CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Fairbanks, Grant. Experimental Phonetics – T21

An Acoustical Study of Vocal Pitch in
Seven- and Eight-year-old Girls *1

Grant Fairbanks
University of Illinois

Esther L. Herbert
Los Angeles Public Schools

J. Marian Hammond
Pasadena City College

The present study is a companion to that of Fairbanks, Wiley, and Lassman
(3), who studied the vocal pitch of seven- and eight-year-old boys,
and differs only in the sex of the subjects. Changes in the vocal pitch of
the female during early life undoubtedly take place, but the extent of the
changes, although presumably much smaller than in the male, has not
been investigated. As a matter of fact, unlike the situation in studies of
male voice, the literature provides no satisfactory information regarding
vocal pitch in average females at any age level. The present study, probably
for the first time, presents acoustical data on the vocal pitch of girls.

The pitch level of speech, because of its relationship to growth of the
larynx, is of major interest, and this study is the first in a program of acoustical
investigations designed to permit the plotting of a curve of pitch level
as a function of age in the female. Comparable curves for males and females
will differ greatly from adolescence on, but the differences or similarities
during pre-adolescent years are not to be predicted on the basis of
present information. The present study, taken in conjunction with its companion
(3), begins to throw some light in this direction.

Voice breaks in the male adolescent are phenomena of common experience,
and have been studied acoustically by Curry (1), and by Fairbanks,
Wiley, and Lassman
(3). In the former study a group of ten-year-old preadolescent
controls were found to have voice breaks similar to those of
mid-adolescent subjects. In the latter investigation boys aged seven and
eight presented similar voice breaks, a finding which suggests that the
phenomena are characteristic of male children generally, and are not
unique to adolescence. The present study afforded an opportunity to test
the additional hypothesis that voice breaks are not sex-linked. Finally, it
was a purpose of the study to compare the two experimental groups.


Since its experimental design was identical with that of the companion
study of boys (3), the procedure of the present study will not be described
in detail and may be summarized as follows.177

Two groups of 15 female subjects each were chosen from the public
schools, selection being at random except for sex and age. Age was controlled
within the limits shown in Table I, which also presents measurements
of height and weight, the means being very similar to those of Meredith
(5). The subjects, individually, read aloud a specially constructed, 52-word,
primary-level test passage, which was presented as the central portion

Table I
Age, height, and weight of experimental groups

tableau seven-year-old group (N = 15) | eight-year-old group (N = 15) | age (months) | mean | range | height (inches) | weight (pounds)

of a longer passage, and was easily read by all subjects. Phonograph recordings
were made, and subjected to phonophotography and frequency measurement.
Each voice break was examined with elaborate care by means of
wave-to-wave measurement.


Pitch Level

Pitch Level. The first portion of Table II presents the data on pitch
level. The means of 24.6 and 24.8 tones above the zero reference frequency
of 16.35 cps (4) are translated into the more usual expression of frequency
in the first line. The difference of 0.2 tone, it is seen, is not statistically significant.
Comparable values (3, 1) for boys aged seven, eight, 10, and 14,
respectively, are 25.0, 25.1, 24.4, and 23.4 tones above 16.35 cps. All six
means are Within a 1.7 tone range and approximate Middle C (24.0 tones
above 16.35 cps). Murray and Tiffin (6) report means for groups of women
with poor, good, and trained voices of 23.7, 23.3, and 23.8 tones above
16.35 cps, respectively, while Snidecor's (9) values for rigidly selected superior
female speakers were approximately two tones below Middle C.
178Adult males (8) and eighteen-year-old boys (1), on the other hand, have
been reported with means close to 18 tones above 16.35 cps, or around one
octave lower. In summary, the pitch levels of the two groups in the present
study do not differ from each other, are similar to those of boys of their
own age, to those of older boys up to age 14, and to those of adult females,
and are higher by approximately one octave than those of mature males.

Table II
Pitch level and numbers of voice breaks

tableau seven-year-old group (N = 15) | eight-year-old group (N = 15) | AM | SD | diff. | t | pitch level | cycles per second | tones above 16.35 cps | number of voice breaks | downward | upward | total

1 Ms — Mr
2 t, 1 per cent, 2.763; t, 5 per cent, 2.048

Voice Breaks

Voice Breaks. Probably the most important finding of the study, certainly
unexpected by the experimenters, was the occurrence of voice breaks.
The lower portion of Table II shows means of 3.1 and 3.4 voice breaks for
the two groups, each divided about equally between downward and upward
breaks. None of the differences is statistically significant. In the companion
study of boys reading the same passage means of 3.8 and 3.7 total
voice breaks were found, while boys at ages 10 and 14 have been reported
with means of 3.3 and 4.2, respectively, in a passage of roughly comparable

