great Dictionary Problem

From Lojban
Jump to: navigation, search

The baseline ends five years after the dictionary gets published. Does that provide incentive never to start the dictionary? And there are some who think the idea of a Lojban Dictionary is lame.

  • rab.spir:
    • It provides incentive for me to help with the dictionary. When it gets published, I'm going to mark my calendar and celebrate "Baseline Day" for the next 5 years. I understand the need for the baseline now, but I will feel so much more free when it's over.
      • Jay:
        • Hrm. Thought the textbook was also required to start the baseline period?
          • nitcion.:
            • Yup: dictionary, and textbook (by which Lojbab still has in mind his textbook, rather than Robin and my lessons; at the minimum, the baseline textbook would have to cover a lot more of the language than our lessons anyway.) The community reserves the right to add to such baseline material. I have mentioned it would be a good thing to have a Lojban And Logic text that did at least the logic stuff more formally than the refgramm does, and Lojbab in response has suggested to pc that he draft something; obviously this is at a very informal level still, and given recent sentiment, I have my doubts it would be included in a baseline.
              • lojbab:
                • Bob LeChevalier still has in mind his textbook, but that could change, and his mind could be overridden by the voting members choosing to adopt what is written as a baseline in order to say we are done. This won't stop me from intending to write my own textbook, since I know that there are people who find what I was trying to do in keeping with what they were seeking. If I could write ad hoc texts and exercises, there might be a textbook already.
  • To people who think that: What about the idea of the dictionary being written in Lojban? Would that kind of dictionary be lame?
    • nitcion:
      • I don't see why a naturalist would find prescription any more congenial if it's written in Lojban metalanguage rather than English metalanguage. Speaking as a hardliner, of course.
      • rab.spir:
        • One thing I'm curious about - would a dictionary of Lojban in Lojban suffer from the same problem of an English dictionary that words are defined cyclically? Would it theoretically, if not practically, be possible to define cmavo in terms of mekso (and previously-defined cmavo), and gismu in terms of cmavo (and previously-defined gismu), if the order was chosen well?
        • Now that I think about this a bit, this would basically be the code for an AI which understands Lojban encoded into mekso - not something that could very well be considered a dictionary anymore.
    • Lojbab:
      • People would be impressed if we could write a dictionary of Lojban in Lojban, but it would not serve the other purposes of a dictionary: to support new language learners who want a handy reference for the language lexicon, to document for outsiders what the language is like, or to serve as part of LLG's product suite that supports the organization financially.
  • Oh, there's more:
  • Should a Dictionary recommend one cmavo usage, when usage has gone one or two ways? Or does it only have descriptive, not prescriptive force? And if it only has descriptive force, then why the hell are we waiting for it in order for the baseline to even begin? And if it's prescriptive, who does the prescribing without every other Lojbanist yelling "Cabal"? (nitcion, speaking as one who's already been yelled at. If it's Elephanted, I'll grant, at least each individual Cabal gets equal time...)
    • xod:
      • If Elephanted at large, the Dictionary will take 10 years - am I really exaggerating? But if done by a Cabal, it will arouse the ire of those outside that Cabal, and take much, much time of those Cabal members so applied, and likely burn them out.
      • The construction of the dictionary includes deciding which of a set of conflicting tanru interpretations gets elected to be the lujvo definition, doesn't it? And the small corpus we have is not big enough to have given competing interpretations and usages sufficient exposure.
        • Jay:
          • Yes, so for concepts which a dictionary really ought to have, somebody or some group has to pick the best one. If it sees a lot of contrary usage later on, well, then the next edition of the dictionary gets corrected.
          • .kreig.daniyl.:
            • No. The dictionary can give two definitions. Dictionaries of natural languages try quite hard to be descriptionist - if both senses are used, the dictionary should list both, with the more common sense first. My views are the same on cmavo - both senses of fa'a should be listed, with directional first as it is supported by more usage.
