type 3 fu'ivla

From Lojban
Jump to navigation Jump to search

fu'ivla that have rafsi prefixed to them to give some sort of semantic hint.

Discussion

  • nitcion:
    I (as self-appointed defender of Type 3 fu'ivla)...
  • jay:
    Are there people opposed to type 3 fu'ivla?
    • Let me rephrase: I (as self-appointed opponent of type 4 fu'ivla) :-)
      • Is that complete opposition to type 4 fu'ivla, or merely opposition to skipping over the Type 3 phase?
        • nitcion:
          Strong opposition to skipping Type 3, reluctance to see Type 4 become widespread, for reasons abundantly argued elsewhere (see the fasxolarto rant, for example: the need to distort the source word occasionally past recognisability, rather than merely misstressing it and tacking on a final vowel and a crunchy start; the ambiguity of which source sense is invoked in case of ambiguity or accidental homonymy; the at times extreme difficulty in telling when the Type 4 loan doesn't break; and the fact that neither 'Tweener nor the Book particularly used them any (and discouraged them before going through Type 3 apprenticeship), so they just don't feel canonical to me.) Default Type 4 make the same impression on me as experimental gismu. (Whether they should or no is another matter.) And I'm sorry, but by the time we come to substitute glico with gli'ico rather than kulnrxinglici or kulnrxanglia (see Experimental cultural gismu), I've got to wonder what the actual advantage of Type 4 fu'ivla is at all...
  • Adam:
    Besides ideological opposition, I think that the major problem with type 4's is that the WordResolutionAlgorithm has not been completely debugged, and it seems to be weakest on picking out Type 4 fu'ivla, so it's hard to tell what forms are valid.
  • .kreig.daniyl.:
    Nick's not the only one. I find ALL fu'ivla un-lobykai, as lojban is designed to work with no other common language, and to understand type 4s you need to know the language it comes from so well (to recognize that it's the same word, a la gli'ico) that you probably have no reason to be using lojban. Would a computer know what a fasxolarto was? No. Nor would it understand mabrnfaskolarto, but then it would know that it was not supposed to - seeing type 4s confuses me, and gives me what la and. once referred to as the Logical heebiejeebielets. If you want lojban, use lojban words and tanru (Which I find more lobykai than fu'ivla, by far]. If you want English, you know where to find it - and if you are reading this, you already have.
  • How do you intend to name all existant animals without fu'ivla?
    • Like Nick, I intend to use type 3s. But they are the one case where I feel fu'ivla to be a necessary evil. I do not care, for example, for terms like pemcrxaiku, which can be written as slaka kancu pemci instead.
  • There are other forms of poetry with exact syllable counts. In fact there is another Japanese poetry form exactly like the haiku except with longer lines. (can't remember the name).(TANKA.) How do you distinguish between the two? tack on more seltau?
  • paci slaka pemci, tho I'm not sure it parses.
    • It does, though it's a sumti "13 syllable poems". It also doesn't refer to haikus. Presumably you were after something more like slaka pacimei pemci (or maybe slaka jaumei pemci? or slaka mupi'ezepi'emumei pemci?) .i co'a ma li mu su'i ze su'i mu du li paci
      • Well, maybe if you changed that to paze slaka, since they are 5,7,5...but there's more to that than the count. You must also have a season-reference, and convey the poetic effect known as "yugen".)
        • .i ko smaji zo'o .i le'e nalponjo punai sanji la'edi'u ca lenu finti lo pemcrxaiku
  • xod:
    fu'ivla are needed to express specific concepts that come from a non-Lojban culture. Haiku isn't just poetry with 17 syllables! It is a specific tradition, with a specific history, and a tanru describing this honestly would take several dozen gismu with a lot of their be places filled. That's not jbokai, that's just silly.
  • There is a difference between the definition of a verse form, and the traditions associated with it. Haiku are the only poemform I am aware of which always has exactly 17 syllables. And yes, I meant to say 17 and not 13 above, I still haven't memorized all my numbers. Haiku is a poemform which your interlocutor is likely to have heard of, but for, say, the game of go, I'd try something like strategic stone-putting game - while discussing the game I would use the Japanese terms, but so as not to alter the morphology of the Japanese I would use type ones, but in passing mention I would use tanru.
