Impartiality is a necessary condition of all scientific observation. We observe because we are interested in the result of our observa- tion: some chance occurrence has suggested to us an explanation of particular events, and we are interested to discover, by system- atic enquiry, whether the explanation is correct. The trained observer, psychologist or physicist or what not, can take the sug- gestion for what it is worth; he does not allow it to affect his observation. But the beginner is exceedingly liable to be led by interest into partiality; and so to see, not what really happens,, but what he desires or expects to see happen.
Impartiality in psychological investigations, however, is pecul- iarly difficult. In most sciences, the danger of partiality begins after a few accidental observations have suggested a certain view. In the case of (i) animal and child psychology, the bias may exist before any observation has been made at all, and all obser- vations, from the very first, be vitiated by it. Mother and nurse find intelligence in the baby when the disinterested observer can see nothing out of the common; and lovers of animals tell wonder- ful tales of the intelligence of their special pets. In the case of (2) adult human psychology, bias may also be prior to any obser- vation. A certain resoluteness and evenness of disposition, a moral steadiness and balance, are required of the introspective psychol- ogist. It is not only that " what ardently we wish we soon be- § 10. Rides for Introspection of Sensation 39 lieve ": the chemist runs that danger equally with the psychologist. It is rather that the objects of investigation are intrinsically elusive, that their investigation demands both quickness and accuracy, and that the observer has to forget all social relations and take up a sturdily independent attitude to facts which are, in part at least (§ 2), of his own making. Many people are too complaisant, too reflective (letting reflection about experience take the place of experience itself), too impressionable, etc., to be impartial.
To get at facts, we must be wholly unprejudiced: interested in the general subject, but not concerned to estabhsh a particular result.
(2) When we introspect, we must have our attention under control. The attention must not be permitted either to flag or to wander.
The reasons for this rule have been given above. The better we attend to an occurrence, the more accurate and lasting is our memory of it.
It is difficult for the beginner to control his attention. In the first place, he has not learned by experience what exactly it is that he is required to attend to, and so is liable to be distracted by what are really accidental and irrelevant stimuli. And when this difficulty is overcome, there is still the danger that the attention may wander or flag. The observer will be apt to interrupt his introspection, asking himself whether he is carrying out instruc- tions, whether his attention is at full strain, what is the meaning of this or that condition of the experiment, etc. Practice is the only remedy for these faults j and even practice cannot secure an unflagging attention, if the observation be too long continued.
(3) When we introspect, body and mind must be fresJi.
Fatigue and exhaustion prevent any sustained concentration of the attention. We cannot attend if we are sleepy, or if we have worked our muscles to the state of pain and stiffness. And if we cannot attend, we cannot introspect.
It follows that we can introspect best in the morning; or, if 40 Sensatioji as a Conscious Element morning hours are not available, in the late afternoon, after re- freshment by moderate exercise. Introspection should not be attempted immediately after eating, /.<?., at a time when we are normally sleepy. It follows also (i) that if a psychological investi- gation promises to be at all long, we should work upon it rather for a short time daily during a number of days than for any length of time together during a few days, and (2) that we should work at the same hour of each day. The first rule provides against fatigue during a single sitting, the second keeps the conditions of freshness and tiredness constant from day to day.
(4) When we introspect, our ge7ieral disposition^ physi- cal and mental, should be favourable. We must feel well, feel comfortable, feel good-tempered, and feel interested in the subject.
Any physical or mental discomfort hinders introspection: breath- lessness, a cold in the head, a too hot room, a strained attitude of the body (as when one sits in a low chair at a high table); or irritation at having to work at this particular time, self-conscious- ness, nervousness (anxiety as to whether one is working correctly, as well as one's neighbour, etc.), impatience, disbelief in the value of the special experiment, dislike of those with whom one is asso- ciated to perform it, the disregard which comes from frequent repetition of a certain act of introspection and the contempt which is bred by familiarity with its conditions, etc. It is difficult to be sure that our results are obtained under the most favourable conditions of physical and mental disposition; but no others are really trustworthy.
