The Open Collections website will be unavailable July 27 from 2100-2200 PST ahead of planned usability and performance enhancements on July 28. More information here.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An investigation of the Luscher colour test Melhuish, Peter William 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-UBC_1973_A8 M44_4.pdf [ 5.9MB ]
JSON: 831-1.0101351.json
JSON-LD: 831-1.0101351-ld.json
RDF/XML (Pretty): 831-1.0101351-rdf.xml
RDF/JSON: 831-1.0101351-rdf.json
Turtle: 831-1.0101351-turtle.txt
N-Triples: 831-1.0101351-rdf-ntriples.txt
Original Record: 831-1.0101351-source.json
Full Text

Full Text

•AN INVESTIGATION OF THE LUSCHER COLOUR TEST by Peter W. Melhuish B.A., University of Br i t i s h Columbia A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Arts in the Department . of , • • \ • \ . Psychology We accept this thesis as conforming to the required-standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1973 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver 8, Canada ABSTRACT The Luscher Colour Test was subjected to examination. A survey of psychological investigation of colour responding precedes the experimental part of the thesis. Colour, preference, the deter-minants of colour preference, and colour meanings are examined specific a l l y . Part Two consists of four experimental studies. The f i r s t was an application of colour measurement techniques to the coloured stimuli on the test. Results showed important differences i n the colours on three published versions of the same test, whereas i n fact they are supposed to be the same. A rationale explaining Luscher's choice of colours to comprise any one version however, was not apparent. A second study examined the applicability of the test with colour-deficient persons.' Certain of the test stimuli i n particular were found to be indistinguishable i n hue to dichromats. No d i f f e r -ences i n Luscher colour preferences between 18 colour-deficient sub-jects and a control group were found, however. Sixteen depressives, 16 nonpsychiatric hospitalized, and 16 nonpsychiatric nonhospitalized persons were administered the Luscher Colour Test in a third study. The test was only able to differentiate between a psychiatric and a nonpsychiatric population when colour preferences were scored v i a the Luscher method of interpretating preferences. Without this interpretation the test cannot make such a distinction. Finally, 15 persons were administered the test in a longitudinal / study. Results showed that colour preferences changed noticeably from week to week, implying that personalities would also change accordingly. The results taken collectively suggested that the Ltischer test i s not a valuable aid to personality diagnosis and research. \ \ i i TABLE OF CONTENTS Page ABSTRACT i LIST OF TABLES v LIST OF ILLUSTRATIONS . .-' v i i ACKNOWLEDGEMENT ix INTRODUCTION 1 PART ONE: THE LUSCHER TEST IN PERSPECTIVE I. DESCRIPTION OF THE TEST 4 History The Task ,The Test Colours Rationale (Theory) Underlying the Test Scoring and Interpretation II. COLOUR PREFERENCE 13 Investigations of Colour Preference Sex and Age Factors Hue, Saturation, and Brightness Methods of Determining Preferences III. DETERMINANTS OF COLOUR PREFERENCE 2 2 The Meaning of Colours Synaesthetic Attributes Connotative Meaning Affective Meanings and Values Rorschach Research on Colour and Affect Affective Meanings of Individual Colours The Physiological Meaning of Colour Symbolic Values Summary of Colour Meaning PART TWO: EXPERIMENTAL INVESTIGATIONS WITH THE LUSCHER TEST IV. OBJECTIVE ANALYSIS OF THE TEST COLOURS V. COLOUR VISION AND THE LUSCHER TEST VI. COLOUR PREFERENCE AND PSYCHIATRIC STATUS i i i Page VII. STABILITY OF LUSCHER COLOUR PREFERENCES 105 SUMMARY CONSIDERATIONS 1 1 3 FOOTNOTES 1 1 8 BIBLIOGRAPHY 119 i v LIST OF TABLES Table Page 1. Meanings and associations of the eight colours in the Luscher test 23 2. Affective values found to be ascribed to various colours 36 3. The different values found to be ascribed to various colours 38 4. Luminance values in nits and derived Y% values for the eight colours of LCTj 57 5. Colour specifications for the three versions of the eight-colour panel 58 6. Colour specifications for additional panels in the Full LCT 61 7. Mean rank ordering of Luscher colours by preference for colour-deficient groups 86 8. Rank order correlations between preference and photometric measurements 87 9. Rank order correlations between preference and subjective brightness . . . . . 87 10. Mean rank ordering of Luscher colours by pref-erence for three groups 97 11. Rank order correlations between preferences of the three groups 97 12. Group differences in colour preference scored by Luscher method 98 v Table Page 13. Summary of ANOVA 100 14. Colour preferences of four subjects selected from scalogram 101 15. Colour preferences of four subjects over six weeks . . 