Let O answer "up" or

6. Introspections. — Kepeat the largest step in Exp. 4 deliberately a number of times and allow O to study and describe the subjective différences of the sounds by which he judges their direction. Kecord in fuU his observations on différences in intensity, quality, distance, tactual sensations, motor tendencies, visual

FiG. 10.

imagery, — in short, ail features which seem to resuit from change in the direction of the sound.

The last five experiments may be discussed together most profitably. Fig. 10 is the record of the localizations of a trained observer within the right half of the horizontal plane at the level of the ears. It is based upon 18,000 measurements under the most favorable conditions and therefore has a high degree of validity. The radii show the directions and the arcs represent the

AUDITORY SPACE 61

distance between the standard and the compared sound which will yield 75 per cent of right judgments.* As the sounds were produced 1 meter f rom the center of the head, each degree corresponds to 17.5 millimeters. Thus the curve shows a limit of .9** at 0° F (our Exp. 2), 1° at 0° B (our Exp. 3), and 4.5° at the right (our Exp. 4).

In gênerai, then, the law which should stand as the prédiction of our results is that the localization in this horizontal plane is poorest at the side, improves by three large steps in front and three behind in passing toward the médian plane ; it is about equally fine for front and back, and it is more than four times as délicate for thèse points as for the side.

Curves of this kind hâve been worked out for représentative planes in the field of space around the head. The results of some of thèse are shown graphically in Fig. 11, which represents the right hémisphère. Observe that the first vertical at the left stands for the front half of the médian plane, and the first vertical at the right represents the back half of the same plane, and that the other vertical lines represent intervening meridians at intervais of 15° on the surface of a hémisphère centered at the observeras head and lying toward

* In laboratory experiments it îs customary to use the différence which la calculated to yield 75 per cent of correct judgments as the measure of the discrimination. In the présent experiments we hâve adopted the much higher standard of 100 per cent correct judgments, partly to shorten and simplify the method, and partly to avoid computations. Of course, the records for our experiments will therefore be correspondingly larger. They are also larger on account of the observer's lack of training, but they should express the above gênerai tendency equally well.

AUDITORY SPACE

his right.* The middle horizontal line represents the one in which our Exps. 4 and 5 are located ; above and

liSSïï'

(HHHHHHHH

MMM)-4-^MHMMM^

FlG. 11.

below this the parallel horizontal lines represent meridians at the respective latitudes marked.

The little figures may be ealled sensory ellipses in auditory space discrimination. They are constructed

*In reality thèse converge to one point at the top and one point at the bottom; the top and bottom lines become single points within the médian plane. They are projected in the présent manner in order to show the relative dimensions of the ellipses.

AUDITORY SPACE 63

in the foUowing manner: At any intersection, vertical and horizontal distances are laid off on the basis of the corresponding measurements. Thus, if the measure of the discriminative ability is 5°, the points are laid off 2 1-2° below and 2 1-2° above the intersection. Two points are laid off on the horizontal line on the same principle. An ellipse is then constructed with thèse points as the termiui of the two diameters.

The ellipses mean that every diameter represents, on the scale of the chart, the degree of séparation of two sounds which would yield 75 per cent correct judgments of direction under the given conditions of the experiments. *

Thus, following the middle horizontal line, 0® F, 0° B, and 90° K represent the points in Fig. 10 which correspond to our measurements in Exps. 2, 3, and 4 respectively. Exp. 5 is represented on the vertical axis at 90° K. This ellipse shows that at this point our discrimination is keener in the horizontal direction than in the vertical.

The psychologist can lay out a plot of the acoustic field, represent measurements of significant distances and directions upon it, trace their laws and distributions and tell why the features represented hâve taken thèse particular forms. And furthermore he can give his chart of the acoustic field to the world as a guide in the attempt to hear direction, and as a criterion of the validity of purported hearing of direction.

The introspections in Exp. 6 bring out a number of

♦ Where there are only two dots and no ellipse, the measurements hâve not been made for the other diameter.

64 AUDITORY SPACE

înteresting subjective factors. Some of them may be predicted.

The soimd seems to be strongest at the aurai axis. This in turn makes the sounds which are at the aurai axis seem nearer than those which are away from it; thèse sounds are also richer and clearer.