These are the most important of the general rules to be followed in introspection. Can we ever be quite sure that we have followed them } Even when we think that all possible precautions have been taken, it must often be the case that certain of the required conditions are left unfulfilled. However favour- able the general disposition, e.g.^ and however trained the § lo. Rules for Introspection of Sensation 41 power of observation, there may still be some unnoticed wavering of the attention, some unsuspected tinge of pre- conceived opinion. While, therefore, we may reasonably hope to get perfectly correct results in many instances, we cannot be sure that the single result of any particular experiment is entirely free from error. Now there is a method, employed both in science and in practical life, which helps us to set up a working standard, a norm with which all individual results may be compared, under the most various and fluctuating circumstances: the method of averages. We make a large number of observations, and take their average. This mean result will not repre- sent the observer at his very best, but will indicate the normal or average performance which may be expected of him when the conditions under which he observes are as favourable as human nature can make and keep them. The average lies, i.e., somewhere between the result obtained under absolutely favourable conditions and that gained under conditions which just fall short of being en- tirely favourable. The more highly trained the observer, the greater his impartiality, and the better his general dis- position, the more nearly will the average approach to the ideal standard.
The method of averages is always employed in psycho- logical experimentation. Oftentimes the average is brought exceedingly near the ideal result by the fact that the errors of the separate experiments cancel one another when the average is struck, — as many falling on the plus side as fall upon the minus side.
The * facts ' of the psychology of sensation are, then, average results obtained from observers trained in intro- spection under experimental conditions of both a general 42 Sensation as a Conscious Element and a special character. The former we have learned to know as impartiality, attention, freshness, and favourable disposition. The latter we shall discuss (under the head- ing of Method) in following chapters.
§ II. The Classification of Sensations. — Every sensation comes to us from a definite bodily organ. We may there- fore divide sensations into groups or classes, according to the various sense-organs which the body possesses, and speak of eye sensations, ear sensations, nose sensations, skin sensations, muscle sensations, joint sensations, etc.
This list, if completed, would be perfectly accurate. But it is convenient to make some further divisions and subdivisions, which are justified by differences in the nat- ure of the stimuli necessary to arouse sensations of certain kinds. Classification by siimnlns^ in addition to the classi- fication by sense-organs, is useful in two ways, (i) Sensations in general fall into two principal groups, according as their stimulus is external (originating outside the body) or internal (originating within the body). Light, the stimulus to vision, is an external stimulus; muscular contraction, the stimulus to muscular sensation, is an in- ternal stimulus. We therefore distinguish between sen- sations of the special senses, which are stimulated from without, and oi'ganic sensations, the stimulus to which consists in a certain state, or change of state, of the inter- nal bodily organ from which they come. There is one, and only one, sensation quality which is common to every department of sense, whether external or internal: the quality of pain. We shall therefore speak of pain as a coninio7i sensation.
(2) The nature of the stimulus may differ within the same sense. Light is the stimulus to sensations from the § II. Classification of Sensations 43 eye. But the physicist recognises two kinds of light, — white or mixed light, and pure or coloured light (light of one wave-length or of one vibration-rate); and we have two corresponding groups of sensations, — sensations of brightness (black, grey, white), and sensations of colour. Sound is the stimulus of hearing: but sound may be pro- duced both by a single shock or concussion of the air and by an air-wave; and we have two corresponding types of auditory sensation, — the simple noise (shock), and the simple tone (wave).
Our list will, then, take final shape as follows. To realise the part played in mind by the different groups of sensations, the reader must compare it with the corre- sponding list of § 22.
I. Sensations of the Special Senses (external stimulus).
1. Visual sensations.
a. Sensations of brightness (stimulus: mixed light).
b. Sensations of colour (stimulus: homogeneous or pure light).
2. Auditory sensations.
a. Sensations of noise (stimulus: sound concussion or shock).
b. Sensations of tone (stimulus: sound-wave).
3. Olfactory sensations (stimulus: odorous particles carried by a draught of air).
4. Gustatory sensations ("Stimulus: the chemical constitution of certain substances, which enables them to excite the organs of taste).
5. Cutaneous sensations.
a. Sensations of pressure (stimulus: mechanical affection of the skin).
b. Sensations of temperature (stimulus: thermal affection of the skin).
II. Organic Sensations (internal stimulus).
6. Muscular sensations (stimulus: contraction of muscle).
44 Sensation as a Conscious Element 7. Tendinous sensations (stimulus: pull or strain upon tendon).
8. Articular sensations (stimulus: rubbing or jamming together of surfaces of joint).
9. Sensations from the ahmentary canal.
a. From the pharynx (stimulus: dryness of mucous membrane).
b. From the oesophagus (stimulus: antiperistaltic reflex).
c. From the stomach (stimulus: dryness of gastric mucous membrane).
10. Circulatory sensations (stimulus: change in circulation).
11. Respiratory sensations (stimulus: change in breathing).
12. Sexual sensations (stimulus: change in blood-supply, or in secretory activity, of the sex organs).
13. Sensation of the ^static sense ' (stimulus: change in the distribution of pressure from the water of the semicir- cular canals of the internal ear). III. Common Sensation (external or internal stimulus).
