Re: Never violate a universal unless it seems like a good idea at the time

>I'd be interested to know how many universals other people's conlangs
>violate, and whether this came about naively or deliberately.
I've just started seroiusly designing a few languages very recently (like
about a month ago.)  However, I received a batchelor's degree in
linguistics 9 years ago, so I knew that there were such things as
universals before I began.  (And there are universals in the realm of
phonology as well, which is where I really knew them from.  I never
understood syntax that well; it was all very theoretical and confusing when
I took it.  The class was also highly geared toward getting you to
subscribe to Chomsky's Unversal Grammar.)

In any case, knowing that there were ways to do unrealistic things with
phonology, and wanting to avoid perpetrating any of those, I dug out a
graduate-level phonology textbook that I had bought before I left college
and started going through it.  (I am now working the excercizes for ch. 3,
and wish me luck, because some of those excercizes are brutal, or at least
confusing.  I'll be asked to do language reconstruction pretty soon.)

My next steps will be textbooks on historical linguistics (which was not
taught at *all* at my school), morphology, syntax, typology, and
semantics.  (Ok, I'll admit that this is going a little overboard for the
purposes of conlanging, but for me it's a fun thing to work on completing
my education.)  Oh, yes, and probably that book on tone that someone
mentioned on the list within the last couple weeks, since one of my
conlangs is a tone language.  (Why I am putting myself through the pain of
a tone language, I don't know, I really don't.  When I was in school, I
could *never* remember whether I was saying 'milk' or 'cucumber' in
Chichewa.  I could get the tones right but never could remember which one I
was saying.  And then at the sentence level there's tone sandhi and things
like that.  I'm going to end up with a conlang that I am *completely*
incapable of pronouncing correctly.)

So right now major development on the languages is stalled while I learn
more about the nature of language itself.  (And at the current moment
further self-education in linguistics is stalled while I am working on
better constructing the cultures who speak the languages.  In my case, the
languages exist because there are people who speak them, and the people
exist because they are in a story.)

So, to answer your question about the violation of universals, I am taking
pains with all of my conlangs not to violate universals nor stretch
credulity, because I want the languages (and the cultures) to be as
plausible as possible.  (But this doesn't mean that I don't put unusual
things in them that may be difficult to pronounce, such as a very nearly
complete series of labialized consonants.)   I had decided that a certain
one of my languages would be OSV (it seemed reasonable to me), and then
someone posted on the list not long ago a breakdown of what percentage of
languages had each of the basic word orders.  OSV was, to my surprise and
dismay, right down at the bottom, with 0.1% of the world's languages having
OSV word order.  So I am rethinking my decision, since it would
automatically make this language one in a thousand, which is pretty
marked.  On the other hand, perhaps I will just leave it marked and know
that it is one in a thousand.  (Only problem here is that I already know
that it's got a bunch of sister languages.  Are they all OSV or just this one?)

Does anyone know anything about that 0.1% of natural languages with OSV
word order?  According to those statistics, there can't be more than a
handful of them in the entire world.  Do they all belong to the same
language family?  If not, then how did the peculiar word order arise?  What
else is distinctive about them as a group?

Isidora
linguist at play

Replies