Henry J. (Henry Jackson) Watt.

Psychological studies online

. (page 1 of 21)
Online LibraryHenry J. (Henry Jackson) WattPsychological studies → online text (page 1 of 21)
Font size
QR-code for this ebook


ytfhuA**^,.



PSYCHOLOGICAL ANALYSIS AND THEORY
OF HEARING.



BY



HENRY J. WATT.



FROM THE BRITISH JOURNAL OF PSYCHOLOGY, VOL. VII. PART 1, MAY 1914.





1*1



CAMBRIDGE
AT THE UNIVERSITY PRESS.



[FROM THE JOURNAL OF PSYCHOLOGY, VOL. VII. P T . 1, MAY, 191h /p v
[All Hights reserved.] Library



PSYCHOLOGICAL ANALYSIS AND THEORY
OF HEARING.

BY HENRY J. WATT.

7. The background of senses on which hearing stands.

Three groups of senses and their problems.
II. Views concerning the quality of sounds.

III. The relations of tones, vowels, and noises.

IV. The distinction of two aspects of tone within pitch.

a. The hearing of single tones.

b. Orthosymphony.

c. Theory of the preceding.

d. Pitch the more precise basis of judgment.

e. v. Liebermann's deep symphonic pseudotone.

f. Binaural mixture ; theory of its occurrence.

g. Vocality.

h. Interval ; further tJieoretical indications ; 'direction'

in the tonal series.
V. The nature of the system of hearing : really correlated tvith

spatial differences, systemic, but not cognitively spatial.
VI. Physiological theory of hearing.

Formulation of a new theory ' on demand ' ; its relation to
the theories of Jlelmholtz, Ewald, and ter Kuile, and to the
main groups of psychological facts.

DURING the last few years rapid additions have been made to our
knowledge of auditory sensations, their relations, and their causes. And
the natural product of the new facts has come forth in various extensions
and modifications of previous theories. One of the most striking
features of these is the eagerness with which inspiration is borrowed

J. of Psych, vii 1



959991



2 /'sf/c/in/oyfcal. A-iHi/i/xi* n<l Theory of Hearhnj

from the visual system of qualities. But we must remember that
vision is only one of the senses and in taking its complexity as a model
\vc may fail to do justice to the claims of the other senses. I have
attempted to satisfy these claims by a general consideration of all the
senses and in doing so I was compelled to advocate a change in out-
treatment of sound 1 . It is commonly held that the attribute of quality
in the sense of sound is represented by what we commonly call pitch.
I proposed to see in this, however, not quality, but the auditory
analogue of what we call 'local sign' or localisation in the cutaneous
and visual senses. The attributes of sound were then to be : quality,
of which, as in the sense of touch, there is only one form ; intensity ;
pitch, generically named ' order ' ; voluminosity, the parallel of extensity
in other senses ; and the temporal attributes. On careful consideration
of all the new facts and theories I venture to say that the extensions of
auditory theory which build upon the common treatment of pitch as
quality and which confine their search for inspiration by analogy to the
sense of vision with its manifold qualities are entirely misleading and
that the best guide to a theory of sound, as of every sensory experience,
simple or complex, is a general consideration of the typical characteristics
or attributes of all the senses in parallel, as far as they go. In this
paper I propose to show that the analysis of auditory sensations I
suggested is not only preferable to any other present theory, but is the
only acceptable line of analysis.

I. The background of senses on which hearing stands.

'Midst the details of our knowledge concerning the complex senses it
must not be forgotten whence we obtain our pattern for their analysis.
We certainly do not find it in vision, which itself needs more help than
it can give. Our standard lies rather in the simplest of sensations. In
them we read both the problem and the solution of all the others.
These may be divided into two groups, making with the first group
three in all. A brief statement of these three groups shows how the
problems of hearing, which belongs to the third group, must be
approached.

The first group consists of the cutaneous and visceral senses.
There is no difficulty in identifying the attributes of all these. In spite
of the existence of different systems of receptors for each sense, only
one quality occurs in each sense. It may differ in intensity. But one

1 See This Journal, 1911, iv. 143 ff.



HENRY J. WATT 3

quality in varying intensities would only constitute one sensation, not
a sense. That is provided by the occurrence of differences of locality or
local sign, which is an ordinal attribute. Along with it goes a con-
tinuitive attribute, here known as extensity. By that is not meant
a variable feature of area or surface, but a constant constitutive of area
of any kind. Only by a variation of order along with constancy of
extensity can a variable extent be thought to occur. In the cutaneous
senses the minimal sensations we actually receive are very small extents
and present practically no room for further differentiation of localities,
at least as far as our phenomenal distinctions go. The temporal attri-
butes may be neglected in this paper 1 .

