From cbmvax!uunet!cuvmb.bitnet!LOJBAN Wed Aug 5 15:49:45 1992 Return-Path: Date: Wed Aug 5 15:49:45 1992 Message-Id: <9208051601.AA27942@relay1.UU.NET> Reply-To: CJ FINE Sender: Lojban list From: CJ FINE Subject: Re: quest for opinion X-To: ash@ACM.RPI.EDU X-Cc: Lojban list To: John Cowan In-Reply-To: ; from "Arthur Hyun" at Jul 30, 92 10:37 am Status: RO > > > I have some opinions, but in the interest of saving bandwidth, > it seemed wiser to first ask if these subjects have been gone over > while I've been off the list (about a year). If someone's archived > any discussion, or can paraphrase such, I'd like to get a copy. > If not, I'll be glad to try to explain my opinions in greater detail. > > 1) Place structures. More specifically, is it appropriate that one > gismu cover more than one relation because of place structures? There was a long discussion last November. Bob's position was different relations iff different place structures iff different brivla I'm not sure how it all ended up - at one point there was discussion of a place-suppressor, ie a cmavo which said "This place is not just empty, but is actually absent from the bridi-relation I am expressing (which is thus a different relation from the one this brivla normally expresses)", but I don't know whether this was actually rejected or just got fed up waiting and went away. > > 2) Emphasis and idiom: There are several sets of words in the gismu > lists I have that seem to me to express the same relation(s). They > appear to me to differ only by a matter of emphasis and/or idiom. > Shouldn't there be just one gismu? It is no longer the case that gismu are (semantically) primitive - that was one of the paths by which a concept was admitted into gismu space, but there are others - particularly, primitiveness in human terms, and productiveness in tanru and lujvo. Thus there may be words that are logically unnecessary - though none, I think that are strictly synonymous once you take their place structures into account. > > 3) Translation vs transliteration: Has anyone yet managed to produce > a concept in lojban that is honestly very difficult or impossible > to express in English, yet is understandable by people in a non- > idomatic way? Here is essential that I emphasise *concept*, not > sentence or somesuch. (Just because it'll probably come up, I'll > try to handle the JL16 example at the bottom of the letter). Don't know > > 4) I haven't had time to really work over the BNF, but I'm wondering > if it's true that you cannot distinguish grammatical from ungrammatical > lojban without knowing the rafsi used? It depends slightly what you mean by grammatical. All you need to *parse* a text is to be able to recognise brivla - you don't need to know anything more about them. However a sentence with a sumti in a place which the selbri lacks, is in a sense not grammatical - though I think most people prefer to treat it as grammatical but nonsensical, like a sentence with two inconsistent values in the same sumti-place (which is perfectly possible with explicit FA tags, and the parser will accept it as valid). Note the the rafsi are nothing to do with it - the place-structure of a lujvo cannot be algorithmically determined from the component brivla or their rafsi (even if some form of dikyjvo were adopted, this would still be true). > > That's that. I thank y'all for your time and help. > > cheers, > arthur > > mi danfu tu'a la artr > > --- > > Now the JL16 example I was talking about: > > On the bottom of p16 of JL16, there's an example of a set of > circumstances that could be transliterated as > "X kept on: kept on hitting the dog too long, too long." > > I think I'd be more likely to render the English as: > "X punished the dog until it died." > > Or iff you want greater specificity: > "X punished the dog so often that it died." --or-- > "X punished the dog by hitting it on so many occasions that it died." > > I think that the concept can be clearly and succinctly expressed in > English when translated, but not transliterated. > In context, these may be perfectly good translations, but note that you have introduced a concept (punish vs hit) and dropped one (punish vs hit-too-often). I think the original translation given is not a good example of something which cannot be clearly expressed in English: it has a sort of pragmatic unacceptability, but is perfectly understandable once explained. I would be much more interested in a text that cannot be expressed in English without a substantial paraphrase.. co'omi'e kolin