On Tue, 21 Nov 2000, Oskar Gudlaugsson wrote:

> I always approach language courses with skepticism. The nature of languages
> is so often misconceived, no less by the various authors of language
> courses. I start by skipping to the pronunciation guide, which says a good
> deal about the book's linguistic accuracy; I've seen so many language
> courses for English people describe [E:] or [e:] as "the long 'a' as
> in 'date' or 'nay", even in serious books written by professors. So that's
> always a giveaway. Overall, I distrust books that avoid using IPA at all
> costs (instead using inaccurate approximations to anomalous English vowels;
> I've seen a book describe one vowel sound with a 7-8 line paragraph, for
> some central vowel in Rumanian). The "'a' as in Southern British 'father'"
> line is the one thing that makes me wish non-English language courses were
> more abundant.

But to someone--say a student or businessman who may not be aware of IPA,
or doesn't have time for it, or can't find sources on it at all (before
the internet, it would've been impossible for me to find anything on it
in *English* while I lived in Korea), why is this so bad?  I would *like*
it if my grammars had IPA values.  But the truth is that a lot of people
seem to use these books who are *not* familiar with IPA, and so the
inclusion of "English approximations" is helpful to them.  All of the
grammars I have *stress* that this will only get you in the ballpark
range of a sound, and that you should listen to native speakers and have
a good instructor help you with learning to produce the sound.  I would
think that this is a function of the book's intended audience (and also
whether or not it's a "stand-alone" book or meant to be used with an
instructor to help you).

I grant you that having an index or something IPA, for those of us who
know it, would be nice.  (It would help me a *lot* with Turkish
pronunciation.)  But I'm not convinced that "English approximations" are
bad per se.  (I do admit that I have to be careful with my book on Irish,
because the approximations come, naturally, from British English...)

Also, face it: there are people like me who *still* can't reliably roll
r's.  (Well, I can manage an approximation to the French uvular? and
German flapped? r's, but the trill is beyond me and my taps are very
unreliable.)  Or produce other sounds "on-target."  I've seen a few
pronunciation guides that tell you what sounds *will* be understand as
the target phoneme by native speakers, even if your "accent is funny."  I
think this is also useful as a stopgap while I learn to produce various
sounds.  (I wish I had been born to a language that *has* rolled r's,

What I'd like to see in a pronunciation guide, and haven't yet, is some
brief discussion of the rhythms and tones of a language--paralinguistic
features, are they called?  The *rhythm* of French, frex, sounds quite
different from the rhythm of Korean or German, even if you're not
listening to specific sounds--the ups and downs of tones.  Perhaps this
is something you just have to pick up by ear, though.

I don't know anything about Greek, alas, but isn't the stress pattern in
English pretty darn unpredictable?  I haven't found one in Japanese or
Korean, which are pitch-accented, though that could be ignorance on my part.

YHL's two cents, centimes, won, or whatever small currency you favor