Re: Language change that complicates the syllable structure

Quoting Isidora Zamora <isidora@...>:

> >What little I've read of historical linguistics suggests that weakening
> >and elision of consonants is much more common than elision of
> >vowels and epenthetical insertion, which by the way are the only two
> >syllable-complicating processes I know of. Can other kinds of language
> >change (ie. non-phonological) influence the syllable structure in a
> >complicating way? The arise of compounding? Cliticization of particles,
> >and subsequent incorporation in an inflectional/derivational system?
>
> My logic tells me that, if you are starting with a language that has a
> strictly (C)V syllable structure, then no amount of compounding is ever
> going to get you a consonant cluster.
>
> There are a couple of suggestions that I can make, and they all involve
> vowel loss.  (In the phonology textbook that I am currently reading for
> fun/research, I have seen a number of examples of loss of word-final
> vowels.  (There is actually a tecnical term for loss of a word-final vowel
> but I can't remember what it is, and I can't find it in the textbook
> quickly, and don't have the time for a more thorough search at the moment.)
That would be "apocope". Loss of a medial vowel is "syncope". Presumably,
there's one for the loss of initials too.

> Loss of word-final vowels, followed by compounding, would give you
> consonant clusters.  However, you would still never see a consonant cluster
> at the beginning or end of a word, only in the middle.  Additional loss of
> some word-initial vowels could give you some word-initial consonant
> clusters.
Another way is to syncopate* some the vowel of initial CV syllables, creating
a cluster with the consonant of syllable no two. This happens alot in the
development from Classical Klaish to Tairezazh; eg _derenu_>_dren_**. This
being Tairezazh, there's also plenty of examples where there already were
clusters; eg _kestrolu_>_kstrol_. And in some cases the resulting clusters are
illegal, and thus mutate to something more acceptable; eg _ñerestu_>_grest_.

Now, of course, this is a conlang. But the process is well attested from
natlangs - it's, f'rinstance, the origing of "f'rinstance", as well as
Georgian's trademark initial clusters.

* I'm never going to understand the average IAList's aversion against deriving
irregular and unnecessary verbs from everything. And that guy Mark Reed posted
a link to thought that consonanty words with haceks all over were ugly ...

** Syllabomangler's rule #1; don't apply a rule that messes up syllables -
apply two!

> You could also have the language undergo a process similar to Havlik's Rule
> in Russian.  (Havlik's Rule was a natural  process that really seems too
> weird to be true.  It seems like someone must have made it up
> artificially.)  Havlik's rule dramatically altered the structure of a lot
> of Russian words leading to some great new consonant clusters.
Isidora, would you care to elaborate a bit on this? I'm not into Slavicistics,
but now you've made me intrigued!

> My last suggestion is not something taken from natural language, but an
> idea mostly out of my own head.  Doesn't Japanese devoice vowels between
> voiceless obstruents, or something like that?  Have a process like that,
> then take it one step further and eliminate those vowels altogether.  That
> should give you some consonant clusters to be reckoned with.
Japanese is very possibly heading that way. In certain dialects those
voiceless vowels are already completely gone, or so I've heard.

                                                        Andreas

Reply