🚀 go-pugleaf

In article <1993Jan27.040154.20592@midway.uchicago.edu>, goer@ellis.uchicago.edu (Richard L. Goerwitz) writes:
> What is a phoneme?  (Not a joke question; I just discovered that
> I honestly don't know, and suspect that nobody else does, either.
> Perhaps this is just sheer arrogance on my part, though, so I ask
> everyone now...).

I'll go way out on a limb and make most of the readership made at me
with this definition:

	A phoneme is the minimal psychological construct representing
	certain sounds which occur in a given language.

This accounts nicely for allophones, which are the actual sounds
corresponding to a given construct, and does not lock us in to stupid
attempts to locate phonemes out there in actual speech, as the
structuralists before Chomsky tried to do.  It also avoids the
Chomsky-Hallean stupidity of *denying* phonemes entirely in favor
of feature matrices, which are fine for analyzing and contrasting
sounds and for constructing phonological rules, but which tell us
nothing about how speakers conceptualize sounds.

Attack if you dare; I'm spoiling for a fight.

--Leo Connolly

connolly@memstvx1.memst.edu writes:

>> What is a phoneme?
>
>I'll go way out on a limb and make most of the readership mad at me
>with this definition:
>
>	A phoneme is the minimal psychological construct representing
>	certain sounds which occur in a given language.

This is intuitively correct.  In fact, I don't think that Chomsky him-
self is really all that concerned about phonemes.  It was Halle that
got into all the matrices.  I also don't believe that either denied the
existence of systematic units called phonemes.  They just made them far
more abstract, and treated them as bundles of features.  The term for
their kind of phoneme is "systematic phoneme."

Chomsky effectively showed that systematic phonemes were more practical
and general that the so-called "autonomous" ones (i.e. ones treated as
irreducable psychological units defined by distributional contrastive-
ness).  If we use autonomous phonemes in biblical Hebrew, for instance,
we end up with way, way more phonemes than we want or need.  The idea of
a systematic phoneme is one of those units that are left when all of the
ostensibly regular phonological rules are reversed (rules which may cre-
ate superficially minimal pairs, but which in fact are simply reflexes
of a much simpler underlying system).

The only criticism I would level at your definition is that it is
vague.  Can you offer us a discovery procedure like the descriptivists
and structuralists had?  Can you offer us a cognitive model and a for-
malism, like what generative phonologists have?  Simply calling a phoneme
a "psychological construct representing certain sounds" doesn't move us
off of square one.

>Attack if you dare; I'm spoiling for a fight.

Tell you what.  I'll believe that your definition is clear if you can
take a standard data set, and show how your definition of the phoneme
offers us a superior way of describing some synchronic phonological pro-
cess.

I'm not so much spoiling for a fight as I am spoiling for some examples :-).

-- 

   -Richard L. Goerwitz              goer%midway@uchicago.bitnet
   goer@midway.uchicago.edu          rutgers!oddjob!ellis!goer

In article <1993Jan27.095659.5244@memstvx1.memst.edu> connolly@memstvx1.memst.edu writes:
>In article <1993Jan27.040154.20592@midway.uchicago.edu>, goer@ellis.uchicago.edu (Richard L. Goerwitz) writes:
>> What is a phoneme?
>
>	A phoneme is the minimal psychological construct representing
>	certain sounds which occur in a given language.
>
>Attack if you dare; I'm spoiling for a fight.

Hey, yeah, a fight over the phoneme!  It happened in the 30s!  It happened in
the 60s!  Now, it's the 90s, and we're well overdue for a good brawl over
mentalism in phonology.  :)

> stupid attempts to locate phonemes out there in actual speech, as the
> structuralists before Chomsky tried to do.

If, by "locate phonemes ... in actual speech," you mean denying the mental
reality of phonemes, you're not reading the same structuralists I am!
 
Actually, Kenneth Pike's perspective turns out to be pretty interesting,
because he contradicts himself:

"The sounds of a language are automatically and unconsciously organized by
the native into structural units, which we call PHONEMES." (Pike, 57)

It seems clear that the sounds of a language *are* automatically and
unconsciously organized by the native, but I don't think I've ever understood
the Aristotelian mania that drives the search for absolutely distinct
"structural units."  Sometimes that makes sense, and sometimes it don't.
You can't expect everything to fit (cf. Feyerabend, _Against Method_).
What's important here is that Pike definitely believes in the psychological
reality of phonemes; he doesn't think of them as things that *only* exist
as linguistic models (and neither do I).
 