There is also a tendency to hear the forward sound as higher and the backward sound as lower in pitch than the standard sound. The most interesting and complicated observations probably refer to misplacements : the observer knows that the sound cornes from a given point, but he hears it in some ojher direction. Thèse misplacements, in a careful observer, foUow definite laws, which may be worked out, as many of them already hâve been.

Most observers hâve distinct visual imagery for the movements of the sounding body. Frequently one is conscious only of this mental vision, and the sound changes as interpreted in terms of sight. The observer says that the sound is forward because he has a mental picture of the sounding body in that position. Nearly ail observers make automatic movements in adaptation to the changes in direction of the sound, and may be conscious of thèse movements and be guided by them without being aware of their auditory basis. Mental images of movement or tendencies to movement are also prominent. One may detect images of strain and movement which vary with the direction of the sound.

Among the numerous tendencies which affect localization are the foUowing:

There is a tendency to locate every médian sound in

AUDITORY SPACE 66

the upper front quarter, by the force of habit. Likewise sounds at the level of the ears are usuàlly misplaced upward, and sounds in the vertical plane through the aurai axis are usually misplaced forward.

The effect of expectation is exceedingly complicated. Thus, we tend to hear a sound from the direction it is expected ; but if the direction is sensibly différent, tbere is a tendency to overestimate this différence. Try it.

There is a tendency to turn the eyes in the direction of attention, and this turning of the eyes influences the perception of direction.

Thi hearing of direction is a neglected capacity. We ordinarily dépend upon sight ; but should the exigencies of life demand it, we could improve very much in the hearing of direction.

Two ears are necessary for the normal hearing of direction, but certain localizations are possible with one ear alone. Fig. 12 is the curve of localisation, in the same plane as is represented by Fig. 10, for a person who hears only with the right ear. It shows that hearing of direction is keenest opposite the ear which is intact and gradually decreases both in front and behind to a point of poorest hearing in a symmetrical point on the other side. Thus, he detects a différence of 4° opposite the good ear, but there has to be a différence of 21.6° opposite the deaf ear. Note the bearing of this upon a theory. The dotted curve is a copy of Fig. 10.

It is probable that pure tones cannot be localized at ail by one ear alone.

Différence in the keenness of the two ears leads to misplacements whenever the décline or restoration of

AUDITORY SPACE

hearing in one ear is sudden. When the loss or restoration is graduai, we adapt ourselves to it.

There are numerous Systems of confusion-points; that is, two points in différent quarters which hâve the same ^^ocal sign" so that, although far apart, the sounds from thèse two sources cannot be distinguished. Thus 45° right-back and 45° right-front are often confused.

Every unnatural position of the body, or part of the body, tends to produce errors in localization. For ex-

FiG. 12.

ample, if the head is turned to the right, a sound will be slightly misplaced in the direction in which the head is turned.

It is usually supposed that a sound cannot be heard as well from behind as from the front. But experiments show that we can hear just as faint a sound from the back as from the front. Still the common supposition leads to many misplacements.

Our hearing is keenest directly opposite each ear and it decreases gradually toward the médian plane both iu front and back. Intensity of sound may therefore be

AUDITORY SPACE 6Ï

interpreted in terms of direction; it is possible to predict an apparent change in direction f rom the mère varying of the intensity of the sound.

Other things being equal, the fainter sound is judged to be more distant.

Timbre is the characteristic quality of a sound and dépends upon the présence and reinforcement of overtones. The différence between the tones of the varions musical instruments is chiefly a différence in timbre. As a rule, the richer the tone the better it can be localized. The human voice in normal speech can be localized better than any other sound. The localization of the sound of a tuning-fork is very much inf erior to that of the voice ; and some contend that a simple pure tone cannot be localized at ail. At any rate, the richer the tone the more ^^ear-marks" of direction it can hâve.

The longer a sound lasts within certain limits, the better it is localized.

By introducing a f unnel into the ear and turning it in différent directions, one can make the same sound seem to come from différent places.

The résonance and similar modifications of the sound by the environment are familiar sources of disturbance to localization. If the experiments are made indoors, they should preferably be made near the center of a relatively empty cubical room.

