Re: Linguistics journals

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

Lojbab:
> >> Well if they've already done all this, why are *we* reinventing the
> >> wheel?  We could just take their scheme, add vocabulary words and a
> >> method of pronouncing the symbols.  I thought what we were doing was
> >> more ambitious than what was already available.  I'd be interested
> >> in hearing more about some of these notational schemes.
> >
> >This is part of what Lojlan did: make the notation speakable. It
> >then did a load more non-logical stuff besides, to enhance usability
> >and achieve various other auxiliary goals.
>
> Ah, but the various auxiliary goals made the primary one meaningful for
> linguistic pruposes.  Thus one can communicate pred-logically (should one
> desire)

Lojban of course forces one to speak in predlogic, though we
can choose to ignore that in communication.

> and incidentally throw in the attitudinals or discursives that
> have no agreed upon logical structure but yet are meaningful for human
> communication.  We also have short cuts that enhance the abbreviated
> expressive power of the logic.  pc has said that using numbers other than
> none/some/all is a hairy mess in logic - a defined mess, but still hairy.
> It thus is convenient to be able to say 3 dogs without having to expand it
> into da poi gerku zo'u de poi gerki gi'e na du da etc.

I agree, though I do think that if we had had more experience
in using minimalist Lojban that corresponds straightforwardly to
ordinary logic notation, then it would have been clearer where
abbreviations were called for.

> >> On the other hand, Lojban does provide some fairly enticing area for
> >> linguistic research (which I may pursue when I get my MA out of the
> >> way). Certainly the creation  of a speech community from scratch
> >> would offer some intriguing possibilities for sociolinguists,
>
> >This is true, though Klingon would be a richer lode in that
> >respect.
>
> Yes, but is a speaking community of Klingon truly possible (especially given
> the proprietary nature of the language)?  mark may have opinuons on this.

It was my outsider's-impression that Klingon has the liveliest
speech-community after Esperanto.

> >> and a discourse analysis of Lojban would be another possibility. At the
> >> moment this is hampered by the small amount of written Lojban (other
> >> than translations) in circulation, and the lack of spoken exchanges,
> >> but as a long-term project it would be very interesting to see to
> >> what exten Lojbanists follow the discourse patterns of their native
> >> discourse communities or create new discourse patterns specific to
> >> Lojban.
> >
> >True again, though maybe Esperanto might be the richer lode this
> >time.
>
> Maybe.  But I think a language of very different structure and expressive
>  capabilities than ones native tongue would be far more revealing.  To me,
>  studying
> the discourse of Esperantists would seem no more exciting than study the
> Spanish spoken by the American emigre community in Buenos Aires as compared
> with the local native Spanish and the native English.  Esperanto seems
> unlikely to evolve anything sophisticatedly interesting in discourse
> patterns.

This is an empirically decidable matter.

> >> Yet another research area would be language aquisition - is
> >> Lojban easier to learn as a second language, and (when we eventually
> >> have children learning Lojban) is it possible to aquire it as a
> >> first language, or does it have features which make conscious
> >> learning necessary?
> >
> >Someone has to risk fucking their progeny up first.
>
> Well Shoulson has reported that someone is already doing this woth their
> progeny for Klingon %^)

I know: D'Armond Speers. I happened to tell this to some linguists
I was having a conversation with, and they were horrified.

> The Loglan/Lojban corrolary to this version is that:
>
> If a language were to have restrictions rmeoved from the language relative to
> other languages, while having other aspects of the language tightly
 constrained
> in areas that other language are not so constrained, this would maximize the
> apparent effects on the thought patterns of the speakers.

How might that be tested?

> For Lojban, added constraints include the unambiguous grammar aspects - the
> logical connectives, the terminators, the unusual lo/le/la distinctions along
> with their mass versions, and the strong distinction between ordinary and
> abstraction sumti both structurally and semantically (i.e. explicit
> sumti-raising).

You should bear in mind that in Lojban-as-actually-used, many of
these constraints are distorted or ignored altogether.

