Contrasts

[From Rick Marken (931005.1300)]

Martin Taylor (931005 13:30) --

Bruce is trying to develop the idea of phonemic contrasts, and
Rick resists the perception.

Not at all. I am trying to determine the referent (the usual meaning,
not Hal's) of the word "contrast". Everything Bruce has said in his
posts on the "pair test" convinces me that he uses the term "phonemic
contrast" to refer to what are called "errors" in PCT.

I think they are talking about different
levels of the talker/listener's hierarchy, and will never agree if they
continue to do so.

This implies that the word "contrast" refers to a perception of some
kind and that Bruce and I are only disagreeing about the type of
perception. If this is what Bruce meant by contrast (a type of perception)
then I have no problem with it; I have many perceptions that I would
be comfortable calling "contrasts"; my perception of the relationship
between bin and pin is one, as well as my perception of the relationship
between the height of a very short and a very tall person, etc. What I
am "resisting" is the idea that "contrast" refers to some special property
of speech which provides a basis for its perception. I reject this meaning
of "contrast" because, as an explanation of speech perception, it is
"dormative", ranking right up there with "affordance".

Bruce used the continuum level (finessing "configuration-transition")
phonetic difference, to try to present the category level phonemic contrast.
Rick used the phrase "word event," which seems to me to be an oxymoron

You are seeing a disagreement between Bruce and I that does not exist.
If Bruce thinks of a contrast as nothing more than a phonetic difference
then I have no problem; the word "contrast" is just being used to refer
to perceived differnces between speech sounds. I think that limits a nice
word a bit (why not include perceived difference between anything -- like
short and tall people) but it certainly makes sense to me. If this is all
Bruce means by "contrasts" then I perceive no contrast between our views.
But I have the feeling (and Bruce can correct me if I'm wrong) that Bruce
thinks there is something more to "contrast" than "perceived difference".
I think he believes that contrasts "do" something to make speech perception
possible. I think Bruce believes that contrasts EXPLAIN perceptions (notably
of speech); I believe that contrasts ARE perceptions.

Best

Rick

[Martin Taylor 931005 16:00]
(Rick Marken 931005.1300)

Bruce used the continuum level (finessing "configuration-transition")
phonetic difference, to try to present the category level phonemic contrast.
Rick used the phrase "word event," which seems to me to be an oxymoron

You are seeing a disagreement between Bruce and I that does not exist.
If Bruce thinks of a contrast as nothing more than a phonetic difference
then I have no problem; the word "contrast" is just being used to refer
to perceived differnces between speech sounds.

You missed my point utterly, which is a pity, since I was trying to "contrast"
your readings with my readings of Bruce. I do see a difference, in that
you insist on removing a distinction Bruce is trying to make, a distinction
that I think is important to make. If Bruce isn't trying to make the
distinction, then I am indeed seeing a disagreement between Bruce and you
that does not exist, but that leaves a disagreement between you and me that
does exist (at least I perceive it to be so).

My point was that "contrast" is a good word to use for differences at the
category level (and perhaps above), whereas "perceived difference" has the
implication that it can have a continuum of values from small to very big.
Categories are not like that. Their differences are more in whether a
pattern is a member of one and *therefore* not of another one. Categories
themselves may be similar or very dissimilar, which is how we arrive at
taxonomies. But if a perception is that a pattern is a member of a category,
the configurations (transitions, events, what have you) that lead to that
perception tend to be ignored.

To follow Bill Powers, today (reference not available, but called "fish
chowder"), the lower level perceptions are easily controlled, and hence
not normally consciously perceived. The categories (words) and sequences
(syntax) may themselves not be consciously perceived, for the same reason,
but they are more likely to be perceived than the lower ones. (How often
have you said or heard someone say "Did I really say that?").

To return to the main point, phonemes and words are categorical, and the
relevant perceptions are whether a pattern belongs or does not belong to
the category, whereas phonetic "contrasts" are perceived differences, the
relevant perceptions being the degree to which this or that pattern exists
in the acoustic signal.

But I have the feeling (and Bruce can correct me if I'm wrong) that Bruce
thinks there is something more to "contrast" than "perceived difference".
I think he believes that contrasts "do" something to make speech perception
possible. I think Bruce believes that contrasts EXPLAIN perceptions (notably
of speech); I believe that contrasts ARE perceptions.

Bruce can speak for himself (and often does). I have never read him as
saying that contrasts "explain" perceptions. Contrasts are a way of
describing categorical differences, so that one can be sure to use
terminology that is not confused with "perceptual differences". Of course
they reflect differences in perceptions, but they do so in a way that is
more logical than numeric. Yes, they do "do" something to make speech
perception (and production) possible. What they "do" is to ensure that
if you gradually stiffen your lips in pronouncing, "mug" until you
find yourself saying "bug", the related meanings do not on the way
involve an insect that carries coffee, or a container with chitin-covered
wings.

I'm not going to say that a contrast IS a perception, so much as to say
that in order to perceive the category, one must control along a dimension
of difference on which the contrast is represented. Since the contrast
is logical, this "dimension of difference" is actually a lower-level
continuum perception or set of perceptual signals, and since (as I am
convinced) it requires hysteresis (catastrophe) to generate categorical
boundaries, the location of the category boundary on this dimension is
likely to shift according to context.

A contrast represents a move across a category boundary. A perceptual
difference, considered widely, incorporates a contrast, but in the context
of the discussion it does not.

Martin