    • As soon as you come up with a concise way to say "strategy", I'll play a game of othello with you. :-)
    • nitcion:
      military strategy's jamna platu, and tactics are jamna ve pruce? In any case, won't something like jinga platu or jinga tcila bo platu or jinga banli bo platu do?
  • How awkward. A type 3 fu'ivla is essentially a name for a concept, except that it's easier to pronounce, more lojbanic, and gives a sense of the meaning. Why go out of your way to replace a name with a different kind of a name and with long and inexact circumlocutions?
  • If I am discussing the game of go, and I say mornrcitco it gets the point across. However, almost everyone who plays go believes either (a) that one should use go terminology from the language in which one is speaking, hence the nice lojbanic tanru serti morna, for a stair-like pattern, or (b) that one should use the japanese term, so it is a shicho - and the Japanese word is not done justice by the lojbanic approximation. The sounds are not the same - Japanese is a pitch-accent language and lojban is not.
    • That is true of any two languages. Do you use the correct pitch-accent when you say the Japanese terms in English?
      • Yes. the Shi is in low pitch, the cho in high, at least as I have always heard it.
  • If I was in a discussion of the game of go, any go player would understand mornrcitco if they spoke lojban - but there is more to a language than communication. The Japanese go terms have tradition behind them (as opposed to behind the tesuji (= patterns which appear in go) themselves).
    • So what's your point? That we should talk about go only in Japanese, in order to get the pronunciation and the tradition right?
      • That if you are of the school that uses the fu'ivla, you should use a type one. If you prefer to use the local language term, you should use it. BTW, the name Go is actually not really borrowed from anywhere, so for lojban an analogy would be my suggested tanru.
  • BTW, would would your fu'ivla for go be? kelcrgo or kelcrigo, with a longer prefix than actual loan? Ick. kelcruetci or kelcrputaki, which would not be understood by all go players, and not by any who don't play?
    • I won't argue about the details of how to say go, since I don't play. Yes, the prefix will be longer than the root in some cases. So? It's still not long. Note also that all roots in a type 3 fu'ivla must begin in a consonant.
      • kelcrgo: Go, English. kelcrigo (which would, I suppose, have to be something else if the root has to begin in a consonant. Hmmm...): Igo, Japanese. kelcruetci (which would have to be kelcrxuetci): Weiqi, Mandarin. kelcrputaki: Budak (pronounced and formerly transliterated as putak), Korean. Kelcrgo is particularly repulsive, as in my pronounciation at least it necessitates either a syllabic r or a buffer vowel, meaning that the root is half as meny syllables as the prefix.
        • .i la lojban na zasti mu'i leka mansa le do sidbo pe leka melbi .i xu do ji'a na pilno su'o gismu mu'i lenu do na se pluka le ri tarmi .i na ka'e go'i gi'e fuzme pilno le bangu
  • xod:
    go is another example of something that needs a fu'ivla. Go is not every strategy game with black and white pieces. It is a particular game with precise rules and again, a specific history duplicated by no other. ALL of this needs to be communicated! That's what names are for; the word "go" is not a long, arduous description. In order to explain and describe go suitably, you would need another monstrous tanru with at least 50 gismu and lots of be places filled. Somebody is really missing the point. And why are you using morn- for a game instead of kelc-?
  • .kreig.daniyl.:
    mornrcitco refers to a specific pattern which appears in the game of go. Also, as I mentioned in my discussion of pemcrxaiku, there is a difference between naming something and defining it.