There are exceedingly interesting phenomena of sound fusions. No matter how many sounds are produced simultaneously, no matter from what différent directions they may come, provided the sounds are of

68 AUDITOKY SPACE

the same kind and pitch, they will be heard as one. A very pretty experiment is to connect two téléphone receivers to the same tuning-fork so that the same tone is produced in both receivers. The two tones will invariably be heard as one, and this one is located according to definite laws. If the two receivers are pressed against the ears, the fused sound will be heard at the root of the tongue — a very weird expérience. The laws for the localization of fused sounds are among the most valuable criteria for testing a theory of sound localization.*

To revert, then, to the questions at the beginning of the chapter, we may sum up the évidence in the f orm of a reply. The foUowing points may be considered in f avor of the intensity theory :

(a) The gênerai variation in the delicacy of localization ; that is, the f orm and distribution of the sensory ellipses.

(b) The existence of confusion-points.

(d) The introspections which reveal our dependenoe upon intensity.

(e)The inferiority of monaural Localization.

From experiments of this kind we may therefore conclude that intensity is the primary f actor in the local sign of hearing.

But intensity opérâtes in two distinct ways: (1) through the binaural ratio and (2) through monaural

* This may be made a very effective démonstration exercise to be developed by two or three members of the class.

AUDITOKY SPACE 69

ratios. With two ears we get simultaneous impressions in which the intensity for one ear stands in a certain ratio to the intensity for the other ear. But, for each ear or with a single ear, a sound in one direction has a certain intensity ratio to the same sound in another direction.*

We hâve also found distinct évidence of another factor, namely tone quality, or the modification of the character of the sound by the form of the outer ear. Among the évidences for this are :

(a) Complex sounds can be localized more accurately than relatively pure tones.

(6) The possibility of improvement with practice, especially within the médian plane.

(d) The curve of monaural localization.

(e) The introspective évidences; i.e., consciousness of quality différence.

Hence we shall conclude in favor of the fifth theory, the intensity-quality theory : sound is localized by means

♦ An întereatîng corroboratîon of this twofold nature of the local sign of direction of sound is found in the "teeth" of the curve in Fig. 10. There are five prominent teeth, namely, 15° RF, 50° RF, 90° R, 60° RB, and 25° RB. The more accurate the measurement the more distinct thèse teeth become. Now, the explanation is to be found in the fact that there is a change at thèse points in the means of localization. At 15° RF and 25° RB there is a transition from the simple balancing of the relative intensity of the two ears to the récognition of quality changes for one ear. At 50° RF and 60° RB there is a transition from this double standard to the monaural localization on the basis of the character of the sound. The large peak at 90° R is due to the fact that this représenta a sort of zéro-point for both modes of localization.

10 AUDITORY SPACE

of binaural and monaural ratios of intensity and characteristic différences in quality. *

Is the ear a space-sense organ ? It is. As we hâve in two eyes a means for the perception of distance, we hâve two ears for the perception of direction. We also hâve on each side, where the two ears cannot well cooperate, ear-funnels which change the quality and the intensity of the sound with référence to the direction in which the vibrations impinge upon one ear.

Movement, tendencies to movement, images of movement, visual or auditory images, tactual sensations or images, and other factors may hâve been observed as secondary factors which influence localization.

What are some of the laws of localization? The graphie représentation in Fig. 11, for example, embodies a whole system of laws.f

*Quality is hère used in the popular sensé, as synonymous with timbre. Strictly, the quality of the tone is its pitch.

fA good survey of this topic is found in Pierce, "Studies in Space-pereeption." The présent chapter is based mainly upon Starch, ''The Localization of Sound/' The University of lowa Studies in Psychology, Nos. 4 and 5.

CHAPTEE VI

TACTUAL SPACE

For Two.

1. Tactual Localization' of a Point. — Mark off an area 50 millimeters wide and 100 millimeters long from the wrist upward, on the volar (inner) surface of the forearm by ink-dots at the corners of the parallelogram. Make a similar plot in the note-book. Working within this area, touch O, who sits with eyes closed, lightly with the point of a pencil and require him to try to locate this spot, eyes opened, by touching with another pencil.* Mark the relative positions of the two spots by dots in the note-book plot, and connect them by a light Une to indicate the magnitude of the error. Label the dots S (stimulus) and L (location) respectively. Make 15 trials. Record O's analysis of what constitutes the "local sign'' — the terms in which he recognizes the location of a touch.f

There are three factors in the record: the measurement of accuracy of localization in terms of the magnitude of the error; the direction of the error; and the analysis of the local sign.

