Message-Id: <m0qRvLP-000023C@xiron.pc.helsinki.fi>
Date:         Sun, 24 Jul 1994 00:37:00 -0400
Reply-To:     Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      Lojban as a Pedagogical Tool?
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 9411
Lines: 172

Subject:      Lojban as a pedagocical tool?

Rollin Thomas <raheem@TYCHO.PHYSICS.PURDUE.EDU> asked back at the end of May:
>Has anyone investigated or come up with some ways in which Loglan/Lojban
>coule be useful in second language acquisition (I mean guided SLA, not
>necessarily spontaneous.)  Any ideas?  I am personally not quite as
>familiar with the structure and certainly not the vocabulary of Lojban
>as I am with TLI Loglan...  But anyway, any ideas?

Sorry to take so long to deal with this question.

Investigated?  Not yet, in any scientific fashion.  Thought about?  Yes,
quite a bit.

First of all, JCB in creating the language, used an algorithm for gismu
(Prim) making that presumed something about the way people might learn
vocabulary for a foreign language, based on cognate-like memory hooks
between words.  He has written that he tested this method against other
possible methods, but never described his methodology in detail, nor
published his data or results.  We have instrumented LogFlash, our
vocabulary program, to gather such data that could be used to detect any
cognate effect based on his word recognition scores, but as yet no one
who has used LogFlash has 'finished' and sent us back the data files
which could serve as the basis for a first-cut analysis.

So one way Loglan/Lojban can be used is as an experimental vehicle to
learn ABOUT SLA.

Second of all, the strong differences between the structure of English
(or any other natural language) and Loglan/Lojban forces a person
wishing to express in Lojban to think closely about exactly what he is
trying to say - and what it really means.  Some phenomena, such as what
we call 'sumti raising' (mentioned by Nick Nicholas in a posting
yesterday) are innate to the structure of all nat langs, but in Lojban
are made explicit.  Thus we think more carefully about the meaning of
such things as "I want the book" vs.  "I want that I own the book" vs,
"I want that I at-least-temporarily have the use of the book."  The
former is vague and can mean either of the latter two as well as a few
other things.  The LOjban structures encourage you to think about which.

Thirdly, Lojban, while having an easy to learn grammar, has hooks in the
language to support some of the more 'esoteric', for English speakers,
structures of other langauges.  Lojban has a complete set of perfective
tenses, whereas English and most other languages have a degenerate
remnant.  Lojban has locative tenses as explicit as languages such as
Finnish.  Thus, in learning Lojban, you learn a superset of the
structures (and their semantics) needed in any language.  It then
becomes easier when learning that other language to map the new
strutures to something familiar - a process which is misleading at best
in English.

Fourthly - word semantics.  When you learn the Russian word for hand you
eventually learn that it means arm as well - to a Russian, that thing
with fingers on it is just the end of your arm and doesn't need a
separate word.  I have found it especially good to map foreign language
meanings into Lojban rather than into my native English, because I am
less likely to map onto the words the peculiar connotations of the
English words that are practically subliminal to me.

Fifthly, and perhaps most important:  It takes a long time to learn a
foreign language.  People denigrate the amount of time it takes - our
schools pretend to teach languages with as little as 3 hours per week of
study for 36 weeks of the year, a ridiculous 108 hours of study.  In our
era of minimal homework, this is all most kids do with the langauge
unless they are interested.

By comparison - as a child you spent 12-16 hours a day every day for
several years to learn your native language.  That is 3000+ hours EVERY
YEAR.  And the native language is the standard people use to determine
whether they 'know' a language.  I have seen numbers like 1500 hours of
study as required to achieve near-native fluency.  That would take 14
years in our current schools.

But therein lies the clue.  If it takes you 15000 hours to learn your
native language and 1500 to learn a second language, how long will it
take to learn a third language.  There have actually been studies.  I
don't know how it translates into hours, but studying a third foreign
language takes even less time than the second.  If it takes 4 years to
achieve a certain level of mastery of German, but if you spend 2 years
of French study and then 3 years of German study and acquire the same
skill level, you have effectively learned that third language for an
investment of 1 extra year.

