LINGUIST List 10.1879

Tue Dec 7 1999

Disc: What Exactly Are Allophones?

Editor for this issue: Karen Milligan <karenlinguistlist.org>

Directory

1. H. M. Hubey, Re: 10.1876, Disc: What Exactly Are Allophones?
2. Dan Moonhawk Alford, Re: 10.1855, Disc: What Exactly Are Allophones?
3. Jorge Guitart, Re: 10.1876, Disc: What Exactly Are Allophones?

Message 1: Re: 10.1876, Disc: What Exactly Are Allophones?

LINGUIST Network wrote:
> 
> LINGUIST List: Vol-10-1876. Mon Dec 6 1999. ISSN: 1068-4875.

> Date: Sun, 05 Dec 1999 18:44:03 -0800
> From: "Patrick Farrell" <pmfarrellucdavis.edu>
> Subject: Re: 10.1855, Disc: What Exactly Are Allophones?

> Now, this doesn't prevent "allophone" from ALSO being used in such a way as
> to designate one of the things that plays a role in an allophony
> relationship. Once we have established that we are talking about the people
> who are members of a particular club, we can talk about counting members,
> seeing members, talking to members, etc. Why? Because "member" can be used
> to designate the things that are in a given membership relation. Similarly,
> "allophone" could be and perhaps is used to designate the phones that are in
> a given allophonic relation with a phoneme (and/or other phones). Once we
> have established that we are considering the phones [x] and [y], which are
> allophones of /z/ (relational use of "allophone"), we could talk about the
> distribution of the allophones [x] and [y] (thing-designating use of
> "allophone"). Thus, there is a sense in which Jorge and advocates of the
> "allophone is thing" stance are surely right. People at least could use
> "allophone" in such a way as to designate something conceived of as a thing.

Indeed this is the difference between everyday talk and precision
required
of science.

Normally we'd say things like:

1. Susan is human.
2. Saturn is a car.
3. Taunus is a Ford.
4. Fido is a dog.

but not

5. An apple is an apple.

The differences are that we do not ordinarily name our apples. To
see
this let's look at what we mean by these words. Those who bring
up
the concept that allophones, phones, phonemes are "abstract"
things
really mean that they are "sets" (more or less). The reason for
this,
we are told, is that nobody hears a phoneme for it is an abstract
thing
and we only hear specific instantiations (sample functions in
probability
theory, or "tokens" in computational linguistics and computer
science).

We see that in (1) Susan is a token/member of the set human. In
(2)
Saturn is a member/subset of the set "car". In (3) Taunus is a
member/subset
of the set "Ford". In (4) Fido is a specific example of the set
"dog". But 
(5) sounds really stupid because it is so trivially true. If on
the other
hand we pointed to some object on a table and said;

6. That thing is an apple.

then again we have asserted that some specific object is a member
of
the set "apple". 

But we should not overdo it. Because we can get stuck in some
nasty
mind loops, for example: Nobody has seen a vehicle because it is
an
abstract concept. There are examples of them, say, like cars,
buses,
trains, trucks. But nobody has seen any of these either, because
they
are also sets, and they have specific examples, such as Saturn,
Ford,
Volkswagen, etc. But nobody has seen these either because they
are
also sets, and they have as members abstract objects like,
Taunus, 
Corolla, etc. But then again, these are also sets, so we
eventually
get down to specific examples/tokens such as "My Saturn", 
"Jane's Honda". These are specific concrete objects. The rest are
not.

Exactly in the same way, only specific realizations, recordings
of
speech sounds exist. Phonemes, phones, and probably allophones
(except
for those that are specific realizations) are all "abstract" in
the
same sense, i.e. they are sets.



- 
M. Hubey, Computer Science
Email:	hubeyhmail.montclair.edu
Backup:hubeyhalpha.montclair.edu
WWW Page:	http://www.csam.montclair.edu/Faculty/Hubey.html

Message 2: Re: 10.1855, Disc: What Exactly Are Allophones?

or, The Allophonic Tower of Babel

[Ahmad R. Lotfi writes:]

<<I guess this theme has been the most popular one on the list for the
whole year (though most probably NOT the most important one). But why is
it so? I doubt it is (as somehow suggested) due to this linguistic
education of ours. It's simply because we linguists today have so little
in common to talk about. This divergence of interests can be primarily a
symptom of a crisis rather than the maturity of the science of language. 
We've become so much involved in our own 'field of specialty' that
linguistics is already on its way to turn into a private science. I think
that's why Linguists have welcome allophones so warmly: they come from the
already scarce terminolgy we have inherited from the 'wise' linguists of
the past. Ironically enough, and in want of a common language, scientists
of *language* are 'babbling' (rather than talking) among themselves with
the very rare occasion of some good old words of the past (like 'phone',
'phoneme', and 'allophone') heard from time to time: does it mean the
tower of Babel may be left unfinished another time?>>

