LINGUIST List 3.783

Sat 17 Oct 1992

Disc: Phoneticians

Editor for this issue: <>


  1. John S. Coleman, 3.769 Phoneticians I
  2. Richard Ogden, phoneticians
  3. , RE: 3.770 Phoneticians II
  4. , Phoneticians

Message 1: 3.769 Phoneticians I

Date: Mon, 12 Oct 92 10:32:54 ED3.769 Phoneticians I
From: John S. Coleman <>
Subject: 3.769 Phoneticians I

Joe Stemberger makes a number of statements about
what phoneticians may or may not in principle be
capable of hearing, e.g.

> Cases of incomplete neutralization where the phonetician's ear is
> incapible of hearing the difference have been reported in child
> phonology and in aphasia.

> In child phonology, it is clear that some children incompletely devoice
> final obstruents, but the phonetician can't detect it

Most phoneticians working in the areas of child language and aphasia
may not (probably haven't) been good enough to hear these things,
(said with no disrespect intended). But please take a look at
Kelly and Local's "Doing Phonology", especially pp. 38-41, which
examines the speech of a child in therapy, using ear-trained
listening techniques. I'm sure it would improve most readers
understanding of what can be achieved by ear-training, even in
this area.

> some of the differences reported for
> final devoicing in Polish, for example, HAD to have been detectable by a
> phonetician.

I believe they probably were, but that many phoneticians do not have
a very strong interest in impressionistic studies.

--- John Coleman
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: phoneticians

Date: Mon, 12 Oct 92 17:00 GMT phoneticians
From: Richard Ogden <RAO1VAXB.YORK.AC.UK>
Subject: phoneticians

Alexis Manaster-Ramer writes:

> One, the whole
> pointless discussion started with a simple query I posted about
> whether anyone knew of published work that tends to show that
> phoneticians, no matter how good, cannot hear certain distinctions
> which languages nevertheless make.

--- I don't think the discussion is pointless. There is a discrepancy between
the things that phoneticians of all kinds observe and what phonologists take
to be relevant and are prepared to describe in their work. (When was the last
time you saw a phonology paper relating the phonological description to
instrumental work?) I think lots of phonologists don't understand what
phoneticians do, and vice versa.
The example about confusing [k] with [x] (the original posting said even good
phoneticians do this) seemed to me to mean that the phoneticians in question
were expecting to hear [k] because they believed there 'should be' a '/k/'
there (whatever this might mean). A later posting said that there were good
reasons for describing a velar fricative in English as a 'stop' because it
wasn't an 'ach-laut' or a 'gamma'. I think the confusion of terminology here
(names of letters and and informal name for a sound in a particular language
for things more appropriately described with IPA terminology) reinforced my
initial scepticism.

--- the initial request (below) was not as innocent as it is made to sound in
the latest posting from Alexis Manaster Ramer:

> There are many examples, I believe, where what even the best
> "ear" phoneticians systematically mishear what is actually
> uttered.

I have consistently replied to Alexis Manaster-Ramer's postings because as an
'ear phonetician' who also uses instrumental techniques, I think the
assumption made in the above statement (and in subsequent postings) needs
challenging and refining; and its implications for linguistics as a whole need
stating. I have asked what is meant by 'best'; what the linguistic relevance
of inaudible distinctions might be; and pointed out that phoneticians are
trained to listen without prejudice to speech - good phoneticians do this, bad
ones don't. Names need not be named in this connection; we are not out on some
witch hunt to separate good from bad. But I am unhappy about the
generalisation that is implied by the statement above, namely: 'Don't bother
to listen, measure it.' A balance is necessary; but at the basis of phonetics
must be an ability to listen and observe in the appropriate ways. Other
contributors have stressed this as well.

I am sorry if this has turned out to be a slanging match. I have tried to ask
questions I believed to be important, but haven't had many answers. Of course
the initial request for information should be respected, and if there are any
appropriate references, I hope they will be regarded critically and with a
proper understanding of the place of transcriptions and instrumentation in
linguistic analysis.

