From LOJBAN@CUVMB.BITNET Sat Mar 6 22:46:26 2010 Reply-To: Logical Language Group Sender: Lojban list Date: Tue Dec 12 16:39:08 1995 From: Logical Language Group Subject: Re: TECH: PROPOSED GRAMMAR CHANGE X6: Simplification of compound tenses X-To: jorge@PHYAST.PITT.EDU X-cc: lojban@cuvmb.cc.columbia.edu To: John Cowan Status: OR X-From-Space-Date: Tue Dec 12 16:39:08 1995 X-From-Space-Address: LOJBAN%CUVMB.BITNET@UBVM.CC.BUFFALO.EDU Message-ID: >> Well, I;ll start by noting that half of your discussion deals with using >> tenses inside selbri (NOTE 2) which is NOT part of your X6 change. > >Yes, I explicitly left that for another day, because that would indeed >constitute a change, as opposed to a simple extension. OK, It was mystifying seeing a long rationale for a different change in themidst of a semi-formal X6 proposal. >> Second, you have mostly been making proposals that deal with things in the >> preparser/lexer, and these changes are VERY difficult to check out with >> any certainty. The risk factor is high, even if we were sure the change is >> virtuous. We cannot afford high risk changes at this point. > >This change is very simple to check, as I'm sure John will agree. It >essentially gives ZAhO, ZEhA and TAhE the same grammar, except for >the NAI and PU attachments. The change is even easier to check for >preparsability than the current state. I would even go for blending >ZAhO and TAhE, since I don't really see why ZAhO NAI is not allowed. It does still have to YACC, and having units that have the same grammar except for minor differences in pre- and post- attachments is harder toget YACC to accept than simoply giving them all the sa,e grammar. >> Third, this is now the 6th proposal you have made in less tha a month, when >> we managed to go almost 2 years with only a couple of proposals. > >There is nothing new about any of my proposals. I mentioned all of them >(except perhaps X5, which is incompatible with some of the others) to >John when I was reviewing the grammar papers, and I'm sure I've discussed >some in the list in the last couple of years. I've certainly used often >the {ba'ibo}s. Rferences to your earlier comment would be useful. We still have them, and I think you argued more effectively in your comments %^) >The reason that I made the formal proposals now and not before is that >I was waiting for the publication of the dictionary. Up until very >recently, the dictionary was going to be the first publication. That's >what I understood was decided during Logfest. No it was at LogFest that we switched and made the refgrammar highest priority. But it was not something that was going to involve me that personally until Cowan declared the thing done. Then we got to the point where Cowan was still not making progress and I wasn't ewither, so we agreed informally that I would keep working on getting orders out and otherwise shift my activities to enhance more immediate publication. This was aided by the commitment of a third party to commit significant volunteer effort if we kept to schedule on publsihing the refgrammar, to make sure we geta quality publication soonest. >> If indeed >> the language was in such bad shape that it needed 6 change proposals in a >> month, then I would again retreat from having a baseline, and forego yet >> again the idea of getting a stable language. > >Who says the language is in bad shape? All the proposals (except X5) are >only extensions. They don't let you say anything new, they only let you >say more easily things that can already be said in a more roundabout way. >If someone had started learning the language three years ago (as I did), >they would require zero relearning with this changes. They shouldn't even >be called changes, as they are only almost self-evident extensions. Because they are labelled "changes", they are seen as such. And of course this is the sort of thing that has to be done to focus attention on preventing change in baseline configuration control. Spolighting even minor changes as CHANGES, means that nothing creeps in unnoticed. The problem is that there have been all manner of chnages and extensions proposed over the last couple of years - indeed, Lojban List is largely dominated by discussions of proposed changes. We can imagine the impression this gives outsiders, when all we can tyalk about is what changes we are thinking about. But from my standpoint, I have to look at all the other changes that were discussed in the last year and wonder what will happen when you get to X75 and catch up tot he current day, if you are still working on changes that you proposed 2 years ago. >John has proposed thirty something similar changes, most of which were >implemented. And almost all of them came specifically out oif his work writing the refgrammar. And they were only considered because it made it easier to write. John's significant volunteer effort in putting that book together gave him a little added clout to get things considered. Even so, 22 of the xchages occurred in the first year after the baseline, 11 in the second year, and we had only 4 or 5 change proposals in the 2+ years since we declared a rebaselining, most of which have remained unadopted (the official parser is still 2.33, and a good thing too, since a parser bug seems to have crept in a later version). BTW, on extension vs change. Yes we are more permissive of extensions, but they still have been justified in general by saying thats omething COULD NOT be said, and not that it simply had to be said in a more roundabout way. >I understand that you will trust John's judgement of what >is necessary more than mine It isn;t that - I have no choice because he is writing the book and mintaining the parser. I HAVE to trust him. Luckily he has proven worthy of all of the trust that the community