The voice breaks in the present study, although not obtrusive, were
clearly audible in the phonograph recordings, and the oscillographic records
were verified by impartial individuals expert in acoustical measurement
of speech. Since the phenomena were found in the studies cited, but
not in studies of eighteen-year-old boys (1), adult males (8), and adult
females (9) which employed identical phonographic instrumentation and
measurement procedures, the possibility that the voice breaks were artifacts
of apparatus or procedure appears unlikely.179

Fairbanks, Wiley, and Lassman (3), having found voice breaks in pre-adolescent
boys, concluded that they “… are not to be attributed exclusively
to adolescence.” The findings of the present study appear to warrant the
additional conclusion that voice breaks are not restricted to male children.
Taken as a whole, the data of the various studies cited indicate that voice
breaks are non-sex-linked phenomena of childhood which typically disappear
prior to establishment of adult vocal habits.

Table III
Extents, upper limits, and lower limits of voice breaks

tableau seven-year-old group | eight-year-old group | N | AM | extent (tones) | downward voice breaks | upward voice breaks | total voice breaks | upper limit (tones above 16.35 cps) | lower limit (tones above 16.35 cps)

For the means presented in Table III the voice breaks were pooled
within each age group of 15 subjects, N varying thus as shown for downward,
upward, and total voice breaks. The mean extents are seen to be in
the neighborhood of six tones, or one octave, agreeing generally with previous
findings with male subjects (1, 3). Although neither the acoustical
nor the neurophysiological bases for voice breaks can be satisfactorily explained
as yet, extents of this order may probably be regarded as characteristic
in view of this repeated finding.

Table III is also concerned with the location of the upper and lower
limits of voice breaks, i.e., the levels from which and to which they take
place. If the mean upper limits are compared to the mean pitch levels
(Table II, line 2), the similarities will be noted, while the lower limits
are distributed about the commonly reported pitch level for the adult male.
In these respects also the present subjects resemble male children. Curry
(1) offered the following explanation for this location in male adolescents:
180“During ages when breaks are occurring, it is conceived that the laryngeal
anatomy of the normal male is undergoing reorganization which will re*
suit in the establishment of a new vocal pitch level at the common level
of adult male speech…. The breaks… are down to and up from the region
of this adult pitch which probably is to be established in a few years for these
subjects, and which adjustment may be assumed to be in progress at age
fourteen. … it has been somewhat surprising to find a large number of
breaks occurring at ten years.” The immature female subjects of the present
study obviously may not be regarded as undergoing laryngeal “reorganization”
leading to the ultimate establishment of adult pitch levels in
the region of the lower limits of their voice breaks (See above, Pitch Level,
for pitch levels of women). This location defies explanation, although the
hypothesis quoted above appears to be untenable. It seems clear, however,
that the location of the lower limits of voice breaks may not be regarded
as forecasting the ultimate location of the adult pitch level, and that there
probably is no relationship between the two.

The finding of voice breaks in these various samples of pre-adolescent
children poses a most intriguing question: Why, since Aristotle, have
voice breaks been observed during and associated exclusively with male
adolescence, and overlooked in the speech of younger children of both
sexes? One explanation might be that they are more readily heard during
male adolescence because they take on other acoustical characteristics, e.g.,
concomitant intensity or wave composition changes. Although direct measurements
of such other variables have not been made, oscillograms and
recordings of voice breaks, however, indicate no obvious differential of
this kind. A more likely speculation would seem to be that the answer to
the question may be found in the relationship between the typical location
of the breaks (commonly between 24 and 18 tones above 16.35 cps, or vice
versa) and the mode pitch level of the individual in question. As long as
the voice breaks are in the lower portion of the pitch range, down from
and up to the mode, they tend to be overlooked as common phenomena of
childhood, although they may be heard if the attempt is made. As the adolescent
male pitch level lowers toward the adult male level, probably sometime
between age 14 and age 18, and probably quite rapidly in most cases,
the voice breaks, if they persist, are, after a time, found above rather than
below the mode pitch level. With such a relationship to the mode they are
then heard as anachronistic returns to the childhood level against a background
of predominantly low pitched, male, quasi-adult phonation, and
thus become more obvious to the listener. 12181

Pitch Variability

Pitch Variability. In Table IV are shown data on measures of certain
aspects of pitch variability. 2 3These aspects are probably more closely linked
to oral reading and speaking ability than to vocal development. In general,
the means differ from those of other samples of children and adults (1, 2,
3, 8, 9) by amounts not worthy of remark, except in the case of the 90 per

