Re: Pronunciation keys

Mark J. Reed skrev:
 > On 1/30/07, Benct Philip Jonsson
 > <conlang@...> wrote:
 >>  STRUT -- which I, oh horror, normally render as
 >>  [3\]! :-)
 >
 > That's not at all horrific. My STRUT is [3], and I'm a dead-
 > normal L1-Anglophonic Murkan.

The thing is that my L1 /3\/ is rounded (note the backslash
in the CXS!) -- as it should be as it descends from Common
Scandinavian short *o( or from secondarily lowered and
centralized *2( (if we disregard some oddball cases where
the etymon had *a( or *E( next to a labial consonant) -- but
that's more like the vowel in colloquial French _comme_. Not
that GAE has anything low rounded that it could be readily
confused with... :-)

 > See the long thread I started last June entitled
 > "Schwa and [V]: Learning the IPA". STRUT epitomizes
 > /V/, which is most assuredly nowhere near [V] IML -
 > or, I daresay, in GAE.

True, and in BrEng it's usually [6]. I *hear* GAE /V/ as
[3\], but that has probably more to do with L1 perceptual
influence, although I hear it most clearly in _love_ and
_above_ with final /Vv/ (how's that for an ugly CXS!) where
the [v] may labialize the [3].

Anyway most phonemes in my L2 English -- which I learned at
around the age of nine from my grandmother who had lived
through her twenties (in uterque sensu, as she was born in
1899) in Chicago -- map to some phoneme or allophone in my
L1. The main exceptions among the vowels are /Aw/, although
my L1 /u:/ and /u\/ are phonetically [uB_o] and [u\B_o] from
which I could extract a phone [B_o] -- [w\] in BXS -- to use
for English /w/ and in /Aw/ [AB_o], and [&] which is an
allophone of /E/ or /a/ (they merge before /r/) in my L1.

The 'Swedish' -- a 'General Swedish as opposed to my Lect1
-- pronunciation key for English which I developed for the
use of kids may be of some interest (numbers in round
parentheses refer to notes below the table):

== Vowels (1) ==
                     (Swedish phone(me)s)
* KIT     i         /i/
* DRESS   e         /e/
* TRAP    ä         /E/ [&] (2)
* LOT     a         /A/ [a] (3)
* STRUT   ö         /2/ [9]/[3\]
* FOOT    u (o)     /8/ (/u/) (4)

* FLEECE  ij        [ij] (5)
* FACE    ej        /e/+/j/
* PRICE   aj        /A/+/j/
* CHOICE  oj        /o/+/j/

* GOOSE   o         /u/
* GOAT    å         /o/ (3)
* MOUTH   ao        /A/+/u/ (6)

* NEAR    ir        /i/+/r/
* SQUARE  er        /e/+/r/
* START   ar        /A/+/r/
* NORTH   år        /o/+/r/
* FORCE   år
* POOR    or        /u/+/r/ (7)
* NURSE   ör        /2/+/r/ [9r]/[3\r]

* FATHER  a         /A/
* THOUGHT a         /o/ (3)

== Consonants (8) ==

* CHURCH  tsh       /t/+/s`/
* JUDGE   dzh       /d/+/s`/
* THING   th        ? (9)
* THIS    dh
* ZERO    z         ? (10)
* SHIP    sh        /s`/
* ASIA    zh        ?
* YET     j/y       [j] (11)
* WET     u         /8/ [u\B_o] (12)
* (WHEN   hu        /h/+/8/ )

_dhö kuikk braon faks dzhömps åvur dhe lejzi dagg_,
though the scheme is intended to show the pronunciation
of words in lists rather than phrases.  To expose grade
school kids (most _gymnasium_ 'High School' kids too,
I'm afraid...) to IPA is IMO useless.  They normal reaction
to not understanding the special symbols is to ignore them.
That goes for many university students too IME, except they
are usually forced to learn to transcribe the language they
are learning.

