Message-Id: <199601171115.GAA27789@locke.ccil.org>
Date:         Wed, 17 Jan 1996 11:31:20 BG
Reply-To: IAD@BGEARN.BITNET
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: Ivan A Derzhanski <IAD@BGEARN.BITNET>
Subject:      Re: TECH QUERY: variant fu'ivla
To: John Cowan <cowan@LOCKE.CCIL.ORG>
In-Reply-To:  Message of Tue, 16 Jan 1996 14:45:23 -0500 from 
              <cowan@LOCKE.CCIL.ORG>
Status: OR
Content-Length: 1473

On Tue, 16 Jan 1996 14:45:23 -0500 John Cowan said:
>Lojban has the feature that there are many ways to write most lujvo [...]

And just as many ways to pronounce them, of course.

>A parallel situation exists in type 3 fu'ivla (those made with gismu-based
>prefixes) [...]: "ricrxacere" and "tricrxacere" are both possibilities for
>"maple" (genus >Acer<).

Not {-akere}?  There is no {c}-sound in Latin, and we shouldn't make
too much of the (necessarily arbitrary) choice of Roman letters to
represent the sounds of Lojban.  (Meaning that I would like to think
that Lojban would sound the same if it had a wholly different spelling
or even a different alphabet from the outset.)

>So far, the Lojban community hasn't taken a position on whether or not
>these are to be taken as equivalent.
>
>I favor declaring them equivalent:

So do I.

>while this limits the theoretical
>size of fu'ivla space, it makes for simplicity: [...]

It doesn't limit the fu'ivla space, because you wouldn't use {cpirdodo}
and {cipnrdodo} to mean different things anyway.

While we're at the subject of fu'ivla and their shapes, what about
fu'ivla starting in {CCV'V-}, where {CCV} is a classifying rafsi and
the original word starts in a vowel (or a vowel preceded by a consonant
that we choose to ignore), say, {cpi'alauda} for `lark' (Alauda),
{cpi'irondo} for `swallow' (Hirondo)?  As far as I can see, such
words don't run the risk of being parsed as something else.  Comments?

--Ivan