[lojban-beginners] Re: Continuing with beginner vocabulary tools ...

[Colin Wright <colin.wright@denbridgemarine.com>]

What follows is heavily edited and I haven't retained
attributions.  Thank you to Chris and Matt for their
contributions.

A summary of my take on this is as follows:

* rafsi and place structures deserve cards of their own.
* They are less important than the gismu itself
* Their cards should be introduced later.

Now read on ...

>> Memory drill systems, including mine, generally require
>> short answers but allow long questions.

> I've included the place structure in all the
> gismu I've been learning.

I don't understand how this can be cast as a question with
a short answer.

> But when I forget one of the higher places, but get the rest
> of it basically right, I create a new item that has only that
> one place, and I make a note on that place in the other item
> so I remember next time I review the gismu.

What system are you using?  How do you "make a note" on an item?
I've wasted 20 hours trying to get SuperMemo running and have
given up in utter disgust.  Maybe other people can get it working,
but I, like some others on this list, use multiple operating
systems in multiple places, and have opted for a simple approach.  
It may be suboptimal in one sense, but something that runs
everywhere is a clear winner.

> I think that a vertical format for gismu places is much easier to
> quickly scan to ensure that you've answered correctly, so I've been
> reformatting all the gismu definitions to look like this:

Are you not checking your answers automatically?


zukte
-------
x1: a sentience
x2: a means, an action
x3: goal of action


> ...I move the verbiage to the line between the places:
> manci
> -------
> x1
> wonders/awes/marvels at
> x2

So, what system *are* you using?


> I think a better way to learn rafsi is indirectly.
> So I propose the best way is to start learning lots of lujvo,

This is worth thinking about - nice idea.




>> My thought has been that one card giving only english->lojban reviewed 
every
>> day is about the same work as two cards, one for each direction, reviewed
>> every other day.  Each card has the same idea but in a different direction,
>> so I wonder if "two cards half as often" is a better solution.

> I asked the Supermemo help about this very question, and they said
> that in general having both directions isn't necessary and ends up
> being more work for you than you'd want.

I've found when using only the one direction that I can't get the other
direction as fluent.  That might just be me - I have known "idiosyncracies"
(to use a polite word)


>> Perhaps for each gismu there should be 10 cards, and that the initial
>> review period is set to 5 days for each card.

> Ack! I can't really explain succinctly why this is a bad idea, but my
> experience with Supermemo has shown that this sort of approach is
> probably very wrong. You'd be much better off spending the extra time
> you save from not having as many flashcards in reading lojban or
> translating things into lojban.

I'm not suggesting they all be introduced at once, that would be severely
suboptimal.  I'm suggesting using only the keyword cards for a while, then
later, several phases later, when introducing some of the less frequent
gismu start introducing some of the detailed places for the more frequent
gismu.

> Alternatively, you could look into branching out into other ways of
> reinforcing the gismu that provide a more natural context for the
> word. For instance, you could use the gismu in a short example bridi
> and convert that into three or four items. The more of your brain your
> items use, the more interconnected your knowledge of lojban will be,
> and the easier you'll find it to actually *use* the language.

This is independent of the task being discussed, which is memorising
vocabulary.  Interlinking by structured usage is another topic that
I have ideas about.



> I'm not sure the jbo-->eng and eng-->jbo of the same word are necessarily
> the same piece of information. For that matter, not all pieces of knowledge
> associated with a given gismu are of equivalent importance. I think it would
> be quite odd to learn the rafsi for word #40 before I learn gismu #200. It
> would be interesting to try to decide how much more often a word is used as
> a gismu than it is used in its rafsi form. Perhaps some means to determine
> this could be devised.

This is why rafsi and places should be introduced later, but perhaps deserve
their own cards.