Notes:

  (1) I deliberately didn't rely on vowel length, which is
      dependent on factors of stress and the openness of the
      syllable in Swedish. In fact vowel doubling may be used
      to indicate stress, if one doesn't like capitalization
      or underlining is unavailable. [@] is essentially
      missing in Swedish; the best respelling is _ö_ or _u_
      since _e_ may elicit something too close to [I].
  (2) [&] occurs for /E/ next to /r/.
  (3) A spelling _åo_ for GOAT is unfortunately likely to
      elicit [u:], and we do want to avoid 'unnatural' merger
      of LOT-GOAT and/or THOUGHT-GOAT, so a deliberate
LOT-THOUGHT-
      FATHER merger is preferable, since the functional load
      of those distinctions is minimal anyway. RP chauvinists
      will go ballistic, but the fact is that an American
      accent is easier to learn for most Swedes. As for LOT-
      THOUGHT merger most young Americans have it and this
      respelling is aimed at the growing-up generation.
  (4) The functional load of the FOOT-GOOSE distinction is
      minimal, and so should be perhaps be neglected in favor
      of _u_ [8]/[u\B_o] always being consonantal /w/ not
      entirely native-like but perhaps less ambiguous once
      you get used to the accent. Cf. note (12).  Also _u_ as
      a vowel (Swedish /8/) may be more fortunate for /@/.
  (5) The main long allophone of /i/ is [ij], and the the
      spelling _ij_ does in fact elicit [i:j] from my test
      subjects; IMO a reasonable deviation.
  (6) The only reasonably common Swedish words having
      anything close to [aU] are _kaos_ 'chaos' (which may
      be monosyllabic) and _paus_ 'pause' which may be
      realized as [pAB_os]. Since _au_ is usually /A/ in the
      very common _automat_ and _automatisk_ we have to
      avoid that graphy.
  (7) The spelling _or_ may elicit [o:r], but the damage is
      slight as many (most?) non-rhotic Englishes have [O:]
      in this set.
  (8) I deliberately avoided the Swedish graphies _dj, sj,
      tj_ which are [j], [x], [s\] for the majority of
      Swedish speakers. The _+h_ graphies are essentially
      foreign, but at the same time suggest something similar
      to _s, z, t, d_, The graphy _ch_ has to be avoided
      since it normally maps to [x], [s\] in Swedish.
      Consonants after checked vowels should be doubled, as
      should _r_ before _t, d, n, l, s, z_, to avoid
      coalescence into [t`], [d`], [n`], [l`], [s`],
      [z`]/[r\].
  (9) Swedish has nothing close to [T] or [D], and its /t/,
      /d/, /n/, /s/, /l/ are post-dental. When I speak
      English off guard I actually get /T/ and /D/ as
      interdental stops against postdental stops for /t/ and
      /d/! My test subjects make passable interdental
      fricatives after little instruction, since anterior
      consonant articulations are so easy to describe! :-)
(10) Swedish has no voiced sibilants even as allophones,
      though instruction and consistent indication in
      respelling helps.
(11) The letter _y_ is [y] in Swedish, although _yoghurt_
      and _yoga_ are resonably common loans with _y_ = /j/,
      and the spelling _y_ may be warranted to avoid
      confusion with the English _j_ = /dZ/, since _y_ isn't
      used for any vowel in the respelling. Tis should affect
      FACE, PRICE and CHOICE, but not FLEECE, since _iy_ may
      elicit [jy].
(12) _U_ for /w/ is preferable to _o_ since we don't want
      'wolf' _uolf_, 'would'-'wood' _uodd_, 'woman' _uOmman_,
      'wound' _uOnd_ to become [ulf], [ud], ['uman], [und].
      We may get [wolf] etc., but that's IMO less grave.
--


/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    a shprakh iz a dialekt mit an armey un flot

                                 (Max Weinreich)

Replies