Complexity in language

[j.guy@trl.oz.au (Jacques Guy)]

Acting under duress (John Cowan urges me to repost this here),
I repeat here what I posted not very long ago on sci.lang
about the old saw, found in the FAQ, that all languages
are equally complex.

-----------------Start of quote---------------------------

From: jbm@newsserver.trl.oz.au (Jacques Guy)
Date: 1 Dec 1994 16:28:12 +1100
Newsgroups: sci.lang
Subject: sci.lang FAQ

This is a good FAQ, all in all. Mind you, since I don't
work in a university, what irks Jon Aarbakke (who does)
so much irks me little because  GB and all that is
so many light years away from me that it has me utterly
unaffected.

However, there's this business about languages being
equally complex. Now really, folks, this is completely,
but completely, wrong. So I took my keyboard, turned it
seven times in my inkwell, and came up with this:


It is far from true that all languages are equally complex. I suppose
that this strange notion is a throwback, or an offspring, whichever you
prefer, of "politically correct" whereby all languages are equal,
complexity is good, so all languages are equally complex. Now, if you
have tried to learn French, and then Spanish, you might have noticed just
a tad of a difference. Spanish has nowhere as many impossibly difficult
vowels as French, nowhere as many abominably irregular verbs, nowhere as
crazy a spelling... But let's not get personal, so allow me to take as
an example two languages which will not make anyone raise an eyebrow, I
am sure. One is called Tolomako, the other Sakao. Both are spoken in the
same village, called Port-Olry, a place on the island of Espiritu Santo
in Vanuatu. Both languages are closely related, by which I mean that
they might have been one and the same perhaps 1000 years ago, probably
much less.

Noblesse oblige, phonology first.

Tolomako vowels: a, e, i, o, u
     consonants: p, t, k, B (<beta>), G (<gamma>), m, n, s, ts, r, l
     syllable structure is (C)V(V)

   Sakao vowels: a  (front unrounded)
                 a^ (back rounded)
                 E  (<epsilon>)
                 e
                 i
                 O  (IPA mirror image of "c")
                 o
                 u
                 oe (i.e. IPA o-with-e as "oe" in French "oeil")
                 o/ (i.e. IPA o-slash, "eu" as in French "peu")
                 y    (i.e. French "u")
                 (i)  (an always unstressed high vowel, unmarked
                       for rounding or backing, as elusive as the
                       infamous French so-called "mute e")
     diphthongs: oeE
                 a^O
     consonants: p, t, k, m, n, ng, B(<beta>), D (<delta>),
                 G (<gamma>), h, s, r, R (unvoiced trill), l
 semiconsonants: j, w
 syllable structure (that is, if "syllable" makes any sense in that
     language, and I suspect it does not): a single vowel, or diphthong,
     surrounded by any number of consonants.
     Example: i       "thou"
              mhErtpr "having sung and stopped singing thou kept silent"
     (m- 2nd pers. hErt "to sing" -p perfective, -r continuous)

Oh, I forgot: consonants are long or short, e.g. oeBe "drum", oeBBe "bed".

A bit of grammar now. Obligatorily possessed nouns for instance:

Tolomako          Sakao

na tsiGoku        oes(i)ngoeG       my mouth
na tsiGomu        oes(i)ngoem       thy mouth
na tsiGona        Os(i)ngOn         his/her/its mouth
na tsiGo...       oesoeng...        mouth of...

na Buluku         ulyG              my hair
na Bulumu         ulym              thy hair
na Buluna         uloen             his/her/its hair
na Bulu...        no/l...           hair of...


A tad of verbal phrases now. Sakao has some holophrastic tendencies there,
Tolomako none.

A simple example, to serve to illustrate how that may have developed:

Tolomako
 mo        losi na    poe  ne    na   matsa
 3rd pers. hit  art.  pig  prep. art. club
 he hits/kills the/a pig with a club (a praiseworthy occupation, leading
 to rising in society)

Notice that I said "prep.". Tolomako has only one preposition to make do
for locative (ubi, quo, unde, qua), whether in space and time, and for
the instrumental. Here it's the instrumental.

Sakao

  m(i)-       jil -(i)n        a-   ra   a-   mas
  3rd pers.   hit +transitive  art. pig  art. club
  Same meaning, same highly regarded sign of social achievement.

  What I called +transitive is a suffix that turns an intransive verb into
a transitive one, and a transitive verb into a ditransitive one. Ditransitive?
I made it up. That's a verb that takes two direct objects, one of them
expressing, loosely speaking, the instrumental. When a verb takes two
objects they can occur in any order. Only the meaning can disambiguate. Thus:

   m(i)jil(i)n ara amas
   m(i)jil(i)n amas ara

mean the same.

Now for a nice example of Sakao's holophrastic tendencies:

mOssOnEshOBr(i)n aDa EDE
he-shoots-fish-follows-continuous aspect+transitive  bow sea
he kept on walking along the shore shooting fish with a bow

You can say mOssOnEshObr(i)n EDE aDa equally well, of course.

Broken down into its bits and pieces:

mO-   3rd pers.
sOn   to shoot with a bow, but because here it takes an incorporated
      object, its initial consonant is long: ssOn
nEs   fish. Note the morphophonology: ssOn+nEs > ssOnEs
hoB   to follow
-r    continuous aspect
-(i)n +transitive

a-    article
Da    bow, instrumental

E-    article
DE    sea, direct object. Of what? of hOB "to follow" of course!

Which of the two languages spoken in Port-Olry do you think the
Catholic missionaries learnt and used?

Could that possibly be because it was easier than the other?

Are all languages, then, equally easy, or difficult?

--------------------end of quote---------------------------

I cannot resist the perverse pleasure of adding a few
things. If you have had a close look at the Sakao
examples, you might suspect that things like

m(i)jilrap(i) amas, namely:

m(i)- jil -ra -p    -(i)n   a-  mas
3rd  kill pig perf. +trans art. club

are quite kosher. And you'd be right!

And further, if you have happened to lurk on sci.lang
and read the FAQ, you may remember that all languages
are equally complex, with the possible exception of
pidgins and creoles. Nonsense again. I have a practical
method of creole ("Le Guadeloupeen Sans Peine", or is
it Martiniquais?) and I assure you that it is far, far
more complex than Tolomako! 

Finally, someone countered on sci.lang that I was
confusing "complicated" and "complex". That not
all languages were equally "complicated" but all
were equally "complex", complexity referring to
their expressive capacity. Stuff and nonsense
again. I can express *anybloodything* in Bislama,
the pidgin of Vanuatu, given the time to explain
what I mean, give examples, build up a specialized
vocabulary, etc. Why, even "fire the photon torpedoes!"
It still is a lot easier to say in Klingon, isn't it?
So this definition of "complexity" is not only contrary
to usage, amounting in reality to "expressiveness" or
"functionality", but is even false once granted.