Intriguingly, Pike seems to reverse himself:

"a PHONEME is one of the significant units of sound arrived at for a 
particular language by the analytical procedures developed from the basic
premises previously presented." (Pike, 63; this whole quote was originally
underlined)

Fabulous, Ken!  Phonemes are artifacts of linguistic methodology!  Good call;
there's definitely something to this.  But then again, what about that 
"Psychological Reality of the Phoneme" stuff?  That's easy to explain;
the organization of the sounds of a language *is* psychologically
real, and Pike's methodology usually tells us how to predict that
organization.  OK, no problem then.  A phoneme is a conceptual by-product
of a linguistic methodology founded on abstract objectivism; it is used to
categorize the sounds of a language in a manner that is, hopefully, congruent
with the way a native categorizes sounds psychologically.

The interesting question is probably, when and why does the phoneme fail?
Pike mentions the "under-differentiation" and "over-differentiation" of
phonemes, along with the "neutralization of oppositions" as special cases, and
he introduces the "ARCHIPHONEME: A special type of phoneme of limited
distribution, postulated to account for under-differentiation of phonemes"
(Pike, 233).  The 'archiphoneme' seems strongly reminiscent of the 'epicycle,'
don't you think?  I guess the mind is a strange place -- one where things
just don't always break down into minimal units (contra Connolly :).

Anyway, now we have an idea about when the phoneme fails (or needs serious
emendation), but what would be really special is to know why.  I've heard
rumors that we're stuck on phonemes as a by-product of alphabetic literacy,
and I've already mentioned the Aristotelian drive to categorize absolutely.
Bakhtin's explanation is much more interesting: "The desire to see language
as a static synchronic system, according to Bakthin, was symptomatic of a
kind of linguistic necrophilia, a nostalgia for deceased languages, whose
systems could be fixed precisely _because_ they were dead.  The fundamental
categories of Saussurean linguistics, which are phonetic (units of sound)
and morphological (units of sense) ultimately derive, he argued, from the
categories of Indo-European comparative linguistics, precisely those
categories most appropriate to a dead or alien [?] language.  Saussure belongs
to a tradition of reflexion on language which Bakhtin labels ABSTRACT
OBJECTIVISM, i.e. a view of language which strives to reduce its constantly
changing heterogeneities into a stable system of normative forms" (Stam,
et al., 13).  In doing this, Saussure is "an heir to the tradition of
Descartes, Leibniz, and Condillac" (ibid.).

Personally, I'd say that was a pretty serious insult.

Pike, not to mention the many other theorists of the phoneme, is clearly
subject to Bakhtin's criticism.  Some of the "subpremises" he founds his
theory on are particularly telling: "(1) A phonemic orthography is the
easiest one for the native to learn to read and write ... (4) There is only
one accurate phonemic analysis for a specific set of data ... (12) Fluctuation
can occur between full phonemes ... (14) Every language has consonants and
vowels ... (15) Certain kinds of segments may be vowels in one language but
consonants in another, and vice versa ... (21) Occasionally a single segment
may constitute a consonant and a vowel simultaneously" (Pike, 62).  From these,
we can infer that writing is very high on Pike's agenda, that ambiguity won't
be tolerated (*despite* "fluctuation"), and that traditional phonological
categories (inherited from Dionysius Thrax, ~100 B.C.!) are to remain
unchallenged, even if there is some confusion about them.

Deriving the fallibility of the phoneme from the inflexibility of writing is
no huge chore, at this point.  Linguistic categories are obviously subject to
the tyranny of the letter.  Comparative linguistics began as a study of
written language, after all (I'm thinking mainly of Franz Bopp), and Pike's
whole project is described as an attempt to reconcile speech with
writing ("to lead the student to arrive at an [written!] analysis which
parallels the vague or explicit observable reactions of speakers to their
own sounds" (Pike, 64; again, all underlined)).  If such a reconciliation
should prove to be fundamentally flawed, I guess I wouldn't be surprised.