Sensations of taste may be added to this group. Their attributes
are identical with those of the cutaneous sensations in all respects but
one, viz. that in taste we find four different qualities. But this difference
presents no real difficulty. We have almost as much reason to see in
the four qualities of taste four different senses as in the four different
qualities of cutaneous sensation. What these reasons are need not be
recounted here. At the same time we may readily admit that the four
gustatory qualities are more akin to one another than are the other
four. Thus arises the suggestion and the problem that the sensory
qualities occur only in forms which differ from one another by a distinct
step. The solution of this problem must be the preliminary to any
problem as to the nature of the difference between qualities.

The second group of senses differs from the first clear and easy
group in being obscure and somewhat difficult in the matter of attri-
butes and in presenting a certain amount of complexity. The senses
included in it are the muscular, articular and organic. The obvious
form of variation of the first is intensity, while locality, extensity, and
quality are all obscure. Quality is easily absolved by the recognition
that, like those of the senses already mentioned, it does not really vary.
Locality or order varies from muscle to muscle, but not for any one
muscle ; and extensity is contained in the variation of the mass or bulk
of the sensations which come from muscles of different size. In the
articular sense we find a similar group of features, differently distri-
buted however. For, while quality is likewise single, it is intensity
which is obscure. The reason for this is probably merely want of

1 For a detailed account of the distinction and justification of these attributes, see my
previous papers in This Journal: (1) " The elements of experience and their integration,"
1911, iv. 135 ff. ; (2) "Are the intensity differences of sensation quantitative?" 1913, vi.
176 ff. ; (3) " The main principles of sensory integration," ibid. 240 ff.

12



JW//X/X mnl 7V/mr// of Henri mj

variation 1 . The only obvious variant in articular sensation is ' position,'
which we may take to be its ordinal attribute. Extensity behaves here
as in muscular sensation. It is implicit in the varying bulk, mass, or
volume of the sensations from the joints, small and large. In organic
sensations we find that obscurity and want of variation is fairly general,
but with a little willingness all the attributes can easily be identified.

Two psychological problems are raised by the sensations of this
group. The first is that of the presence and cause of the obscurity of
an attribute. In these cases it may be referred readily enough to the
absence of variation in the physiological correlate. Thus the organs of
the articular sense are stimulated at all times by forces of practically
constant intensity, while the organs of the muscular sense, if there are
enough of them in each muscle to form a small system there, are all
stimulated at once, not one at one moment and then another.

The second problem is that of compound sensations. In all the
senses of this group we have good reason to suppose that many receptors
are stimulated at once, as there are many in each muscle, round each
joint, and in each proprioceptive mechanism of the body. This fact
confirms the psychological conclusion that must be drawn from the
mass, bulk, or volume that is inseparable from the corresponding sensa-
tions. Apparently we never get in isolation a single ' spot ' of articular,
muscular, or organic sensation, but only a mass or area of it. Area is
familiar to us in cutaneous sensation and there is every reason to
suppose that area consists there of a large number of neighbouring
minimal (as we get them) sensations, which fuse to a continuous whole
in virtue of the extensity that each possesses. There is in area no
accentuation or discrimination of orders, unless by means of concomitant
differences in the variable attributes of intensity, as also sometimes of
quality (vision) 2 .

This theory of compound sensations may be extended to apply to
the minimal sensation from the ' spot ' of cutaneous sense, which may
well be supposed to be only minimal for us, because the units of the
receptor organs or their neural attachments make a smaller particula-
tion of sensation impossible, and not because the ' spot ' sensation does
not really consist of still smaller psychological particles fusing into
a tiny area in which no orders can be discriminated 3 .

1 Cf. This Journal, 1911, iv. 159.

2 Cf. my paper on " The psychology of visual motion," This Journal, 1912, v. 32 ff.

8 Cf. the psychological theory of intensity offered by F. Brentano, Untersuchungen zur
Sinnespxycholoyie, Leipzig, 1907, 53 ff.