[moonhawk]

I welcome this as one of the most brilliant and insightful comments yet
put forth in this thread. Is there indeed a linguistics field any more, or
are there only linguisticses? Have we gone beyond mutually understandable
to mutually unintelligible lects of linguistics? As we delve, each in our
own way, into the meaning-full heart of the human cosmos, we find we can
no longer report our findings to each other intelligibly, lacking a common
metalanguage with which to speak of language -- as shown by our dear,
confused allophone, now not knowing, in our collective professional mind,
what to be: whether to be a "state" or a "process", a "thing" or a
"relation", and if a "relation" whether that will ultimately be seen as
another "state" or "thing" by those so inclined.

[Jorge Guitart writes]

<<PPS Message to Dan (who thinks that allophones are relations): Can
relations have phonetic features? We speak in English of the aspirated
allophone of /p/ as opposed to the unaspirated one. In Spanish we have the
approximant allophone of /b/ as opposed to the plosive one, etc.>>

[moonhawk]

Here's a radical idea: what if "allophone of" is a generic way, from the
phone point of view, of saying it's related to one or more phonemes, while
from the phoneme point of view we specify that same relationship by terms
like assimilation, complementary distribution, free variation. I.e., when
we say "flap is an allophone of /t/ in American English" or "flap is an
allophone of /r/ in British English," they are shorthand ways of stating
that there is a relationship between phone and phoneme without having to
specify exactly what relationship it is, in which contexts it will be
found, etc.

I'm not sure what it would mean for allophones to relate to one another,
or why one would need that, when allophones are seen from a thing
perspective, though when seen from a process perspective I can, as rules
sequencing with each other.

So do relations have phonetic features? Well, do rules "have" phonetic
features? I think not. They USE the phonetic features; they ASSUME them;
but they don't HAVE them -- only the phones they're working with have
phonetic features, just as the phonemes have phonemic features.

Well, except that "have" is tricky: we must remember that it is WE who
project all of these distinctions onto both what we think we hear and
those realms of the unseen and unheard which we try to make visible.

And WE can come to the issue from dramatically different places. My
experiences in the field with Cheyenne in the '70s -- developing an
orthography, dictionary and K-4 curriculum -- nearly force me to come from
a different place on the allophone issue than someone who has worked
exclusively with phonological theories and prepared data, for instance. In
physics this is known as putting the observer back into the equations.

So maybe there's room for two different definitions of allophone in the
linguisticses after all -- one, the classical one, which is good for
actually doing fieldwork and analyzing the data, and another, more modern
or post-modern, which is good for ... forgive me -- I guess I'm still not
clear what use an "allophone as a thing" is, so I won't be able to
complete that sentence. Maybe someone can complete it for me. ;-)

The above was composed before seeing Patrick Farrell's comments, which I
heartily enjoyed, of course -- and especially his final points:

[Patrick Farrell writes:]

<<In fact, is there any real issue? Words are often polysemous. One of the
ways they can be polysemous, apparently, is in alternating between
designating a certain relation among things and the things that are in
that relation, much like "hammer" alternates between designating a process
involving a certain kind of instrument and the instrument used in that
process.

Of course, I take it that one of Dan's points is that it is ONLY the
English-contingent fact that "allophone" is a noun that leads us to
associate a superfluous thing-designating concept with it and maybe even
to consider this concept its central or primary sense and to think that
the word is needed to designate some kind of thing that exists in the
world independent of our conceptualization of it. I don't dispute this
point.>>

[moonhawk]

'Nuff said.

warm regards, moonhawk

Dan Moonhawk Alford <dalfordhaywire.csuhayward.edu>
CSU Hayward, CIIS, JFKU

Visit Moonhawk's webpage at 
<http://www.sunflower.com/~dewatson/alford.htm>;
for recent presentations and hard-to-find classic articles.

Message 3: Re: 10.1876, Disc: What Exactly Are Allophones?