CHAPTER III The Quality of Sensation I. Sensations of Special Sense § 12. The Quality of Visual Sensations. — The stimulus to vision is light. Physical theory regards light as a wave movement in the ether with which space is filled. Light is either mixed or pure (homogeneous): mixed, if it consists of waves of every possible length, travelling together; and pure (homogeneous), if its waves are all of the same length. Mixed light always excites the sen- sation of brightness; a single pure light, the sensation of some colour.
( I ) Sensatio7ts of Brightness. — We have only five names, in ordinary conversation, to indicate different kinds or qualities of brightness: black, white, grey, dark grey and light grey. When put to a rigid test, however, the eye is found to be capable of distinguishing 700 bright- ness qualities, varying from the deepest black to the most brilliant white.
The method by which we determine the number of brightness qualities is as follows. Four circular pieces (discs) of cardboard are prepared, two of dead black and two of white. A cut is made in each, from outer edge to centre, so that a black and a white can be fitted together, and a white sector, of any desired width, laid over the black surface. The backs of the discs are divided up into degrees and fractions of degrees, in order that the 45 46 The Quality of Sensation amount of white which replaces the original black in a particular experiment may be accurately measured.
For purposes of experiment, the discs are mounted in pairs — a white behind a black — upon 'colour wheels,' which allow of their rapid rotation. When the discs are combined in this way, presenting a black and white sur- face, and are rapidly rotated, they give rise to a sensation of grey the first experiment is Fig. 3. — Two discs, cut for mounting upon the colour wheel, ihe rigure shows the ' •' way in which they are fitted together for parison of an all-black rotation. disc with a surface in which there is a shght trace of white, what mixture of black and white is just different in sensation from dead black; the object of the second, to discover what amount of white must be added to black to make a grey just different in sensation from the grey of the first mixture; and so on. Each of these 'just different ' brightness sensations is a conscious element.
The sensation of brightness (black, grey, white) is pro- duced id) by the action of mixed light upon the retina or nervous network (Latin rete, net) of the eye. It can also be produced (p) by the mixture of certain pairs of pure lights, — red and bluish green, orange and blue, yellow and indigo-blue, greenish yellow and violet (the ' comple- mentary ' colours); and (<:) by the mixture, in proper pro- portions, of three pure lights, properly selected (red, green and violet, or red, yellow and blue, etc.).
(2) Sensations of Colour. — The colours of the solar spectrum, which may be taken as standard or normal § 12. The Quality of Visual Sensations 47 colours, are named: red, orange, yellow, green, blue, indigo-blue and violet. To these may be added purple, a colour compounded of red and violet, and lying for sen- sation midway between these qualities. Purple is com- plementary to spectral green; and also produces the sensation of brightness when mixed with yellow and blue. We speak further of reddish orange, greenish blue, etc. There are thus some twenty-five words and phrases in ordinary use as names of the spectral colours.^ But here again, the eye, when placed under experimental conditions, is found to distinguish far better than language does: we can discriminate, if the purples are included in the sum, about 150 spectral colour qualities.
Method. — Two entirely similar spectra are thrown upon a wall, one lying directly above the other. The upper spectrum is then shut off from view by a black screen, in which a narrow upright slit is cut; nothing is seen of it, therefore, but a single line of colour, — say, of red, — coming through the slit. A similar screen, with a similar slit, is placed over the lower spectrum. The ob- server moves this second screen to and fro, until the line of red appearing through its slit is just different in sensation from the hne of red seen through the slit of the upper screen. The lower slit is then left in place, and the upper screen moved until another just different red is obtained. The lower screen is thereupon moved again, — and so on, until the whole band of colours has been passed over; each in turn, as it shows through the slit, be- ing compared with the next preceding colour quality. Every ^ just different ' colour is a conscious element.
Colour sensations, of a quite simple nature, but different from the colours of the spectral series, are produced by the 1 Red, reddish orange, orange-red, orange, orange-yellow, yellowish orange, yellow, yellowish green, greenish yellow, green, greenish blue, bluish green, blue, indigo-blue, bluish violet, indigo-violet, violet-blue, violet, purplish violet, violet-purple, purple, purplish red, reddish purple, violet-red, reddish violet.
