Re: I was wondering...

From: "Arthaey Angosii" <arthaey@...>
> >sounded dumb, there wasn't enough variation, i.e. I seemed to have favorite
> >sounds that became too predominant.  There are around 72,000 potential CVCVC
>
> But, at least in English (the only language I speak fluently), all
> potential combinations are _not_ distributed evenly. For example, of all
> digrams in English TH occurs 3.15% of the time compared to LY's 0.47%. So
> it seems to me that having a "favorite" sound would make a language _more_
> natural, not less.
Of course, in LangMaker you can set up percentages.

For Proto-Trentish I have:

Patterns
CV 30%
CVCV 50%
CVCVCV

C =
t n s r 7%
pr tr c cr kr J ?r 3%
p k m N  sr ? h

V =
a e i o u 7%
la le li lo lu l l: 3%
a: e: i: o: u:

My general method when making Trentish roots is to create a small number of
roots (maybe 3) and choose an equal set of meanings, then choose which fits
which best, which I guess is a good enough trade-off between words sounding like
what they "ought" to sound like and the fact that words generally don't have to.

For example, this page here has
*sete  "kill"
*kre   "all"
*kra   "earth"

*mi:Ne:      "tail"
*ke:sri      "claw"
*sra:cru:mo  "person"

(These, theoretically, come out as Trentish *set'e, *xe, *xa, *mjiNje, *khesli,
*SatKumo, respectively.)

This is not reflected in any of the forms on the Trentish website yet, which
hasn't been updated lately.

    *Muke!
--
http://www.frath.net/