Double Articulation

[Dan Parvaz]

Quite aside from Ms. Caswell's ravings, the discussion of double
articulation is useful, as can be seen from Adam Schembri and Scott
Liddell's postings to the list. A good introduction to double articulation
as used in semiotics can be found here:

http://www.aber.ac.uk/media/Documents/S4B/sem08a.html

It isn't Oxbridge (it isn't even Yalevard), but it'll have to do. :-)

The question on duality of patterning is an interesting one. When I teach
this to undergraduates, the usual strategy is to trot out three English
phonemes -- /t/, /a/, and /p/, to pull some out of a hat -- and show that
"pot," "opt," and "top" can be formed. Once we leave that level, we are
constrained by what the words mean; reordering the morphemes in a word or
the words in a sentence may be permissible or not, and may change the
meaning or not, but no one would argue that it is the same as re-ordering
the segments in a word, beyond some rather superficial formal sense.
Duality of patterning then, happens somewhere at or below the level of a
word/morpheme.

On to signed languages. In many signed languages, GIVE has some sort of
agreement (not that we would all call it that, but if I could count on
your charity for a moment) between Agent and Beneficiary. Re-ordering the
time slots, which in some models requires reversing the direction of the
movement, can produce a different well-formed sign.  However, there are
questions about how much of this is really reordering?  If you change the
movement, you have a different "segment" if that's what you want to call
it. Furthermore, even if this were a strict permutation (no
substitutions!) then the change in the sign is based on the underlying
semantics of the locations.

So duality of patterning in a signed language cannot rest on the kind of
permutability arguments that we are used to using. We can talk about signs
having some kind of arbitariness without sacrificing their iconicity, but
I'm not convinced that that's the same thing.

Any takers? I know Sherman has thought long and hard about this, and I
know he isn't the only one.

____________
DAN PARVAZ
Computational Linguist, CSI, Inc.
PhD student, University of New Mexico
dparvaz@{mac.com,csi-inc.com,unm.edu}