From LOJBAN%CUVMB.BITNET@UBVM.CC.BUFFALO.EDU Fri Nov 17 01:03:08 1995 Received: from VMS.DC.LSOFT.COM (vms.dc.lsoft.com [205.186.43.2]) by locke.ccil.org (8.6.9/8.6.10) with ESMTP id BAA16883 for ; Fri, 17 Nov 1995 01:03:03 -0500 Message-Id: <199511170603.BAA16883@locke.ccil.org> Received: from PEACH.EASE.LSOFT.COM (205.186.43.4) by VMS.DC.LSOFT.COM (LSMTP for OpenVMS v1.0a) with SMTP id 44C87845 ; Fri, 17 Nov 1995 1:55:43 -0400 Date: Fri, 17 Nov 1995 00:35:30 -0500 Reply-To: Jorge Llambias Sender: Lojban list From: Jorge Llambias Subject: Re: ni, jei, barely, almost To: lojban@cuvmb.cc.columbia.edu, jorge@minerva.phyast.pitt.edu Status: OR And: > 1. I haven't had the original message yet so don't know what was proposed. > I'm amazed at the idea that duu is a subtype of ka. In general, I wish > the existence of NU be forgotten. I think there are three useful NUs: du'u, nu and ka. I adhere to the idea that the rest be forgotten. (I don't understand how du'u is a subtype of ka either.) This is my current understanding of them: {du'u} is the most basic one. It abstracts the actual relationship. For example {mi djuno le du'u do klama le zarci}. I know that the relationship {klama} holds among the arguments {do}, {le zarci}, {zo'e}, {zo'e} and {zo'e}. {nu} is the most used, and it is the physical realization of the relationship. For instance {mi viska le nu do klama le zarci}. What I see is not the relationship {du'u} but its embodiment {nu}. I am tempted to say that {le nu broda} is {la'e le du'u broda}. {ka} is very different from those two and requires a blank argument place, because it really gives a function. I missed John's latest proposal for the lambda variable, so I can't really comment, but a PA in that role doesn't make much sense to me. What is needed is a KOhA to keep an argument place open. I will comment further when I have a chance to read the actual proposal. (As an aside, I don't think that something like {mi nelci le ka do melbi} makes any sense. It should be {mi nelci le nu do melbi}. I don't think {ka} and {nu} should be thought as the English -ness and -ing, because those correspond to the endings for adjectives and for verbs respectively, but Lojban makes no such distinction. As for the remaining NUs: {za'i}, {zu'a}, {mu'e} and {pu'u} are all kinds of {nu}. I don't think I've seen any example where using any of them makes anything clearer than simply using {nu}. ({za'i} is the most popular, perhaps because the word "event" makes people think that {nu} has to be something dynamic, but this is not really the case. {nu} is perfectly fine for even the most static of states.) {jei}, {su'u} and {ni} are generally used as indirect questions. {jei} has two meanings: by official definition it is a truth value. By usage it is the yes/no indirect question "whether", equivalent to {du'u xukau}. I find the definition as a truth value totally out of place among the NUs. There should be a lujvo that means "x1 is the truth value of x2". The usage definition is ok, but redundant to {du'u xukau} and not so frequent that it is worth the trouble to have a short form. I don't agree that {du'u xukau} in any way suggests that there are only two possible answers corresponding to truth values 1 and 0. The question {xu ko'a melbi} can perfectly well be answered with "sort of", in which case {mi djuno le du'u xukau ko'a melbi}, "I know whether koha is beautiful" means that I know that koha is sort of beautiful. I don't see why {xu} should in any way suggest that there are only two possible truth values. {su'u} has seen little to no use. The two or three times I met it it was used for the indirect question of manner "how", as in "look how they run". I would say that can be taken care of by {tai makau} or {ta'i makau}. {ni} is not too clear, but it corresponds to the indirect questions of quantity, amount and degree: how many, how much, how far, how fast, how long, how pretty, how hot, how red, how . For example: mi do zmadu le ni ke'a citka Which could possibly mean: mi do zmadu le ka ke'a citka xokau da I surpass you in how many things we eat. mi do zmadu le ka ke'a citka pixokau da I surpass you in how much we eat. mi do zmadu le ka ke'a citka xoroi I surpass you in how many times we eat. and who knows how many other things. In this case context helps a lot. Something like {le ni klama} is much more mysterious. Finally, {li'i} and {si'o} are the subjective abstractions. Again, there is little to no usage of those, and I'm not convinced that they add anything significant to using {du'u} or {nu} or {ka}. The subjectiveness seems more to depend on the predicate being used rather than on the abstraction itself. In summary: I recommend using {du'u}, {nu} and {ka} and forgetting about the others. > 2. There are two distinct kinds of gradience in truth values. The first > concerns the fuzzy boundary between true and false: we take T & F to > be points 0 & 1 with nomansland between them. The second concerns degrees > of truth and of falsity: how much would the world have to change for > some state-of-affairs to become the case (if it is false) and to cease > to be the case (if it is true). If you expressed this in numbers, then > you'd use the full scale (of (I think) real numbers), with negative > numbers for falsity and positive for truth. You talked about this before, but I think I'm starting to finally understand it only now. I would also add that those two gradients are distinct from the scale provided by {to'e}. A negative value in the -inf to +inf scale does not imply a positive value in the {to'e broda} direction. In other words, "almost beautiful" does not mean "slightly ugly". > I think I once suggested that {jei} denote the former type and {ni} the > latter (though I'd prefer to use selbri+duu). Here I'm lost. How do abstractions enter into it? The idea of the scales is a nice way to picture it, but I hope you are not suggesting that we use actual numbers for it! {li ni'ureze ni mi melbi} = "-27 is the amount of I'm beautiful". I would find something like that ridiculous and rather useless. What information does the number -27 convey? > 3. To what extent, I wonder, do we have ways of expressing these varieties > of truth gradience? > We seem not to have anything intermediate between {na} and {jaa}. I think > that's the sort of thing Steve has been saying we should have. > As for the gradable T & F, Jorge has proposed additions to NAHE: I like your proposals, so I withdraw mine. > jaacai very true > jaa(sai) (fairly) true > jaarue slightly true (true, but only just) - BARELY > narue slightly false (false, but only just) - ALMOST > na(sai) (fairly) false > nacai very false > > - these I think are quite good. I agree. I would like to see {naru'e} in the dictionary entry for "almost". > For indeterminate, fuzzy: > nanaicai near 1.0 > nanai(sai) > nanairue > jaacui = nacui 0.5 > jaanairue > jaanai(sai) > jaanaicai near 0 > > - these are less satisfactory, but they're a start. I prefer: je'ucai je'u(sai) je'uru'e je'ucu'i je'unairu'e je'unai(sai) je'unaicai I find nanai and ja'anai somewhat confusing. > I'm less sure about the following NAhE forms. If we used them then it would > be nice to have rafsi for {cai} and {rue}. > > jeacai to a large positive extent > jea to a positive extent (unspecified or ungradable) > jearue to a small positive extent > naerue to a small negative extent > nae to a negative extent (unspecified or ungradable) > naecai to a large positive extent > > What do rodo reckon? > I like them. The difference between {naru'e} and {na'eru'e} is of course the same as that between {na} and {na'e}. The first one would say that there is almost a relationship between the arguments, while the other claims that there definitely is a relationship, which is almost but not quite the one corresponding to the selbri in question. Jorge