summary of responses re: pitch discrimination

["Robert J. Zatorre" <MD37@xxxxxxxxxxxxxxxx>]

Dear List

Many thanks for your helpful responses to my query. I've excerpted the most
relevant responses below. The consensus seems to be that there is little
evidence to support the belief that very many normal listeners would have
trouble with an up/down pitch task, unless the frequency difference was
very small or some other parameter of the task made it very difficult. In
our study the maximum difference was 20% (1200 vs 1000Hz), so this seems to
be large enough that all normal subjects could handle it.

Several people pointed out that difficulties are often more apparent than
real, and involve misunderstandings about the words up/down or high/low.

In response to some people's request, I am attaching the abstract of the
paper in question at the end of this e-mail, for those who are interested
in seeing what we were up to in the first place.

Once again, thanks very much for your insights.


W. Jestead wrote:
>Bob Bilger and I made those paradigm comparisons for both intensity
>and frequency discrimination in the same subjects and found they worked the
>same way.  High-low judgments in a 2IFC task were better than Same-Different
>for both intensity and frequency discrimination.  The reference is W.
>Jesteadt and R. C. Bilger, J. Acoust. Soc. Am. 55, 1266-1276, (1974).

Roy Patterson wrote:
>When naive listeners are first asked to choose the higher, they sometimes
>(maybe 30%) do not immediately know which way the scale goes; that is, how
>to use the two words 'high' and 'low'. But a quick demonstration gets 99% of
>them on to it in less than 1 minute. So I do not know what the reviewers
>reference means.

Brian CJ Moore said:
>In my experience, nearly all normally hearing subjects can tell which is
>the higher of two sequential pure tones when the difference in frequency is
>reasonably large (say a semitone of more), but some have difficulty with
>smaller (but still detectable) differences in frequency.  This difficulty
>can usually be overcome by practice.  Two relevant papers of my own are:

>Moore, B. C. J. (1976). "Comparison of frequency DL's for pulsed tones and
>modulated tones," Brit. J. Audiol. 10, 17-20.
>Sek, A., and Moore, B. C. J. (1995). "Frequency discrimination as a
>function of frequency, measured in several ways," J. Acoust. Soc. Am. 97,
>2479-2486.

A Houtsma had this to say:
>The only documented case I know of is the paper by Tanner and Rivette
>(JASA 36, 1465-1467, 1964) who claimed that Punjabis could not discriminate
>high from low for pure tones. This was essentially debunked in a subsequent
>paper by Burns and Sampat (JASA 68, 1886-1888, 1980) who showed that this
>phenomenon was not an inability to perceive, but an ordinary linguistic
>problem that could be dealt with by proper instructions.

Chuck Watson wrote:
>I would bet that "10 to 30%" is way too high. Virtually everyone can
>discriminate pitch differences of 2-3% or less, at least in our experience
>with close to 1000 normal-hearing (by audiogram) college students.  However
>we normally use a trial structure in which the listener responds that either
>sequence <AAB> or <ABA> was presented, so we cannot say for sure whether
>they can do more than recognize differences.  I have heard various reports
>over the years that suggest that a few people don't appreciate what "high
>and low" mean with respect to frequency.  Can't recall a relevant reference
>right off...wasn't there some talk about cultures in which "pitch height"
>was not such a common concept?  It is an issue in "semantic psychophysics",
>right?


One person to cite a reference indicating that some large  proportion of
people might indeed have trouble with such a task was John Neuhoff, who
wrote as follows:
>In a control condition for another experiment, Mike McBeath and I tested
>38 listeners and found that 30% of them could not identify which of two
>tones were higher in pitch.  Tones had 10% frequency differences but
>also differed by 16 dB in intensity.

>See:
>Neuhoff, J. G., & McBeath, M. K. (1996). The Doppler Illusion: The
>influence of dynamic intensity change on perceived pitch.  Journal of
>Experimental Psychology: Human Perception and Performance.  22
>(4)970-985.


And Sandra Trehub also pointed out the following:
>Shepard (JASA, 1964 36, 2346-2353) reported that, of 50 unselected
>(seemingly normal) colleagues and assistants at Bell Telephone
>Laboratories, only 62% could consistently report whether the second of two
>sinusoidal tones was higher or lower than the first, even when the tones
>differed by a little more than a semitone (1/10 of an octave). Several
>listeners performed at chance levels, but those who were interested in
>music (whether or not they played a musical instrument) were near perfect.
>Perhaps this is the "well established" finding in question.

+ + + + + + + + + + + +
Here's the abstract from our paper, which, we hope, we will soon be able to
say is in press.


Functional specificity in right human auditory cortex for perceiving pitch
direction

Ingrid S. Johnsrude,  Virginia B. Penhune, and Robert J. Zatorreb
ABSTRACT
Previous lesion and functional imaging studies in humans suggest a greater
involvement of right rather than left auditory cortical areas in certain
aspects of pitch processing. In the present study, adaptive psychophysical
procedures were used to determine auditory perceptual thresholds in 14
neurologically normal subjects, and in 31 patients who had undergone
surgical resection from either the right or left temporal lobe for the
relief of intractable epilepsy. In a subset of the patients, the lesion
encroached significantly upon Heschl’s gyrus or its underlying white matter
as determined from MRI analysis. Subjects were asked to perform two
different perceptual tasks on the same set of stimuli. In a pitch
discrimination task, the subject had to decide whether two elements of a
pure-tone pair were the same or different. In a task requiring the
judgement of direction  of pitch change, subjects decided whether pitch
rose or fell from the first tone to the second. Thresholds were determined
by measuring the minimum pitch difference required for correct task
performance. Mean thresholds in the pitch discrimination task did not
differ between patient groups and control subjects. In contrast, patients
with temporal-lobe excisions that encroached upon Heschl’s gyrus in the
right hemisphere (but not in the left) showed significantly elevated
thresholds when judging the direction of pitch change. These findings
support a specialisation of function linked to right auditory cortical
areas for the processing of pitch direction, and specifically suggest a
dissociation between simple sensory discrimination and higher-order
perception.


Robert J. Zatorre, Ph.D.
Montreal Neurological Institute
3801 University St.
Montreal, QC
H3A2B4
Canada

phone: 514-398-8903
fax 514-398-1338