Here are Nick's dictionary notes from 2001.

Jay's proposed dictionary database will go ahead, and is a good and wondrous thing. In the interests of some sort of public standard arising beforehand, if possible, I suggest we group-write a couple of definitions of some of the thornier words, as we would like to see them defined in the dictionary. I am particularly thinking of cmavo with gaps left in the Refgram. I object to group-writing definitions. As I pointed out in the fa'a (Dictionary) page, we should let usage decide. I frankly agree with Jay that one can't push for something without also using it. - mi'e kreig.daniyl.

  • *sigh* So be it. I am now writing the fa'a definition Elephant-style, But I want to say, unambiguously and exasperatedly, that I object to people objecting to the very authoring of the Dictionary. --- nitcion.

For instance:

I [Nick] now believe the right infrastructure for the cmavo section of the dictionary (as opposed to the lujvo section, which will be inherently complex and controversial), is the Elephant.

I think that the Elephant might serve as a good backstop for the debates that may be necessary to solidify definitions of some cmavo. I do not believe that most cmavo will need to be Elephanted, if people come up with usage examples which do not provoke controversy. Even where it is necessary to use the formal structure of the Elephant to lay out the pros and cons for each choice clearly, I think that there is a way to preserve the reference nature of a dictionary by excerpting only the usages and the definitions accompanying each, with annotations indicating controversy, and which definitions are based on "formal analysis" vs. based on "historical design" or "early usage" (I suspect that most of the controversies can be summed up as competition between each of the above). — Bob LeChevalier

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. Elephant is good inasmuch as it lets users see the conflicted state of opinion and select from the clearest renderings of each opposing view. But upon consideration, a dictionary is not as appropriate as a database that's open to the community, not unlike this Wiki. People can add entries for usages they have seen. It should be descriptive and not prescriptive. And there should be a voting system so people can get a glance of the widespread opinion on usages. And this voting needs positive AND negative votes, so that an issue that's never been voted on has the same voted weight as one with 20 supporters and 20 denouncers. --xod

This really does suggest a basic mistrust that there is any Cabal that exists and will act against the wishes of a consensus of the community as it understands that consensus.

Yes, any decisions that the editor(s) of the dictionary make that go against the opinions of any member of the community, run the risk of arousing the ire of that member. But if we are not to have dialectization, then at some point, someone (or some usage) will have to decide/set an example for the rest of the community to follow, and there will be just as much likelihood of arousing the ire of the "losers" in any decision then.

While the Elephant may be a good method for debating these issues (I find the Wiki too chaotic to be readable much less usable), it is way too cumbersome to serve as a reference. A dictionary is first and foremost a reference, and not a forum for points of view. The bulk of the community wants answers, as simple and clearcut answers as possible; let others debate, so long as they can "use" it. Furthermore, the Elephant as a data base would not be printable, and as a text file laid out like Nick's example in fa'a (Dictionary), it would be way too long - a cmavo dictionary alone might run several hundred pages. The dictionary has to be printable, and has to serve the needs of the non-online Lojban community. — lojbab

    • Lojbab, you seem to think a Lojban dictionary can conceivably not be hundreds of pages long. A dictionary not hundreds of pages long is in my view not worth waiting for. — nitcion

If Elephanted, it might take ten years. The rate we're going now, with no Elephant, it'll take till Doomsday. And without Elephant-like structure, no body of knowledge and lore concrete enough to be called a dictionary can emerge. The Wiki certainly can't generate it.

The reason it would take till Doomsday would apply both with and without an Elephant. The Elephant will speed nothing up, and hopefully would slow nothing down except insofar as people enjoying the debate too much might not do as much as they could otherwise in coming up with the examples I've asked for. The real delay in dictionary publication has been a lack of lojbab's time and a lack of money to pay for publication even when it is done. The cmavo have presented a further problem in that most everyone is dissatisfied with the "definitions" that comprise the cmavo list, and no one had any good ideas how to write useful definitions for them (John Cowan at one point thought the task impossible and favored using the cmavo list as it is for baseline purposes, and I was close to conceding this.). But I think that Nick has shown that problem solvable, provided that we have the examples. Without the examples, there is a considerable risk that the existing cmavo list with all its failing will become the sole baseline standard. --lojbab

The Elephant can incorporate votes; and the way I at least see it, usage examples would be a database/stock independent of positions --- so positions can say "Usage 3 is dumb-assed", "Usage 5 is spot on", and so forth. But while it'd be largely open to the community (noone's preventing anyone from posting a position, as long as it's well-formed), it still needs a Clerk to tidy up, wrap up, and generally take editorial responsiblity. Which is ultimately not the same model as the Wiki. (And that's A-OK by me.)