> Removed restrictions include the elimination of noun/verb/adjective
 distinction
> except as revealed through the syntax.  (This alone justifies research in
> Lojban, IMNSHO, since studies of brain damaged people apparently has shown
 that
> syntax acquisition, as well as the noun/verb distinction, involve different
> parts of the brain.  can a person who cannot use verbs due to brain damage
> use Lojban predicates?  Likewise sumti for those who cannot use nouns?)

Interesting questions.

> pc and I once planned to do some simple measurements on logical thinking
> in a group of Lojban students before/after.  If we got significant results,
> we might follow up with a more careful control, looking at populations of
> consistent initial logical knowledge, with a control group, a group explicitly
> taught predicate logic, and a group taught Lojban with only incidental
> predicate logic as came out in the teaching  (eventually we might try
> teaching Lojban explcicitly avoiding the predicate logic notation and
>  terminology, using ONLY Lojban terms).
>
> Alas we could not get a group of students.

What a shame. Couldn't pc have used his students? I'm sure I
could donate some... But I'd have thought a linguistics or
phiolopgy or psych - i.e. a Cog psych - department would
cooperate in a well-designed experimnet.

> >True. However, in principle, Lojban might have something to offer.
> >If Lojban was committed to (a) being able to translate into
> >predicate logic, and (b) attempting to express everything natlangs
> >do, then the Lojban community would have built up a body of
> >at least approximate ways of translating various natlang
> >constructions and meanings into predicate logic. Since only
> >some areas of this translation have been worked on at all
> >intensively, the Lojban project, if not so much the final
> >Lojban product, might have made a useful contribution.
> >
> >However, I do feel that the Lojban project has not worked in
> >this way in practise.
>
> I think you are premature in this judgement.  My main opposition for the
> finely honed semantics  arguments that on occasion overwhelm this list
> is that they alienate newcomers.

Quite so. You basically had to choose between (a) baselining and
recruiting users, or (b) refining and developing the lg, and these
two goals were incompatible, and you chose (a) in consultation
with the LLG membership.

> WHEN we have 50 people who can speak
> and write the language fluently, then a subgroup can go off in more
> siazable numbers than we have in current discussions and hammer out these
> trnaslation issues.

True, but there is no need for there to be 50 fluent people. I'm
not fluent in Lojban, for example, but that hasn't hindered me
insuperably. Indeed, it is my impression that only Jorge can
write Lojban both fluently and in approximate accord with the
grammatical prescription.

> maybe they'll even have the discussions in Lojban,
> though that might be a pipe dream.  But that will leave plenty of other
> people to test their results on, which people won't have to suffer through the
> argumentation that is way over their head.  meanwhile these test subjects
> for finely honed translations will actually have some skill in the language.

I don't understand what you are envisaging here. What do you mean
by "test their results" and "suffer the argumentation"?

> Right now, I think only Nora is both a skilled Lojbanist and almost totally
> isolated from the semantic discussions so that she can read the results
> without remembering the process.  (and she isn't interested, of course %^).

No disrespect to Nora's eminence, but I would tend to doubt that
she is a skilled Lojbanist, unless that means "one of the most
skilled Lojbanists". I may be wrong, but it seems to me relatively
unlikely that someone who has not thought through the issues we
discuss could have as much skill as they would if they had
thought through the issues.

> When you and Jorge get going, you lose even Cowan and me by sheer volume.

Not these days, though.

> The point is that one can acquire a pretty solid expressive capaibility in the
> language using ad hoc methods for communicating fine logical distinctions.

Probably correct, though why should one want to do that in Lojban,
of all languages?

> And you NEED that understanding of the language's capabilities before you
> can meaningfully tackle transaltion problems.

No you don't. Rather, if one undertakes some translation one comes
across problems ("How does one express this in predicate logic?")
and then goes off in search of solutions for them.

> So let's bootstrap a speaking/writing community and THEN tackle the body of
> predicatese.

I grant that that's the best way to build a speech community.