HENRY J. WATT f>

The third group of senses raises all the difficulties and problems
already mentioned and some new ones. The senses included are vision,
hearing and smell. We find in the matter of attributes special cases of
obscurity, difficulty, and complexity. The obscurity of intensity in
vision raises a general problem regarding the primacy of intensity as
an attribute and regarding its place amongst experiences, if it be
secondary or derived 1 . The manifold and continuous variation of colour
quality, as surveyed in the colour figure, raises anew the problem of
compound sensations. Both of these appear again in the study of
auditory sensations, which adds as a third the problem of a non-spatial,
systemic order and continuity. In this paper I shall confine my atten-
tion to the problems of hearing.

II. Views concerning the quality of sounds.

The feature of auditory sensation which till very lately has been
generally classed as quality is pitch. As it varies continuously from
lowest to highest tones, the number of qualities equals the number of
distinguishable pitches. The smallest number of primary qualities to
which this series of continuous variations can be reduced is two, of
which each occupies one end of the phenomenal series. Such a reduc-
tion which treats the series of tones as the analogue of the series of
visual brightnesses, was suggested by Mach 2 . But it is rejected by
Stumpf and is not generally admitted. Each distinguishable tone is
rather considered to be a simple elementary sensation, requiring, as in
Helmholtz's view, a specific sense-organ. The number of these is there-
fore very great. Ewald has pointed out that the assumption of this
correlation between a vast series of qualities and a vast series of sense-
organs puts an enormous strain upon our conception of the biological
evolution of hearing 3 .

Brentano 4 thought to remedy the deficiency of Mach's theory, while
retaining its postulate of tonal primaries, by extending the analogy with
vision. He therefore recognised in hearing a series of ' saturated ele-
ments ' the tones that lie within an octave. The repetition of octaves

1 Cf. F. Brentano, op. cit., and the physiological theory of intensity proposed by
C. S. Myers, This Journal, 1913, vi. 137 ff.

2 Beitrtige zur Analyse der Empftndungen, Jena, 1886, 122. In this edition Mach uses
the analogy of the series leading from red to yellow. This is in principle the same as the
series leading from black to white.

3 J. B. Ewald, Arch. f. d. ges. Physiol. 1899, LXXVI. 155.

4 Op. cit. 101 ff.



6 Psycholoyi-cal Analyzw and Theory of Hear in <j

was to be explained by reference to variation of the admixture of the
brightness components of sound. But this theory, like Mach's, could
not hope to be accepted on its own merits. For, although it explains
various matters well enough, as any such theory may, it does not add
enough to our insight and to our knowledge to make itself compelling, and
it is not founded upon a mass of observations, as is that lately propounded
by Revesz and von Liebermann 1 . Their theory is practically identical with
that of Brentano, except that what he calls difference of brightness, they
call difference of Hohe". But the observations made by von Liebermann
on his own auditory sensations, as modified by a chronic paracusis, allow
these authors to go far beyond the range of Brentano's views and make
their theory much more convincing. For they have established the
independent variation and recognition of these two aspects of tone.
Their observations seem to me to be thoroughly consistent with all the
known phenomena of hearing and therefore prima facie correct. Only
their classification and theory of these aspects of tone appear to me to
be impeachable and misleading. Their distinction explains much, of
course ; but as its basis in fact seems to be correct, proper classification
and proper deductions therefrom should enable their theory to explain
still more.

It is to be noted that Max Meyer 3 , whom Revesz also quotes 4 ,
distinguishes the same two aspects of tones, but he calls them by
reverse names. Revesz's quality is his pitch and Revesz's height is
his quality.

In the hands of Kohler 5 and more especially of Jaensch 6 , Stumpf's
generally accepted views have taken a different line of development.
The chief influence here has been the observation of the resemblance
between pure tones and vowels and also in the case of Jaensch con-
sideration of the nature of the relations between the stimuli of tones,
vowels and noises. The pure vowel sounds are for Kohler the sole
qualities of hearing. For Jaensch they are only the qualities of the
sense of noise. He recurs to a modified form of the analogy between
vision and hearing in identifying vowels and brightnesses as less
differentiated sub-senses. The stimulus for the former is a rate of

1 Cf. the works quoted below under their names.

2 This is the ordinary German word for our ' pitch' ; for the sake of clearness, however,
I shall translate it by its wider meaning ' height.'

3 " On the attributes of sensations," Psyclwl. Rev. ix. 83, esp. 95 ff.

