"Voiced obstruent" mistakes

[Gary Holton]

James,

I think only a very naive linguist would make such a mistake.
Practical orthographies are just that -- practical -- they are not
intended to represent either IPA or English conventions. I can think
of orthographies which use <a> to represent schwa, and <x> to
represent a pharyngeal fricative. I'm sure others could add to this
list. So I'm curious about your assertion that Athabaskan
orthographies have "tripped up" other linguists. While a linguist
might make casual assumptions in passing, I would hope that any
linguist who actually makes use of data from Athabaskan (or indeed any
other language) would be more careful. This could explain why
anecdotes about the misinterpretation of Athabaskan orthographies are
easier to find than actual citations in published source. Still,
there's some pretty shoddy linguistics out there, so there are bound
to be a few published examples :(

Roy's point about the effect on learners is an interesting one. As
first language speech communitoies continue to shrink, we should
expect L2 learners' pronunciation to become more like English. Just
how much of this convergence is attributable to orthography is an open
question.

Gary


On Wed, Feb 17, 2010 at 6:05 PM, James Crippen <jcrippen at gmail.com> wrote:
> On Wed, Feb 17, 2010 at 15:47, D Roy Mitchel IV <droymitchell at yahoo.com> wrote:
>> What I would share is that I've encountered many students of Tlingit and
>> Dena'ina over the years who pronounce the orthographys' d as a voiced stop,
>> not only because it's spelled like the English voiced stop, but many
>> guidebooks for learners state explicitly that these sounds are "the same" in
>> English as they are in Tlingit and in Dena'ina as they are in English.
>> This is one of my pet peeves.  Admittedly, it doesn't lead the learners
>> mistakenly to create a different word, and English-speaking learners of
>> Athabascan and Tlingit will have pronunciation difficulties, but I cannot
>> agree with steering them in the wrong direction by design.
>
> This is an issue that has concerned me as well, but it's something
> I've kind of written off as inevitable. Interestingly, since General
> American English tends to have fully devoiced initial stops (and
> perhaps affricates), in initial position the orthography leads English
> speakers to produce something rather like the actual sound. This
> doesn't work in other positions, however, and of course English
> speakers will allow voice to spread across word boundaries and thus
> spoil the effect.
>
>> When I talk to learners about the consonant system of Tlingit, for example,
>> I identify seven consonants that may be pronounced consistently the same in
>> English and in Tlingit, three in which certain English allophones may match
>> with Tlingit phones, and 32 in which English has no exact match at all.
>
> That's a good point. It's misleading to assume that e.g. <j> really
> represents the same sound in both Tlingit and English. Unfortunately
> most teachers aren't linguists and thus don't try very hard to be
> precise about these things.
>
> But what I was really looking for was an example or two where
> linguists - people who ought to be more careful - have gratuitously
> assumed that the orthography and/or transcription represented voiced
> sounds and thus they came to incorrect conclusions about phonological
> rules or the like.
>
> James
>