  • What is the difference between the role of this Clerk, and the role of the Cabal in the above alternative? The Cabal has the option to limit itself to a role of no more than minimal Clerking, just as the Clerk has the option to dictatorially act as a one-prenu Cabal. — lojbab

    • No essential difference. The difference is in the mechanisms by which power is negotiated, and the sense of accountability not in the fact that there is exercised power in the first place. — nitcion

Why isn't this Wiki-ish? Lots of little comments followed by an occasional editor actually seems like a pretty common Wiki pattern! See MoreAboutHousekeeping on the original Wiki, for example.

  • Unique editor, editorial responsibility, enforced rather than voluntary tidying up, strict structure of entries. These may appear commonly in Wikis, but they aren't mandatory --- and aren't exemplified all that much here...

The bottom line, I suppose, is no longer "Do you want a dictionary for yourself?" (a lot of you clearly don't), but "Do you think a dictionary should not be written by anyone at all?"

  • Regardless of the will of the participants in this forum, who are by nature computer literate and even computer-oriented, or they would not be here, the bulk of the potential Lojban community will never read a Wiki or an Elephant. The standard for literate languagehood has for a couple hundred years been a dictionary of some kind. Lojban won't really "exist" for many people, until we can hand them a volume labelled "Dictionary" on the cover. One of the first things many new people ask is "where can I get a dictionary?", and the biggest excuse people have had historically for not learning the language has been "There is no dictionary yet". They don't want to make up lujvo, nor to choose amongst grammatical controversies. They want simple answers, and they want those answers to be unchanging (hence the baseline). (Once they learn the language of course, they tend to lean more towards one of the various axes that have been described - see hardliners.) --lojbab
    • Your assessment that most Lojbanists will not be computer literate and will never see an Elephant is, I think, 20 years out of date, and unrealistic. — nitcion
      • I consider myself computer-literate, but I still do not even understand the idea of an elephant, and am unlikely to read it until I have some idea what it is. Elephant has only been described in terms that make it seem like you must not only be computer-literate but also well-read. Perhaps someone could clarify the elephant page? - mi'e. kreig.daniyl. (PS: will the elephant be from Betelgeuse?)
        • The non-flame response is, wait and see. I had a huge flame response here, instead of which, I respectfully suggest you reread the top half of the page, whch is quite clear (a forum like the wiki, but with strict tree-like structure); and if you object to people using words like Ankh and Ankus in the bottom half of the page, which is clearly delimited as a digression, kindly keep it to yourself. You know what they say about not being part of the solution... — nitcion
          • I was dissapointed upon purchasing The Complete Lojban Language to discover it contained no reference dictionary to gismu and cmavo. I bought it in order to learn Lojban, but you can't learn the vocabulary if you don't have it. I am learning Lojban vocabulary from a database of 2,000 flashcards containing no lujvo whatsoever, and I consider that database to be the dictionary I sought. Why not? It should be published in pamphlet form with the word "Basic Learning Dictionary" on the cover. Then later when (or if) you folks work out the lujvo, publish a Lujvo Dictionary. And this strictness is coming from someone who has been an enthusiast for six years; the man on the street would laugh in our faces if we say we can't give him any dictionary at all because we don't have the whole thing yet. Why the all-or-nothing approach? — Eppcott (Matt Arnold)

As for the Dictionarists' time and burnout, well frankly, that's the Dictionarists' business, and the Dictionarists' call; noone's forcing anyone to be a Dictionarist, but noone should be discouraging anyone either (at least until they find out for themselves how hard it is.) It can be argued (and I'm not doing this to single him out or denigrate him; quite the contrary) that time and burnout is what happened to John with the refgrammar. Does that mean he should not have gone ahead with it?

  • Remembering that I became active in Loglan intending to serve as Dictionarist back in 1985, and am still interested in the job 16 years later, though I too have suffered periods of burnout, I am not likely to become discouraged about the need and the possibility of producing a good dictionary. I am more concerned with making it small enough to print, while at the same time having been distressed because I did not know how to define cmavo without examples (and coming up with ad hoc examples has always been my biggest weakness in teaching the language). — lojbab

At any rate, let's not second-guess the Elephant's structure until we see it lumbering into action... — nitcion

If Elephant is good, it's good at providing a structure for these debates. It's not a dictionary database, and it shouldn't be; that's asking too much of one app. So it should probably be its own structure (which Jay is building?) with links to appropriate Elephant discussions. --xod

Indeed. Look at jbovlaste before you say that Elephant is the only way to do it. My current pause in jbovlaste development is to take into consideration the recent growth in Wiki usage, and now to also consider how Elephant might interoperate. It looks as though I might simply provide a mechanism for jbovlaste to point to an Elephant argument or arguments which are relevent to the word in discussion. --jay

. *shrug* By all means, you and John should be (and presumably are) talking. I still think that cmavo definitions are a rather different kettle of fish than lujvo definitions, but nothing intelligent can be said until both mechanisms are up and running. — nitcion.

See also Great Dictionary Problem.