Message-Id: <m0jdGsA-0002rhC@snark.thyrsus.com>
Date: Wed, 15 May 91 04:06 EDT
From: lojbab (Bob LeChevalier)
To: lojban-list
Subject: lerfu words
Status: RO

Cross-posting to an answer I wrote to someone on the LINGUIST list:
>However, some Turkish speakers (my informant is
>a Turkish Cypriot), in reciting the alphabet, call this letter /Ge/,
>where G stands for a voiced velar fricative, although the sound [G]
>otherwise does not occur in their speech!  (It does in some other
>dialects of Turkish, low-prestige dialects, but that's another matter.)
>Does anyone know of other such instances?

The artificial language Lojban was designed with 'letter names' outside of
the regular morphological system, putting somewhat of a strain on that
system in the process.  As a modern era invented language, Lojban needed
to support the trend towards acronyms, abbreviation, etc. that have arisen.

However, even the consonant and vowels on Lojban alone took up a lot of
'critical' short word space.  Supporting intenrational scientific language
which uses Greek characters, IPA, and various diacritical distinctions
from natural languages, would have further strained a language that was trying
for one word-one concept.

The current design put all consonant names in a 'crack' of otherwise undefined
morphology:  where 'y' is pronounce schwa, these are 'by', 'cy', 'dy', etc.
For all other letter words, of any language, there are various shifts to
indicate upper/lower case and which alphabet - these are 'regular words'.
But the letter names themselves are created by taking any other legal Lojban
word-form (including Lojbanized names, which have a distinct word form) and
appending the word 'bu' that turns it into a letter name.  The vowels thus
become 'abu', 'ebu', etc.  But where "zo'o" is the Lojban indicator of humor
"zo'obu" means " :-) " and its various network permutations, and numerous other
'creations' sprung up almost instantaneously from this rather elegant solution
to what had been a messy problem.

-lojbab