There are, however, many psychologists who would re- fuse to accept these statements. Some look upon affection as an attribute of sensation, on a level with quality or in- tensity; they speak of the feeling-tone or the affective tone of sensation, and not of a separate affective process. Others identify affection with a certain kind of sensation; pleas- antness, they say, is a diffuse sensation of tickling, or a weak sensation of lust; unpleasantness is a low intensity of cutaneous or organic pain. Others, again, think that the same processes may appear sometimes as organic sensations and sometimes as affections, according as they are isolated in consciousness, analysed out by attention, or are given in a mixed medley of unanalysed experience.
The first view, that affection is an attribute of sensation, is easily disposed of. For affection has attributes of its own, — quality, intensity, duration; whereas the attributes of sensation are ultimate phases of an elementary mental process, and cannot by any effort of abstraction be factor- ised into simpler components. Moreover, if any single attribute of sensation is reduced to zero, the sensation dis- appears: a sensation that has no quality, or no intensity, or no duration, is not a sensation at all; it is nothing. But a sensation may be non-affective, perfectly indifferent, and still be far removed from disappearance. This view, then, may be dismissed from consideration.
Nor does the second view seem to be more tenable. All these sensations, tickling and lust and pain, may become clear, may be made the object of attention; indeed, they 238 Affection are, so to say, naturally clear, precisely the sort of sensa- tions that we are obliged to attend to. But we cannot at- tend to affection. Again, all three sensations have theii own quality as sensations. We have not fully described tickling or lust when we have dubbed it pleasant; we have not fully described pain when we have dubbed it unpleasant. Tickling has a ticklish quality, that is peculiar to it as sensation; lust has a specific quality, which differentiates it from tickling; pain has a peculiar quality of itch or ache or thrill. An experience which is both ticklish and pleas- ant, or is both itchy and unpleasant, cannot be identified with pleasantness or unpleasantness; it is something more. And finally, tickling and lust may, with change of cir- cumstances, be either pleasant or unpleasant, and the same thing holds of pain; the scratching of an irritated area of the skin may be at once painful and distinctly pleasant. While, however, these arguments appear unanswerable, it is nevertheless true that the theory which they oppose is held by a large number of modern psychologists.
The third view tries to reconcile the belief that affection is a second form of elementary process with the belief that it is only a certain sort of sensation. We have, it says, vari- ous obscure and confused experiences which, ordinarily, we take as they come, without asking what they are made up of. So long as we do this, the experiences present them- selves as pleasantness or unpleasantness; but if we scruti- nise them, and analyse them into their elements, then they reveal themselves as complexes of organic sensation. The reply is, surely, that experiences which may be analysed into organic sensations are complexes of organic sensations, and nothing more or less; but that, if an experience is pleasant or unpleasant, then the pleasantness or unpleas- § jo. Other Views of Affection 239 antness remains, no matter how many organic sensations are found along with it. Unless affection is organic sen- sation, — and we have declined to admit that affection is a sensation of any kind, — the third view is untenable.
It is impossible here to discuss or even to mention all the views that have been held by psychologists concerning the nature of affective experience. Affection has been described, for instance, as a relation, whether between sensations or between a particular sensation and all the rest of consciousness. There is some ground for both opinions. Certain combinations of tones are pleasing, certain others displeasing; certain architectural proportions are beautiful, certain others are ugly. These facts tell in favour of the first theory. Again, a sensation that now enters consciousness enters into relation with all the contents of consciousness, and must, so to say, make its terms with those contents. If it fits in easily among them, it will probably be pleasant; if it disturbs or sup- presses them, it will probably be unpleasant. All this tells in favour of the second theory. Nevertheless, both theories alike are to be rejected. Against the first we must urge that affection is not always dependent upon the relation between sensations, but may attach to a single sensation; and, besides, that affection, even if dependent upon sense-relations, need not on that account be itself a relation. Against the second we must urge, in the same way, that the fitting-in of the new sensation, or its reverse, is purely a matter of the behaviour of sensation; it is not necessarily ac- companied by affection; still less is it to be identified with affection.
Affection has also been described in terms of mental attitude. It has been defined as such an attitude: affective experience, we are told, is made up of one's own attitudes, and sensory experience of the impressions that one receives: and it has been defined as a character of mental process that depends upon and represents one's personal attitude. In its cruder forms, this theory implies the existence of that permanent mind, that mind-substance or mind-animal, which we rejected in §3. But even at the best we 240 Affection may well enquire whether it does not put the cart before the horse. Does feeling represent our mental attitude, or is our mental atti- tude the expression of feeling?
