> > My question is, do we know
> > anything specific about *when* it happens, i.e. whether one particular=
=20
language,
> > given the arrangement of this, instabilities of that, and similarities =
to
> > the other, is *more likely* to undergo a certain change.
>=20
> Well, this is actually a rather complicated question. In short, the answ=
er
> is yes, with ifs.
Ah. On to the if's, then...
Basically, you have to look at the empirical evidence about
> individual cases, which happen more often, in what phonetic environments,
> and so on. Statistically speaking, there are quite a few phonetic change=
s=20
that
> seem to happen very frequently in just about every language family around
> the world. Rhotacism (when [s] changes to [r]), for example, seems to oc=
cur
> a lot.
What if the language in question already has an [r] -- will the pre-existen=
ce
of such a sound tend to hinder this change? How about de-voicing of [r]s?
One particular hypothetical dialect of Aroven does that, giving a sort of
dry fluttering sound in place of the [r] which sounds a lot like an [x];
/r/ is probably the most "unstable" letter in the language...
Exactly why this is, I don't think any linguist really knows. You can
> describe the process (usually the [s] becomes voiced to [z], and then acq=
uires
> retroflex qualities), but in terms of why that should occur, where other=
=20
languages
> are perfectly happy *not* changing their phonetic form (here, rhotacizing=
), we
> (the scientific community) just don't know. So, probably at best, one c=
an=20
talk
> about statistical likelihood, but nothing is dead set. What we do know,=
=20
however,
> is that once a change has occurred, the very strong tendency is for speak=
ers
> (whether when children or not) to spread that rule across the entire spec=
trum=20
of
> the phonology, in every word where the rule can apply, to the extent that=
one
> can call such changes laws (much as Newtonian physics provides a very ver=
y
> close approximation of reality).
Yes, I imagine that if there *is* a way to predict such changes, the whole =
of
the phonology would have to be taken into account, among other things...
>=20
> The conlanging motive
>=20
> > behind this question is the possibility of altering a conlang so that, =
even=20
if
> > it were actually used over a longish period of time by a large speakers=
hip,=20
it
> > *will not change*, the possibility of a "terminal dialect" of the langu=
age,=20
so
> > to speak.
>=20
> Really, well, the problem, as alluded to above, is that some people find =
such
> statistical changes good (in that they will change where change is possib=
le),
> while other speakers will find such changes unnecessary. So, you can't r=
eally
> do anything about that fact. Somewhere along the line, someone will find
> something about your phonology that they find hard or difficult, and will
> act accordingly. Probably the best thing you can do, then, is to design=
your
> phonology in such a way that you avoid as many of the possible statistica=
l=20
changes
> as possible. This will, however, be difficult, if not impossible, as the=
re=20
are
> statistically likely changes that have exactly opposite effects. Take, f=
or
> example, syncope (the loss of medial vowel sounds). Syncope has given ma=
ny
> English dialects the form /plis/ for <police>, because to those speakers,=
they=20
found it
> easier to pronounce the consonant cluster at the beginning than to pronou=
nce=20
the
> only very slightly pronounced vowel in between the segments /p/ and /l/. =
On=20
the
> other hand, many languages detest such consonant clusters (like, say,=20
Hawaiian)
> and will try, as much as they can, to eliminate them.
So a "stable" phonology would depend on the phonological tastes of the
particular speakership (Aroven would be one to immediately shorten to /plis=
/ :)
... so if those tastes change than even the most "stable" phonology is doom=
ed.
Ok, so now the question remains whether a terminal form can be reached for
a language with a fixed set of such habits...
So, you see, you could make
> a language that will change more slowly, statistically speaking, but ther=
e is=20
no way
> for you to engineer people's attitudes centuries from now, when there mig=
ht be
> very different (internalized and unconscious) feelings about what the lan=
guage
> should be like.
Yes, that's unfortunately true. Damn. :-)
> > I've been putting a lot of effort lately into guiding my conlang down
> > such a path, though I've never really been sure if such an expectation =
is=20
even
> > reasonable; certainly you can deliberately use particularly "stable" so=
unds,
> > e.g. no "ps-"s, "pn-"s or whatnot, but is it possible to actually make =
the
> > entire language phonetically "stable"? I know Icelandic hasn't changed=
much=20
for
> > hundreds and hundreds of years, but I wonder if this isn't due more to
> > linguistic conservatism and the small number of speakers than anything=
=20
else...
>=20
> I have a feeling that the rate of language change is closely tied to many=
=20
other
> social changes occurring in any given society at any given time. Take th=
e
> Great Vowel Shift, e.g., which occurred in the English language sometime=
=20
around the
> time of Chaucer (ca 1350s). This was a time of great change in England=
=20
itself,
> when trade with the continent (in things like textiles and raw goods like=
=20
wool)
> was expanding at a fast rate, and thus the influx of foreign ideas and=20
influences
> was increasing, paralleled by increasing urbanization and the innovation =
of=20
new
> agricultural methods, which during the first part of the century brought =
about=20
a
> burgeoning population, barely able to be fed by even these new methods.
> Of course, there also followed the Black Death, which then reduced the=20
population
> of Europe by about a third, which even further revolutionized the social=
=20
arrangements of
> society (empowering those serfs who had previously been strictly tied to =
the
> land and thus were better able to demand better living conditions from fe=
udal
> lords and other such things). So, in general, you're talking about a time=
of=20
truly
> great change.
>=20
> I don't think you can point to any one of these events as the source of w=
hy=20
the
> language changed, but I would be willing to bet that the change itself is=
=20
related
> in some way to the total impact of these changes, if only indirectly. It=
must=20
be
> noted, however, that much of this (in this particular problem) is just=
=20
speculation.
>=20
> Now, compare that to Iceland: relatively isolated from the rest of the w=
orld,=20
to
> the point that it would be affected by social changes only many decades i=
f not
> centuries after the original development of the change back in Europe. =
=20
Icelandic
> has been able to remain pretty much the same because the population has a=
lways
> been relatively small (even today there're only about 300 thousand people=
=20
there)
> and homogenous, thus leaving the innovators few and far between (as great
> innovators will always make up an extremely small minority of any populat=
ion).
> Icelandic's "conservatism" doesn't have anything to do with what the peop=
le=20
there
> think about language (as someone else noted recently, most people everywh=
ere
> think their language is exactly the same as that of their youth, and of t=
heir=20
forefathers,
>=20
> but this is selfevidently not the case); it has to do with the social for=
ces=20
that have
> affected the Icelanders since their first settlement in the late tenth ce=
ntury=20
(ca. 950s,
> IIRC), which were themselves, not the people, tending to cause few change=
s in=20
the
> language.
>=20
> So, getting the point: here too you see that you can't really foreordain=
what
> your language will be like many centuries or millennia hence, because all=
=20
sorts
> of social events will intervene which you could never have foreseen. (Yo=
u'd=20
have
> to be a psychohistorian like Asimov's Hari Seldon to do that).
I guess the only thing to do, then, is to design a language which is
self-consistent, relative to one's own particular mindset -- seeing as the =
shift
to any other particular mindset will inevitably set linguistic changes into
motion...
Thanks a lot for your input
Josh
_/_/ _/_/ _/_/_/_/ Joshua Shinavier =20
_/ _/ _/ Loorenstrasse 74, Zimmer B321=20
_/ _/ _/_/_/_/ CH-8053 Z=FCrich =20
_/ _/ _/ _/ _/ Switzerland =20
_/_/_/_/ _/_/_/_/ _/_/_/_/ http://www.delphi.com/aring