Unilang: the Phonology
From: | Oskar Gudlaugsson <hr_oskar@...> |
Date: | Tuesday, April 17, 2001, 23:54 |
In the (highly theoretical) design of a unilang's phonology, I'd start by
by identifying the basic qualities required:
1. Ease. All sounds should be roughly approximable by any speaker of any
human language. Distinctions between sounds should be as basic as possible.
2. Renderability. If we were to take only the first point into account, we
might successfully design a language with only the most basic sounds, we
might have to tell the Frenchman that he lives in "Pilanisi", or the
Englishman that he speaks "Inilisi", or something like that. Not quite
acceptable.
The renderability principle obviously is in direct conflict with the
principle of ease; optimality requires a very careful selection of
phonemes, plus some other tricks we might think of.
One such "trick", at the very heart of my phonological scheme, is what I
might call "the flexibility trick", or "the approximation solution".
Leaving grandieuse terms out of it, here's what I mean:
Instead of defining an absolute value for each phoneme (along with the
usual tired old allophones, such as /n/: [N] __velar), define a _range_ of
sounds; describe the compulsory characteristics of the phonemes (all
vowels, for instance, might be [+syllabic] and [+voice]), leaving free that
which is irrelevant (for example, speakers would be free to assign voicing
to /s/, whichever way they find natural).
At least two aspects of phonology must be considered while arriving at the
definitions: articulatory (primarily), and acoustic. To explain the
importance of acoustics, let's consider this: with a hypothetical /h/
phoneme, which could be pronounced as any posterior fricative, leaving
voicing free might not be a good idea, because a voiced glottal fric might
not be heard at all (or might at least not be perceived as /h/) by many
listeners. Generally, articulatory considerations win out on acoustics, but
sometimes one would have to stop and consider the acoustics, certainly.
Now that I've lined out these basic ideas, I'll continue with another post
in the same thread, with my own phonological scheme, and the definitions.
Óskar