Patrick Farrell wrote

Two points of view concerning "allophone" are expressed in the following:
> 
> >>JORGE SAID
> > > Saying that an allophone is a relation is negated by saying that X is an
> > > allophone of Y in language Z, since X is ***something****.
> 
> >DAN RESPONDED
> > You've totally misunderstood me. I specifically put brackets around the
> > [X], which you've abandoned; too bad, since the brackets say, universally
> > to linguists, that the [X] is a PHONE. Perhaps I should have written it
> > out more clearly: The phone [X] is an allophone OF the phoneme /Y/. I was
> > trying to say that "allophone" by itself is misleading, assisting us to
> > imagine that it is a thing, whereas "allophone of" points more directly to
> > its use as a relation between two things, phone and phoneme. "Allophonic"
> > would be even more preferable for seeing its relationship role.
> 
> It seems pretty clear to me that there is a sense in which Dan is surely
> right. It would be pretty weird to say "The word I just heard her say
> consisted of three allophones," whereas it would be OK to say "The word I
> just heard her say consisted of three phones/phonemes." Why? Because
> allophony is fundamentally a relational concept. "Allophone" is quite like
> the word "member". If someone asked me to list all the things in the world,
> I wouldn't say "whales, pine trees, members, ..." Why? because "member" is
> an inherently relational concept. It designates the relation that exists
> between a group and its constituents. By the same token, if asked to list
> the things that make up languages, I would say, perhaps, "phones (or
> phonemes), words, sentences, etc.", but not "allophones, words, sentences,
> etc." "Allophone" (which is used generally in the predicate phrase "be
> allophone of") primarily designates the relation that exists between phones
> and phonemes.
> 
> Now, this doesn't prevent "allophone" from ALSO being used in such a way as
> to designate one of the things that plays a role in an allophony
> relationship. Once we have established that we are talking about the people
> who are members of a particular club, we can talk about counting members,
> seeing members, talking to members, etc. Why? Because "member" can be used
> to designate the things that are in a given membership relation. Similarly,
> "allophone" could be and perhaps is used to designate the phones that are in
> a given allophonic relation with a phoneme (and/or other phones). Once we
> have established that we are considering the phones [x] and [y], which are
> allophones of /z/ (relational use of "allophone"), we could talk about the
> distribution of the allophones [x] and [y] (thing-designating use of
> "allophone"). Thus, there is a sense in which Jorge and advocates of the
> "allophone is thing" stance are surely right. People at least could use
> "allophone" in such a way as to designate something conceived of as a thing.
> 
> In fact, is there any real issue? Words are often polysemous. One of the
> ways they can be polysemous, apparently, is in alternating between
> designating a certain relation among things and the things that are in that
> relation, much like "hammer" alternates between designating a process
> involving a certain kind of instrument and the instrument used in that
> process.
> 
> Of course, I take it that one of Dan's points is that it is ONLY the
> English-contingent fact that "allophone" is a noun that leads us to
> associate a superfluous thing-designating concept with it and maybe even to
> consider this concept its central or primary sense and to think that the
> word is needed to designate some kind of thing that exists in the world
> independent of our conceptualization of it. I don't dispute this point.
> 
> Patrick Farrell
 
 
Jorge Guitart replies

Thank you, Patrick, for your very clear distinctions. I wonder if you
would be opposed to my telling students that at the observable level, a
word is in 'allophones' (rather than in phones) while at the mental level
it is in 'phonemes'. It is in allophones because every sound represents a
phoneme. An allophone is a representative of some phoneme. 

Let's say you have a social club and there are several committees that are
in charge of the different activities of the club. Only club members can
be committee members. Now, member is, as you say, a relational concept (cf. it being a relation, a la
Dan). I can say that every committee is composed solely of
members and I can say that every committee has three members. I don't have
to say 'of the club'. It is 'understood'. How is 'member' different from
'allophone'? 

I look at a transcription of English 'pop-up' as in baseball. The first p
is marked as aspirated and the last as unreleased; the middle p is not
marked at all, by convention. Let's say the vowels are somehow marked with
some kind of detail,e.g. length (the main-stressed vowel is a bit longer
than the unstressed one, say) Do you mean to tell me that I cannot say to
someone to whom I am trying to teach the meaning of allophone "Look, this
is all alophones"? Notice the singular, 'this is', not 'these are'.To me
it is like saying' This form I am pointing to is all "in allophones".'

If I remember correctly the original query was how to define 'allophone'
for someone who doesn't know what they are. 

I would go for the thingy thing: For any phoneme, its allophones
are the sounds that speakers perceive as identical even
though they are physically different. Substitute phones for sounds if you
wish. And of course a given sound/phone can in one instance be the
allophone of phoneme X and in another instance the allophone of phoneme Y
since real people do not respect biuniqueness. To insist: since the
important thing is the identity, 'allophone' is a psychological entity,
not a physical one. I have no quarrel with saying that it (whatever 'it'
refers to) enters into a relation. Surely it does. But it is not a
relation itself.

Best

Jorge



Jorge