109 v i LIST OF ILLUSTRATIONS . Figure Page 1. Energy distribution of laboratory-built Illuminant *C 51 2. Reflectance characteristics of the four primary colours in LCTj, 55 3. Reflectance characteristics of the four auxiliary colours in LCT} 56 4. Loci of the eight colours in the three versions of the Liischer Colour Test 59 5. Loci of the colours in the additional panels of the Full LCI ..... .- 62 6. Confusion lines relevant to the eight colours in the Random House version of the Liischer Test 77 7. Confusion lines relevant to the eight colours in the Pocket Books version of the Liischer Test 78 8. Confusion lines relevant to the eight-colour panel in the Full LCT 79 9. Confusion lines relevant to the additional chromatic panels in the Full LCT . . . 80 10. Comparison of luminance values for six colour-deficient subjects vs normal on the four primary colours in LC^ 81 11. Comparison of luminance values for six colour-deficient subjects vs normal on the four auxiliary colours in LCTj 83 v i i Figure Page 12. Distribution of Luscher preferences for colour-deficient and colour-normal subjects . . . . 84 13. Scalogram depiction of colour preferences for three groups 99 14. Scalogram depiction of colour preferences for fifteen persons tested over six weeks 107 v i i i ACKNOWLEDGEMENTS I am grateful to my advisor, Dr. Romuald Lakowski, for his encouragement and assistance given to me throughout the course of this investigation. His suggestions and advice at every stage of its development are greatly appreciated and will not be forgotten. I would also like to thank Drs. Edro I. Signori and Dimitrios Papageorgis for their helpful suggestions and criticism in the last two years. To Dr. Winston Mahabir for his help and assistance in organizing testing schedules at Crease Clinic, Riverview Hospital, special thanks are given. I would also like to express my appreciation to the staff of the Visual Laboratory, in particular to Mr. Keith Oliver. Finally, the author would like to thank his wife, Suzanne, for her invaluable help in editing, and for her patience throughout. ix INTRODUCTION There are many types of investigations which purport to assess or evaluate an individual's unique personality — some methods are ac-corded more scientific basis in fact than others. For example, many people believe that a man's signature or his birthdate are sufficient to supply information of a kind which allows for evaluation, pronoun-cement, and prognostication. Others hold that events in the first six years of a person's l i f e can be analyzed to determine his psychological status fifty years later. S t i l l others maintain that intelligence and ability and interest tests allow for the best evaluation of an individual Since the development of the Rorschach Inkblot Test (Rorschach, 1942), which makes use of chromatic cards as part of it's testing material there has been increasing interest and research into the area of colour response and colour response evaluation. The Liischer Colour Test, published six years later, represents a product of investigation into such matters. In view of the relatively recent emergence of Colour Psychology as an area of experimental psychology, any technique which claims to diagnose personality exclusively on the basis of colour pref-erence should be examined carefully. The Liischer test claims to do just that, however, i t has not yet come under the critical eye of scientific inquiry. Studies using the Liischer test have been conducted, but almost ex-clusively with German speaking populations and with few exceptions, the results have been published in German journals. No experimental work of 2 any significance has been forthcoming from either British or North American sources. This thesis deals in part with presenting what I hope will be regarded as constructive criticism of the test. The Luscher test has been chosen as a focal point of investigation not only because i t illustrates the types of problems encountered generally in the psychol-ogical study of colour, but also because i t professes to aid persons in-volved in clinical and personality research. Part One of the thesis surveys some of the more general problems associated with the test and is intended to provide a framework within which the Luscher test may be evaluated more specifically. Part Two consists of four exploratory studies — the f i r s t , a colorimetric analysis of the test colours. The second study examines the relative applicability of the test to colour-deficient persons. The third looks at the success or not of the test to distinguish between mentally dis-turbed individuals and 'normals'. And the fourth is a longitudinal study of colour preference. PART ONE THE LUSCHER TEST IN PERSPECTIVE I. DESCRIPTION OF THE TEST History The Luscher Colour Test was invented by the Swiss psychologist and colour consultant, Max Luscher, who f i r s t presented the test in 1947 at the International Congress of Psychology in Lausanne, Switzerland. The manual accompanying the test asserts that i t is widely