Table IV
Measures of pitch variability. All values in musical tones

tableau seven-year-old group (N = 15) | eight-year-old group (N = 15) | AM | SD | diff. | t | total pitch range | 90 per cent pitch range | extent of inflections | downward | upward | total | extent of pitch shifts

1 Ms — Mr
2 t, 1 per cent, 2.763; t, 5 per cent, 2.048
3 95th percentile — 5th percentile

cent range. Here the values resemble those for children aged seven, eight,
and ten, and are smaller than for older subjects. This restriction of 90 per
cent of the pitches used to a relatively small range probably describes in
part the impression of pitch monotony heard in the recordings. In this
regard, the companion study of boys (3) observes that “… this reduced
variability is not characteristic of the speaking voices of children, but may
well be typical of their oral reading during the early grades.” As will be
182seen in Table IV, the differences between the means are small and, for the
most part, not statistically significant In the last two items of the table,
extent of upward and total shifts, the differences are significant at the 5 per
cent level, but are perplexing because of their direction. 3 4In general, the
phonograph recordings revealed differences in oral reading ability in favor
of the older group. Data on the durational characteristics of the performances
will be reported later, and will show large differences in the expected


Two groups of female subjects, one consisting of 15 seven-year-old girls
and the other of 15 eight-year-old girls, were selected at random except
for age. Under laboratory conditions the subjects read aloud a 52-word primary-level
test passage; phonograph recordings were made and subjected
to phonophotography and frequency measurement, with the following

1. Pitch levels of both groups were located close to Middle C, thus being
similar to those of adult females, and to those of boys at ages seven, eight,
10 and 14. The values were approximately one octave higher than similar
values for mature males.

2. Voice breaks were found in both groups, comparable to those of male
adolescents and pre-adolescents in frequency of occurrence, extent, and location,
indicating that they are not exclusively sex-linked or adolescence-linked

3. In such aspects of pitch variability as pitch range, extent of inflections,
and extent of pitch shifts the two groups were similar to male

4. In all important respects the differences between the two groups were
not statistically significant.


1. Curry, E. T. The pitch characteristics of the adolescent male voice.
Speech Monog., 1940, 7, 48-62.

2. Fairbanks, G. Recent experimental investigations of vocal pitch in
speech. J. acoust. Soc. Amer., 1940, 11, 457-466.

3. Fairbanks, G., Wiley, J. H. and Lassman, F. M. An acoustical study
of vocal pitch in seven- and eight-year-old boys. Child Develpm.,
1949, 2O, 63-69.

4. Fletcher, H. Loudness, pitch, and timbre of musical tones. J.
acoust. Soc. Amer.
, 1934, 6, 59-69.183

5. Meredith, H. V. Stature and weight of private school children in two
successive decades. Amer. J. phys. Anthropol., 1941, 28, 1-40.

6. Murray, E. and Tiffin, J. An analysis of some basic aspects of effective
speech. Arch. Speech, 1934, 1, 61-83.

7. Pedrey, C. P. A study of voice change in boys between the ages of
eleven and sixteen. Speech Monog., 1945, 12, 30-36.

8. Pronovost, W. An experimental study of methods for determining
natural and habitual pitch. Speech Monog., 1942, 9, 111-123.

9. Snidecor, J. C. Studies in the pitch and duration characteristics of
superior speakers. Ph.D. Diss., State University of Iowa, 1940.

Manuscript received August 2, 1949184

1* Reprinted from Child Development, Vol. 20, 1949, pp. 71-78.

21 Pedrey (7), in a non-instrumental study of the oral reading of boys aged 11 to 16,
reported only four voice breaks as he listened to his 1014 subjects for a combined total
of 84 hours. While this figure, secured from unrecorded and, hence, unrepeatable speech,
observed subjectively by one listener, cannot be taken as a valid measure of the frequency
of occurrence even of obvious voice breaks, it does suggest that they vary in their
obtrusiveness, and arc, on the whole, relatively unobtrusive.

32 “The total pitch range is the difference between the highest and lowest fundamental
frequencies measured in a given sample and is expressed … in tones … an inflection
is defined as a frequency modulation, either upward or downward, without interruption
of phonation, while the term shift refers to a change in pitch which takes place
between the terminal pitch of a given phonation and the initial pitch of the subsequent
phonation.” (2) The 90 per cent pitch range is “The range between the 95th and 5th
percentiles of the frequency distribution of pitches used; it is expressed in tones.” (3)

43 Computation of the F statistic showed that the variances do not differ significantly.
F, 1.70 and 1.69, respectively; F. 10%, 2.48.