LINGUIST List 2.238

Friday, 17 May 1991

Disc: Phonology and Orthography

Editor for this issue: <>

Directory

1. , Re: Phonology and Orthograpyh
2. Joe Stemberger, Re: Queries

John Coleman writes:

>...The reasoning behind the argument
>that orthographic practises support phonemic theory seems to be blind to the
>fact that phonemic theory is historically based on alphabetic orthographies.

Actually, this is begging the question. Historically, linguists have assumed
that alphabetic orthographies were based on the phonemic analysis of a dialect
of a language. The reverse hypothesis is hardly a fact. Linguists have also
felt that phonemic theory was necessary in order to capture generalizations
about allomorphy, not just spelling. Baudouin de Courtenay, the primogenitor
of phonology, analyzed psychophonetic alternations as involving two distinct
phonemes and physiophonetic alternations as involving one. That is the histor-
ical basis for all subsequent theories of phonology and morphonology. Far
from being the basis of phonemic theory, phonemic theory served to explain
the evolution of writing into an alphabetic stage.

John Coleman's point about there not being a unique phonemic analysis for a
language is correct but irrelevant. Phonemic theory was originally grounded
in psychological function, and then (under Saussure's influence) it came to
be grounded in social function. The fact is that constructing an alphabet is
a practical matter, which dictates that social convention force a single
phonemic system on a group, even if the individuals comprising the group don't
all share the same psychological system. It would simply be impractical for
a society to adopt graphological conventions on a dialect-by-dialect (or
individual-by-individual) basis.

Finally, John has argued, confusingly I think, about the efficiency of 
alphabetic vs. morabetic writing. He fails to notice that radically different
phonologies require radically different alphabetic conventions. You can't
compare the number of symbols needed for Hebrew with the number of symbols
needed for Greek, and claim that that tells you anything about the efficiency
of morabetic vs. alphabetic writing. For any given language, the number of
morabetic symbols required to represent that language will always be greater
than the number of alphabetic symbols needed. This is so because moras are
always describable in terms of combinations of segments. In a pure CV language
you need enough symbols to cover all possible CV combinations. (BTW, 
devanagari is not purely syllabic or morabetic. It has vowel symbols for
cases where vowels begin words.)
 -Rick Wojcik (rwojcikatc.boeing.com)

Two comments on orthography and phonology.

1) Cambodian has an orthographic distinction between velar /k/ and
uvular /q/. The consultant we had for a field methods class swore
up and down that they were still distinct, but they were clearly all
pronounced as velars in his speech (by ear, by spectrograph, and by 
looking into his mouth when he talked). However, the letter of the
alphabet still had a clear /q/. (Interestingly, the letter name had the
/q/ word-initially, but all examples of words spelled with it had it
in word-final position.) I have no idea whether the two sounds are still
distinct in other dialects of Cambodian.

2) As far as letter names being outside the normal phonology... In Slovene,
all consonant letter names are the phoneme followed by shwa, as in
/b^, d^, z^/, etc. As I understand it (and my familiarity with the
language is from pedogogical grammars and a brief visit to the country
after high school), shwa occurs only as an epenthetic vowel, where it is
inserted BEFORE unsyllabifiable word-final consonants (so you get
alternations like nom. /p^s/ 'dog' vs. gen. /ps-a/, and nom.masc.sg.
/slad^k/ 'sweet' vs. fem. /slatka/). And the shwa is otherwise
unstressed. So, the shwa is in entirely the wrong place in letter names. 
Interestingly, when my children (monolingual English speakers, other than 
two words of Swedish and three of Slovene) were first figuring out how to 
segment words into parts, they gave the sound as [b^], with a stressed shwa, 
and my wife and I helped them out with exactly the same sort of 
pronunciation. But stressed shwa doesn't otherwise occur in English, 
either (even in the kids' speech at the time).

These examples are probably rare across languages, but it's not clear that
letter names are always well integrated into the phonology of the language.

---joe stemberger

[End Linguist List, Vol. 2, No. 238]