Re: more help, this time on sound change

En réponse à The RipperDoc <ripperdoc@...>:

, but I wonder if I should do the sound change with
> inflected words? Isn't that what makes grammar irregular, that sound
> change
> affects some inflections but not all? Am I supposed to inflect _all_
> words?
The Neo-Grammarians (the School of Prague) described sound change as a "blind
force" which applies in any case (i.e. only restricted by phonology, but
absolutely not by the part of speech of the word, whether it's a root form or
an affixed one, and in whatever grammatical form it is). In other words: sound
change is absolutely regular (appears in all situations where a certain
phonological context is met, whatever kind of word it is) and may very well
ruin grammatical regularities and create plenty of irregularities. To counter
this brute force though, there is the so-called phenomenon of analogy, which
consists of replacing an irregular form by a more regular one (that's what for
instance changed the plural of "cow" from "kyne" to "cows", analogy restored a
regular plural). But analogy is not a regular phenomenon (in the same kind of
example, it didn't "correct" plurals like mouse-mice, child-children or ox-
oxen). So in the Neo-Grammarian point of view, you have the "disease": sound
change, a brute force, blind to any grammatical features (so you have to apply
you sound changes to all *surface forms*, that's to say to all inflected forms,
even if it ruins the sense of root - French suffered that, i.e. words which in
Latin were obviously from the same root look now in French remotely similar,
but you cannot make a single root out of them. An example is the verb "aimer":
to love, stem "aim-" /Em/ vs. the corresponding noun "amour": love, stem "am-
" /am/. Though the orthography helps to see the connection between them, in
sound the two stems don't seem any more related together than with any other
one-syllable stem -) and creating irregularities, and you have the "cure":
analogy, which restores regularities but acts irregularly and usually leaves
some irregular forms unchanged.

Nowadays this extreme point of view is not hold anymore. Sound change is still
seen as regular (though its "absolute regularity" is questioned, but still
taken as a good hypothesis as long as it's not proven wrong), but not
restricted only by the phonological context, but also by grammatical features
(like the part of speech or the grammatical form itself, like for instance a
feature like "past tense"). Still, sound change applies to fully inflected
form, not roots and affixes separated, and still mostly creates irregularities.
It's just not that powerful anymore, since grammatical features can restrict it
and thus keep grammatical systems from being ruined. The view about analogy
didn't change though. It still produces regularities by abandoning irregular
form for more regular ones, but doesn't always apply to all irregular forms.

As a conclusion, yes you must apply your sound changes to fully inflected words
(not roots and affixes separated), but you may restrict your sound changes to
certain parts of speech or certain grammatical forms (or rather, have certain
grammatical forms prevent the sound change, as if the pressure of the system to
remain was strong enough to prevent a sound change which would have broken it)
to keep some systems working, and don't forget that even after the sound
changes, you can use analogy to replace some irregular forms by regular ones,
remembering though that analogy usually leaves a few irregular forms unchanged.

> I also have an other problem with the sound change, but I guess you
> can't
> help me on that. The problem is that most of my words doesn't change at
> all,
> but some words are affected by lots of rules and therefore change
> quite
> much. Is that a realistic approach, that some sounds change but others
> don't?
>
Of course! The French initial /p/ for instance can be traced directly back to
the Latin initial /p/! Remember though that sounds usually have different
histories depending on their position in a word (for instance, vowels tend to
change little in the first syllable or under the accent, but a lot - often
until disappearance - in other positions and especially post-accentual.
Consonnants also tend to be more preserved in word-initial position than in
word-medial or word-final position), which may add more variety to the sound
changes, and that normally a majority of words change (at least a little) after
a few sound changes. But it's absolutely normal that some words don't change at
all and others become completely unrecognizable (so much that the language has
to invent new ways to express what those words used to express).

Christophe.

http://rainbow.conlang.free.fr

Take your life as a movie: do not let anybody else play the leading role.

Reply