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

Re: [lojban] Conversational ability test.



Of course, we can only guess at the type of mistakes that'll be made, but if the listener can't understand, the speaker can expect a [ki'a] and then start over with [sa] or [su], as required. This test seems like a good idea to me too. (I mean, I did come up with it practically simultaneously with Robin) 

From my point of view, zoi-quotes shouldn't be allowed in such a way that they quote another language. All test subjects should pretend to speak exclusively Lojban. This way, it'd be necessary to describe what is meant using only Lojban.

On a side-note, I'd expect zoi-quotes to be used only to quote non-linguistic material, such as hyperlinks (as clearly demonstrated on IRC) and onomatopoeia. 

Back on topic, it's clear that a certain amount of vocabulary is necessary, at least in terms of gismu. (In fact, knowing only gismu, it's possible, I believe, to avoid using all the TAG cmavo at the exception of [fi'o], should [fi'o] function exactly as does TAG itself in all contexts.) Otherwise, a already demonstrated grasp of the grammar would obviously be required. Typically, I observe, rather skilled Lojbanists do clearly understand the grammar, yet are simply lacking in the department of vocabulary. (I myself am a victim of vocab-laziness.) I estimate that knowing the 700-or-so most frequently used gismu is a sufficient amount when it comes to solving these various puzzles. 

As for the puzzles themselves, I'm not certain as to what could potentially be implemented. Throwing switches in a coordinated fashion could test for comprehension and correct use of ordinals, as I believe was pointed out by Robin in #lojban. To test for directions, perhaps, we could construct a sort of maze or obstacle course. The testee is required to escape, but is blindfolded and can hear nothing but the voice of the tester, who omnisciently observes the maze from above, or through a camera affixed to the head of the testee. Following only the tester's instructions, the testee would have to escape.

Any other thoughts for tests?

mu'o mi'e la tsani

On Wed, Jan 4, 2012 at 6:02 PM, Pierre Abbat <phma@phma.optus.nu> wrote:
On Wednesday 04 January 2012 13:40:36 .arpis. wrote:
> Are "sentences" made up of lojban words which are incorrect
> grammatically lojban? What about if both people misuse (by implicit
> contract) certain words (I can't think of a good example now)? What if
> they habitually forget to separate the main bridi from the sumti
> before it? What about lojban with a lot of "zoi" and "zo'oi" and "me
> la"? Where is the line drawn between incorrect lojban, bad lojban, and
> not lojban?

I think that using bare foreign words should be penalized more heavily than
quoting them with ZOI, which should be penalized more heavily than
grammatical errors. It may be difficult or impossible, in certain cases, to
avoid foreign words. For example:
(alis) .i ko lebna le .arxokuna le tricu
(bab) le moki'a
(alis) .arxokuna
(bab) lo .arxokuna cu mo
(alis) me la'o sy. mapache .sy.
(bab) ua .i mi punji le .arxokuna ma
I'd ignore, or score very lightly, the sorts of errors that everyone makes
when speaking, even in his native language. For example, I've heard a Spanish
speaker say "otro", then realize that the noun he's about to say is feminine,
then say "otra" and the noun. Of course, until we have some native Lojban
speakers with enough proficiency, we can only guess what sorts of errors
those will be , but here's one example. You mean to say "jmina", but get the
nasals mixed up, and half devoice the "j", resulting in something
between "cnima" and "jnima". You then say "si", but you don't know which the
listener heard. The preceding word is a CV cmavo like "mi". So you don't know
whether the hearer is erasing one word "cnima" or two "mi jnima".

Pierre
--
li fi'u vu'u fi'u fi'u du li pa

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.


--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.