accepted by clinicians and physicians in Europe as a diagnostic aid. There are two editions of the test, one "abridged and one unabridged edition. The unabridged edition, or 'Full Luscher Test' as i t i s called, i s available in German only, and is published by Test-Verlag in Switzer-land (originally in 1948, republished i n 1969). The abridged edition i s available in English, and was translated and edited by Ian Scott. There are two versions of the abridged edition, one published by Random House (1969) , the other by Pocket Books (1971). Some research has been conducted with the Luscher test, published mostly in medical journals. Inspection of the bibliographic entries included in the test manual shows that age, sex, and cultural differences in test performance have been examined. Some of the references are obscure however, and a search for particular studies revealed that"they simply do not exist. No research has yet been published i n Canadian, American, or British journals, in spite of the increasing popularity of the test in these English speaking countries. The exception to this i s an art i c l e by Pickford (1971), which reviews the test and points out certain problems associated with i t , but which contains no experimental supporting data. 5 The Task The Luscher test is a forced choice colour preference task. With the abridged edition, a subject is presented with eight differently coloured cards, arranged in front of him side by side and presented in a random order. He is instructed to indicate which of the cards he likes the best "without trying to relate these colours to anything else". The card so indicated is then removed, and he is then instructed to indicate the colour he likes best from the remaining seven, which is also removed, and placed face down next to the most preferred. This procedure continues until the subject establishes his order of preference for the eight colours. The order of preference is recorded. Following this, the cards are shuffled, and presented to the subject again but in a different random order from the first t r i a l . The subject then repeats the preference task, but is instructed not to "make a conscious effort to remember or duplicate what you had last time". It can be appreciated that this is somewhat difficult to do with only a short intertrial interval, and some investigators, Morgensen and Juel-Nielsen (1962) for example, regard the second t r i a l as "rather super-fluous as most subjects."simply repeat their performance quite closely." In the Full Luscher Test (Full LCT), a subject is presented with the eight colours and is asked to indicate his preferences, however the colours are a l l printed on a single page in two columns of four colours each. Thus there is no opportunity to remove any of the stimuli from a subject's visual field once he has selected them. He remains exposed to a l l eight throughout the task, which brings in an additional memory 6 factor to the task, because a subject on making his sixth choice for example, must remember which five of the eight he had previously selected. In addition to the eight-colour panel i n the F u l l LCT, there are six other panels, each contained on a page of i t s own. On one of these, a subject makes a similar type of selection from neutral colours, arranged in a diagonal cross. On the other panels a subject i s confronted with paired comparison tasks. A neutral grey plastic cover i s provided which allows only two colours to be exposed at a time. The F u l l LCT contains 73 colours, consisting of 25 different hues and shades, and requires 43 selections to be made. The Test Colours The eight colours of the Random House abridged edition (which is the version used in the experimental investigations) are printed i n a matte finish on thick paper. Each card is 11.0 cm high and 7.2 cm wide and is rounded at the corners. They thus resemble playing cards i n size and ih thickness. Each colour is identified by a number printed on the back, and i s named as follows: 1 - Dark-blue, 2 - Blue-green, 3 -Orange-red, 4 - Bright-yellow, 5 - Violet, 6 - Brown, 7 - Black, 0 - Grey. The cards give the appearance of having a slightly rough texture, but are smooth to the touch. The eight colours of the Pocket Books version are similar i n shape but smaller in size, measuring 7.0 x 3.8 cm, and have been printed with a glossy fini s h . The printing i s of poorer quality and some of the cards look 'splotchy'. 