4 Zur Grundlegung der Tonpsychologie, Leipzig,. 1913, 42.

6 Ztsch. f. Psychol. 1910, LIV. 241 ff. and 1911, LVIII. 59 ff.
6 Ztsch. f. SinnesphysioL 1913, XLVII. 219 ff.



HENRY J". WATT 7

vibration which varies irregularly but not too extensively about a
certain average, while the stimulus for the latter, although not neces-
sarily irregular, is at least in the normal, light-adapted eye usually
provided by certain pairs of lights or any mixture of these pairs.
The more differentiated and presumably later developments tone and
positive colour are evoked by steady rates of vibration and are like-
wise to be identified. The relation between tones and vowels, like
those between colours and brightnesses, varies, sometimes being closer
than at other times. Thus, presumably, we should have to admit a
whole spectrum of qualities in the tones, and a series of qualities in the
vowels as it were octaves of brightness. Why there are different
qualities in this series is not explained. And as Jaensch distinguishes
vowels from noises, I can see no reason why we should not admit three
sub-senses in hearing instead of merely two 1 .

The analysis into sub-senses is hardly attempted at all by Revesz 2 .
The only indication of it we get is the suggestion that the height of
tones would need two psychophysical processes, while quality, as the
evidence of binaural mixture of small pitch differences seems to show 3 ,
would also probably require two. If we add to these the differences of
vocality which Revesz recognises 4 , still more processes will be required
and we shall quickly exhaust the resources suggested by even the most
complex of visual theories.

The enormous influence of the analogy with vision upon these
theories is obvious. I shall now proceed to summarize the facts and
observations included in them and to interpret them according to my
own theory of the attributes of sound stated above. It will then be
evident which of all theories is the more systematic, that is free of
difficulties and full of promise.

1 Cf. op. cit. 240, 255, " Weichen die Schwingungazahlen der einzelnen Eletnente eines
Kurvenzuges allzu stark von einander ab, so wird aus dem Vokal ein Gerausch....Das
undifferenzierte Geriiusch ist somit die eine Klippe, welche bei der Herstellung eines
Vokales vermieden werden muss; die andere Klippe ist der Ton." If there are only two
cliffs, there must be a valley between them, namely vowels. Or are there really three
cliffs?

2 Zur Gmndlegung der Tonpsychologie, 41 ff.

:{ Nachr. d. Gesell. Wiss. GiHtingen, Math.-phys. Kl. 1912, 676 ff.

* Zur Grundlegung der Tonpsychologie, 89, " Es gehort eben jeder Schwingungszahl
eine Qualitat, eine Hohe, nd eine Vokale zu."



8 Psychological Awifi/xi* <m<l T/H'wy of Hrarimj

III. The relations of tones, vowels, and noises.

There are pure tones and there are compound tones. The latter
consist of many simultaneous tones which harmonize more or less with
one another and with their fundamental component. If a tone does not
last longer than the time of two vibrations, it is heard as a noise. Many
tones of neighbouring pitch or generally inharmonic in relationship,
sounded together, form a noise. The pitch of single noises is not very
evident. But many noises contain distinguishable tones, and if noises
are sounded one after another, their pitch becomes easily noticeable.
Thus we obtain the propositions : (what is presumably) tone is some-
times heard as noise; some noises consist of (what are presumably)
tones; some noises contain tones. The net result of these is the
proposition : many (phenomenal) noises are, or consist of, (real) tones.
In reliance upon this, it has been usually inferred that all noises consist
of tones or are tones of very indefinite pitch or are not yet tones, so
that there is no special sense of noise 1 . This conclusion seemed to be
confirmed by the rather vague and conflicting statements made about
the tonal nature of vowel sounds. The synthesis of sounds by experi-
mental means did not admit of a direct examination of the proposition :
are there any unanalyzable noises of indefinite pitch ?

This experimental question has been answered by Jaensch by the use
of a selenium cell placed in the circuit of a telephone and illuminated
by an arc lamp whose light was varied by the revolution of an obstruct-
ing disc. The edge of this disc was cut out so that the variations in
the length of its radius corresponded with the variations in height of any
desired vibratory curve, pendular, periodic, or irregular. The results of
these experiments are most acceptable. A constant rate of vibration
produces a tone. The same average rate of vibration produces a vowel-
like tone, if the mean variation from the average is still small. As the
mean variation increases, the sound passes gradually into a vowel, then
it takes on a noisy character, and when the mean variation is great
enough, it may finally appear as pure noise. The average rates of
vibration of the vowels m, u, o, a, e, i, and s are approximately octaves
of one another 2 . These vowels do resemble certain tones, as Kohler
maintains 3 . But they are not to be identified with them, as he pro-
poses. The resemblance is close only between the lowest and highest