              • nitcion:
                • Noone says the dictionary has to be descriptive. The understanding in the Dictionary being a baseline item is in fact that it is prescriptive. But I'm growing tired of defending hardlinerism and all. If you want that kind of dictionary, that's the dictionary you'll get.
                  • lojbab:
                    • Believe it or not Nick, I understand what the hardliners want. I also understand that the battle between descriptivism and prescriptivism in dictionaries has gone on for a couple hundred years, and the descriptivists seem to always win in writing the book, only to have their description treated as a cast-in-concrete prescription. Therefore I favor a descriptivist dictionary (with suggestions as to what the prescriptions are by careful annotations in order to meet the inevitable fate in a constructive manner), a prescriptivist reference grammar and a primarily prescriptivist textbook. The prescriptions I will tend to favor in print will probably be somewhere in between my public positions and those of the hardliners, and those who think that usage has decided things otherwise from what I intended (even though I feel that there is too little usage to claim any such thing).
  • la xod:
    • I think people are afraid that some bad metaphysics will get encoded into the dictionary by an unfortunate success of a misguided cabal. Then the views of a small handful of people able to outnumber, outargue, or outshout another handful, at such an early stage in Lojban's development (we hope), could confine the people of the future to a bad usage with the shibboleth of Officialdom. This is my fear. I am not sure I still want a dictionary. I would like a clear record of our opinions and our alleged agreements, but I no longer think official decisions should be issued at this time of primitive development. As to those, like And Rosta, who think the language is only 25% complete, they should develop their views and weight them with the force of usage in a body of work.
      • ... which will likely be ignored, anyway. But I give up, xod. Any notion of a dictionary as completing the definition of Lojban, you are implacably opposed to. Reading between the lines, the same is apparently true for most Lojbanists. I'm sick of spitting into the wind, and I'm sick of being accused of masterminding a cabal. If this dictionary is not to baseline anything in the language, and if people seriously think (beyond my understanding) that the language truly is complete, then I see no reason to include a dictionary in the baseline.
        • nitcion.:
          • Or for me to contribute to such a dictionary.
  • If there is no dictionary, then those who wish to learn the language, especially those who attempt to learn outside of the Lojban mailing list/wiki communities (which seems to be the vast majority), will use the printed gismu and cmavo lists as their basis for learning, with all of its failings. The "cabal" will be Lojbab of 1990 or whenever I wrote most of those 100 character abbreviated definitions for use in LogFlash. I think you would prefer what I (or Nick) or anyone else would come up with, to having the bulk of the community learn Lojban as it is in the inadequate-for-the-present-and-even-more-for-the-future word lists. I also hope that you want more people willing to commit to serious learn the language, and large numbers will not do so until we can hand them something labelled "dictionary".
    • lojbab:
      • You should not be concerned about the potential power of a cabal, because most Lojbanists, like you, And, Nick, and many others, once they acquire a basic knowledge of the language, will make it their own, and will do with it what they will and not as the cabal might dictate. Americans in particular (and I think the sort of people who would learn Lojban worldwide) are not the sort to bow to authority. (One reason I am not a hardliner is because I believe that this would cause a hardline prescription, which necessarily would be academically technical, and as a result all usage will fall to the lazy and sloppy side of the prescription, rendering the prescription moot. As it is, a hardliner can be stand up for truth, justice and logical correctness against the evil naturalist cabal, and motivate some to like-minded good behavior. Meanwhile the learner can aspire to meet some minimal standard of good Lojban that, while not logically perfect, is communicative and satisfying.
  • xod:
    • I am not implacably opposed, simply worried and unconvinced. Lojbab claims he wants a dictionary, but also seems unwilling to conclude any decisions concerning ka and ce'u. He says this, which I cannot argue with, and to which I repeatedly find myself returning: "The less that we can get away with prescribing about semantics, the more suitable Lojban will be for a Sapir-Whorf test, because the way of thinking in the language will (perhaps) be determined by the language, and not by academics arguing about the language in English."