So, in summary, I agree that sound patterns *are* organized mentally;
however, that organization is *not* minimally constructed, as Connolly
would say.  Rather, it has a socially, historically, and idiosyncratically
constructed dynamism to it.  This is not exactly ... 

> how speakers conceptualize sounds

because I'm afraid I agree with Pike in locating these sound patterns in
the unconscious.  Anyway, I'd say the desire for phonemic minimalism results
simply from linguistics' positivist, superliterate heritage.

[Disclaimer: I'm only guessing. :) ]

--
References:

Pike, Kenneth L., _Phonemics: A Technique for Reducing Languages to Writing_.

Stam, Robert, et al., _New Vocabularies in Film Semiotics_.
-- 
Christopher Pound	+	"Homer deserves to be thrown out of the
Anthro Grad Student	+	 contests and flogged." 
pound@ruf.rice.edu	+	 	                  -- Heraclitus

In article <1993Jan27.040154.20592@midway.uchicago.edu> goer@midway.uchicago.edu writes:
>What is a phoneme?  (Not a joke question; I just discovered that
>I honestly don't know, and suspect that nobody else does, either.
>Perhaps this is just sheer arrogance on my part, though, so I ask
>everyone now...).

Maybe it's not a joke, but it's certainly a bit coy.  You wouldn't ask if you
didn't already know what answers you were going to get.

Let's cut to the chase:  Sure, the phoneme isn't the end-all be-all of 
linguistics.  This doesn't mean we don't know what one _is_.  I offer the 
following evidence that I know what a phoneme is:  Give me a reasonable amount
of data about any language, and I'll give you back at least one phoneme of
that language.  Obviously, this isn't a spectacular claim about the 
descriptive power of the concept, but it does show that I know some definition,
even if it's only a procedural definition (how to find one).

Just what's your point?

Bob
r-knippen@uchicago.edu

Christopher Brian Pound writes:
>>
>>	A phoneme is the minimal psychological construct representing
>>	certain sounds which occur in a given language.
> 
>Actually, Kenneth Pike's perspective turns out to be pretty interesting,
>because he contradicts himself:
>
>"The sounds of a language are automatically and unconsciously organized by
>the native into structural units, which we call PHONEMES." (Pike, 57)
>
>It seems clear that the sounds of a language *are* automatically and
>unconsciously organized by the native, but I don't think I've ever understood
>the Aristotelian mania that drives the search for absolutely distinct
>"structural units."  Sometimes that makes sense, and sometimes it don't.
>You can't expect everything to fit (cf. Feyerabend, _Against Method_).
>What's important here is that Pike definitely believes in the psychological
>reality of phonemes; he doesn't think of them as things that *only* exist
>as linguistic models (and neither do I).

You're a genius.  This hits the nail on the head.  The fact is the in some
ways we operate with a system that looks a lot like structuralists' autono-
mous phonemes.  In other ways, though, we are quite capable of operating in
ways that are more abstract, more amenable to generativists' systematic
phonemes.

The two models are not mutually exclusive.

>Intriguingly, Pike seems to reverse himself:
>
>"a PHONEME is one of the significant units of sound arrived at for a 
>particular language by the analytical procedures developed from the basic
>premises previously presented." (Pike, 63; this whole quote was originally
>underlined)

In a recent conversation with Patricia Donegan, one of the participants
in this forum, she mentioned that it simply never occurred to some of
the mid-century structuralists that the distributional and psychological
definitions of the phoneme were different.  Generative phonologists don't
seem to show signs of taking the distinction to heart, either.

Pike was a really good mind, I believe.  IMHO.

>The interesting question is probably, when and why does the phoneme fail?
>Pike mentions the "under-differentiation" and "over-differentiation" of
>phonemes, along with the "neutralization of oppositions" as special cases, and
>he introduces the "ARCHIPHONEME: A special type of phoneme of limited
>distribution, postulated to account for under-differentiation of phonemes"
>(Pike, 233).  The 'archiphoneme' seems strongly reminiscent of the 'epicycle,'
>don't you think?  I guess the mind is a strange place -- one where things
>just don't always break down into minimal units (contra Connolly :).

