LINGUIST List 2.293

Friday, 14 June 1991

Disc: Turkish, Register, Jewish names, Phonology/Orthography

Editor for this issue: <>

Directory

1. John Phillips, Re: Responses - Turkish morphological parser
2. , Register
3. Daniel Radzinski, Jewish surnames.
4. John Coleman, Phonology and Orthography

A colleague of mine here at Umist has a Turkish morphological
parser, part of a prototype machine translation system. He is
Jeremy Carroll, jeremyuk.ac.umist.ccl

				John Phillips

Anyone who is really interested in register should read
Halliday on it if they haven't already. Studies inspired
by Halliday's early work didn't win many friends, but his later
approach is more interesting, in my view. He defines register
as "the configuration of semantic resources that the member of
a culture typically associates with a situation type" (Language as
Social Semiotic, Edward Arnold 1985, p111), the situation type
being defined in terms of values of field, tenor and mode. As I've
said elsewhere, I think there's a lot of work still to do on
firming up these ideas, but they're worth looking at. For a critical
discussion, though a bit out of date, see my Systemic Linguistics:
Theory and Applications, Batsford 1985, especially Ch 5.

Chris Butler
Dept of Linguistics
University of Nottingham
Nottingham NG7 2RD
UK

Ellen Prince indicates that "it is typical for jews to take names that are
phonologically and even apparently morphologically consistent with the
languages of the countries in which they reside." This is certainly true.
Consider the following variations on (Ha)Levi:

Levin, Levine, Levitus, Leefsma, Horowicz, Hurvitz, Gurevich, Levitz(?)

I guess one definitely must consider this factor in the context of
acronym etymologies.

-- Daniel Radzinski

> (2) Does anyone understand why John Coleman thinks that the length of
> some (unspecified) bit-encoding of the 2D character patterns is a
> suitable measure of how difficult it is for people to learn and use a
> writing system?

Allow me to make myself clearer. A number of people have argued that
alphabetic writing is "superior" to, say kana-like writing systems because
fewer alphabetic symbols than kana symbols are required for the orthography
of a language. This property has been called "efficiency".

However, a number of important considerations are being overlooked in this
argument. Firstly, learning a writing system is not just a question of
learning a set of symbols. It is also necessary to learn the way in which
combinations of symbols are interpreted. Alphabetic writing costs more
than kana-type systems on this score. Secondly, a simple comparison
of the number of symbols is no use. Some alphabets have more symbols
than they "need" from a phonemic point of view, e.g. positional variants
or historical distinctions no longer preserved (such as in Thai orthography).

Since the original discussion what not just about orthography, but phonology,
I proposed that the efficiency of a writing system could be assessed less
prejudicially to any particular system in terms of how efficiently the
phonological distinctions [that's where the bits come in] are encoded,
and how well redundant information is left UNrepresented. The assumption
I am making is that the orthography which is most phonologically efficient
is one which encodes all the phonological distinctive oppositions of a
language, and no redundant information. By this measure it can be seen
that kana, and yes, even syllable-based systems may be more efficient
encodings than an alphabet. I make no claims as to how this relates
to learnability.

Margaret's comment that
> However, it follows by the same line of reasoning
> that an ideographic system (e.g. Chinese) uses storage space even more
> efficiently and should thus be an even more popular (stable, adopted
> by other languages, easy for children to learn) method of writing.
is a non-sequitur. Chinese has a great many different characters for each
syllable, and is thus an inefficient means for encoding the phonology
of Chinese. A true syllabary of a couple of hundred characters might
be most efficient, given the pervasive order-redundancies of Chinese.
In general, because there are two places where consonantal oppositions
occur in syllables, demisyllable systems are the most efficient.

As well as this technical defence of my claim, I would like to close
by adding that the tenacious defence of the "superior efficiency" of
alphabetic writing that this discussion has engendered has at times
been accompanied by a Eurocentric tone. The comment about not being
able to read Arabic aloud until you know what it means is equally true
of alphabetic scripts, yes even Finnish.

--- John Coleman

[End Linguist List, Vol. 2, No. 0293]