From arj@nvg.org Mon Oct 18 15:43:47 2004 Received: with ECARTIS (v1.0.0; list lojban-list); Mon, 18 Oct 2004 15:43:47 -0700 (PDT) Received: from sabre-wulf.nvg.ntnu.no ([129.241.210.67] ident=[7f938Umv6c0A3Lab5ITqY2bwZtToLuBb]) by chain.digitalkingdom.org with esmtp (Exim 4.34) id 1CJgDk-0002XC-NW for lojban-list@lojban.org; Mon, 18 Oct 2004 15:43:17 -0700 Received: from hagbart.nvg.ntnu.no ([129.241.210.68]:1690 "EHLO hagbart.nvg.ntnu.no" ident: "NO-IDENT-SERVICE[2]" whoson: "gunnarre") by sabre-wulf.nvg.ntnu.no with ESMTP id ; Mon, 18 Oct 2004 23:28:56 +0200 Date: Mon, 18 Oct 2004 23:30:06 +0200 (CEST) From: Arnt Richard Johansen X-X-Sender: arj@hagbart.nvg.ntnu.no To: lojban-list@lojban.org cc: hedybos@hotmail.com Subject: [lojban] Re: [hedybos@hotmail.com: Feedback (long, sorry)))] In-Reply-To: <20041018165014.GI22470@chain.digitalkingdom.org> Message-ID: References: <20041018165014.GI22470@chain.digitalkingdom.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed X-MailScanner-Information: Please contact the ISP for more information X-MailScanner: Found to be clean X-archive-position: 8752 X-ecartis-version: Ecartis v1.0.0 Sender: lojban-list-bounce@lojban.org Errors-to: lojban-list-bounce@lojban.org X-original-sender: arj@nvg.org Precedence: bulk Reply-to: lojban-list@lojban.org X-list: lojban-list On Mon, 18 Oct 2004, Robin Lee Powell wrote: > I am not anywhere near enough of a linguist to handle this. > > -Robin > > ----- Forwarded message from Hedy Bos ----- > [snip header] > > Dear Sir(s), > > I have the following feedback on Lojban: > Thank you very much for your insightful message. I will reply to your points in turn. But first a general remark: In most of the cases where we made Lojban different from all known natural languages, it was deliberate. This could make it easier to test if the universal in question actually is due to how the human language faculty works, or if it is simply a statistical fluke that we have not yet observed any such languages. > -The denotation is hell. A period expressing a phoneme? Why? And what?s with > the phobia for the symbol ?h?? If your defence is ?it is used only to > separate vowels?, then there still would be no reason why the symbol ?h? > could n?t be used. Another reason is that after the introduction of the h phoneme, it became very frequent, and it was changed to ' because it was more visually lightweight. That's not the best of reasons in my personal opinion, but there you have it. > The first basic rule in lexicography is ?be consistent?. If you choose to go > roman, go roman all the way! Don?t use diacretic (and illogical) symbols > that will chase beginners away instantaneously! It?s pretty much illegible > like this. The apostrophe I can live with, but don?t incorporate something > like a period in the system. Many languages express a normal phoneme such as > a glottal stop with an apostrophe (or an alif). Use an ?h? for all I care. > Don?t use a dot. You can drop the period entirely, and not be misunderstood, so long as you keep all the words separated (by spaces). In fact most of us do, when writing informally (such as on IRC). I am surprised that you didn't mention Lojban's system of obligatory pauses, which is very unnatural indeed. This is the precise reason that we use periods in texts aimed at beginners: to remind them where the language's self-segregation rules demands a pause. > -it?s either under- or overspecified. If it really is such a logical > language then let logic do it?s job and dare to make a choice for either > system. Especially the deictic/locative marking is hopelessly overspecified > in an innatural manner. Languages naturally have either an unspecified form > (mi klama) or a specified (mi su klama) form. Nobody needs an superspecified > form. If it does, it will come naturally, when it?s not a pidgin anymore. > Indonesian manages well (with ?saya pergi? and ?saya sudah pergi? as the > two only options) I don?t think Lojban needs to specify everthing in such a > frantic way. It?s logical, people can think for themselves. > Common people have no use for such specifications as mentioned in the > section about deictics. They will not use such forms. It could be useful as > a jagon maybe, but then I still think that people will fill in the gaps in a > language for themselves. For example: people living in mountainous areas > tend to modify their deictic system to be able to express referrals to > objects according to their relative position concerning height (?that goat > above me? ?that house on the same level as me?). These things come > naturally. No pilot/mountaineer will spontaneously use a form as predicted > in your grammar in a situation where it is needed (?enemy at twelve o?clock > above...?), nor will he look up in a grammar what form he should have used. > If the situation he would like to describe keeps occuring, he will come up > with a ?patch? himself. The only chance that correct Lojban will be used, > would be when pilots and mountain people learn the propre terms in avance. > The greengrocer