> >First, let us recognize two ways of using Lojban.
> >
> >CRITERIA OF SUCCESSFUL USAGE:
> >I. The speaker successfully communicates with the addressee.
> >II. The utterance actually encodes the meaning the speaker
> >intended it to.
> >
> >(I) has been advocated by Lojbab, and to me seems a bit pointless,
> >but to other people is highly appealing.
>
> I advocate I asd the only possible first step.
>
> I am also not all that sure II is meaningful.  If I have an in-mind meaning,
> then successful communicate requires that the meaning be encoded in the
> utterance plus any/all extralinguistic information.  For Internet
 communications
> there isn't all that much extralinguistic information that can be passed.
> We instead rely on the shared extralingusitic information previously
> communicated  (i.e. common understandings  of "the real world").
> Actually encoding all of that real world knowledge is impractical in
> human speech (Cowan's maxim of precision vs. infinite length here inserted).

I think you miss the point. (II) is not a matter of encoding
everything one wants to communicate to the addressee. Rather,
(II) involves the following process:

a. Speaker wants to encode thought T1.
b. Speaker encodes T1 in sentence 1.

The challenge is: given (a), how can (b) be achieved?

(I) involves the following process:

a.  Speaker wants to communicate thought T1 to Addressee.
b.  Speaker utters sentence S1. S1 encodes thought T2, which
    the speaker is not aware of. T2 has a resemblance to T1.
c.  Addressee has a rough idea that Speaker wants to
    communicate something like T1, and partly decodes
    S1 only to the point sufficient to confirm that something
    like T1 is intended.

It is (I) that is going to build a speech-community, but
academically it's of interest to pragmaticians rather than
logicians. (I) is all about the pragmatic skills of the
speakers - which are considerable (my son has brilliant
pragmatic skills, not noticeably inferior to mine, yet can
hardly communicate anything verbally). (II) is about the
Lojban prescription and logic.

> >(II), though, forces you think through the logical structure of
> >what you want to say.
>
> Only if you limit yourself to the pure predlog encodings of Lojban.

??? How do you reach that (seemingly illogical) conclusion?

> If I use the attitudinal ".ui", it is  possible that it will be understood
> adequately per I, without either person thinki8ng through the logical
> structure of the emotions, if such a thing exists.

I think an utterance "ui" would pass through the filter of (II)
unchallenged. It has no logical structure.

> Indeed, to the extent
> Lojban is successful, its speakers will not HAVE to think through the logical
> intricacies, but will merely uise the language and thereby be logically
> correct.  They don;t have to be conscious of the logic in order for the
> loghic to be valid.

What you don't seem to understand is that since we do not already
have a community of speakers who speak Lojban that is in perfect
conformity with the prescription, the only way you are going to
get people who speak Lojban that is in conformity with the
prescription is by their thinking through the logic and other
intricacies. After a while that becomes ingrained and then you
have fluency.

> I think this can work well compared to English MERELY by eliminating the
> inconsistent useages of some and all andany with and withoiut negation.
> No Lojbanist may really understand how this is supposed to work, but
> its quite likely that two Lojbanists can manage to communicate a consistent
> interpretation of a given construct.  that construct then indeed actually
> enocdes the interpretation.  (What may be lacking will begeneralizable rules
>  that allow people to predict how other interpretations might be encoded,
> but I think again that this is premature.  Nick had to analyze 3000 lujvo
> before he came up with his version of dikyjvo.  When we have people who have
> made 3000 different expressions involving "any" and successfully communicated
> them, we MAY be able to figure out ruiles for "any".

This could indeed happen, though it doesn't hold much interest
for me. What would one learn from it?

You think of Lojban as a kind of ugly duckling pidgin that one
day will become a swanlike creole. For you, true Lojban is that
swan in the future. For me, true Lojban is in the Woldemar Codex,
or future editions thereof.

Anyway, we've had this discussion numerous times before, and I
respect your position, and admire the way you have shepherded
Lojban through creation-by-committee to publication, so there's
no need to prolong the discussion.

--And