48 The Quality of Sensation mixture of mixed with pure light. Thus pink results from the mixing of red and white; brown from that of black and yellow; mauve from that of purple and white. Each of the 150 spectral colours may be combined in this way with each of the 700 brightness qualities. But the more white or black we mix with a colour, the harder does it become to distinguish that colour from other colours. Hence, instead of having 150x700 colour sensations of this mixed origin, we have only about 150x200, or Method. — Two spectra are thrown on a wall, as before. The two screens are so adjusted as to show precisely the same line of colour — say, of red — in both spectra. The light from which the lower spectrum is taken is then slowly brightened, until its red line is just lighter (pinker) than the red line of the upper spec- trum. Thereupon the upper light is brightened, till its red line is of a just lighter quality than the pink of the lower spectrum. The process is repeated, until one of the originally red lines is so highly illuminated that the red is lost, and only white is visible. The reverse method is then followed, each red line being darkened in turn, until one of the reds becomes black.
Every one of the 150 spectral qualities must be lightened and darkened in this way, until the full number of ^ just lighter ' and * just darker ' colours has been made out.
We possess, therefore, about 700 + 1 50 4- 30,000 quali- ties of visual sensation: we have to search in the structure and function of the eye for the conditions of some 30,850 conscious elements. For the qualities are all equally elemental as sensations, however different the physical processes (stimuli) with which they are connected.
Various explanations have been offered, by physiologists and psychologists, of the changes set up in the eye by the action of light. It is impossible, in the present state of our § 12. The Quality of Visual Sensations 49 knt)wledge, to say decisively what the conditions of visual sensations are. We may, perhaps, suppose that light which falls upon the retina gives rise to two processes of chem- ical decomposition: one of them (the achromatic process) dependent upon the intensity of the light (the ' height ' or amplitude of the light-wave), and the other (the chromatic process) upon its wave-length. To the former corre- sponds the sensation of brightness, to the latter that of colour. We must further suppose that when waves of certain different lengths are mixed, — waves of all possible lengths, waves of three lengths (properly selected and proportioned), and waves of two lengths (if these are com- plementary),— their undulations cancel one another; so that we have no sensation or colour (connected with the chromatic process), but only a sensation of brightness (connected with the achromatic process).
It must be carefully noted that although the achromatic process is excited by the intensity of a light stimulus, the sensations con- nected with it — black, white, grey — - differ in quality. We shall return to the point later (§ 24).
The various sets of three pure Hghts (red, green and violet; red, yellow and blue; orange, green and violet; purple, — we haVe admitted purple to the rank of a spectral colour, — yellow and blue, etc.) which, when mixed in certain proportions, arouse the sensation of brightness, will, if mixed in certain other proportions, arouse that of colour: and it is possible, by varying the propor- tions, to get from each set of three all the different colour quali- ties. But red, green and violet give far richer and more brilliant colours, when rightly combined, than do any other three qualities. Hence these are known as the primary colours of the spectrum.
There are four spectral colours which are of especial interest in practical life: green, blue (§ 5), red and yellow. Red is the colour of blood (Skt. riidhira, Icel. rothra, blood); yellow is the pale colour of young vegetation (Gk. y\oy], ^(XcDpos), and thus has 50 The Quality of Sensation the same significance, derivatively, as green. These four colours are known as the principal colours of the spectrum. We are no longer interested in them for the reasons which drew the attention of primitive man to them; but they are of such great impor- tance in painting that the term ' principal ' is still in place. They are no simpler, as conscious processes, than the other colour qualities.
The word '■ mixture ' in this Section means always mixture of lights, and not of pigments (paints). Mixture of blue and orange hght gives a sensation of grey: mixture of blue and orange paints gives a sensation of green. The blue pigment crystals let through blue and green light, the orange crystals orange, yellow and green light. The blue and orange cancel each other, and only a yellow- ish green is left to be seen.
§ 13. The Quality of Auditory Sensations. — The stimulus to hearing is sound. From the physical point of viev^, sound is a movement of the air particles. The movement may be continued and regular (sound-wave) or momentary (shock or concussion); or it may consist of mixtures of waves and shocks, or of successions of shocks. A sound- wave excites the sensation of tone; a mere concussion, or a wave movement of less than two complete undulations, the sensation of simple noise. The other forms of air movement are connected with complex auditory processes.
What we commonly call the ' pitch * or ' height ' of tones is their psychological ' quality.' Differences in the quality of simple noises may also be conveniently ex- pressed by these terms, although they are ordinarily em- ployed only with reference to tone.
(i) Sensations of Tone. — We speak of tones as 'high' and 'low,' 'harsh' and 'clear,' 'shrill' and 'mellow'; and we distinguish ' thin ' tones from tones which possess ' volume,' etc. We also have symbolic names for the § 13. The Quality of Auditory Sensations 51 twelve tones comprised within each of the seven octaves of the musical scale: (ffl, ^*, bV, d^^%, a^^, etc. When accurately tested, however, the ear is found to be master of a much wider range of tones than language indicates; we can hear about 11,000 different tones.