Accuracy of localization varies for différent parts of the skin ; in gênerai, the portions of the skin which are

•Keep the résulta secret from 0.

f For définition of local sign see Ch. IV^ p. 45.

72 TACTUAL SPACE

used most in tactual perception are most sensitive, Hence we find but a crude localization for the large areas for which tactual localization is relatively unimportant, and a fine ability for those portions which are small, movable, and readily accessible. This should be illustrated in the chart just made by large errors in that portion which is toward the middle of the forearm, and small errors at or near the wrist.

The probable direction of an error is predictable according to a gênerai law ; the point touched is generally misplaced toward the nearest joint or actively used portion of the skin in the localization. .This should show in the chart by a tendency of the L-dots to be nearer the wrist than the S-dots. The largest errors are along the axis of the arm, unless the point touched is on or very close to the wrist. The errors in the radial and ulnar directions are small.

It is difficult to discover by introspection just what constitutes the local sign. The localization of the tactual impression, like the projection of the retinal image, is automatic. The mode of the accompanying consciousness varies for différent parts of the skin and with différent individuals. The record might mention visual imagery, association with blood-vessels, ridges, wrinkles, curvatures, relative callousness of the skin, tendencies to movement, etc. The foUowing account of the local sign of the skin is apposite.

"Not only îs the skin, physiologically regarded, a localizîng organ: the organism îs endowed with reflex localizîng movements. If a spot of the skin is iriitated, hand or foot moves to it re* flexly, in obédience to purely physiological laws. Out of thia nnconscious localization the conscious local mark arises, by the following stages: (a) The movement of the hand or foot, though

TACTUAL SPACE 73

reflexly set up, occasions organic sensations in the joint, tendon, etc.; so that definite groups of organic sensations become connected with pressures upon particular parts of the body. The local sîgn may consist, therefore, of remembered organic sensations. (6) The reflex movement toward the irritated spot will usually be seen; so that the local sign may contain a visual sensation, a picture of the part touched, as well as organic sensations, (c) The organic sensations may pass unnoticed, owing to the habituai nature of the movement. The local sign of a pressure will then be a sensation of a quite différent order — a sensation of sight. {d) Finally the visual picture itself may disappear, and its place be taken by a word, the name of the part of the body pressed. Often enough, when we say that we remember an occurrence, we remember only the form of the word which describes it. So now, when I am touched upon the arm, there flashes up in my mind the word *arm,' and this word is the local sign of the pressure." — Titchener, "Outlines of Psychology," p. 167.

Tactual space is inseparably bound up with visual space. In this experiment it is extremely diflBcult to make a purely tactual localization ; there is almost invariably at least a visual image of the arm and particularly the portion stimulated. To test the effeet of the visual image upon the accuracy of localization, experiinents hâve been made upon trained observers who had the power of practically inhibiting the visual images, of neglecting them, or of utilizing them in locating the pressure sensations. The results show much greater accuracy with the employment of visual imagery than without it.

Blind persons, however, who cannot rely upon the visual imagery, exhibit no larger errors in localization than seeing persons. This is of course due to the great relîance which they must place upon pressure sensations, and to the practice effects in delicacy of discrimination which foUow. This would tend to show that, although vision constitutes for normal persons an im-

74 TACTUAL SPACE

portant élément in the local sign of pressure, ît îs not an indispensable élément.

The right arm and hand of the observer hâve been used to earry the pencil in locating the pressure sensations. This has added a whole group of muscle sensations further to complicate matters. It is well known that distances are not properly judged by movements of the arm and hand when flexed, as required in touching a part of the other arm. Much of the constant misplacement distally has been found to be due to this f actor. This suggests again the interconnection between tactual and muscular space.*

2. The Two-point Space Threshold. — ^To measure the threshold of space discrimination for two points upon the skin, détermine the minimum distance that two spots lying in the longitudinal axis of the volar surface of the forearm, stimulated simultaneously, may be separated and still be perceived as two.

