora" the time in spanglish); Being is not Sein (or is it).
Americans may be the most linguistically inept of any earthlings, however. Hearing a few words of spanish (or worse, chinese, or arabic..imagine doing philosophastry in mandarin, JLster), the average yokel starts reaching for his semi-auto and his rebel flag....
Well, -- in the heyday of ordinary-language philosophy, there was a flurry of articles (I think A. N. Prior joined) about the 'ontology' that English commits one to. Very few of those discussions would quote from Grice, because this was the time of the 'unwritten' or unpublished time', and I am not really interested in the debate unless it touches on what Grice, only here and there, says about the deep categorial structure of a language. Since he was a pragmatist at heart, and would accept that a categorial scheme may be thought of as primitive for a purpose but that another one -- which is describable as derived from the former -- may be thought to be prior for yet a different pupose. I can think of his reference to 'atom-electron-quark' versus 'chair-table-person' along those lines; but there are other more purely philosophical distinctions: material-object (or thing) language versus sense datum language. He also played with 'legal right/moral right', and various other dichotomies. Consider 'pscyhology' for example. He also touches on them, psychological predicates. A psychological predicate (or concept) is it primitive in the scheme of English, or is it derivable from terms for 'behaviour'. Don't our terms for behaviour already presuppose some element of psychologism? And if the physicalist is right and it all reduces to matter, does that mean that the categories of agentivity, etc., have to be understood as 'material'? Or consider his formal devices. He speaks of x, y (Ex), Ax, ix -- In Log. & Conv. but he says nothing about the range of such variables. Can a 'number' fall within the domain? Should a 'sense datum'. It seems he is interested with common-sense interpretation of the calculus. With B as bald, and K as king, say -- (to yield (Ex)Kx & -Bx (the king of France is not bald). But what is "x" -- we don't need to know. But if we play with predicates on the line of 'king' and 'bald' the quarks are left someplace behind! -- Or something.
ReplyDeleteI should learn some Hopi.
Yes, it's ALL about Sapir/Whorf! Thanks for
input!
Even my ref. to Whitehead in another post, on process, was motivated by that, and by Quine's idea that "Gavagai" may not refer to a rabbit but to a rabbit-part (or rabbit process). There is some inscrutability of translation, these empiricists think, as we translate. But my point refers to the availability IN ENGLISH of a more or less adequate translation. I thought R. Wallis was being hyperbolic about the 'doom to fail' thing -- English seems richer than most writers on Sapir/Whorf concede, and we are still talking English (even if meta-lingusitically) when discussing the categorial structure of a lingo OTHER than English, I expect.
Quine's thoughts on language (or "language-behavior) seem a bit sketchy in parts. In his extreme example of an anthropologist in the bush encountering an unknown tribe speaking an unknown language, perhaps the Anthro. person would be required to relate the words with action/behavior a word "elicited" (in QuineSpeak), but even oral traditions have lore or myths, probably some fixed definitions which the Anthro. would have to learn. At any rate the Gavagai example seems a bit inadequate. There's a visual aspect to language learning, at least initially--just relating words and sentences to objects. But that's hardly the whole of it, at all (what do the words "democracy" or Justice or history relate to...). In that I still agree, slightly, with Chomsky's points on the creativity of language use (that is, assuming he hasn't repudiated that...), and the limits of empiricism in a sense....
ReplyDeleteThe point on variable-ization is somewhat interesting. (x) may work in an equation, describing objects in motion, particles, etc.--but the logicians' use is quite different, a bit ad hoc in a sense. Though, if one defines domains and objects (persons, or polynomials?) then the variable seems adequate. Or not, at least in Gricean terms--but I am of the opinion much "conv. implicature", at least in informal settings, is more akin to induction and inference, however quotidian--and relates to agency as well. Intelligent sane people it's not that problematic (tho' could be...in a Finnegan's Wake type of liter-rary situation); with plebes, it might be. The implicatures of say attorneys or professors are not those of mobsters or mormons.
Process...yes that's it! The map is not the territory, JL.
Good. Perhaps we should revise why R. Willis thought of "Hopi". I mean, his example was rather simplistic. We would need to know the 'grammar' of Hopi to see if the inadequacy (in some use of that term) is due to --- what? It cannot just be lexical or an overlap of lexical categories. It cannot be morphosemantic. It may be syntactic. It CAN be pragmatic. We need a lot of detail, etc.
ReplyDeleteYou are right about the wider context behind the "Gavagai" example.
And the point about the variables x, y, z, etc. as used by Grice. He uses variables in Logic and Conv. when giving the translation
(Ax) "all"
(Ex) "some" ("at least one") --- he gives the
----------------------------two glosses.
(ix) "the"
--- but it's not clear what stands for 'x'. Of course he doesn't NEED to provide the domain. But one wonders if the same quantifiers are realised in ALL languages.
And also, if we do introduce a predicate, "... is bald", and another "... is a king of France", we seem to be on more or less safe ground. Apparently, the logician should not require that there is consistency in the predicate choice. One may bring in psychological predicates, observable predicates, physical predicates, quantum-theoretical predicates, Hopi predicates, whatever. The logic is only concerned with x having any property P (or rather phi, instantianted by Predicate P), and so on.
But if the categorial system -- or quantification system -- is OTHER in the object-language (unlike English which has the three things) we may be at a loss. Bennett has considered how a language (in his "Linguistic beaviour") would evolve from very simple utterances to all the details of predicate calculus. This in the last chapter of that book which was his second one after a good analysis of "Rationality" in the eponymous book. Bennett IS a Good Griceian. Etc.