On 5/20/2014 5:49 PM, 'John E Clifford' via lojban wrote:
> This discussion reminds me of the obscurity or paradoxicality of the
> whole baseline project. The obscurity
is:What is it to do? The paradox
> comes from the usual answer: to determine the present state of the
> language. But, of course, the present state of the language is -- by
> definition -- just that specified in CLL (after typos are corrected,
> discovered inaccurate statements clarified, and the definitive version
> of xorlo added). So, where is the research?
I'm not sure "research" is needed, per se. If it is, then it consists of:
Finding the typos, discovering the inaccurate statements, and figuring
out how to clarify them in ways that don't muddy some other point, and
coming up with a definitive statement of xorlo (which despite being
supposedly well-defined, keeps coming up as a topic of highly technical
questions).
Others might call this "simple editorial work".
> The escape from the
> paradox is to say that what is sought is a report on what is
currently
> accepted as Lojban (by whom is left somewhat open -- perhaps the
> committee or Robin?). But this goes against the whole point, since it
> involves a constantly shifting target and standard: what goes
> uncriticized nowadays was once rejected or at least suspected, so both
> the language as used and the tastes of the referees are in flux (and not
> guaranteed in agreement if there is more than one).
Not guaranteed, which is why we did not require unanimity in voting.
> While no changes
> can be suggested, let alone approved, until the baseline is done,
> countless changes might be (and surely have been) made since the process
> began, almost all without discussion or approval by any body.
And all of them are unofficial, though a few might sneak in if they are
close to consistent with the wording so far, or if they apply to byfy
sections as yet
unanalyzed.
> So, in
> what sense is what is to be described the Lojban(s since some changes
> have gone equally in several directions)?
If several directions, then one must be chosen. Otherwise, we have
schism, which most find unacceptable.
> In retrospect, it might have
> been better to allow changes through a careful process from the get-go,
> building up a cumulative description as we went along.
That is what was happening up until 1997 or so. But we couldn't produce
a dictionary or even a good selma'o catalog, because we couldn't come up
with dictionary-quality definitions of the cmavo. While I worked on the
gismu/lujvo portion of the dictionary, Cowan started to work on the
selma'o catalog which in turn would hopefully lead to the cmavo
definitions. He started with my tense and negation papers, which each
covered a small number of selma'o,
and went on from there From
1994-1997 he wrote corresponding essays on all the chunks of the
language, incidentally generating 20-odd formal change proposals to the
grammar to respond to ambiguities and limitations (a couple of these
were changes actually proposed by others, but codified by Cowan). Those
change proposals, as they were accepted/approved by the community of
Lojban-List readers, were incorporated into the formal YACC grammar as
well as into his textual essays. Then the essays were collected into
CLL, with a cross-indexing "selma'o catalog" added as the last chapter.
But this still did not lead to definitions of each cmavo suitable for a
dictionary. And a couple of years after CLL was published, once the
community of users who had studied it grew to critical mass, people
started finding typos and undefined holes in what was described. But no
one followed
Cowan's demonstrated procedures for formally proposing
changes, and no one wrote cmavo definitions. When I realized we had
passed 5 years of so-called baseline without even finalizing the cmavo
definitions, I got the Board (with membership approval) to pass the buck
to the BPFK. It's been more than 10 years now, I think.
Under Nick, the BPFK got the job partly done, but there was too much
pressure to finalize certain things before all the other sections were
done, and no one volunteered to shepherd the boring parts of the
language, which never did get done. Robin moved things much further
along, but still couldn't get people to tackle the most boring stuff.
We seem to be lacking the compulsive completists that Cowan, Nora,
Athelstan, Nick and myself were (and I should probably include you among
those, since in many ways this whole project - of completely codifying
the language -
got started with your attempt to codify all of the Loglan
cmavo and selma'o on a set of pages that still graces my historical files).
> There would then
> at least have been some control (beyond the annoying scolding that we
> can't suggest that yet,
People can suggest anything they want, but they cannot be formally
decided without a formal change proposal which of necessity has to
include a statement of what the status quo is, which in turn requires
the status quo specification to be complete.