From a calling-card eut a«strip 15 millimeters wide and as long as the card ; eut this diagonally. Take the two triangular "points/' and improvise a pair of compassés by holding thèse between the thumb and first finger. Set the points at the desired distance by measuring on the millimeter scale. Adopt the foUowing scale of steps in séparation: 1, 2, 3, 4, 7, 11, 16, 22,

* Other methods of localizing hâve been designed to elimînate some of thèse complicating factors: (a) hold the pencil just over the point stimulated, but do not touch the skin; (h) mark the localization upon a photograph or plaster-of-Paris model of the arm; (c) describe in words as nearly as possible the location. Thèse methods may be oombined and other methods derlved from them.

TACTUAL SPACE 75

29, 37, and 46 millimeters. Press the two points simultaneously, gently and evenly, for about 1 second and require O to report whether he feels one or two.* Start with the smallest step, and if O reports "One" try the next larger step, and continue thus until O reports "Two/^ With this step give successive trials, interspersing control triais,! until O makes a mistake or has perceived the two points correctly ten times in succession. If a mistake is made, take next higher steps in order until ten successive correct judgments are given on one step*. Record the number of correct judgments for each step tried; the last may be considered the threshold required and a measure of O's power of space discrimination under thèse conditions.

3. Relation to Direction on Limbs. — Measure, in a similar manner, the two-point space discrimination with the compass-points applied in a transverse direction on the same région as in Exp. 2. Compare results with those of Exp. 2.

4. Relation to Part of Body. — ^Measure the two-point •space discrimination, as in Exp. 2, (1) on the tip of the first finger of the left hand and (2) in a vertical Une on the back of the neck. Record and compare results.

* Avoid ail unnecessary fatigue of the skin area selected for exploitation. Take great care to apply and to withdraw the two points simultaneously. Do not press hard; a dearly defined sensation of contact is ail that is necessary.

f Control triafs may be made by touching with one point only in about one-third of the trials; if 0 answers "Two" in a control trial, this is an error indicating that he has not set himself a sufficiently high standard of oertainty and he must be tried on a next higher step.

76 TACTUAL SPACE

The two-point space threshold, which is measured in Exps. 2, 3, and 4, ia an index to spatial sensitiveness.

Exp. 2, the results of which may seem startling to the uninitiated, demonstrates the fact that the two points may be separated some distance and yet be perceived as one. This is to be explained upon the assumption that the local signs for the two points stimulated are indistinguishable and therefore there is no basis for a differentiation of the impressions. The two-point threshold for this direction upon the arm is large.

In Exp. 3, with a transverse direction npôn the arm, the threshold is smaller. This is in Une with the fact, found in Exp. 1, that the radial and ulnar misplacements in localization are small. The explanation for this is based in part upon the anatomical fact that the pressure-spots lie doser together transversely than in a longitudinal direction.

In Exp. 4 the spatial sensitiveness is found to vary

with the différent parts of the body ; upon the fingers it

is very délicate, upon the neck very obtuse. The fol-

lowing table * gives a list of représentative threshold

values for various parts of the body in longitudinal

direction :

Tongue-tip 1.1 m.m.

Palm side of last phalanx of finger 2.2 "

Red part of lips 4.4 "

Tip of nose 6.6 "

Back of second phalanx of finger 11.0 "

Heel 22.0 "

Back of hand 30.8 "

Forearm 39.6 "

Sternum 44.0 "

Back of neck 52.8 "

Middle of back 66.0 "

* James, Psychology, Briefer Course, p. 62.

TACTUAL SPACE 11

The spaoe discrimination in any given portion of the skin may be représentée! by so-called sensory cireles, more properly called sensory ellipses, on the same plan as we represented auditory space discrimination in Chapter V, Fig. 11. Thus, the data in this table would furnish the major axes for ellipses.

The différence in distribution of the pressure endorgans and nerves, the unequal amount of practice occasioned by exposure and use of certain parts of the body, and habituai protection of others by the garments, are probably suffîcient to account for the wide range of the two-point threshold on différent bodily régions.

There is a graduai transition f rom oneness to twoness. Titchener quotes from Henri the following introspections for a séries of increasing distances :