[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Phoneme versus word recognition.



>A graduate student at McGill has done some research with narrow band noises
>of ambiguous pitch, showing that people are good at identifying melodies
>made of these types of notes, even though they are poor at matching the
>pitch of the individual notes.  I thought that there might be some relation
>to the "word superiority effect" in reading (we have many references) and in
>speech perception.  On the latter topic we have been able to find only one
>article, Norris & Cutler (1988).

If I understand correctly, when you refer to the "word superiority
effect", you are interested in the issue of topdown
influences from the lexical level on phonemic perception.
If this is the case, here are a few references:

- Samuel AG. Phonemic restoration: insights from a new methodology.
J Exp Psychol Gen. 1981 Dec;110(4):474-94.

- Pitt MA, Samuel AG. Lexical and sublexical feedback in auditory word recognition.
Cognit Psychol. 1995 Oct;29(2):149-88.

- Samuel AG. Lexical activation produces potent phonemic percepts.
Cognit Psychol. 1997 Mar;32(2):97-127.

- Norris D, McQueen JM, Cutler A. Merging information in speech recognition:
feedback is never necessary. Behav Brain Sci. 2000 Jun;23(3):299-325; discussion 325-70.

- Pitt, M.A. & McQueen, J.M. (1998). Is compensation for coarticulation
mediated by the lexicon? Journal of Memory and Language, 39, 347--370.


You'll see that the debate is not over.
IMHO lexical superiority effects on phoneme
identification are postperceptual.

But I am not sure that the situation with melodies is similar (are the
melodies known in advance, or unknown to the listener?)

>Does anyone know of other examples of the
>superiority of recognizing a larger unit, even though there is a lot of
>uncertainty about the component units?

The 'duplex perception' phenomenon (Liberman & Whalen) comes to mind:
the third formant transition can be unaudible by itself, yet sufficent to
cue the contrast between two consonant.

The following references may also be relevant:

- Day & Wood (1972) Mutual interference between two linguistic
dimensions of the same stimuli. JASA, 52.

- Tomiak et al. (1987) Integral perception of phonemes: evidence
for a phonetic mode of perception JASA, 81, 755-781.

They are not exactly what you are looking for, but
they claim to show that phonemes within a syllable are perceived in an
"integral" fashion, that is not as independent units. [The second chapter of my dissertation dealt
with describe this research and follows it up. (Role de la syllabe dans la perception de la parole. in French,
available at http://www.pallier.org/papers/thesis/thesis.html).]

Concerning phonemes, there are several reasons to doubt
that there are independent perceptual units;
Among others, we have found that people can perceive phonemes
that have absolutely no acoustic correlates in the signal,
(see, e.g., Dupoux, E., Pallier, C., Kakehi, K., & Mehler, J. (2001).
New evidence for prelexical phonological processing in word recognition.
Language and Cognitive Processes, 16)

Hope this helps,


Christophe Pallier
www.pallier.org
www.unicog.org