Method. — One of two precisely similar tuning-forks has its tone lowered or ' flatted ' a little by the attachment of a small weight to each of its prongs. The forks are struck, one after the other, at an interval of a few seconds, and the listener is required to say whether their tones seem to him alike or different. The flatting is increased, until he finds the tones just different. Then the weighted fork is taken as standard, and the other weighted a little more heavily, — until its tone appears just different from (flatter than) the slightly lowered tone of the first experiment. The tests are repeated upon a large number of forks of different natural pitch, so that no tone quality which can possibly be sensed is missed. It will be found in this way that tone sensations furnish the 11,000 conscious elements mentioned above.
(2) Sensations of Simple Noise. — Language has sev- eral words, to express different kinds of simple noises: snap, rap, tap, puff, pop, shock, crack, flick, thud, etc. We also use various adjectives to indicate various noise qualities: * sharp ' crack, ' dull ' thud, etc. As usual, how- ever, the sense-organ can draw finer distinctions than are drawn by language: we can discriminate some 550 quali- ties of simple noise.
Method. — If a tuning-fork is struck, and more than two of the air-waves which its vibration occasions are allowed to reach the ear, we hear a tone. If less than two complete undulations are heard, — the rest being cut off from the sense-organ by the drop- ping of a sound-proof screen between it and the fork, — we get a sensation not of tone but of simple noise.
Two forks are taken, as in the previous experiment. Let us 52 The Quality of Sensation suppose that one vibrates 128 times in the i sec, and the other (the flatted fork) 126 times. The former is sounded for some- thing less than yf^ sec, and the latter for something less than yf g- sec.; and the observer notes whether the two noises are of like or different quality (height or pitch). If they are alike, the experimenter flats the flatted fork still further, until the observer remarks a difference in noise quality. A record is then made; and the experiment proceeds as in the case of tones.
Most of the noises which we are accustomed to hear (crash, roar, hiss, rattle, splash, clatter, etc) are of a complex nature, comprising several simple noises or a number of these mixed with tones.
Sound is received into the outer passage of the ear, conveyed inw^ards by a series of vibrating bodies (elastic membrane, chain of small bones, etc.), and finally produces a movement in the w^ater (endolymph) of the cochlea of the internal ear. The cochlea is a hollow tube, through the v^hole length of v^hich is stretched a membrane, the basilar membrane. The cross-fibres of this membrane are arranged like the strings upon the backboard of a piano; they are very short at the beginning (treble strings) and gradually increase in length as the membrane continues (bass strings). Each cross-fibre carries sensitive cells, v^ith which the fibrils of the auditory nerve are connected. A movement of the water in the tube excites the cells standing upon particular strings or cross-fibres. Only those strings are affected, in a given case, whose vibra- tions correspond to the sound outside the ear which causes the movement of the water. Every string may thus be said to be ' tuned ' to a certain sound-wave.
We may illustrate this process by supposing that the top of an upright piano is turned back, and a word shouted into the body of the instrument. The piano is * set ringing '; certain strings are § 14- The Quality of Olfactory Sensations 53 thrown into vibration, because the speaking voice contained cer- tain tones to which those strings are ' tuned.' Different strings respond, according as the word is pitched high or low.
The sensation of tone arises when a fibre of the basilar membrane is made to vibrate regularly by more than two successive wave movements of the endolymph; that of noise, when a fibre gives a jerk or twitch, in answer to a single push or less than two complete wave movements of the endolymph. Physically, therefore, a simple noise is merely an imperfect tone. There is no sharp line of division between the stimuli: two complete wave move- ments will give rise now to one and now to the other sensation. The sensations themselves are, however, quite distinct.
§ 14. The Quality of Olfactory Sensations. — Smells are ordinarily classified as agreeable and disagreeable, and named after the objects which give rise to them (musk, violet, etc.) without regard to their likeness or unlikeness in sensation. It is impossible, in the present state of our knowledge, to say how many qualities of smell the nose can distinguish. In all probability the number is very large. It is also probable that smells fall into groups of similar qualities, and that the members of each group form graded series, like those of tone or brightness.
Method. — Prepare a number of solutions of odoriferous sub- stances. Close one nostril with cotton-wool, and sniif at a solution until you can smell it no longer. This will soon happen, as the nose is easily fatigued. Then sniff at another. If this is smelled, its quality differs from that of the first solution; if you cannot smell it, its scent is the same as the scent for which the nose has been fatigued. If you exhaust the sense of smell by tincture of iodine, you will find that you can still smell oil of lavender, but 54 The Quality of Sensation