LINGUIST List 10.1824

Tue Nov 30 1999

Disc: Re: Issue 10.1778 What Exactly Are Allophones?

Editor for this issue: Karen Milligan <>


  1. Dan Moonhawk Alford, Disc: What exactly are allophones?
  2. Jorge Guitart, Re: 10.1817, Disc: Re: Issue 10.1778 What Exactly Are Allophones

Message 1: Disc: What exactly are allophones?

Date: Sun, 28 Nov 1999 17:43:17 -0800 (PST)
From: Dan Moonhawk Alford <>
Subject: Disc: What exactly are allophones?

What a bizarre set of postings on this topic! "Grown" linguists treating
allophones as "things," abstract entities just like phones and phonemes!

Gee, I remember long ago learning from Peter Ladefoged and Vicki Fromkin
at UCLA that an allophone (which we have Benjamin Whorf to thank for,
according to John Carroll's Introduction to _Language, Thought and
Reality_) is a RELATION, not a THING. Phone, allophone, and phoneme,
listed, compared, and contrasted as three things, as some posters have
done, makes me fear the type of training our grad students are getting
these days.

I was either taught or learned by experience that "allophone", written by
itself, is a giant "STAR!" The only safe way to use the form is in the
phrase "[X] IS AN ALLOPHONE OF /Y/ IN LANGUAGE Z", as in "flap is an
allophone of /t/ and /d/ in English". So there is no such "thing" as an
allophone, only phones being in ALLOPHONIC RELATIONSHIP to phonemes.

But why is it left to me to say this? Why haven't the distinguished elders
of our profession weighed in on this yet? Perhaps they dropped off the
list when we stopped having important discussions and it turned into a
professional bulletin board a while back.

(See my webpage, below, for other actual contributions by Whorf -- things
he actually wrote, rather than crackpot Hypotheses falsely attributed to
him by non-linguists and linguists alike, which were fabricated by them
from his self-named "*principle* of linguistic relativity," which he
forged from his understanding of both von Humboldt and Einstein.)

warm regards, moonhawk

Visit Moonhawk's webpage at 
for recent presentations and hard-to-find classic articles.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 10.1817, Disc: Re: Issue 10.1778 What Exactly Are Allophones

Date: Mon, 29 Nov 1999 11:11:53 -0500 (EST)
From: Jorge Guitart <>
Subject: Re: 10.1817, Disc: Re: Issue 10.1778 What Exactly Are Allophones

In response to the following by Jorge Guitart
> << Coming belatedly to the matter of allophones:
> It seems to me that any theorizing on the allophone has to consider at all
> times the fact that allophones are not really physical
> entities but rather psychological ones. You can analyze a
> given pronunciation phenomenon following strictly phonetic criteria and
> discover that it is exactly the same phenomenon that occurred at, say, two
> other previous times but miss out on the fact that the language user
> thought that the first two times it was the 'same' sound while the third
> time it was a 'different' sound. Conversely, three different
> pronunciation phenomena were adjuged by the language user as the 'same'
> sound. >>

John Tiels wrote:

> Jorge,
> I think you are also raising an important point, not only about psychological
> versus physical phenomena (which is one way to read the type/token
> distinction mentioned in some previous postings) but also that the
> description is at the level of psychological reality to the user, and not to
> the researcher...I'm not very well read up on current phonological theory,
> but I believe that phonological theory may be diverging in terms of
> explanatory goals--whereas generative phonology has been concerned with this
> level of psychological reality to the user and with the language user's
> overall judgments of "same" and "different " sounds and how these constitute
> a phonological system, more recent approaches may in fact try to theorize at
> the level of supposed cognitive structures inaccessible to speaker's normal
> awareness, although this may be a simplification...does anyone have
> information about this?
> I'm interested in this partly because of the difference in explaining
> phonology to linguists or phonological specialists and explaining basic
> principles to, for example, second language learners who are not especially
> trained in linguistics--if we perceive a difference between "voiced" and
> "unvoiced" for example, does it help a learner to know that the unvoiced is
> actually voiced, but with a later onset time?
> Isn't it really the perceived contrast between the two that makes a
> difference to the language user (if not the language specialist who can
> identify the "real" processes or the "real" allophones)? If that process is
> normally below conscious awareness, is that level of detail useful or even
> "real" if what is being heard is "unvoiced" as a phonemic contrast?
> A discussion about this point on the Linguist List a couple of years back
> raised this question for me specifically in relation to ESL and the kinds of
> explanations that teachers were giving their students--considering that I
> worked with non-linguist but educated and literate populations who had
> virtually no interest in that level of information.
> It seems that some important aspects of this discussion are:
> - the relationship of "type-level" description (pattern) to token-level
> description
> - the meaning of psychological reality and its relationship to the
> description of a phonemic system, specifically, is one discussing speakers'
> habitual awareness, trained awareness, or cognitive structures and processing
> - the relationship of phonological theory to phonetic theory (and perhaps the
> differential use of the concept "allophone" in both areas)
> - the relationship of sound differences related to meaning to other kinds of
> regular, sound patterning which are not related to propositional,
> denotational meaning (to put it crudely)
> Perhaps some discussion of what kinds of explanation for whom (and by whom)
> would open up some topics on phonological theory only implicitly touched upon
> here.

John, thanks very much for your comments. A propos of explaining phonemes
and allophones to nonspecialists (e.g. students in a Spanish phonetics
course, present and prospective teachers in an intro course on second
language acquisition) I have always used a sort of popularized version of
generative phonology in which there are invariant underlying forms that
are in 'phonemes' and are unobservable (we can only theorize about them)
and phonetic forms which are in 'allophones' and are observable and can be
described physically. I make it a point to tell them that in normal 
communication [as opposed to talking about phonetics] sounds are never
considered in isolation by speakers(and listeners!) but
are always inside the meaningful units that we call WORDS, and one
important thing to take into account is the fact that normally users
perceive the different pronunciations of a word as the same 
word, even though it sounds very differenly from the word as pronounced in
isolation. One good illustration of this is what happens to English 'you'
in 'See you later' and 'Where did you go this morning?' (Also what happens
to 'did' in the latter sentence!) With the same examples you can show the
kind of distortion that happens at the physical level because of phenomena
like assimilation, vowel reduction, etc. But the crucial thing, I tell
students, is what happens inside your head. Despite all the distortions,
you still hear the word 'you' in every case. I also use American English 
t/d flapping to illustrate the fact that people perceive as different
sounds that are physically the same. (Again, allophones are psychological
units allophones are not physical units but psychological ones.) From
there we go to the mediation between the phonemic and the phonetic (a set
of principles (you can call it rules) that the speaker applies and the
listener 'undoes'.


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