phonological saliency

[Ray Weitzman]

In case you missed it, this is the text of the message I sent to Katie and
Info-Childes in reply to her message:

>'Saliency' is a term that is used in speech perception to refer to the
>relative control a particular acoustic parameter (duration, intensity,
>formant frequency, pitch, etc.) has in discriminating one phoneme from
>another, one syllable from another, one word from another, one phrase from
>another, or even one sentence from another.  Comparisions of different
>acoustic parameters as cues to in speech perception are referred to as the
>relative perceptual salience of the parameters or more technically as
>auditory cue weighting.  As far as I know it has nothing to do with why a
>child might show a particular preference for particular words or sounds,
>much "explain" those preferences.

By "phonological saliency" I take it you mean the relative control a
particular acoustic property has in distinguishing one phonological unit,
like the syllable or phoneme, from another phonological unit.  If so, then
I'm not sure what you mean by "[t]he saliency value of a particular
phonological feature".  Phonological features are generally highly abstract
and are manifested in various ways acoustically.  It is difficult to imagine
how one might measure the saliency of one phonological feature with another
phonological feature.  In the case of acoustic properties, experimental
conditions can be created for comparing the relative influence of two or
more them.  Any acoustic property may have some control of distinguishing,
for example, one syllable from another.  It will all depend upon the degree
of difference.  For some, a relative small change may lead to a change in
perception; for others, a relatively large change has to occur for
perception to change.  Perhaps this might be the basis for measuring degree
of saliency, but keep in mind that saliency of one acoustic property may
also depend on the values of other acoustic properties, as well as
individual and group (i.e., dialect.) differences of language-specific
listeners.  Thus, determining measures of saliency can be quite complex.

The studies that I am familiar with all seem to indicate that saliency of
acoustic features change over the time course of acquisition.  At least some
differences in acoustic features seem to become less salient as children
start to learn to discriminate utterances in their native language.  This,
of course, doesn't mean that children lose the ability to make these
discriminations, for with the appropriate learning protocol, they can
"re-learn" these distinctions again, if not for their native language,
perhaps for use in another language.  Furthermore, I wouldn't be surprised
if saliency also changes over the life-span of the individual, although I
haven't heard of any studies that examine this possibility.

Ray Weitzman


----- Original Message -----
From: "Hua Zhu" <Hua.Zhu at newcastle.ac.uk>
To: <info-childes at mail.talkbank.org>
Sent: Tuesday, May 25, 2004 3:55 AM
Subject: RE: phonological saliency



Alcock is right in saying that the notion of acoustic/phonological
saliency, though alluded to frequently in
the literature, lacks clear, testable definition.

In my work on the acquisition of Chinese phonology, I define
Phonological Saliency as a syllable-based, language-specific concept.
The saliency value of a particular phonological feature is determined
primarily by its role within the phonological systme of the language. It
could be affected by a combination of factors, e.g the status of a
component in the syllable structure, especially whether it is compulsory
or optional; the capacity of a component in differentiating lexical
meaning of a syllable; the number of permissible choices within a
component in the syllable structure.

For further information, see:
Zhu Hua & Barbara Dodd (2000).The phonological acquisition of Putonghua
(Modern Standard Chinese). Journal of Child
Language, 27 (1), 3-42.
Zhu Hua (2002). Phonological development in specific contexts:studies of
Chinese-speaking children. Clevedon: Multilingual Matters.

Currently I'm working on evaluating the above definition using
cross-linguistic data:
Zhu Hua & Barbara Dodd (Eds.). (Planned publication date: Dec. 2004).
Phonological development and disorders: A cross-linguistic
perspective.Clevedon: Multilingual Matters.

I will appreciate your views on the role/definition of phonological
saliency, especially with regard to acquisition.

Zhu Hua, PhD
Lecturer in Language & Communication
School of Education, Communication & Language Sciences University of
Newcastle upon Tyne NE1 7RU, UK

>
>I have a question about this concept which seems to be bandied
>about a lot in the language acquisition literature.  While
>there do seem to be some empirical studies, for example of the
>amplitude of particular phonemes compared to other phonemes,
>it also seems to be a concept that many assume in studies -
>for example, assuming that initial syllables or phonemes, or
>final ones, or stressed ones, will be more salient to children
>learning language.  I am using this concept in the field of
>literacy (spelling, in particular) and although I can find
>many papers in spoken language acquisition which draw on the
>concept of saliency to explain children's preferences for
>particular words/sounds, I can't seem to find any discussion
>of the concept per se, or measurements, either acoustic or
>behavioural, of some aspects of salience.
>
>Does anyone have any ideas - is this lost in the mists of
>time, or something that linguists take in with their mothers'
>milk and I missed out in my neuroscience education? Or am I
>confusing two different concepts?
>
>thanks
>
>Katie Alcock
>
>
>
>
>