On Thu, May 29, 2008 at 1:36 AM, Steve Bett <stbetta@...> wrote:
>> Is there enough distinction between "cot" and "caught" to warrant two
>> separate letters? What about "pen" and "pin"?
>
> SB: An orthographer is free to merge phonemes and use one grapheme for two
> phonemes but is this a good idea?
>
> We have a definition of phoneme. If the difference shows up in one word
> pair, that is sufficient ot identify a phoneme or phonemic distinction.
True.
However, different varieties of English have different numbers of
phonemes/phonemic contrasts, due to splits and mergers. This makes it
difficult to come up with a set of symbols that will treat all
varieties of English phonemically, and the question was where to draw
the line (or, put differently: which varieties of English will we
consider and which will we ignore).
> What is the proposal to replace the minimal pair approach? How will an
> insignificant difference be defined?
One person's insignificant difference (e.g. "horse/hoarse", which has
no difference at all for me) is another person's phonemic contrast
(e.g. "pen/pin", which is different for me and alters the meaning of
the word, but are pronounced the same for quite a few people, who
distinguish between the two as "writing pin"/"ink pin" and "stick
pin", or similar devices).
> What major dialect of English fails to distinguish pen and pin?
See the diagram at
http://en.wikipedia.org/wiki/Phonological_history_of_English_high_front_vowels#Pin-pen_merger
-- large parts of the south-eastern United States have this merger.
> The easy solution is to leave it to the professional lexicographers.
*nod* that's what it also says in the preface to _Androcles_.
> Just
> stick with a respected dictionary or pronunciation guide. The only problem
> is there is no pronunciation guide for World English. Dictionaries are
> published for each major broadcast dialect.
*nods* The professionals can then decide which dialect(s) to base
their pronunciation guide on.
>> Since I merge "horse" and "hoarse", I don't know how speakers with the
>> distinction feel about the difference, but I can imagine that it's
>> similar.
>
> Dictionaries are not of one mind when it comes to distinguishing /ôr/ and
> /oUr/.
> hoarse might be /hoUrs/ while horse might be /hOrs/ (SAMPA-IPA notation).
>
> I think that a majority of lexicographers represent both words as /hOrs/.
Then such dictionaries might be a good fit for my variety of English,
but not so good for people who have a distinction there.
> What about forty-four? Is or pronounced the same as our?
I don't know - as I said, NORTH and FORCE (to use Wells's lexical
sets) are the same sound in my speech.
>> For example, I'd be for a "cot/caught" distinction but think that a
>> "wait/weight" distinction is less likely to be useful and would lead
>> to more people having to learn spelling rules (or look up words) than
>> "necessary".
>
> SB: What dictionary makes a wait-weight distinction. Usually both are
> /weIt/.
I don't know of any dictionary that makes a distinction, probably
because the varieties that make it (see
http://en.wikipedia.org/wiki/Phonological_history_of_English_fricatives_and_affricates#Wait-weight_merger
) are fairly regional and not typically represented in dictionaries
(except, I suppose, specifically local dictionaries, which would
probably also capture local vocabulary and not merely a local
pronunciation of standard English).
>> > Every school
>> > child would have to know some obscure symbol they may never encounter
>> > because they might want to read a book written by a foreigner who
>> > pronounces
>> > some letter or something slightly differently. We don't have a letter
>> > for a
>> > rolled R as in Spanish, which would be different from one used in French
>> > or
>> > Japanese. Should we include these as well?
>
>> No, because those are not phonemes in English.
>
>> If someone from Spain pronounces "r" with a trill, then they're
>> pronouncing the English phoneme /r/ rather differently from most other
>> people, but it's still the same phoneme.
>
> SB: ?? The Spanish r is not the same phoneme as the English r
As I understand the term "phoneme", it only makes sense within one language.
Or put another way, two sounds are separate phonemes if there is a
minimal pair of words differing only in those sounds and with
different meanings.
Comparing a Spanish word and an English one in order to find a minimal
pair doesn't make sense for me.
> See the SAMPA analysis for the phonemes of English and the phonemes of
> Spanish.
You can talk about "the phonemes of English" or "the phonemes of
Spanish", but I don't think it's meaningful to compare an English
phoneme with a Spanish phoneme.
Now, if you were to say that the most common phone_t_ical realisation
of the English phoneme /r/ is different from the most common
phonetical realisation of the Spanish phoneme /r/, I'd be with you --
but you can't say that the distinction is phonemic since the two
sounds belong to different languages; likewise, the term "minimal
pair" makes no sense to me when talking about words from separate
languages.
For example, English "feel" and German "fiel" sound similar except for
the sound of the "l" at the end, but I don't think that establishes a
"minimal pair" which lets you say that "the two sounds are separate
phonemes" -- the term just doesn't make sense to me here.
Especially if you're aiming at making symbols for all English phonemes
-- the pronunciation of sounds in other languages should be irrelevant
to that end.
Cheers,
--
Philip Newton <philip.newton@gmail.com>