§ 71. The Methods of Investigating Affection. — There are two chief difficulties in the way of an experimental treat- ment of the affective processes. We cannot attend to pleasantness or unpleasantness; and we can describe our affective experience only in a roundabout way. The first of these difficulties needs no further discussion; we have seen that affections lack the attribute of clearness. The second difficulty arises from the fact that spoken language I lis a language not of feeling butof idea. If I say 'I am * very angry,' you know that I am angry; I have given you the idea of my anger; but I have, after all, merely indicated my feeling, and not described it. And even if I should attempt a detailed description, my account — so far as it referred to the pleasantness or unpleasantness of the anger — would still be an account at second hand; I should be obliged to translate my affection into an idea of affection. / There is, it is true, a language of feeling: the language of \ exclamation and gesture. But we have learned, in the course of civilisation, to repress our emotions; we rarely use this language; if on occasion we wish to do so, we are apt to make ourselves ridiculous; and the language itself is but little developed in comparison with speech.
Partly on account of these difficulties, and partly for other and historical reasons, experimental psychology has, until recently, neglected the study of affection. The prob- lems that were at first attacked in the psychological labora- tory were suggested by physics and physiology and astronomy: the problems of mental measurement (§ 63), of § 71. The Methods of Investigating Affection 241 the number and nature of the sensory qualities and of their relation to the organs of sense (§§ 14 ff.), and of the duration of mental processes. Feeling had no place in the pro- gramme. Now, however, that psychology has gone far enough to experiment on its own behalf, without suggestion from its neighbours, the study of feeling has begun. There are two experimental methods at present in use: the method of impression and the method of expression.
(1) The method of impression has taken various forms, (\j the most promising of which is the method of paired comparisons. In this, a series of similar stimuli is laid out, and" tlie stimuli are presented to the observer two at a time; care is taken that every member of the series is paired with every other member. We may have, for instance, a series of squares of differently coloured paper, numbered in order 1 to 50. We cut two square windows in a sheet of neutral grey cardboard, and show the colours in these win- dows; the series of observations is so arranged that colour 1 shall be shown along with 2,3,4,...UP to 5° 5 colour 2 with 3, 4, 5,...; and so on. In a first experiment, the observer may be asked, as the successive pairs are exhibited: Which of the two colours is the more pleasant? In the next experiment: Which of the two is the more unpleasant? And the series may be repeated as often as seems necessary. The introspective task is extremely simple: the observer has merely to be passive, to let himself go, to allow the stimuli to take affective possession of him. He need not even speak; all that he has to do is to point towards the window which contains the more pleasant or the more un- pleasant colour; and the experimenter then records his choice.
At the end of an experiment, every colour has received a J 242 Affection certain number of choices, proportional to its affective value; a very agreeable or very disagreeable colour will have been preferred many times over, a colour that is hardly more than indifferent will have been chosen but rarely. If, now, the names or numbers of the colours are arranged in order, 1 to 50, along a base line, and the number of preferences is indicated in every case by the length of a vertical line erected upon this base, then the smooth curve which joins * the tops of the verticals may be termed the affective curye for the particular observer; it is a curve whose course directly expresses his affective response to colour stimuli.
The method may be put to various uses: here is an illustration. We have argued, in § 69, that pleasantness and unpleasantness are opposite and incompatible. If we have recourse to the method of paired comparisons, we can prove that they are opposite. Ex- periments made with series of coloured papers, series of musical tones, and series of rhythms (rates of metronome beats) give affec- tive curves for pleasantness that are just the reverse cf the same observers' affective curves for unpleasantness: stimuli that receive a large number of choices in the experiment on pleasantness re- ceive a correspondingly small number in the experiment on unpleas- antness. To prove a thing like this may look like proving the obvious. But, in fact, nobody knew for certain, until these experi- ments had been made, that pleasantness and unpleasantness were opposite; psychologists might believe, but they did not know. And besides, things that look obvious are oftentimes, in science, the things that very particularly call for proof.
The method may also be used to measure the individual differ- ences in affective response to stimuli. We say that there is no disputing about tastes; and we tend to think, in general, that sen- sations are experienced in the same way by all normal persons, but that affections are experienced differently, individually, personally. In reality, sensations are less stable and affections are more stable than we suppose. For sensations are changed and shifted by the I § 7i. The Methods of Investigating Affection 243 varying conditions of adaptation, contrast, attention; while the affective curves of a number of different observers are sufficiently alike to show that, under the same circumstances, all normal per- sons react affectively in very much the. same.way.