With Lojban, we seem to get about double the learning productivity of a
natural language, or even slightly better.  The Lojban grammar is so
simple and regular that people regularly post good non-trivial Lojban on
their first or second attempts.  In recent weeks, Matthew Faupel, Chris
Bogart, and Bruce M.(?) have all done so.  Thereafter, skill in the
langauge seems to be directly proportional to how many words of
vocabulary you have at ready recall.  To write fluently at the level I
am in this article takes 20-50K words of on-call vocabulary or more.  In
English - or in any foreign language, or in any constructed languages.
There are that many meanings and the words associated with them - to be
learned.

A productive derivative morphology like the suffixes of Esperanto and
Russian, or the affixes of Loglan/Lojban can help, as can cognates
(though semantic errors result from too much of the latter).  But you
still need to learn the words.

We have found informally that procedures like LogFlash are extremely
efficient at teaching the first 1000 words or so.  I'm a poor language
learner, judging by my vocabulary learning (without LogFlash) in 3
natural languages.  But I averaged better than 100 words per week 'cold
turkey' with LogFlash right after we invented the new words, spending an
hour a day for 8 weeks or so.  Few systematic vocabulary methods claim
more than 50 words per week mastered, so we have a lot of room for
individual variation.

(After that first 1000 words, experience seems to be that you learn
words by using them.  Nick Nicholas has now surpassed me in both written
and oral skill in the language 'merely' by his extensive translation
work.)

Thus a major benefit of Lojban in SLA might be to spend 1 year learning
Lojban and then a couple of years learning some real 'target' language.
It is my as yet untested claim that the avergae person doing this might
achieve higher proficiency in French after a year of Lojban and 2 years
of French than they would in 3 years of French alone.

Sixthly, learning Lojban is FUN.  Not quite 'mind candy', 'cause it's
good for you.  If learning languages takes a long time, being able to
make it fun, can make that time pass more quickly.  And the quick basic
proficiency in Lojban gives an early feeling of accomplishment that can
sustain one for the long haul of mastery.

Now for the pipe dream.  Since reading a book about language by Anthony
Burgess, I have long thought about language learning and kids.  Having
two Russian-native kids has heightened my awareness of language interest
and language learning in kids, as well as my awareness of how those
interests evolve with time.

I firmly believe that our schools need an integrated language study
program - not one that necessarily teaches kids a specific foreign
language - if they don't use it, their skill won't develop and they will
lose most or all of what they have learned.  Rather, people need to
learn 'about' language, both their native language, and the wide
spectrum of languages that exist around the world.

I envision a currciulum that would have some coursework during each year
of school up till 12th grade that would teach this, possibly but not
necessarily integrated with English/language arts studies.  The goals
out of this study would be to have a fundamental understanding of how we
turn sounds and symbols on a page into a language, to provide the tools
and skills in learning to expand vocabulary in both the native and in
foreign tongues, and to learn how to learn and maintain a foreign
language for the most part on your own.  Only after completing much of
this curriculum might students select a foreign language to study
seriously.

I envision that Lojban might be useful in such a 'learning about
language' curriculum.  As a simple model of a language, the basics can
be taught simply, and the insights come fast.  Using ideas I've
expressed above, the model Lojban can be used to explain ideas from
other languages that don't convey well in English.  And kids with
relatively little proficiency can produce reasonably good Lojban with
little advanced knowledge.

(I note in fairness that Esperanto could be used equally well for these
purposes and might offer advantages to kids following up with studies in
a European language due to its generic European root vocabulary.  But
Lojban can be used to support learning about all languages equally, not
just the popular-because-easy European languages.)

Hope this was worth the delay.

lojbab
----
lojbab                           Note new address:    lojbab@access.digex.net
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA                        703-385-0273
           Ask me about the artificial language Loglan/Lojban.