If I say "bad" in my pronunciation to many of my neighbors here in Chicago,
they understand it, but repeat the word back in their own dialect.  They
don't notice anything other than a slight accent.  They immediately map the
form to a phonemic structure of some sort, and when they repeat it, they
essentially ignore the phonetic nuances of my speech.  Clearly a psychological
phoneme here.  Note, though, that when I talk about "root beer," they often
use my pronunciation, and often use theirs as well.  My mother-in-law uses
the tense o pronunciation except in the phrase "A & W Root Beer," in which
cases she uses [u].  The fact is that this dialectal difference is perceptible.
It crosses phoneme boundaries.  Given that they have no rule that maps the one
sound to the other, they simply treat the two forms of "root" as synonyms.
Sound seem to be mapping to abstracts units of some kind.

I have this crazy theory that morphology can condition only rules based on
psychological phonemes.  Subphonemic processes are never part of the grammar,
never conditioned by higher-level considerations.  I've had enough conversa-
tions with David Stampe that I suspect I've derived this notion from him
somehow.  I have some interesting data to back it up if anyone is *really*
interested.  It's not as circular as it sounds.

Or then maybe it is.

>Pike, not to mention the many other theorists of the phoneme, is clearly
>subject to Bakhtin's criticism.  Some of the "subpremises" he founds his
>theory on are particularly telling: "(1) A phonemic orthography is the
>easiest one for the native to learn to read and write ...

My experience with Semitic scripts indicates that they don't even represent
all phonemes.  What is more, when diacritics are introduced to fill in the
gaps, these diacritics do not represent systematic phonemes; nor do they
represent autononous phonemes, as discovered by traditional, distributional 
methods.  What do they represent?  Let's call them psychological phonemes
for the fun of it.

Patricia Donegan has promised to send me a paper on a related topic, and I
am eager to read it.

>So, in summary, I agree that sound patterns *are* organized mentally;
>however, that organization is *not* minimally constructed, as Connolly
>would say.  Rather, it has a socially, historically, and idiosyncratically
>constructed dynamism to it.  This is not exactly ... 
>
>> how speakers conceptualize sounds...
>
>[Disclaimer: I'm only guessing. :) ]

So are all of us.  But at least we're admitting it.


-- 

   -Richard L. Goerwitz              goer%midway@uchicago.bitnet
   goer@midway.uchicago.edu          rutgers!oddjob!ellis!goer

rmk4@midway.uchicago.edu writes:

>>What is a phoneme?  (Not a joke question; I just discovered that
>>I honestly don't know, and suspect that nobody else does, either.
>>Perhaps this is just sheer arrogance on my part, though, so I ask
>>everyone now...).
>
>Maybe it's not a joke, but it's certainly a bit coy.  You wouldn't ask if you
>didn't already know what answers you were going to get.

Honestly, I didn't know what answers I'd get.  Every now and then it's fun
to ask a question that will separate true students of linguistics from the
gabby pack.  This time I have some real concerns that, despite supposed ad-
vances in phonology, nobody knows what a phoneme is, despite its widespread
use (in various forms) in the literature.  I'm just fed up with my own shabby
attempts at defining the bloody thing, and thought I'd use my frustrations
to spice up sci.lang.

>Let's cut to the chase:  Sure, the phoneme isn't the end-all be-all of 
>linguistics.  This doesn't mean we don't know what one _is_.  I offer the 
>following evidence that I know what a phoneme is:  Give me a reasonable amount
>of data about any language, and I'll give you back at least one phoneme of
>that language.  Obviously, this isn't a spectacular claim about the 
>descriptive power of the concept, but it does show that I know some definition,
>even if it's only a procedural definition (how to find one).

What's your procedural definition, in brief.

Note:  I'm being coy, but I'm not fishing for any specific answer.  I really
want to know if anybody can tell me what a phoneme is, or how to find one.