(People have made reference to software development in this discussion,
and the concept I am describing is taken from procedures that were
standard in most parts of the industry when I was working in it, and so
far as I know are still standard, albeit more automated.)
> when the end of yet was clearly nowhere in sight
The end has been in sight for a long time. The will to cover
the
intervening distance hasn't been there since Cowan and I stopped leading
the effort. It seems to take a completist attitude.
I had the rather extreme ideal of Sir James Murray of OED fame, and a
couple of books on lexicography to give me guidance, but I simply
couldn't do that kind of job while doing all the rest of the Lojban,
jobs I was trying to do, especially while also parenting two young kids
(something Robin has similar discovered %^). Cowan did a splendid job
on CLL but has seemed to suffer burnout since then. Lots of others have
taken up the slack, but no one seems to have the drive to get it all done.
> Lojbab has presented a different scenario, in effect. CLL was not a
> specification, he says, and what the baseline is is a specification (it
> is not perfectly clear what that all means in the light of several --
> unfortunately non-equivalent --
grammars).
The baseline requires a specification defining ALL of the cmavo, because
that is what was still missing in 1997. The YACC grammar as published
in the appendix to CLL remains the standard for resolving any
non-equivalencies.
>That is, apparently, that the
> baseline is to fill out -- with examples and other commentary -- all the
> details that CLL merely sketched in broad strokes (not a description
> that applies to CLL very handily).
And in particular, to come up with dictionary-quality cmavo definitions,
so that we can combine cmavo and gismu and lujvo definitions along with
a formal grammar into a well-defined set of what people understood
originally as the "baseline" of words and grammar (with any many
questions about ambiguities resolved as possible).
> that is, the baseline project is to
> find out what CLL meant -- or, rather, has come to mean --
to people who
> claim to using Lojban.
Perhaps, but more limited than that. Dictionary-quality definitions
(that are "definitive" %^) is the key. The BPFK sections were the
method of writing those definitive definitions.
> The problem is one of control again: if CLL was
> not specific, then there are a range of possible meanings and how are we
> to decide which one is right?
That was the job of the BPFK, and the "how" was basically "whatever was
needed to achieve consensus".
> That is, establishing the baseline is
> already doing what the baseline was to serve as basis for: establishing
> changes in the language. And so the circle goes on. The most we can
> actually do is report on what users actually do (somehow skipping
> mistakes of various sorts) and come up with a description of a Lojban
> (or several), but not of Lojban.
A
single "Standard" Lojban, whether anyone actually bothered to follow
that standard. (It was envisioned that the BPFK would evolve into a
group that would certify the degree of compliance with the standard of
submitted works)
> Until someone declares that one of the
> things come up with is the real thing, which will ot be a unanimous
> decision, of course,
It must be a consensus of whoever constitutes the BPFK at the time.
(Consensus being defined so as to not allow individual vetoes).
> since it will not be either CLL Lojban (which it
> turns out did not exist as such) or a new version arrived at in an
> approved way.
I think it did and does exist, because people are capable of deciding
whether something is consistent with CLL for the most part. We know
what things people are proposing as "changes" to CLL, such as xorlo, and
which are "corrections" and
"clarifications".
> All that being the case, some body, consisting of people who actually do
> the work, needs to do something along the lines laid out over the
> decades -- or along some other effective lines. Or we can go on as a
> squabbling group, boasting about a language we don't have, with
> properties what we do have doesn't have, recruiting (under somewhat
> false pretences) ever more people who deviate ever further from a
> nonexistent norm. Well, that is a language, after all, though not quite
> what anyone seems to have in mind.
The standard in most lay people's minds is that a language isn't really
a language until it has a dictionary (and for most that means a physical
book, because until we commit to fixed print, it remains a nebulosity).
We never produced one; jbovlaste is a data base, and not a dictionary.
And the baseline was originally
defined as a dictionary and a grammar
(CLL can fulfill the latter, even if it is not a
"specification" sufficient to produce dictionary definitions of cmavo).
lojbab
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+
unsubscribe@googlegroups.com.To post to this group, send email to
lojban@googlegroups.com.Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.