The method of impression can do us further service. The simple introspective task which it sets is repeated, over and over again; the introspective experience ac- quired during a series is cumulative, all of the same kind. Hence the observer is able, in the intervals between suc- cessive series, to make a full report upon the affective con- sciousness: he can describe the course of the sense-feeling, take note of anything like a mixed feeling, state whether the agreeableness or disagreeableness of the stimuli was always of the same kind or showed differences of quality, and so on. It is, indeed, to this method that we must mainly look for a settlement of the vexed questions of affective psychology. The work is laborious, and con- sumes a great deal of time. But the method offers the advantage of a twofold control: an external control by the affective curve, the objective record of the distribution of preferences; and an internal or subjective control by the accompanying introspection.
(2) The method of expression seeks to record the bodily changes which accompany the passage of an affection through consciousness. Just as we express emotion by smiling or frowning, by laughter or tears, by clapping the hands or shrugging the shoulders, so do we express the simplest affective experience by a change of various bodily functions. Tests have been made of pulse, respiration, the volume of a limb, involuntary movement, muscular strength, and the response of the bodily tissues to the elec- tric current.
Affection Affection Physiology has long been in possession of instruments that fur- nish the records required, and psychology has borrowed these instruments and has adapted them to its own purpose. Take pulse, for instance. We know that the pulse can easily be felt in the wrist, over the radial artery. Now suppose that a small tin funnel is covered, at the wide end, by thin rubber sheeting; and that the sheeting is laid over the artery, and the funnel tied to the wrist. At every beat of the pulse, the rubber will be lifted, and a puff of air will be sent out from the small end of the funnel. Con- nect this end, by a piece of stiff- walled rubber tubing, to the small end of another funnel, whose wide end is similarly covered with rubber sheeting. Hinge to the lip of the second funnel a light splinter of bamboo; let the splinter lie across the rubber sheeting, Fig. 29. Marey tambour (rubber-faced funnel) with writing lever, rubber tub- ing, and air-cock. The tubing is continued from the air-cock to the tambour on the wrist.
and let it rest in the middle on a tiny disc of cork cemented to the rubber. Whenever a puff of air is sent from the first funnel, by the pulse beat, the cork disc of the second funnel will be lifted, and the splinter will rise with it. Now take a piece of glass, that has been smoked over a gas flame, and set it up, in the vertical position, so that the free end of the splinter rests lightly against it. It is clear that, as the pulse rises and falls, the bamboo point will rise and fall on the glass, and will trace a clear line in the soot. And if the glass be drawn slowly forward at a constant rate, a pulse-curve will be traced which shows the rate and the height of the successive beats. Then the glass may be sprayed with varnish, and we have a permanent pulse record.
This is a rough indication of what is called the graphic method. In practice, the instruments are much more complicated and in- §71- The Methods of Investigating Affection 245 genious. The bodily change may be transmitted to the bamboo stylus not by air but by a system of rigid levers, or by the electric current. And in place of the glass plate we use a kymo- graph, a brass drum covered with smoked paper, which may be revolved at varying rates. To ensure accuracy in the interpretation of the curve, time- markers are employed, which trace a time curve (in seconds, or half sec- onds, or fifths of a second, or even in much smaller units) below the curve of pulse or breathing. In principle, how- ever, the tin funnels and the smoked glass plate represent the method.
The instrument which registers the height and rate of pulse is termed the sphygmograph. Similar instruments, tied round the chest or abdomen, register the course of thoracic and abdominal breathing: they are called pneumographs. Volume is registered by the plethysmo- graph, a large glass jar, partly filled with warm water, in which the hand and forearm of the subject are placed. A glass tubule leads off from the jar, by way of rubber tubing, to the recording funnel or tambour; as the arm swells, the water in the jar rises, the air above it is compressed, and the stylus is thrown up. Involuntary movement is registered by the auto- matograph, a form of FIG. 31. Franck's Volumetric Sphygmograph ^ ^anchette or ouija. (mechanical transmission by means or a sys- * tem of levers). board which still figures FlG. 30. Clockwork Kymo- graph.
