Further discussion of ka'e nai is to be found under bi'ai.

ka'enai and ca'anai are two words which sound perfectly reasonable, are probably used on numerous occasions, and anyone could tell you what they mean - yet they are not grammatical. For some reason, CAhA is a separate selma'o from PU and ZAhO and all the other numerous tense selma'o, and unlike all of them, CAhA does not allow nai to be attached. You need to put na'e before the CAhA word instead.

I have yet to see any reason given for this. So do I - could this be explained by LLG??? It's things like that causing natlang frustration, and even more than natlangs!

  • Cowan explained this in http://www.lojban.org/lists/lojban-list/msg02131.html. There's a good reason for this not being allowed. -mi'e .djorden.
  • It's not clear to me that he's explaining that there's a good reason. He's explaining that the state of nai is pretty messy and that *CAhA+NAI is not a solitary arbitrary exception. If there's a good reason for the status quo, perhaps you could spell it out a bit more clearly? After all, ka'enai is becoming something of a testcase-cum-causecelebre.
    • Did you read what he said? The nai on pu and fa'a were left because it had been that way, and was better to leave it. The CAhA stuff was new, and thus could be shielded from that crap. His (and my) preference would be to eliminate PU+NAI and FAhA+NAI rather than adding CAHA+NAI. Furthermore, if it is 'the same as tenses' (which it isn't — and note that by 'tenses' they actually mean PU and FAhA, ignoring the multitude of other tense cmavo for which NAI is disallowed) as is claimed by the proponents of ka'enai, it would be contradictory negation and not the scalar/polar negation we see when people (ab)use it. I, for one, can't believe this ka'enai crap has gone on for so long with people claiming there's "no reason" for it, when there clearly is (or perhaps they're just being disingenuous and actually understand and just disagree with the motivation). --mi'e .djorden.
      • My argument is that nai should be allowed to modify any word that can be modified by a UI. Selecting which words can or cannot be modified by nai is extremely arbitrary. So it is more general than that some tags can take nai and others can't. --xorxes
        • What kind of modification do you want? nai means different things when applied to different words... — mi'e .djorden.
          • Eliminate selma'o NAI and put nai in selma'o UI. This is a simplification of the grammar: less rules to learn. nai in general converts a word into one with opposite meaning, but in some specific cases (notably logical connectives) it is related to na. --xorxes
            • I meant what kind of modification should the nai do ;P Logical connectives and tenses also are contrary neg with nai; I dunno what else.
              • Well, the effect of nai on TAhEs is interesting. The official definition with PU tenses is not very useful, it should not be propagated to other words if possible. --xorxes
    • Did you read what he said? (Yes of course you did. And yes of course I did.) Regarding the semantics of NAI, my impression, not based on systematic analysis, is that it doesn't consistently encode a single kind of negation; it would be interesting to see a systematic study of this. Regarding the way to undo the mess, the comment someone made below about treating it like a UI (or CAI) distributionally probably better approximates the grammar people acquire inductively. (Regarding this last point, I am about to post a new page.)

(pu'i and nu'o are also members of CAhA, but much less used. na'epu'i is actually in the cmavo list separately, for a fairly useful concept. na'enu'o would be a bit more obscure - it would mean "no unrealized potential", meaning that, whether we have seen the subject perform this event or not, that's all he/she can do.)


Usage is drifting away from baseline!

Well, it has to happen. In a language where we're used to not seeing arbitrary exceptions, and those exceptions that do exist tend to have a logically justifiable reason, CAhANAI compounds should naturally arise.

This creates a trickier situation than simply changes in usage (the "ce'u" ordeal) and experimental cmavo/gismu/rafsi, though. If everyone started saying ka'enai, shouldn't a parser (not the official parser, but certainly something like jbofi'e) parse it? However, to do so would require modifying a rule in the YACC grammar, one part of the baseline which has been sacred.

Though I respect the baseline, and will not use things like ka'enai in Flatland, I might just start saying ka'enai in ordinary usage and see what happens.


nai should have the same grammar as UI. This will eventually be necessary for Three-value logic.

I have always used ka'enai, and this is a dumb exception that, inasmuch as there is any significance to "usage decides", was decided against 10 years ago. I hold with Rob's position.

(One can consider ka'enai the "ain't" of Lojban...)

Created by admin. Last Modification: Friday 30 of November, 2001 12:31:04 GMT by admin.