Choosing Phonemes

adapted from a 1 May, 1997 post by Kapitano Engelfo

Pat Dunn wrote:

I've got a couple more questions for my paper.
1. How do you decide which phenoms to use?

Assyming that's 'Phonemes'... :)
Probably everyone designing auxlangs uses the same three criteria, which may be in comflict:

Maximal differentiation of Articulation. This means that you choose phonemes which are not easily confusable with one another. They may be confusable because they have almost the same place and/or manner of articulation (eg, the 'light' and 'dark' L, as in LET and FALL respectivly.
As I see it, currant thinking is that Place-of-Articulation is better at distinguishing phonemes that Manner (voicing and aspiration mainly). This /p/ is most destinct from /t/ than /p/ is from /b/. /p/ is roughly as destinct from /f/ as it is from /v/.
Maximal differentiation of reception.
Phonemes may be confusable because they sound similar, even if their articulation is different, eg The Spanish-R is made with the tongue-tip, but the German-R with the tongue blade [that is right isn't it, everyone?]. Also the latter may be confused with the 'gh' sound in klingon, which isn't even a trilled/rolled R-sound.
So, Auxlangers try to avoid phonemes that sound similar. Some say that Zamenhof made a mistake in esperanto when he included both the /h/ (as in hot) and /x/ (ch as in loch, bach).
Ease-of-pronunciation, for the majority of possible (or probable) speakers. Usually this means that sounds common to major languages are used.
/N/ (ng) is strongly differentiated from /n/ by place-of-articulation, but most natural languages do not distinguish between the two. TH is (IMO) a highly destinct sound, but most speakers would find it difficult to pronounce, because it is a rare sound, even as an allophone.

Allophones (sounds the are noticably different by are assimilated into the same phoneme, as for the two Ls above) are important in auxlang design, and sometime tricky.

In some natural languages /v/ is an allophone of /f/, but for some it is an allophone of /w/, and for others of /b/. /b/ may or may not be distinguished from /p/, which may or may not be distinguished from /f/.

In my embryonic auxlang Mono, I distinguish between the following phonemes, most of which have several allophones. There is nothing unusual in this list, though most would disagree with some points:

/p,b/
/t,d/
/k,g,tS,dZ/
/f,w/      [/v/ not permitted]
/s,S,z,Z,ts/
/x,G,ks/
/m/
/n,N/
/l,j,r/
/a/
/e/
/i/
/o/
/u/

[Actually it's more complicated than that, but I'll leave that for another time. I hope this hasn't been too patronising for the old pros on the list] PD:

2. Do you have any interesting anecdotes, stories, or blurbs about conlangs that illustrate, well, anything?

Carrie A Schutrik (caos+):

Well, um, er...I remember being amazed when I discovered that my then-boyfriend had an invented language too...and then he was

Well, last night MY boyfriend announced that (after 10 months) I had convinced him that Esperanto is not a nerdish waste of time, and that he wanted to join me in learning it!! This morning he decided that our relationship was a failure and he wanted me to move out.

I don't plan to try to get back with him, but I do plan to continue to learn Esperanto.

Return to Conlang-related topics|Back to FAQ page