Message-Id: <199601261547.RAA01049@xiron.pc.helsinki.fi>
Date:         Fri, 26 Jan 1996 10:44:27 -0500
Reply-To: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      Re: LogFlash and LESSY (!)
To: Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 5645
Lines: 122

>>This is basically how I haveuse LogFlash to learn Russian vocabulary.  But
>>I quickly found that it was more work to create the vocabulary lists than
>>to simply memorize them, so I stopped at around 200 words - not enough to
>>do anything with in the language.
>
>That hasn't been my experience with using FlashCards (aka Mac LojFlash and
>Mac LogFlash) to learn Dutch vocab.  I've accumulated a file of about 2200
>Dutch entries, starting from the word lists in a Dutch textbook and then
>adding a few words every time I do any reading in Dutch.  FlashCards allows
>editing the data file with any text editor to add new words or to correct
>existing entries, without forgetting the user's status.  It also has some
>flexibility about the input format (driven mainly by the need for it to
>work with data files from several independent sources).


Well, it might also have been the difficulty with working with Russian
keyboard %^)

But seriously, I found it difficult when preparing word lists from textbooks and

and dictionaries, to get an ideal set of keywords.  The dual use of the
program for recognition and recall, and the requirement that recall have a
unique prompt (or you will truly have no way to tell which meaning is
intended unless you have reduced things to a word-word memory link rather
than a concept-concpet one), seemed to make it hard.  maybe I just seemed to
run into a lot of word sthat had multiply identical keywords.  Of cour
se,
working in Russian, a lot of early words are what in Lojban are cmavo,
but in Russian they are not necessarily so.  Example
s are Russian "tut",
"zdec'" both meaning "here", used almost id
entically.  How do I come up
with a distinct keyword?

I also faced the problem that I was still learning Russian declensions.
It is fine to add in the infinitive of the verb, but then you need the
nth person singular/plural, and for the nouns (especially the irregular ones

the instrumental plural (or whatever), and you end up having MANY entries for
a few words, or you give up on typing in all the declensions, and have to
learn which declension you have typed in (and whether you have coded thi
s
declension into the answer you have to type.

In othger words, a lot fo the problem I have had with creating word  lists
is making clear and consistent keywords, with a matching style needed if
use a phrase.

Let me give a couple of examples from our cmavo data set

>ki'u      BAI      because of reason                         krinu modal, 1st
>place (justified by)             justifiably; because of reason ... mu'i
>BAI      because of motive                         mukti modal, 1st place
>because of motive ... ni'i      BAI      because of logic
>nibli modal, 1st place                            logically; logically
>because ... ri'a      BAI      because of cause
>rinka modal, 1st place (phys./mental)             causal because ...


Try that again:
>ki'u      BAI      because of reason
>mu'i      BAI      because of motive
>ni'i      BAI      because of logic
>ri'a      BAI      because of cause
>ki'unai   BAI*     despite reason
>seki'u    BAI*     reason therefore
>seki'unai BAI*     reason nevertheless
>mu'inai   BAI*     despite motive
>semu'i    BAI*     motive therefore
>temu'i    BAI*     motive of person
>semu'inai BAI*     motive nevertheless
>ni'inai   BAI*     despite logic
>seni'i    BAI*     entails therefore
>teni'i    BAI*     under logic system
>seni'inai BAI*     entails nevertheless
>seri'a    BAI*     causal therefore
>teri'a    BAI*     causal conditions
>seri'anai BAI*     causal nevertheless
>ri'anai   BAI*     despite cause

Loglanists even may recognize from the jeywords that tehse are the 4 classic
Loglan causal roots, though the Lojban words for them are different and we
have added to the set of 4.  BUt having 4 words meaning "cause", and each of
them having a systematic "declension" that turns "because" into "nevertheless"
it was VERY difficult to come up with keyowrds that were brief enough to be
typed error-free, clear and unique, and yet no misleading or too clumsy in
English.

Thus for logical entailment, we could not merely
say "entails", because fot
the other 3 causals, there was no verb like that which could stand on its
own
and we have found that we are asking either for errors, or for Lojbanists
not to learnt he patterns (or both), idf we did something like:
>seki'u    BAI*     reason therefore
>semu'i    BAI*     motive therefore
>seri'a    BAI*     causal therefore
>seni'i    BAI*     entails

They will usually write the "therefore" on the last one, so we made it part

of the keyword.

Well that is fine, but then the same thing holds true for:
>seni'inai BAI*     entails nevertheless
which is a comfotable English phrasing and includes the key "nevertheless"
expressed by the seXXXnai structure.  But then you have:
>seki'unai BAI*     reason nevertheless
>semu'inai BAI*     motive nevertheless
>seri'anai BAI*     causal nevertheless
which are fairly clumsy in English - but the best we could come up with.

Playing this kind of games with keywords took a LOT of time with LogFlash
data files, and Nora and I both checked the files the hard way by learning
them and changing them when they turned out to be difficult (and changing
one of those keywords after you have memorized it with LogFlash is VERY
difficult.  It has been 10 years or so since Nora last used LogFlash with
TLI Loglan words, and yet every once in a while she will dredge up the TLI
Loglan word for a keyowrd rather than the Lojban word, especially when
using LogFlash after a lapse, but also (though more rarely) in speech.

lojbab