Re: more epistemic perversity (was Re: .i .uepei mi jai sel

[And Rosta <a.rosta@UCLAN.AC.UK>]

Lojbab:
> >(i) Is there a list of which gismu places are presumed to be subject
> >to consensus and which are presumed to be subject to intersubjective
> >disagreement?
>
> Since I have been a strong opponent of semantic analysis of Lojban before
> usage determines the meanings, I certainly have not even considered such a
> list.

OK, but I at least think that any semantic analysis is suspect if it
does not at least in overview consider the entire vocabulary. For
example, if you could show me a systematic and principled basis for
deciding which gismu places are presumed subject to intersubjective
agreement (100% in my view, but probably at least 99% on anyone's
view) and which are presumed subject to intersubjective disagreement,
then I would be prepared to at least consider your proposal
seriously. As it is, your opinions are based on as wobbly foundations
as those of someone with scarcely any acquaintance with the language.

Contrast this with Jorge, who not only endeavours, in his analysis,
to discover or advocate consistencies and regularities, but is also
at pains to establish paradigms of near synonyms, as in his very
neat:
      jinvi   belief
      krici   justified belief
      djuno   justified true belief

--which to me is too elegant to resist.

> >(ii) Your position seems to lead to strange results. For example, the
> >truth of {ti mlatu} is presumed to be determinable against some
> >consensual metaphysics, but the truth of {"true" fa le du`u ti mlatu}
> >is presumed to not be determinable consensually. It seems utterly
> >nonsensical to me.
>
> We have epistemological places on those gismu for which I (and others)
> recognized there is and has been significant argument regarding the
> metaphysics/epistemology.  For other places, where there was no particular
> evidence of subjectivity being "metaphysically necessary" to the definition,
> we have the capabiliyt of adding such considerations using a BAI such as
> vedu'o or perhaps du'o itself.  Thus you can have a subjective mlatu if
> you wish.

I would dispute that any subjectivity is metaphysically necessary,
and that therefore no gismu should have an inherent place that makes
the BAI redundant.

I think we have already established (to my satisfaction at least, and
I thought I was agreeing with you) that x4 is to do with the knower's
epistemology and thus is entirely unrelated to this dispute about the
subjectivity of the x2's truth.

> There was no attempt to make Lojban semantics "systematic" or "logical" in
> the way that you seem to be seeking (regarding the gismu, that is).  We
> were satisfied to make sure that we could cover things that we knew were
> "talked about" with sufficient robustness that new ideas for gismu were
> analyzable in terms of old ones.

I find too much patterning in the gi`uste to find it credible that
this is entirely true, but I do find enough irregularity to believe
it partly true.

> Place structures were analyzed only to
> try ot keep like places on the same order (e.g. to from per klama).  At
> one time we drifted towards very fat gismu, and then at a later time we
> pared them down.  Until pressured by the community, I never felt that the
> details of the place structures were all that important - I was quite content
> not to even baseline them, but was outvoted.

Be this as it may, you are describing the process by which the gismu
were added to the language, and not the process by which their
meaning comes to be decided upon. When you start trying to tell is
what djuno should and shouldn't mean, and what you tell us in
flagrant conflict with all other evidence bearing on the matter, why
should one pay any attention to you? In my case, it would only be if
your case was intrinsically persuasive, and to me it would be more
persuasive if shown to be applicable to the entire brivla vocabulary.

> In my opinion, it is sufficient merely to insist that, whatever the place
> structure, that all usages be consistent with the full place structure whether
> places are elided or not.  This alone would ensure enormous improvement over
> JCB's version of the language, where gismu were used in tanru as straight
> loan-translations of their English keyword.

I wholly agree. For every place I am initially inclined to elide, I
try to consider whether or not I should instead fill it with zi`o.

> I am reasonably convinced that meanings of djuno which accept the existence
> of all 4 places (and hence recognize that knolwedge is tied to epistemology)
> are a sufficiently small range of meanings that usage during the early years
> of the language will serve better to define the word than any anount of
> analysis.

I am in this instance reasonably convinced of the same thing. It is
equivalent to the status quo continuing, with your strictures going
unheeded.

> While I clearly have opinions about the meanings, and can state
> my intent when I wrote the definitions, usage will determine the meanings
> and NOT debate.  There is nothing in the place structure (as opposed to the
> choice of keywords) that requires se djuno to be "true" from the standpoint
> of the speaker; thus I will have no compulsion about using it to report
> "knowledge" that I feel is-or-may-be invalid.  If this makes the keyword an
> inopportune choice, so be it - the keywords were never intended to be
> defining.

You seem to mistakenly believe that the place structures alone,
independently of the keyword, are sufficient to define the meaning.
It is patently the case that the keywords *are* intended to be partly
defining.

The point we've reached, it seems to me, is that (a) you have established
that you are not in conflict with the letter or spirit of the
baseline if you use djuno with x2s you believe false, and (b) you
have failed to persuade any of those who have participated in this
debate that your desire to use djuno in the way you want to has
anything at all to recommend it.

Presumably, judging from what you've said in the past, you're quite
happy with this situation, for only future usage will decide the
matter, so this thread can end here.

--And