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

Re: sometimes behave so strangely

Dear Diana, Al and list,

You also raise the point that the transformation of the phrase from speech to song endures - so that when people listen again to the full sentence, I appear to burst into song. This , to my mind, is a particularly puzzling aspect of the effect. People have told me that the effect is still present, unattenuated, even months later - and this was certainly my experience. As you point out, perhaps the most important question raised by this demonstration is why people don't always hear speech as song. After all, the vowel components of words are harmonic complexes - yet the pitch characteristics of speech are usually suppressed in perception.

true, unless you speak a tonal language, in which case you will focus on pitch patterns in order to figure out the meaning of words. In fact, I wonder whether speakers of tonal languages might be more resistant to transforming pitch into singing since for them focusing on pitch patterns has become part of the speech schema.


Professor Diana Deutsch Department of Psychology University of California, San Diego 9500 Gilman Dr. #0109 La Jolla, CA 92093-0109, USA

858-453-1558 (tel)
858-453-4763 (fax)


On Dec 12, 2006, at 10:08 PM, Al Bregman wrote:

Dear Diana (and List),

Yes indeed! The repetitions do seem to comvert spoken speech to singing.

I wonder whether your phenomenon isn't related to that of "semantic
satiation", in which a word that is repeated over and over tends to
lose its meaning and to be perceived as a sequence of sounds.  The
meaning is not lost in an all-or-nothing fashion, but tends to get
weaker and weaker.  (See

In the case of your demonstration there is a strongly modulated F0
(intonation contour) over the phrase.  It may be that when we hear
ordinary speech, which contains pitch modulation (an acoustic
phenomenon that is present in both music and speech), the speech and
musical schemas are both evoked and compete with one another.
However, when the cues for speech are dominant (i.e., continuous and
non-repeating modulation of F0, without pausing on particular
pitches), the musical interpretation is suppressed.  But when the
phrase is repeated many times, a satiation and weakening of the speech
interpretation occurs (as in semantic satiation), thereby allowing the
musical interpretation to become more dominant.  Of course it doesn't
become completely dominant, or else we wouldn't hear speech at all.
Rather there is an intermediate form of activation in which we hear
both speech and music (i.e., singing).

In your demonstration, immediately after hearing the phase as melodic,
when we listen to the whole sentence again, we still maintain an
association between the phrase and the melodic interpretation.  I
wonder how long this aftereffect lasts.

Your demonstration raises the fascinating question of why we don't
ALWAYS hear speech as singing.  It may be that persons with absolute
pitch come closer to this than the rest of us do, or at least can turn
it on when they want to.  A number of years ago, I asked Poppy Crum, a
graduate student of mine who had absolute pitch, whether she could
assign musical note values to my intonation pattern as I said a
phrase.  She replied that this was easy, and gave me a sequence of
note names.

Whatever the explanation of your phenomenon, it is truly interesting,
and raises some challenging questions.  I hope you yourself, or some
of our colleagues, will be able to shed light on the phenomenon
through a series of analytical experiments.

All the best,


Albert S. Bregman, Emeritus Professor
Psychology Department, McGill University
1205 Docteur Penfield Avenue
Montreal, QC, Canada H3A 1B1.
    Tel: (514) 484-2592, (514) 398-6103
    Fax: (514) 484-2592

On 12/12/06, Diana Deutsch <ddeutsch@xxxxxxxx> wrote:
Dear list,

I've had a number of requests for the sound demonstration I presented at
the recent ASA meeting in Honolulu, in which the spoken phrase 'sometimes
behave so strangely' following several repetitions, appears convincingly
to be sung rather than spoken - though there is no musical context and no
physical transformation of the sound. This demonstration, together with
spoken commentary, is on my CD 'Phantom Words and Other Curiosities'
(available from Philomel Records - http://www.philomel.com).

The sound demonstration is also posted  on the website:


and it's described in the booklet accompanying the CD, which is posted at:


Happy Holidays!

Diana Deutsch



Valter Ciocca, PhD
Associate Professor
Division of Speech & Hearing Sciences
University of Hong Kong
Hong Kong SAR, China

phone: 852 28590581
fax:   852 25590060