1 Cf. C. S. Myers, Textbook of Experimental Psychology, 2nd ed., Cambridge, 1911
25 f.

2 Jaensch, op. cit. 234 fif. :! Op. cit. esp. LVIII. 91 ft*.



HENRY J. WATT 9

vowels and the corresponding tones ; in the middle of the scale, from
below the vowel o to above the vowel e, tone and vowel are easily
distinguishable. Jaensch, therefore, ascribes vowels to a separate and
older sense of noise, of which he supposes them to be primary qualities.
Average rates of vibration that are greater than that of any vowel and
less than that of the next higher one, form a series of mixed vowel
sounds, which show a decreasing resemblance to the lower vowel and an
increasing resemblance to the next higher vowel, as the average pitch
rises. The changes from pure vowel to pure vowel thus obtained are
parallel with the changes encountered as we pass from red to yellow,
etc. 1

Two statements of great psychological importance are involved in
these views : (1) that hearing contains two psychologically independent
sub-senses tone and noise ; (2) that u, o, a, and the rest are pure
vowels, forming a series of qualities in the sense of G. E. Miiller 2 .

The following objections have to be urged against the distinction of
two sub-senses. It is supported by nothing more than analogy, and, at
its best, that analogy is the analogy of stimuli, not of experiences. The
stimuli of vowels and noises are irregular, those of tones are regular ;
the stimuli of colours are regular, those of neutral greys irregular. But
it is to be noted that, while the former vary round an average, the latter
go in pairs those of the complementary colours. It is known that the
sub-senses of vision exist independently ; but there is no evidence that
the sense of noise can exist without the sense of tone. And if there
were such evidence, it would not be clear of ambiguity ; for noises are
not only, ex hypothesi, excited by tones, as brightnesses are excited by
colours, but noises, when given, ex hypothesi, alone, resemble tones. It
is true, as Jaensch propounds 3 , that each positive colour has an affinity
to a neutral brightness, but it is not true, as his diagram suggests, that
each neutral brightness has a resemblance to a colour, qud colour;
whereas each tone resembles (or according to Kb'hler is) a vowel, and
on Jaensch's analogy must resemble it, because it excites it, and
also each vowel resembles (or according to Kb'hler is) a tone or has
the pitch of a tone, as Jaensch 4 has shown experimentally, and as all
those who have attempted to find the component tones of vowels have
observed. Moreover increase of intensity of light modifies a colour in

1 Kohler, op. cit. LVIII. 99; Jaensch, op. cit. 258 f.

2 I.e. "Eine Reihe von Empfindungen, in welcher sich die Qnalitat geradlaufig [d. h.
in konstanter Ricbtung vor sich gehend] und stetig iindert," Ztsch. /. Psychol. 1896, x. 33 ff.

s Op. cit. 264 ff. Op. cit. 288.



10 Ptychologiceil Aiui/yxix (tinl Theory of Hearing

the direction of greater neutral brightness, but increase of the intensity
of sound does not bring a pure tone nearer to either vowel or noise.
The sound-figure given by Jaensch 1 should be made tri-dimensional
to suit the double parallel between the similarity of low and high tones
to vowels and the varying similarity of different hues of different bright-
ness to neutral greys, just as the tri-dimensional colour figure does. But
this cannot be done for tone. The analogy of vision and hearing, tempt-
ing though it be, is both incomplete and misleading.

In any case, what sort of statements do the facts precisely warrant ?
The stimuli used show that, as we pass from tone through vowel to
noise, the average rate of vibration remains constant, but its mean
variation increases ; in other words, the pitch wobbles ; not markedly
and noticeably, but none the less truly. We become less able to indi-
cate the pitch by singing, or by naming it. But we can approximately
find it (with a little circumspection and comparison) the more easily,
the more it is isolated from accompanying (not from successive) tones,
vowels or noises 2 . I see no reason to depart from the substance of the
formulations regarding the relations of tones and noises referred to above.
A noise may then be said to be a simple sound whose pitch is not yet
audible or a complex sound of many pitches which make each other


1 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Online LibraryHenry J. (Henry Jackson) WattPsychological studies → online text (page 1 of 21)