    • I am continually tempted to debate the semantics of Lojban in English, with the two excuses that I cannot use what I don't understand, and the fact that my attempts to converse about these things in Lojban with you (Nick) and Jorge always end up awry after several exchanges. But the synthesis is that these discussions serve to illuminate issues for ME (or the other participants), but the results, however elegant, should not be imposed by us upon others with the imprimatur of officialdom. And that's what an official declarative dictionary would do.
    • Therefore, let's debate, and let's use. And let's let le ba jbopre .a'o do likewise.
    • lojbab:
      • You are missing out on the fact that 90% of the list isn't debating, and a lot of them aren't even reading the debates because there are so many of them. And a few like me spend so much time reading what we do of the debates that we have no time or energy left to use. Give people simple answers without debates, and more of them will advance into userdom. Then you can debate them in Lojban.
  • (I'm glad I toned my original response down; that's why we're still able to talk.) I don't those are excuses, I think they're inevitable, and welcome pragmatism. You cannot make grand claims about si'o being what the x1 of sidbo is, to take one recent example, unless you've defined sidbo in English first - otherwise, you have a humiliating circular definition. It's very easy to do this quasi-mystical appeal to "thinking in Lojban"; but without a formal specification, it is well-nigh impossible to avoid the thinking in Lojban merely being relexified thinking in English. And different peoples' different Englishes, at that. Hence, the debates are far from illegitimate. I'll go further: if someone ignores them, then I have no reason to assume that their Lojban is not just English in Lojban words.
    • xod:
      • Interestingly, I think the opposite: If somebody gets all their Lojban wisdom from English debates, they are undoubtedly using interpretations that fit easily into English. How do you respond to Lojbab's comment concerning Sapir-Whorf vs. English debates?
  • nitcion:
    • Just because something is reasoned about in English does not mean something fits grammatically easily in English. The conflation of language and metalanguage this presupposes is bogus. ka was argued about in English, and I don't see ka as being particularly malglico in the majority view. (In fact, it's Lojbab's ka, as a mapping of -ness, that I see as malglico.) I respond that Lojbab's comment is simplistic, and is in a rush to idealise Lojban-internal thought, without making sure first that the basics of Lojban are immune from English.
      • I find that the best way to immunize my thinking from English is to free myself from English words and semantics. When we argue in English, I argue for my ka. But if my ka is not properly Lojbanic, then I won't be able to make it work when writing in the language, and I will be forced to adopt some if not all of the thinking of others in order to make myself understood. I will undoubtedly use ce'u a lot in my ka constructions when I use the language, in the way that you argue for, assuming that it makes as much sense as you say, and I will do so because you are providing the examples that I will learn from and respond to, and I will have to think your way in order to understand you, and hence will tend to think your way when I write to you. I suspect that is what the language acquisition people say is the way people learn languages best, by patterning after successful in-language examples.
        • lojbab:
          • But when you argue in English, I think in English, tie concepts back to English words, and I think about history and my design efforts that in many cases have gone for a dozen years unwritten, and I'll be the first to admit that this can only result in sloppiness.
  • The results of the interminable debates, I would further contend, are likely to be the only game in town for many of those issues (and I include ka). And inasmuch as this is a language with a history (something much too easy to lose sight of), and usage is guided by those debates, the decisions of the debates will get cemented, and will take a lot of work to get uncemented. So in many instances, description and prescription will converge anyway.
  • So this business of stripping the dictionary of any prescriptive force will not make the dictionary worthless (as I'd initially fulminated.) But, that said, this represents a real policy change (inasmuch as Lojban Central ever had a policy), that must be debated, and that I'm concerned about - a non-prescriptive dictionary is pointless to hold the baseline up for;
    • lojbab:
      • None of the baseline is being written so much to be "prescriptive" as it is to be "descriptive of the design as of the start of the baseline period". There are no language police that will arrest people for using an experimental cmavo, a Resurrected gismu, or a si'o that looks like something other than Lojbab had in mind. The need for the documents is to clearly record where we stand insofar as it is possible, so that people will have a starting point in learning the language - and hence will start to learn the language in greater numbers, and outsiders that we want to gain respect from will see the appearance of a complete effort on first look (as opposed to the 25% that And sees after years of arguing).