-- 

   -Richard L. Goerwitz              goer%midway@uchicago.bitnet
   goer@midway.uchicago.edu          rutgers!oddjob!ellis!goer

In article <1993Jan27.183540.15379@midway.uchicago.edu>, goer@kimbark.uchicago.edu (Richard L. Goerwitz) writes:
> connolly@memstvx1.memst.edu writes:
> 
>>> What is a phoneme?
>>
>>I'll go way out on a limb and make most of the readership mad at me
>>with this definition:
>>
>>	A phoneme is the minimal psychological construct representing
>>	certain sounds which occur in a given language.
> 
> This is intuitively correct.  In fact, I don't think that Chomsky him-
> self is really all that concerned about phonemes.  It was Halle that
> got into all the matrices.  I also don't believe that either denied the
> existence of systematic units called phonemes.  They just made them far
> more abstract, and treated them as bundles of features.  The term for
> their kind of phoneme is "systematic phoneme."

I don't have their book at hand, but it is organized into several sections
that, if I remember correctly, are often simply incompatible.

That having been said, I remember that they indeed talk of systematic
phonemes and even introduce a notation: |x| is a "systematic" phoneme
while /x/ is a conventional one.  By pointing out some interesting
problems with the formulation and application of phonological rules,
they show (to their satisfaction) that conventional phonemes are
superfluous, at which point Ockham decided to shave.  The trouble is
that the same arguments apply to systematic phonemes, which in fact
they hardly use.  And the systematic phonemes are *not* tantamount
to feature matrices, since the actual matrices they posit for
words such as _spit_ do not correrspond to *any* sort of phoneme.
Since English words do not begin with consonant + voiced stop,
the matrix for the [p] is unspecified for the feature [voice].  So
the actual matrix would have to correspond to that monstrosity, the
archiphoneme, rather than to any phoneme at any level.  And worse,
since the only consonant allowed ahead of a stop in English is [s],
the matrix for this consonant is simply [+cons].  What sort of phoneme
is that?  Not even one of theirs!  And unless we want to posit an
"arch-archiphoneme" (quod Deus avertat), we have no sort of unit
at all, just a very imperfectly specified matrix.

> The only criticism I would level at your definition is that it is
> vague. 

Conceded.

> Can you offer us a discovery procedure like the descriptivists
> and structuralists had?

The structuralist procedures are pretty good.  It's the constructs
that are wrong because of their hide-bound premise that a given
phone can correspond to only only phoneme, not two or more.  Alterna-
tions such as _relate : relation_ can be blamed on "morphophonemics",
but that's not an explanation: it's an attempt to avoid saying that
some instances of [s^] are /s^/ while others are /ty/.  German _Rad_
'wheel' was analyzed as /ra:t/, while the genitive _Rades_ was
/ra:d@s/.  Heck, the structuralists' procedures could show perfectly
well that /t/ and /d/ were different phonemes; they simply couldn't
bring themselves to analyze _Rad_ as /ra:d/, contrasting with _Rat_
'advice' (genitive _Rates_), which was obviously /ra:t ra:tes/ (no
need for /@/ in German if you play your cards right).

>  Can you offer us a cognitive model and a for-
> malism, like what generative phonologists have?

A cognitive model is asking a bit much, and anything offered by
generative phonologists or anyone else is partial.  I am saying
that phonemes are represented as indivisible units (not bundles
of features), and that these units do not always correspond to units
in speech on a 1:1 basis.  We linguists can and do break these
units down into features and write very plausible phonetic rules
which are difficult even to express any other way.  This may or
may not be a good thing, and I don't know if such rules correspond
to any mental processes.  It might be better merely to note the
correspondces between the abstract units and their varying realization
in speech.

>  Simply calling a phoneme
> a "psychological construct representing certain sounds" doesn't move us
> off of square one.

No, but it brings us there if we've been off chasing ghosts.  Seriously:
setting a plausible starting point is worth something if we haven't been
able to agree where to start.

> Tell you what.  I'll believe that your definition is clear if you can
> take a standard data set, and show how your definition of the phoneme
> offers us a superior way of describing some synchronic phonological pro-
> cess.

I think _spit_ is a least a standard datum.  Does what I said above about
it clarify some of the issues?  Isn't it better than messing with archi-
phonemes or with segments represented merely as [+cons]?  Haven't I
put my phonemes in the right place?  Let's see what others have to say
before I elaborate any further.

--Leo Connolly