[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: What are we after?

Yes, I agree. I rarely use the IRC channel, but would someone here ask
kribacr to join the mailing list? -Eppcott

On Fri, Sep 25, 2009 at 2:11 PM, Mark E. Shoulson <mark@kli.org> wrote:
> Hash: SHA1
> OK. ÂSince I wasn't paying really paying attention during the flurry of
> emails about this (a case of TL;DR, which I totally do NOT promise won't
> happen again), I went to the googlegroups site and read through the
> archive. ÂHere are some of my thoughts in response to everything being
> discussed:
> I think we sort of need to get a little more back to earth on this
> project. ÂNot that the discussions have not been brass-tacks practical:
> you can't get more reality-based than actually starting to put together
> the database. ÂBut I think there is a lot of very lofty, far-reaching
> ideals that Lojbanists are so fond of, getting in the way of the grubby,
> impure task ahead.
> So, to start with the impure: *certainly* we should start by designing
> the lowest level first. ÂEven if we have the language proficiency for
> higher levels (which itself is questionable, depending on how high we're
> talking), we don't have enough (any) experience with how testing goes
> and how this will work in the community to handle anything higher than
> MAYBE three levels. ÂEven three is a little iffy.
> That's why I started where I did in the googledoc, to give an idea of
> what kind of proficiency we should be looking for at each level. ÂBut I
> went about it the wrong way. ÂLike everyone else, I was too caught up in
> designing the tests to pay attention to the more basic requirement of
> getting a handle on wtf we're doing here.
> So that's what we should do, I think. ÂTry to agree amongst ourselves in
> a more informal (but therefore understandable!) way what it is we expect
> a "Level One Laureate" to be like. ÂWhat should such a one be able to do?
> And we should enlist help with that also. ÂEspecially because, as
> old-timers, we forget what newcomers are like and what it's like being
> one. ÂWe need to talk to kribacr from IRC; he's spent a lot of time
> doing teaching duties on the IRC channel, and working on his order of
> presentation of concepts. ÂIf that's an order that people seem to learn
> comfortably, then that's also a good order to test people with. ÂAnd
> maybe later talk to lindar, as one of his successful students (and mine:
> I started lindar off and kribacr took over).
> So, to start things off, when I think of what I would see in someone who
> had completed Level One certification (and just numbering them 1, 2, 3,
> is probably best), I think of somebody who can manage short, simple
> sentences, involving simple sumti (mi, do, ti...), basic brivla (mainly
> gismu, but chosen by utility and frequency, so if there's a
> commonly-used lujvo, no problem) and 2-piece simple tanru of them, NU
> abstractors, but only one level deep, and some conversationally-relevant
> UI and COI. ÂI tried to express that in the document, but was thinking
> more in terms of test design. ÂTry looking over it with testee
> description in mind and maybe we'll be able to discuss things more
> meaningfully.
> There's other stuff I probably would comment on, but this post is
> already way too long. ÂLet's start with the basics and work our way up.
> ~mark
> Version: GnuPG v1.4.9 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
> iQEcBAEBAgAGBQJKvQfhAAoJENKD8527nYua1h0H/0+yXlU7HT+VyZg2I7NlcjAM
> 8WWfL1awl/DWTGgi2pql2U8CPz/ytJK021L4TxQuviG3DHJZQnQ8NSLFH08wgMU0
> X36J0qW7CvoCmxP+aTuYwWDPrWHuBhUma4Vj4bu93af/8t2K7+qlde9NFIzLsknU
> tG7svV1pBN6e9aw3o6NnuA3CvtqCVEoNxJvsOmdCJnzNNyWFHIvhzLtYf1rDKcpE
> 1zX4h7iniOETMkLEaXukQke+j/u2B6YyA8L673Oi+U5kS3lRJ0GmKQj0guXXTGFk
> TsaPeA801c0fCuCXPL5X8Ki9kxMoox/jijwhFFR27dqBAVHro4/b+i0wEUPAPow=
> =nLR3