8.1529, Sum: Predictability

[The LINGUIST List]

LINGUIST List:  Vol-8-1529. Sun Oct 26 1997. ISSN: 1068-4875.

Subject: 8.1529, Sum: Predictability

Moderators: Anthony Rodrigues Aristar: Texas A&M U. <aristar at linguistlist.org>
            Helen Dry: Eastern Michigan U. <hdry at linguistlist.org>
            T. Daniel Seely: Eastern Michigan U. <seely at linguistlist.org>

Review Editor:     Andrew Carnie <carnie at linguistlist.org>

Associate Editor: Ljuba Veselinova <ljuba at linguistlist.org>

Assistant Editors:  Martin Jacobsen <marty at linguistlist.org>
                    Brett Churchill <brett at linguistlist.org>
                    Anita Huang <anita at linguistlist.org>
                    Julie Wilson <julie at linguistlist.org>
                    Elaine Halleck <elaine at linguistlist.org>

Software development: John H. Remmers <remmers at emunix.emich.edu>
                      Zhiping Zheng <zzheng at online.emich.edu>

Home Page:  http://linguistlist.org/


Editor for this issue: Martin Jacobsen <marty at linguistlist.org>

=================================Directory=================================

1)
Date:  Sat, 25 Oct 1997 16:55:24 +0800 (CST)
From:  wzyem at pubms.pku.edu.cn
Subject:  Predictability (fwd)

-------------------------------- Message 1 -------------------------------

Date:  Sat, 25 Oct 1997 16:55:24 +0800 (CST)
From:  wzyem at pubms.pku.edu.cn
Subject:  Predictability (fwd)

Dear Linguist,

On Oct.5th I put on a query entitled "predictability and redundancy."
I owe thanks to the following people for their answers: Peter Daniels,
Daniel E.Collins, Andreas Mengel, Richard Wright Vincent Jenkins and
Paul Boersma.

Peter Daniels said Predictability == Redundancy, it's a tautology.
Daniel Collins cited "Preliminaries to Speech Analysis" by Jakobson,
Fant and Halle to illustrate the relations. He concluded that "the
difference between vowels is predictatable from the palatalization of
the preceding consonant. Therefore, except in special cases, it is
redundant, i.e. not needed to make distinctions. Redundant features
are invoked (play a discriminative role) chiefly when the message
cannot otherwise be interpreted.

Paul Boersma presented an answer in terms of algorithm. He holds that
"predictability is a property of a linguistic analysis while
redundancy is a property of a speaker's grammar and a *possible*
result of predictability. The redundancy that JCH (Jakobson, Cherry
and Halle) refer to, has to do with minimal linguistic description,
and is a *necessary* result of predictability."  Richard Wright seems
to disagree with the above answers. His is a phonetician point of
view. I cite some of his lines

Among phoneticians it has long been common knowledge that
"redundancy" and "predictable" phonetic differences are neither.
	
Given a particular set of phonological contrast, it is difficult
or impossible to predict phonetic details without some knowledge
of those phonetic details.
	
Given this lack of phonetic redundancy, phonology seems to have
two choices: 1. to propose two modules (phonology, phonetics)...
2. to incorporate all of the non-predictable phonetic detail into
the phonological grammar...

Vincent Jenkins recommended some important article of Information
Theory.

Shannon C. E.,1948. A Mathematical Theory of Communication. Bell
		Technical Journal 27
	
Shannon C.E., 1949, Communication in the Presense of Noise. Proc.
                I.R.E. Vol.37 pt 1 (1949), pp 10-21
	
Miller G. 1981. Language and Speech. Freeman
	
Herdan, G. 1966, The Advanced Theory of Language as Choice and
		Chance.

Andreas Mengel also holds that the predicatability will result in
redundancy from the view point of information theory. He produced a
very interesting illustration with the drawing of a face.


Best regards
Wang Zhenyu
Room 1056, Building 46
Peking University, 100871
Beijing, China










---------------------------------------------------------------------------
LINGUIST List: Vol-8-1529