Affection FlG. 32. Pneumograph.
largely in spiritistic seances. A board is slung from the ceiling, so that it lies horizontally just over the surface of a table; a pointed glass rod stands vertically in a hole pierced at its forward end. On the table, under the point of the rod, is spread a sheet of smoked paper. If the arm is laid carefully on the board, and left to itself, the glass point traces on the paper the record of its involun- tary movements. Muscular strength is registered by the dynamograph or ergo- graph: in the former, the hand grips a steel spring, the compression of which sends a puff of air through the trans- mitting system to the stylus; in the latter, the finger pulls re- peatedly against a weight or a spring, and the cord upon which it pulls carries a stylus which writes directly upon the kymographic surface. Lastly, in order to measure what has been termed the psychogalvanic reflex, two electrodes, which are connected with a battery giving a constant current and with a sensitive galvanometer, are applied to the hands or to other cutaneous surfaces; an affective stimulus is then presented to the subject, and the resulting deflection of the needle of the galvanometer is recorded.
In every case, the procedure of the experi- ments is the same. First of all, a normal reading is taken from the instruments: the experimenter assures himself that the subject is in an indifferent frame of mind, and without applying any stimulus takes a curve of pulse, breathing, or volume; or allows the stylus of the automatograph to record the involuntary tremor of the arm; or obtains an ergographic record; or notes the constant position of the galvanometer needle. Then he applies some agreeable or dis- agreeable stimulus, and the record is made while the sense-feeling FlG. 33. Franck's Plethysmograph.
§ *]\. The Methods of Investigating- Affection 247 is running its course. The experiments are repeated again and again; and finally the experimenter tries to discover, from his tracings and notes, the exact correlation be- tween affective quality and the bodily change which expresses it.
The method of expression was introduced into ex- perimental psy- chology with great expectations; and the results of the first experiments seemed clear and FlG 34# Automatograph.
definite. All the bodily functions were heightened and strengthened in pleas- ure, and all were weakened and depressed in unpleas- FlG. 35. Mosso's Ergograph.
ant states of mind. A leading psychologist compared the method to 'an extraordinarily delicate chemical rea- Affection gent ' for the detection of pleasantness and unpleasantness. Very soon, however, these dreams were dispelled. The l+j-j- Fig. 36. Part of a tracing obtained with the.Franck plethysmography one- third actual size. The small waves indicate pulse, the large waves breathing. Change of volume is indicated by the varying height, above the time-line, of the entire tracing. An unpleasant stimulus was applied at the moment marked by the left-hand cross, and removed at that marked by the right-hand cross. The unit of the time-line is 1 sec.
changes in the curves are not correlated, unambiguously, with changes of affective experience: they depend in part upon purely physiological factors, and in part upon other psychological processes, — for instance, upon attention.
FlG. 37. Von Frey'sSphygmograph. The entire apparatus is bound upon the wrist, the button at the extreme right resting on the artery. The movements of the writing-lever are recorded on a very small, light drum, driven by a clockwork which is housed in the square box to the left. The clockwork also actuates a time- marker, shown in the Fig. at the lower edge of the drum.
Indeed, the simple application of even a weak stimulus appears to throw the whole body, as it were, into reverbera- tion; you cannot show the observer a wall-paper pattern without by that very fact disturbing his respiration and § 71. The Methods of Investigating Affection 249 circulation. If the automatograph proves that the organ- ism involuntarily expands, reaches out, in pleasure, and involuntarily withdraws, shrinks back, from the disagree- able, it proves also that you reach out when you think of a house across the street, and shrink back when you think of the person behind you. There may be a general rule that pleasantness goes with quick and shallow breathing and slowed pulse, unpleasantness with slow and deep breathing and quickened pulse; such an antagonism would well express the opposition of the affective qualities. But there can be no doubt that these changes in pulse and respiration may be otherwise occasioned, and that the appearance of the affective correlation depends upon con- ditions that are, as yet, only imperfectly understood. Pulse, for instance, has been found to quicken when the stimulus is a pleasant taste, and to slow when it is a pleasant tone or colour; and breathing appears to vary in its behaviour with different individuals.
So far, therefore, we have gained little from the method except a number of divergent results, and the conviction that, before these results can be psychologically inter- preted, we must know a great deal more than we do of the bodily functions involved.