  • nitcion:
    .i mi so'ore'u je ca tugni la xod zo'o
  • This reminds me of someone ages ago on the conlang mailing list, who was building a language in which all the words were constructed out of some small set of simple forms. I remember he had grape as "small round green food." Or maybe that was pea. And that is exactly the problem! For any description you come up with, you'll find it is woefully vague and ambiguous. "Strategy game played with black and white pieces"? That's chess, right? Oh, a Japanese one? Chinese, actually, but then again so is chess.(Chess is actually Indian--elephants and all; Chinese chess is a modified borrowing, rather late by the modifications) But there are others I could probably find, and so on. It's a Zen sort of problem: you can't call a short stick "a short stick" because it's also so many other things (a dog toy, whittler's raw material, half a set of chopsticks) and because so many other "short sticks" are out there that aren't this one and aren't even much like it... and are still "short sticks." And you can't not call it a short stick, because, well, it is, after all, a short stick. There's a koan about that, I believe. This is what fu'ivla are for. There is no Lojban phrase that means "the game of go" with any decent degree of specificity and accuracy, that's less than several heavy sentences long. There just isn't one. If your interlocutor doesn't know beans about go, maybe you're better off using one of those long explanations. But chances are he'll have heard of the game, even if he doesn't share a language in common with you. That's what we use cmene and fu'ivla for. We regularly assume/hope that our audience will be able to match our cmene with its real-world referent (not a name, but the actual thing to which it refers). Fu'ivla are much the same, except that they're brivla, with perhaps more elaborate place-structures than mela CMENE constructions. Whether you prefer type-3 or type-4 fu'ivla is another matter (I'd stick with type-3s if you're hoping for this kind of recognition, and only use type-4s if you were damned sure of your audience, that they'd catch it without the hinting rafsi, or for cei-binding). --mi'e mark.
  • cei is how I think it ought to be done. It occurs to me that my experimental gismu proposal (Using all unassigned gismu as if they were broda, I'm sure I'm not the only one whose thought of it but I haven't seen anyone else suggest it yet) is actually not a baseline violation, as the Book does not say what should happen to those valsi be noda. So, I could say Oriental stone-putting strategy game (I happen to find long tanru to be rather more lobykai than lujvo or poo'ivla) then append a "cei putka" and it does not matter whether my interlocutor recognizes the name Budak (which has been mangled to the point of oblivion to get it into CVCCV form, but unlike a type 4 it doesn't have to sound right) because I have just explained what it refers to. If they wonder whether it means go or Hasami Shogi, they can simply ask something like le putka mo and get an explanation. - .kreig.daniyl.
  • Okay, let me explain myself again, with non-Japanese based examples this time. Suppose I wish to express the concept of Jeans in lojban.
    1. I can say it with a type 4. This maintains all the ideas people have about Jeans. However, they have to recognize the word, and it is not particularly lobykai - if we want to speak English, why are we using lojban? I wonder how far one could take that opinion. Remember that all gismu come from the six source languages. If we see fit to borrow 1350 words, why not 1351? fu'ivla are almost always nouns that are very specific concepts for which borrowing is completely justifiable. What, you mean like to refer to taxons that could be described by gismu they fall under? Besides, if you can justify one fu'ivla, you can justify two. If you justify two, this justifies another two. And so on, until you start saying Well we "borrowed" 1350 for the gismu, if 1350 is ok why not 11350? and other bullkalci like that. And are the gismu really borrowed? I don't think so, anymore than english borrowed its entire lexicon from PIE.
    2. I could instead use a tanru. This way, people can figure out what I mean. While it does not carry any baggage with it from the connnotations of the word Jeans, the ideas are in the mind of the listener. I therefore do not have to use 500 gismu, as I don't need to tell my interlocutor all my views on wearing comfortable clothing just to mention that I did so yesterday.
    3. I could, using my experimental gismu proposal, explain exactly what jeans are - but without any of the ideological baggage - then add cei jinzi and use jinzi to mean jeans for the rest of the conversation. Ideally, while doing this I would remember that jinzi already means something and use something else, like djizi.
    4. I could do it instead without violating the baseline by using cei broda instead.
    • .kreig.daniyl.:
      All but the first are nice and lobykai. Why do you need fu'ivla?
  • How do you say that when type fours are in the Book but nonce gismu are not? Of course the proper answer to your question is a type three!
    • As I pointed out above, I have realized that using free gismu space in any way DOES NOT violate the baseline, as the book does not say what should happen OR NOT HAPPEN with them. As for what kind of a fu'ivla to use, I feel that a type 3 is MUCH better than a type 4 - the lower the number the better, IMNSHO.
        • Then consider experimental gismu a "type 5 fu'ivla", because I assume that one should only arise if it is shown to be used so often that it needs a CCVCV or CVCCV form.