TurquoiseB wrote: > --- In FairfieldLife@yahoogroups.com, cardemaister <no_re...@...> wrote: > >> --- In FairfieldLife@yahoogroups.com, TurquoiseB <no_reply@> wrote: >> >>> --- In FairfieldLife@yahoogroups.com, Bhairitu <noozguru@> wrote: >>> >>>> Supposedly some translations engines use Sanskrit as an >>>> intermediate language because it is unambiguous. The program >>>> will take text in a language and translate it to Sanskrit and >>>> then from Sanskrit to the target language. >>>> >>> I´m sorry, but this sounds like bullshit to me. >>> >>> I know very little about Sanskrit, but everything >>> I ever heard talked specifically *about* its >>> ambiguity. They talked about poetry forms in >>> which every word in the verse could have several >>> meanings, and the whole *art* of the poetry form >>> was being able to put a whole series of these >>> words -- *each* of them having four or five >>> meanings -- together in such a way that no >>> matter which meaning of any of the words you >>> pick, the whole verse still makes sense. >>> >>> Plus, just looking at the definitions Card posts >>> here, words often have *more* than four or five >>> completely different meanings, right there in the >>> definitions he posts. >>> >>> So I´m thinkin´ that this stuff about using >>> Sanskrit as an ¨intermediate language¨ for trans- >>> lation engines is just someone´s True Believer >>> bullshit. >>> >>> If you want an unambiguous language, choose French. >>> That is why all international treaties use it as >>> the ¨master language¨ for the treaties. There is >>> a copy in the language of each country, but the >>> master is in French, because it is so precise. >>> Everything I´ve ever heard about Sanskrit presents >>> it as just the opposite. >>> >>> Card or others can correct me on this if I´ve heard >>> incorrectly. I´m not trying to knock Sanskrit or >>> anything; it´s just that Bhairitu´s claim sounds >>> the opposite of everything I´ve ever heard about >>> the nature of Sanskrit as a language. >>> >> I love Sanskrit, but not because I'd think it's unambiguous. >> Just as a simple example, the inflectional form 'yoginaH' could >> be either ablative/genitive singular (e.g. from/of) or >> nominative/accusative plural ([many] yogis, either as a subject >> or an object of the sentence), depending on the context. >> > > Thanks for weighing in, Card. As I said, I'm > not questioning the idea of Sanskrit being a > cool language, just the idea of it being a > cool "intermediate" language for translation > because of its "unambiguous nature." Literally > everything I have ever read about it mentioned > that it was one of the *most* ambiguous lang- > uages on the planet. > > Ambiguity *is* the issue when it comes to trans- > lation, whether by humans or by software. It's > captured in the classic example from English: > > "Time flies like an arrow; fruit flies like > a banana." > > This sentence only makes sense, even in English, > once you have "parsed" it and realized that the > words "flies" and "like" have very different > meanings in one phrase than they do in the other. > *As I understand it*, that is the problem with > Sanskrit, and in spades. Each word has *many* > meanings; as Vaj puts it, not just "double > entendre" but "multiple entendre." Thus it is > hard for me to conceive of it as a suitable > language with which to address the many problems > of machine language translation. > > As for the website Bhairitu pointed to, all that > you have to do to see its True Believer nature > is to do a mental "search and replace" on the > text in it and replace every mention of "Sanskrit" > with "Hebrew." Then you'll see what the site is > really about. > > It's attempting to present a case for learning > Sanskrit based on its supposedly spiritual nature, > and its supposed status as the "mother of all > languages." Sanskrit may *be* both. For all I > know, God, all the gods and goddesses and angels > sit around discussing Monday Night Football in > Sanskrit, because it's the most suitable lang- > uage for doing so. Maybe it even has magical > abilities to heal the sick and raise the dead > and fix the game during Monday Night Football. > > I don't know, and I don't care. The only relevant > piece of information in this context is whether > it is an *unambiguous* language. Given a sentence > in Sanskrit, can that sentence be parsed one and > only one way? > > Everything I've ever heard is that the answer to > that question is a definitive "No." And that > unambiguous answer rules out Sanskrit as the > basis of an experiment in machine translation > that is based on the notion of that base lang- > uage being unambiguous. What else would Sanskrit scholars be looking at if it wasn't spiritual texts, Turq? That seems to be the only stuff that survived. Maybe if you learn Sanskrit and maybe Pali while you're at it you can go do some digs in India and see if you can find any texts from the sports section of some ancient daily scrolls or palm leaves. :-D
So to label an organization that studies Sanskrit spiritual texts as TB'ers is a bit off the mark I think. As I believe Vaj already pointed out Sanskrit the root of a lot of languages. That is an additional point to make it an intermediate language. And while we're at it what was the first word in my first post on this topic?