won?t. He doesn?t care. He does not need this kind of > specifications. If people like him would keep ending up in a situation > requiring the specifications they will come up with them. I predict it will > be something else then what you produced. Let this go. Or create > jargon-vocabularies for every professional group only. Still you will be > left with inconsistencies I?m afraid. I don't understand this objection. Every speaker can use as little or as much of the tense/locative system as he or she wants. > -the assumption that natural languages are inadequate aggravates me. I am > convinced that if something does not occur in a natural language this must > be for a (simple) reason: it doesn?t work! You're probably right. I can think of at least a few parts of the grammar of Lojban that are likely to be unstable in native language learners.[1] But maybe some features of Lojban don't exist in the wild simply because they don't? > So the following irritations come > forth out of this opposition: > > -why so many arguments for one stem? Why not express these with > prepositions? This would be more logical and would cause the number of > illogical arguments one would have to learn in order to be able to use a > stem to it?s fullest extense to decrease tremendously. There is no language > that would refuse to express prepositional phrases in such manner. Because > this is more logical ?nd it decreases complicated lines of thoughts when > using a stem. For example: > > barja bar x1 is a tavern/bar/pub ?serving x2 to > audience/patrons x3? > The x2 and x3 can be expressed more logically with some basic prepositions > in my opinion, which would save everybody a lot of puzzling. Lojban does not "refuse to express prepositional phrases". Lojban does have prepositions, but they are (for historical reasons) called "modals".[2] The position of the makers of Lojban, however, was that prepositions are inherently underspecific. When you go "to" Boston, or you send mail "to" Boston, the relationship between yourself and Boston is only superficially similar. Natural languages cope very well with this underspecification, and so does Lojban. But if you want to be very exact, you can do so in Lojban, with only a marginal increase in prolixity. > -why use separate stems for compounding? It does make the entire utterance > shorter, but it also increases the number of forms to be learned. Even more > so, there is no unambiguous way to derive the compound-stem from the > original stem! Why not choose for a simple head-dependent construction? It > will increase the entire compound with one syllable only. Most languages > work just fine with this system. Some do use a possessive marker, such as > Roman languages, or a clitic such as Bantu-languages. The only languages > (consistently) changing a stem are root-based languages with a skeletal > character (Afro-Asiatic) . Either switch to that (for the sake of logic > and/or consistency) or accept that a compound will end up with four > syllables. Never hurt nobody. This system is indeed difficult, but many (me included) have managed to learn a substantial part of it -- not by rote, but by exposure. The intention of this rampant allomorphism is to keep lexicalised and non-lexicalised compounds grammatically separate. To take an example, "minra sance" translates roughly to "mirror sound", and the relation between 'mirror' and 'sound' is left deliberately unspecified, where as the corresponding lexicalised compound, "mirsna", we now agree means "echo". > Nevertheless I will check out these pages regularly. I do think you are on > to something. Nice! You may also want to join the Lojban mailing list; this can be done at http://www.lojban.org/lsg2/ . Volume is mostly less than 5, but sometimes up to 20 messages in one day. > By the way: how does one derive causative forms, stative forms etc.? Are > they derived from the same root or do they have separate forms? I think the > first option would be more logical. Lojban does not have derivation per se, only compounding. Hence, compounding with -gau (gasnu; 'bring about') as the final stem will in many cases correspond to causatives. (But not always; as I implied earlier, the semantics of affix compounds have no prescriptive grammar.) As regards stative forms, if you assume that brivla are verbs, some of these could be called stative, since they predicate nouny things. For instance, in "mi nanmu", 'I am a man/we are men', "nanmu" is a word that for most intents and purposes acts like a verb, but means "is man". --- [1] For a discussion of how a part of Lojban grammar is used differently to the designers' expectations, see Nicholas, Nick 2002. Folk Functionalism in Planned Language: The Long-Distance Reflexive vo'a in Lojban. Journal of Universal Language. 3:1. 133-167. PDF at: http://www.unish.org/unish/DOWN/PDF/Nick_Nicholas(133%7E167).pdf [2] http://www.lojban.org/publications/reference_grammar/chapter9.html -- Arnt Richard Johansen http://arj.nvg.org/ Let's have some real examples from a real, non-English language.