        • It's a policy I do not yet embrace: I still think there is such a thing as a correct and incorrect way to use gismu, and most if not all cmavo, and the dictionary is the place to say it (never mind cukta; consider senvi again); however tiptoe-on-eggshell-ishly the dictionary is authored, it's not going to stop those yelling "Cabal!" (Something, I am obliged to say, I find offensive in the extreme).
  • Keeping peace between naturalists and hardliners in any of this is not going to be easy, because ultimately, naturalists and hardliners do not want the language to go in the same direction. It may yet prove impossible to keep that peace. So you can count me, too, as worried and unconvinced.
  • I've been apprised by Robert McIvor that TLI has had some of the same sorts of issues we have had. I think we can reasonably agree that TLI Loglan as described in Loglan 1 is no more logically and semantically rigorous than my undocumented version of Lojban is, even though JCB sometimes seems to talk as if he was trying to be. They have had a formal Academy, and a tendency towards trying to suggest more formality than was actually present. As such, it has intimidated people, many of whom have shied away from learning their language (to the benefit of Lojban). So in recent years before JCB's death, a movement started which continues for 'Easy Loglan', which is a quickly learned malglico version of Loglan that someone can supposedly pick up in a weekend and not get criticized for making errors.
    • I find the concept of Easy Loglan or for that matter Easy Lojban patently absurd. Such a thing would be what I fear Anglan would become if it were ever created: relexed English with a few cute Loglanisms thrown in. I think that the basic form of Lojban that is taught should be moderately rigorous in scope, but not too far beyond the typical logico-linguistic education level of its learners, and also that speakers should be as tolerant as possible of others' errors. I think that there would be no problem with such a language coexisting with a hardliner academic dialect that (And Rosta) would prefer, with the latter being the seed for what JCB called Loglan Mark II, the more logical version of the language that could only be developed by people who have mastered Loglan, and seen where it needed improving to be more the ideal "logical language".
    • lojbab:
      • But if the main or ideal language requires too much sophistication in order to become respectable, people will shy away from it in favor of Klingon which is merely challenging to the language-learning novice, whereas hardline Lojban is so challenging that its most skilled speakers like Nick Nicholas think it too difficult to contemplate translating Hamlet into. We don't need Easy Lojban, but we do need something that people think that they can learn and eventually master, at the standard that is nominally prime.
  • nitcion:
    • I'd rather a Lojban impossible to translate Hamlet into than one easy to translate Hamlet into.Let me illustrate with an anecdote of Mark's. At a past Logfest, a children's tale is to be translated into Lojban. A goose, ko'a, lays an egg and walks away. The egg is hatched, and the gosling starts looking for its mother. The knee-jerk reaction is to say that it is looking for ko'a. Not so, points out Mark: if the gosling saw ko'a all of a sudden, it wouldn't necessary know it's found who it was looking for. It's looking for X, such that X is its mother. The search is intensional, not extensional: not sisku ko'a, but sisku leka ce'u mamta vo'a. (That's why the place structure of sisku is like that in the first place.) Do you see why Lojban is hard? You have to keep stuff in mind like this all the time. Every second sentence. If you're unlucky, every single sentence. You don't in Klingon. Or English. And no, we don't need Easy Lojban. In fact, anyone who advocates Easy Lojban, I have no hesitation in regarding as selling out Lojban. And I have no problem in holding people learning Lojban to a higher standard than those learning Klingon. I shall not be dissuaded that the ability to express yourself clearly, and without trivial metaphysical confusion, is a core point of Lojban. I shall not be dissuaded that any development of Lojban which allows you to say sisku ko'a in the example above is inimical to that point, and is something I am thus obliged to militate against.Normally, I agree that there's a place for a Formal Lojban and a Colloquial Lojban. But if Colloquial Lojban is to allow things like sisku ko'a, it truly defeats the point of Lojban.