has placed in him. >They are >not so outlandish as you seem to think. They do not affect at all anybody >who has learnt the language as it is. John and I have found, and presume that others find it as well, that the image of instability affects those of us working on learning or using the language FAR MORE than the changes themselves. The feeling that I have to pay attentions to a long-winded inclonclusive debate like we spent a year on dealing with lo/le/"any" issues. As of yet,not one single proposal has emerged to deal with the issues raised - at least not any formal proposals. Do I really think that no proposals will come ou tof it - no, there were clearly one or two ideas that withstood debate. But I don't know what they are. And because of that, I don't feel comfortable attempting to express things in the language that involve an "any" concept. Yousee, *I* amd one of those who cannot learn the language while it is changing, and I no longer am sure what the language "is". If *I* have this problem as leader of the community, how can I sell the language to others as stable. So perhaps your changes are the most minor ones. My question is this - when will we address the major wones that you and And have held forth on for so long? John got his proposals through largely because he wrote them up formall AS SOON AS he proposed them, so we knew and could keep a record of what was and was not decided. Until you posted X1-X6 no one else had ever even tried to follow his model, and here you are saying these are things proposed 2 years ago, and which you have been using in text and speech since you proposed them. How many changes will the other active users of the language have once they get to the point where they decide to post formal proposals. When I said that I could not work on the dictionary because of grammar propsoals, this DID NOT mean that the proposals should be held up and then submitted. What it meant was that, as long as the threat of a change modifying the language exists, then I get distracted by the change. If I had finsihed the dictionary and published it, and you had made these proposals then they would have been flat-out rejected. Wait until 5 years after the books are publsished. That will be the same thing with the refgrammar. The day it is published, the baseline of what it covers goes into effect. Luckily the refgrammar covers a slightly smaller amount of the language than the dictionary, so new lexicon could be proposed that has no grammatical significance. But when the dictionary is baselined, a change in the meaning of a cmavo or the types of constructs it is used in becomes unconsdierable. >As much as it makes sense, not as much as possible. Besides, I'm not adding >anything new, as I explain in my proposal. You can currently fool the parser >by using {nana} glue. My proposal is to make this roundabout unnecessary, >so that I don't have to say {co'a nana ze'a broda} instead of {co'a ze'a >broda} I guess I don't follow this, but if you are saying it needs "na" to be accepted why couldn't you just use "ja'a" instead of "nana". >Compound tenses have not seen much >actual usage, but when and if they ever do, X6 will be adopted by default. >Notice that superficially you can already say {co'a ze'a broda}. My proposal >is to tell the parser that this means what humans will think it means, >a compound tense, and not {co'aku ze'a broda}, as the parser currently >believes. Except that BOTH mean "co'aku ze'aku", at kleast they used to %^). >> Yet you have also proposed that we ADD structure to >> MEX quantifiers (PA-strings) for the same reason. > >I have never proposed that. What I wrote about the substructure of PA >was never a grammar change proposal, it was just to help understand >how the combination of PAs work. It was very incomplete to be a serious >grammar proposal. > >Besides, the two cases are clearly different. The new permitted tense >combinations are things that clearly make sense. Many of the free >PA combinations make no sense at all. I could perhaps test your proposal by putting in unlikely values and seeing how much nonsense I can generate %^). But if it works, it is only because the set of tense words in any given selma'o is much more narrowly confined semantically than the words in PA. >What is the problem with: {mi ta'e ze'u bajra} = "I usually run for long >times". How do you currently say it? Or is it that pc didn't say that it >needed to be said with tense? (I find that the argument "we have everything >that pc said we needed" is not really compelling. I really respect pc's >opinions, but he is not infallible, and neither is your interpretation of >what he says.) Indeed %^) Especially when he changes his mind %^) But pc covered his and our tail on this with his claim that all (or almost all) of tense was abbreviation for subordinate metalinguistic bridi about the untensed bridi. As such, we do not feel compelled to abbreviate all possible things that can be abbreviated. I'll have to think off-line about whther there is another way to say >{mi ta'e ze'u bajra} = "I usually run for long >times >> In the cas eof MEX, there >> is no one person we could use as a standard for "all the things that need >> to be said". > >Well, perhaps that's why the MEX part of the language is not really up to >the standard of the rest (in my opinion, of course). The real reason is that it doesn't get tested in day-to-day usage. >Ok, I will try to do that. I never really bothered to study the YACC rules >because the EBNF seems so much clear, but if that's what it takes... SOMEONE (usually John) has to code up YACC changes and get them through YACC for any grammar change that gets proposed. The standard has always been 0 s/r errors 0 r/r errors, and you cannot get those magic words out of an EBNF formulation. lojbab