[lojban-beginners] Re: anti-Zipfian gismu rant

["Jorge Llambías" <jjllambias@gmail.com>]

On 8/9/07, turnip <turnip@bcpl.net> wrote:

> >{natmritaliano} is not type-3. Type-3 would be {natmrxitaliano}.
>
>  Please explain why?

All type-3 are of one of these forms:

  CVCC-r-C...
  CCVC-r-C...
  CVC-r-C...

(where -r- becomes -n- if one of the adjacents C is an r,
or -l- if one is an r and the other an n.)

{natmritaliano} still happens to be of fu'ivla form, but it's an ordinary
type-4, not a type-3. The reason is that CVCC-r-V... is not guaranteed
to be a fu'ivla. For example, natmriblanu, from a hypothetical nationality
"iblanu", is a lujvo nat-mri-blanu.

So when the borrowed word to be type-3-fu'ivlaized starts with a vowel,
an x is preappended in order to make it start with a consonant.


> > I use the unofficial {talno} for "Italian".
>
>  I would use it, too.   But the gismu list was baselined, so that means no
> more additions, right?

No more official additions, right. But we are talking of using the language,
not of making official additions.


> >{ricyratcu} does not strike me as too long for "squirrel" either,
>
>  Yes, that's true, but squirrel is the lowest frequency word of the high
> frequency group.  I could have used raccoon or rodent, as well, both having
> longer lojban translations, but slightly less frquent in the British corpus as
> well.

That was a frequency list for English words. I don't think I ever talk
of squirrels much, since there are no squirrels where I live. So for me
at least, having "squirrel" in the gismu list would be almost as odd as
having {sfofa}.


>   My point of the exercise was not so much to ask why certain words are
> gismu, but rather why certain others are not, despite their high frequency in
> natural languages,

High frequency in English, in this case. Some concepts will be high
frequency in all/most/many languages, but others, especially animal
names, foods, etc, will probably vary quite a lot from language to
language.

> and (hand-in-hand with this) their basic conceptualness
> (outside of squirrel or zebra, but including things like rodent, which has a
> relatively low frequency, but is a pretty basic concept that could be used,
> and which could replace/augment the ratcu and smacu we have now, and allow
> lujvo for squirrels, beavers, capybara, or any of the 29 families (2277
> species) of rodents).
>
>  (Hey, warned you it was a rant)

I know, I find many oddities in the gismu list too, but on the whole I find
that the range of concepts is covered pretty well. When something basic
seems to be missing and no simple lujvo seems to work, the best policy
is to coin a type-4 fu'ivla (not one of the monster type-3) or even an
unofficial gismu, but other than for names of animals/plants and cultures
this is very rare, which shows that the official list is pretty comprehensive .
I have much more trouble in general with weird place structures than with
missing concepts though.

mu'o mi'e xorxes