      • Not that it matters much, but some filler on the anecdote... The story was already translated: John showed me a cute book with the drawings from the original and the new Lojban text. It was Are You My Mother? by P. D. Eastman, and in Lojban xu do mamta mi. My first reaction was "Actually, it would be better as do xu mamta mi" (and even now I tend to forget it was actually done the other way 'round). Anyway, we're introduced to the mother bird, who is bound to ko'a, and the egg hatches, and sure enough, the baby bird co'a sisku ko'a. And that bothered me; it sounded wrong. He wasn't looking for ko'a, he was looking for his mother! I don't think I quite had the terms "intensional" and "extensional" at the time, but that's what it was. We spent until something like 2am discussing it... all from the simplest of children's books! Yes, everything in Lojban requires thought.
  • nitcion:
    • How the new Lojbab missive fits in, I cannot yet fathom. I don't think it ultimately puts on the table anything we didn't already know would have to happen. But with things as they are, I am not optimistic a dictionary can be written at all.
  • lojbab:
    • If people produce the examples, there will be a dictionary. It may not meet your ideal, nor mine. But it will get done. (We would like a better English to Lojban dictionary than the KWIC-based one that serves as the current draft, but realistically that is what we will have to go with, and it will be better than nothing. A cmavo dictionary with examples will be an order of magnitude better than incorporating the as-is cmavo list and calling that the "cmavo dictionary", and that is the current alternative.
  • Even if my own usages are usually in line with the majority of the opinionmakers, I am not sure that even my own ideas are jbokai enough to be enshrined as Policy that constrains the future. I would rather allow competing theories and approaches to flourish, and hope that the better one survives in the marketplace of ideas.
  • xod:
    • To treat a language as nothing but a bunch of EBNF statements is ridiculous! This is the hardliner attitude! And such an opinion has no place in any project that is looking for any Sapir-Whorf effects, which are subtle not explicit, experiential not logical/mathematical, and are by definition difficult to discuss in a different language. (these English debates!)
  • nitcion:
    • First, thank God you are well.
    • Second, you're mischaracterising the hardliner position grievously: there is nothing mystical or ineffable, and thus unformalisable, about the semantics of ca. But this is arguing over phantoms. If I am still involved in Lojban and have time, I will essay formal definitions where I think they will not do damage. Any such definitions must be accepted or rejected case-by-case, and with the understanding that they do not supplant other modes of explication.
    • Last, we both know what you hope to get out of Lojban, and what I hope to get out of it, and that they are incompatible. What has no place or hasn't in Lojban is a subjective criterion; so what "has any place" in Lojban is not something we're going to agree on. If you don't want anything in the dictionary to be enshrined as official, I certainly am entitled also to not want an anti-formalist approach enshrined as official.
    • Since (once again) this disagreement isn't going anywhere, let's drop this, anyway.
      • Lojbab:
        • Competing theories will persist no matter how firmly prescriptivist the baseliners (and I am hardly prescriptivist). Look how ineffectual my "prescription" has been at constraining the usage of xorxes, (And Rosta), Michael Helsem, or for that matter nitcion. Whatever I would write as "official", it will be ignored by some, and perhaps many. This doesn't bother me, and because it doesn't bother me, it should not be a threat to the language and community (as it has been for many other languages whose inventor was possessively prescriptivist); I intend to take all steps necessary to defuse schismatic forces by having a prescription for those who want one, and a "big tent" for dissenters who do not want one. Hardliners will be welcome, especially if they argue in Lojban, and naturalists will be able to use the language without any official criticism of the fact that they may be doing it in a way other than lojbab wanted. The result will probably be somewhere in between, where xorxes, xod, nitcion, and and are presently able to write in Lojban and (usually) be understood by the others.