Danny Yee >> People >> Mike Salovesh
Date:         Mon, 6 Mar 2000 04:08:34 -0600
From:         Mike Salovesh <t20mxs1@CORN.CSO.NIU.EDU>
Subject:      Re: English spelling

Ron Kephart wrote:

> And, as Jack has pointed out, English has around 12 vowel phonemes, > so we really need 12 "letters," but our own alphabet is deficient, > having come from Latin which had the basic five-vowel system (but > with a length contrast). This is probably our biggest stumbling block > to developing a rational orthography for English- it's very > difficult, if not impossible, given our traditional alphabet.

I have two problems with what Ron says.

1) It isn't absolutely necessary to have 12 distinctive letters in your alphabet to create a good phonemic writing system for a language with 12 vowels.

Writing systems get into trouble when two different phonemes are represented by the same graph (like the c in "case" and the c in "grace"), and where one phoneme is represented by two or more graphs (like the s in "case" and the c in "grace"). You can avoid both kinds of error and still produce a proper phonemic writing system using fewer characters than you have phonemes to represent. The way out is to use combinations of letters, rather than single letters standing by themselves, to represent distinctive phonemes.

Example: Consider the contrasting vowels in the words "hat" and "hot" in most varieties of English. If all I had available were five letters to represent vowel sounds, I would try writing the first as /a/ and the second as /ae/. (It might be somewhat more elegant to think of the first as "a plus zero", the second as "a plus e".) Five letters plus a zero marker would be enough to give unique written representation to 16 separate vowels. All you need is a rule saying that it takes either one letter plus zero or two letters combined to represent a vowel phoneme.

That's just a way to get unique combinations of a small number of distinctive letters to represent a larger number of phonemes.

2) Ron's statement of the English vowel problem conceals a very important feature of English dialect variation. It's too easy to read his "English has around 12 vowel phonemes" to imply that there is the same finite number of vowel phonemes in all dialects of English. (Presumably, that number is either 12 or something close to 12.) Ron knows better, but the facts are so familiar to him that he didn't think of repeating them for the benefit of those who haven't studied dialect variation.

Back when I was doing research on dialects of U.S. English, I found people who spoke 11-vowel English. Other people's speech had to be analyzed as having 12 or 13 vowel phonemes, and I think I remember hearing about someone who spoke English with 14 vowel phonemes. The 11-vowel English I analyzed was internally consistent and perfectly intelligible English; it simply got along with one less vowel than the more common 12-vowel variety.

Suppose someone who spoke 11-vowel English tried to read a phonemicly accurate transcription of 13-vowel English, or vice versa. Reading would not be as simple and direct as Ron reports for people reading his transcriptions of Grenadan or Carriacouan creole.

That's one of the stumbling blocks that hold back adoption of full spelling reform for English. English dialects which just aren't in one-to-one correspondence with each other. You can't develop a single standardized phonemic alphabet and have it fit all the variation that goes on between separate dialects.

I've seen what happens when using a phonemic representation of one English dialect to teach reading to kids who speak another dialect. When our kids hit kindergarden/first grade they went to a school that used someting called "ITA", the Initial Teaching Alphabet. ITA was designed to be a solid, consistent phonemic alphabet, and in fact it is pretty good that way. Trouble is, it's a good phonemic alphabet for what looks to me like a variety of English spoken by members of the middle class living somewhere in England. ITA is NOT a good phonemic representation of standard Midwestern U.S. English.

Since the school my kids were in was here in De Kalb, Illinois, ITA was not that much better as a teaching tool than standard dictionary spelling would have been. Unfortunately, the teachers who were using and demonstrating ITA didn't know enough descriptive linguistics to realize that that their saw had a lot of missing teeth, or their straightedge had kinks in it, or the head was loose on their hammer. They tended to blame the smartest kids in their classes for taking an oppositional stance when those kids had trouble reconciling ITA with the way people speak around here. (Our kids were spared such problems, because they had taught themselves to read -- with a lot of help from Sesame Street -- before they started school. They were, therefore, mostly excused from having to deal with ITA. When they couldn't avoid it entirely, they treated ITA as one of those dumb things that make teachers happy but don't mean anything in the real world.)

Back to the larger context of Ron's point and my quibbling: This thread seems to show that there may be lots of people with strong interests in anthropology who have not been exposed to an adequate introduction to linguistics.

That's exactly the kind of gap in anthropological education I've been moaning about in other discussions here. I know that most anthropology departments do not require their students to gain basic competence in linguistics. I think that is a disservice to our students, and I wish we were doing better by them.

-- mike salovesh <salovesh@niu.edu> PEACE !!!