It has recently been suggested that the methods of impression and expression might profitably be combined, the registration of physical expression serving as an objective check upon the observ- er's introspective report, and the introspection in its turn being used for the interpretation of the objective record. It is, of course, feasible to register the course of pulse or breathing during an entire series of experiments by the method of paired comparisons. But the registration would hardly be worth while, unless the intro- spective task set to the observer were greatly complicated; and then we should lose one of the chief advantages of the method.
Affection It therefore seems best to continue work by the method in its present form, and to wait patiently for the time when increased knowledge shall give us the required understanding of the physi- ological processes.
It may be added that the psychogalvanic method, though in principle it dates from 1888, has only within a year or two been brought into the psychological laboratory. Hence we can hardly judge of its merit or promise as a means of affective investigation, though the indications are that it will offer the same difficulties as the other forms of the expressive method.
§ 72. The Tridimensional Theory of Feeling. — In 1896, Wundt propounded a theory of feeling that differs radi- cally from the view taken in this book. Pleasantness and un- pleasantness, he de- clared, are not simple affective qualities, but general names for a very large number of different qualities. And even so, the terms pleas- ant and unpleasant are not adequate as intro spective descriptions of our affective experi- ence. That experience is, so to say, tridimen- sional. Feeling moves, first, between the oppo- site poles of pleasantness and unpleasantness; secondly, between excitement and depression; and thirdly, between FlG. 38. Diagram representing the course of a typical sense-feeling, according to Wundt's theory. The feeling sets in as a mixture of excitement and unpleasantness, to which ten- sion is soon added. It then drops into the region of pleasantness and calm, takes on a tinge of relaxation, and so ends at the indif- ference-point from which it started.
§ 72. The Tridimensional Theory of Feeling 251 tension and relaxation. Excitement and depression, ten- sion and relaxation are, again, general names for a very large number of ultimately different affections. Indeed, the full list of affective qualities, distinguishable under the six headings, is far longer than the total list of sensations.
This theory has found many adherents, and has natu- rally also aroused a great deal of controversy. It must evidently be met, if our own two-quality view is to be maintained. Let us see, then, what is to be said for and against it.
The first thing to ask about a scientific theory is whether it is logically constructed. Wundt assumes three catego- ries or dimensions of feeling. Pleasantness and unpleas- antness depend, he says, upon the intensity of the stimuli that affect us: a moderate amount of stimulation is pleas- ant, while too much or too little is unpleasant. Excite- ment and depression depend, similarly, upon the quality of stimulus. Finally, tension and relaxation depend upon time, upon the temporal aspects of stimulus; as we wait and expect, we become tense, and when the event happens, we relax. In other words, our sensory experience varies in degree, varies in kind, and varies in time: and every such variation corresponds to a distinct category of affec- tion. But now the critical question arises: Does not our sensory experience also vary in space? Do not spatial perceptions and spatial ideas bulk as large in conscious- ness as intensive or qualitative or spatial? Is not the world of space as native and as important to us as the world of time? And if this is the case, as it undoubtedly is, then Wundt's theory is illogical. There should be four affective dimensions. «ot three; and the fourth sort of 252 Affection 252 Affection feelings should depend upon the spatial aspects of stimu- lus. The objection is, indeed, more than logical; it is also psychological; for a view of feeling which ignores the part played by space in our mental life must be psycho- logically inadequate.
Now consider a second point. Pleasantness and un- pleasantness are opposite, in name and in nature. But relaxation is hardly, in the same way, the opposite of tension; relaxation is rather the minimum, the zero point, of tension. And what is the opposite of excitement? Sometimes Wundt says depression; sometimes calm; sometimes arrest. But these three feelings are surely different: to feel depressed is not the same thing as to feel calmed down, and to feel calmed down is not the same thing as to feel checked or baffled. Logically, the pairs excitement-depression and tension-relaxation are not of the same order as the pair pleasantness-unpleasantness. And again the objection is more than logical; for the direct opposition that is felt in pleasantness-unpleasant- ness cannot be felt in the other two cases.
It appears, then, that the new theory is not logically constructed. More than that, its logical weaknesses point, pretty clearly, to psychological defects. Psychology, how- ever, may now speak for itself: and the first thing it has to say is this, — that excitement and depression, tension and relaxation, are never simple, elementary processes; they are, on the contrary, experiences of some complexity; and they invariably contain organic (especially kinaesthetic) sensations. On the sensory side, they represent — to put it roughly — different muscular attitudes; on the affective side, they may be either pleasant or unpleasant. Pleas- urable excitement may be the opposite of depressed mel- § 72- The Tridimensional Theory of Feeling 253