Richard Ogden
Experimental phonetics laboratory
University of York
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: RE: 3.770 Phoneticians II

Date: Tue, 13 Oct 92 9:38 GMT RE: 3.770 Phoneticians II
Subject: RE: 3.770 Phoneticians II

A little late into the fray, I know, but, as a (perhaps-not-as-paragon-as-I-
would-like-to-be) linguistic 'ear phonetician' and phonologist (who's done
work on the kinds of things that Alexis is talking about, and who's also
seriously into instrumental work) here goes maybe this will alleviate Alexis'
angst a bit - some references (from memory, but if desperation rules then I
can sort out details)

1. The following ear phoneticians have all described differences between the
 so-called 'neutralised' word-final 'voiced' and 'voiceless' portions
in languages such as German: Ellis, 1876, Pronunciation for Singers (and in
Ealy English Pronunciation); Sweet, 1877, A Handbook of Phonetics; Vietor,
 1884, German Pronunciation; Jesperson, 1933, Lehrbuch der Phonetik;
Heffer, 1950, General Phonetics (American one this!!). Typically they have
noted differences between the pressure of final plosivity, differences in vocal
fold activity going into and coming out of the closure portion (eg whisper
versus voicelessness) and differences in duration of closure portions (nb Sweet
also discusses this in his treatment of differing durations of nasality and
laterality in such English words was 'built' 'build'. More recently Kelly and
Local, 1989 Doing Phonology, document and discuss in reasonable
impressionistic phonetic detail a large number of such observable
(by ear and eye) cases (including the dreaded d/t tap stuff; incidentally, have
close listen to the resonance and phonatory qualities your and other people's
vocalic portions in the first syllables of pairs such as 'latter' and 'ladder'
 or the first syllable vocalic portions in Spanish words such as 'pero' and
- it may be revealing).

2. The stuff about 'plosives' being done sometimes with 'fricatives' is well
known enough to get into G Brown's 'Listening to Spoken English, 1977 - a
book for non-native learners. Any student of SOAS who attended Eileen
Whitley's definitive lectures on the Firthian Prosodic Analysis of English
would be well acquainted with this kind of observation (the Firthian archive
material is houses here at York), Ellis in the 19 century in Early English
Pronunciation in discussing varieties and dialects of English also makes
reference to such phenomena (including eg the lack of friction during the
dental portions of many speakers' versions of words such as 'father' and
'this'). More recently see Sprigg, 1974, 'The lexical item as a phonetic
JIPA; Kelly, 1970(?), On the phonology on an Urban English Accent, JIPA;
 any discussion of Liverpudlian English; chapter 11 of Catford's Fundamental
Problems in Phonetics, 1977;

3. I note with interest that Richard Ogden's point about ranges of variability
has not really been taken up. This is a crucial observation - of particular
importance to those engaged in the 'experimental' investigation of such

4. It's bizarre that in the latter part of the twentieth century the
'neutralization' stuff should still be around seeing as it's essentially the
of a segmental orientiation to phonology in which phonological contrasts are
deemed to be instantiated at only one point in the phonetics.

5. As will be apparent I agree with John Coleman's and Richard Ogden's
central distinction between good and not-so-good 'ear' phoneticians; perhaps
the problem lies at the feet of underarticulated phonological theories which on
 the whole don't know what to do with what decent phoneticians can observe -
but that's something else all together. Of course, as in so many things
phonetic, Sweet got it right: 'The test of a good ear is the power of
discriminating sounds.' 1877, A Handbook of Phonetics.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 4: Phoneticians

Date: Tue, 13 Oct 92 21:58:46 EDPhoneticians
From: <>
Subject: Phoneticians

I hate to prolong this debate, but I believe I have finally found
out where some of the confusion on the part of my critics came
from. One of them (who unfortunately won't say this on LINGUIST)
has sent me a list of quotes from illustrious phoneticians such
as Sweet referring to consonants which they describe as "half-voiced"
and such. Apparently, some people think that final devoicing in
a language like Polish, for example, has some connection with this.

Nothing, I repeat nothing, could be further from the truth. Everybody
who has ever worked with Polish, including some of the world's phoneticians,
has always reported that the final consonants in words like, say, 'grad'
are indistinguishable from those in a word like 'grat'.

There have, of course, been recent studies which claim to find an
instrumental detectable difference in such cases, but even these
have been disputed by equally respectable phoneticians, who claim
that there is no measurable difference whatsoever.

But not even the few instrumental phoneticians who claim that
there is a difference have said anything remotely like the statements
that have appeared on LINGUIST suggesting that the final consonants
in words like 'grad' are "half-devoiced".

"Half-voiced" may be a way of describing Dutch initial 'v' or the
way that certain German dialects pronounce all stops, but this has
no relation at all to the literature which claims that very small
but nevertheless measurable differences exist between such
pairs as Polish 'grat' and 'grad'.

Of course, I could be wrong: perhaps I have missed something
in the literature that does says this, but repeated requests
for references to where anybody would say anything about "half-
voiced" finals in Polish (or Russian, or Catalan, or Standard
German) have been met with totally irrelevant references to sounds
in other languages and in other environments than those under
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue