Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!portc02.blue.aol.com!howland.erols.net!usc!news.cerf.net!mvb.saic.com!eskimo!news
From: Richard Wojcik <rickw@eskimo.com>
Subject: Re: Any one-way linguistic evolutionary trends?
X-Nntp-Posting-Host: tia1.eskimo.com
Content-Type: text/plain; charset=us-ascii
Message-ID: <32BCD00A.7B9A@eskimo.com>
Sender: news@eskimo.com (News User Id)
Reply-To: rickw@eskimo.com
Content-Transfer-Encoding: 7bit
Organization: Eskimo North (206) For-Ever
References: <32BB00C2.154E8E4C@ocrat.com>
Mime-Version: 1.0
Date: Sun, 22 Dec 1996 06:07:06 GMT
X-Mailer: Mozilla 3.01Gold (Win95; I)
Lines: 58

Ochlocrat wrote:
> 
> A few other recent threads have touched on (and generally
> dismissed) the question of whether there could be
> any aspects of languages that evolve in one direction only.

Sorry I missed those threads, but this is a complicated question.
The directionality of language change is familiar and widely
accepted among linguists.  It is also well-known that languages 
to "drift" (as Sapir put it) in the same direction over time. 
Witness, e.g., the levelling of noun declensions across Indo-
European languages.  The problem is that no trends are absolute.
The interesting thing is not the observation that such trends 
are exceptionless, but that they are trends at all.  Even if you
can find a few exceptions, that should not lessen your curiosity
about the directions of language change.  Exceptions are data that
help explain the nature of the directionality.  If your theory
can't account for the exceptions, then it needs more work.

> Eg, an alleged trend to evolve from agglutinative to
> inflectional languages, an alleged trend to evolve from
> SVO to SOV order (or was that the other way around), both
> apparently decisively refuted.

Remember that exceptions do not disprove trends.  They only prove
that the direction of change is not absolute.  To disprove a trend,
you have to show randomness.  You would expect affixes to disappear
over time through phonological erosion, but not all affixes are
phonologically vulnerable, and not all phonologies are out to pick
a fight with them.  :-)

[...]
> So are there any trends that evolve in one direction only?
> How about "difficult" consonant clusters?  In Slavic languages,
> many of the most difficult consonant clusters seem to occur
> in the "fundamental" (presumably original) vocabulary, eg:
>         pSenitsa "wheat"
>         ptSela "bee"
> 
> So, a priori, this seems like a trend away from difficult
> consonant clusters.  Are there any counterexamples in
> Indo-European or other language families?

Speaking in very broad terms, you need to understand that phonology exists
for a reason--it coordinates articulatory gestures during speech.  It lets 
you speak sloppily or carefully.  It provides a range of articulation for
you.  The point is that different phonological operations apply to different
speech tempos and styles.  Assimilation and syncope are natural in fast or
sloppy speech.  Dissimilation and epenthesis are more natural in speech
that favors phonetic clarity.  Those Slavic vowels were able to disappear,
but largely because they left phonetic traces on preceding consonants, causing
the C's to undergo major reinterpretation.  Some other languages might prefer to
break up consonant clusters with epenthetic vowels.  It all depends on the
needs of higher level phenomena such as syllables and morphemes.  

-- 
Rick Wojcik                                       Bellevue, WA
rickw@eskimo.com                                  http://www.eskimo.com/~rickw
