Re: [lojban] idea: selma'o fu'ivla

On 24 July 2013 20:17, Wuzzy <almikes@aol.com> wrote:

Problem: To refer to a selma'o in Lojban.

Solution: There are already some of these. But none of these does the
most obvious thing: Creating new brivla to fill the x2 place of “cmavo”.

Well, the experimental solution that is used almost throughout the entire community is not one involving {la'oi}, but rather {ma'oi}, which is at least not a member of an experimental selma'o. (ma'oi is in ZO, whereas la'oi is in ZOhOI.)

My solution: Create a bunch of stage-3 fu'ivla which refer to a selma'o
each.

I like this idea.

It works like this: Use “cmavo” as gismu and append the lower-case’d
name of the selma'o you want to translate.

Examples:
“BAI” → “cmavrbai”
“DOhU” → “cmavrdo'u”
“ROI” → “cmavnroi” (rare case of “n” hyphen here)

The only purpose of a selma'o fu'ivla is to refer to a selma'o, nothing
more. Thus, the definition of a selma'o fu'ivla is like “x1 is the
selma'o BAI.”.

What's important when making up definitions is not what they are in English, but rather what they are in *Lojban*. How would you define these fu'ivla in *Lojban*? (Also, the more correct term in this case is {zi'evla}. It doesn't make sense to call {cmavr<whatever>}, a fu'ivla as it isn't a loanword by any stretch.)

I'd probably define these in Lojban with {x1 se cmavo zo bai .e zo ...} and list all the elements + {fo lo lojbo}. The concise definition would be be {x1 se cmavo zo bai ra'u fo lo lojbo}.

Remaining problems:
fu'ivla do not allow “y” nor do they allow a vowel right after the
gismu part.
Thus, there has to be some kind of hacking around the names of the
selma'o “A”, “BY”, “I”, “Y” and “UI”. My current attempt is to use “lr”
as seperator if the selma'o name does not contain a “Y”. If it does
contain a “Y”, I append a valsi which best describes the selma'o.

Customarily, when the zi'evla-head begins with a vowel, we prefix [x], not [lr], e.g. the fu'ivla for "anime" {skinrxanime}..

Here’s the list of irregegular selma'o fu'ivla produced with above
rules:
A cmavlra

zo cmavrxa .e'u

BY cmavrlerfu from {lerfu}

.i'e

I cmavlri

Sadly, adding [x] doesn't work here as it clashes with selma'o XI.

Your proposal for the l-hyphen works, but in my opinion, it should be limited as much as possible to the case where it is necessary, i.e. between r and n.

UI cmavlrui

the [u] in {ui} is a consonant, so {cmavrui} is fine.

Y cmavrdepsna from {depsna}

.i'e

Using “lr” as seperator is kind of a hack and personally do not like
it very much since there is an “l” hyphen
used although it is not neccessary here. However, the “l” makes clear
that the following “r” is not part of the derived selma'o name.
“cmavnra”, for example, could be misunderstood as “RA” instead of “A”.

However, the vast majority (>110 out of 120) of the selma'o are
completely regular. Most of them have the form “cmavr” + selma'o name.

Example sentence:
“zo dau cmavo lo cmavrpa” translates to “‘dau’ is in the selma'o ‘PA’.”.

Of course, you could also just say, in perfectly ordinary grammar, without the use of any vocab more advanced than gismu, {zo dau cmavo lo se cmavo be zo pa}, or with slightly fancier grammar, both {zo dau cmavo zo'e ne zo pa} and {zo dau cmavo lo me zo pa moi}. :-P

I even wrote a proposal on:
<http://mw.lojban.org/index.php/selma%27o_fu%27ivla>
It is a bit more complete than this e-mail. It also contains some
algorithms (spaghetti code!) to create a list.
But beware, it’s still the very first draft.

Comments? Ideas?

Overall, I like this idea very much. Moving things as much as possible away from strange grammatical magic (e.g. {ma'oi pa}) and into simple predicates is a good practice. Furthermore, we gain the ability to create tanru with these new selbri, whereas before we were required to use {me}, a forethought device of sorts.

.i mi'e la tsani mu'o