7 Colours 1 to 4 are regarded by Liischer. as the Primary Colours, and were chosen because of their particular affective significance, and because they represent "fundamental psychological needs". Colours 5 to 0 are the . Auxiliary Colours, and are included in the test to give some breadth to the choice available, thereby increasing the sig-nificance placed on the position each card holds in the order of preference. Physical characteristics of the colours are not given. It is interesting (and perplexing to some subjects taking the test) that Liischer does not include a 'true red'. This alone makes i t unique among tests which use colour (compare the reds in the Rorschach Inkblot Test, the TAT, the Pfister Colour Pyramid Test, the Stroop Naming Test, for example) and among other colour preference studies. Visual inspection of the Luscher orange-red shows i t to be subjectively more orange than i t is red, and is not at a l l similar to what is usually referred to as a primary red. Furthermore, in his discussion of the psychology of colours, Luscher associates the same affective and symbolic values to the orange-red that other studies suggest are associated to a true red. The colours printed on the pages of the Full LCT are each 2.8 cm square, and are also printed in a matte finish. Texture is similar to that of the Random House version. The first page consists of five neutral colours: three greys of varying brightness, a black and a white. The white square is pasted on the page, not printed like the others. The second page contains the eight colours that make up the abridged edition. The third page consists of the four primary 8 colours arranged into six paired comparisons. There is then a page of four blue hues, arranged into the six possible paired comparisons, f o l -lowed by similarly arranged pages of greens, reds, and yellows. Thus on each of the last four pages a primary colour has been paired with three other similar hues, which in turn are paired with one another. Like the abridged edition, each colour is identified by a number or letter for scoring purposes. No physical characteristics are included in the descriptions of the colours. Rationale Underlying the Test While there would appear to be definite underlying assumptions to the test, Luscher's presentational techniques are not adequate. He does not present his presumed discoveries as a thesis to be defended. It is only from the contextual circumstance of presentation, from random sta-tements, and from his methods of interpretation, that i t is possible to construct his thesis. I suggest that the following paraphrases can be thought to account for the rationale underlying the test: Each colour affects man in a characteristic manner, both psycholo-gically and somatically, and such effects can be investigated. The effects in question should be the same for a l l men and women at a l l ages and in a l l cultures throughout the history of man, i.e., they are normative. Such effects, according to Luscher are the result of the 'Structure' of colour, or its objective meaning. Structures can be discovered and described in adequate formulations, and resultant asser-tions are then testable and confirmable. Blue, for example, has a structure (an effect) of peacefulness, quiet, and tranquility, and this is independent of whether i t is liked or disliked. 9 But each man in accordance with his personality, his character, his feelings, his thoughts, and his motivations responds to a given colour-stimulus in his own way, and in ways significantly different from the responses of other men. Each man prefers some colours, dislikes others, and feels neutral about s t i l l others. This subjective attitude towards a colour is formulated by Luscher to be the 'Function' of that colour. The function is determined by the ordinal position that a colour holds in the order of preference. Each position also has a special significance. Position 1 repre-sents the modus operandi of the individual; position 2 the goal towards which he is striving; positions 3 and 4, the actual state of affairs; positions 5 and 6, indifference or inappropriate behavior; and positions 7 and 8, the inhibited or rejected needs. Thus i f dark-blue is pre-ferred f i r s t , i t signifies a peaceful means of dealing with l i f e . If i t is chosen second, i t means that that individual, is striving for peace-fulness and tranquility in his l i f e , and so on. Hence the personal response of each man — the colour Function — towards what has been discovered about the effects of colour and their effects upon a l l men — the colour Structure — allows for an interpretation of a man's personality, feelings, motivations, conflicts, etc. In fact, the interpretation of personality is purely manipulative, once the. Luscher code is understood. No further 'insight' is required Scoring and Interpretation Scoring of test results is relatively straightforward. The order of preference for the eight colors (abridged edition) is divided into four pairs: the two most liked colours are called 'the + functions', signi-fying the desired objectives; the next two most liked colours are called 'the x functions', signifying the existing situation; the following two colours are called 'the = functions', signifying indifferences; the two least preferred are called 'the - functions', signifying rejections. Tables are then consulted to obtain interpretations for a l l pos-sible two colour combinations for each of the four functions. An inter-pretation for a fifth pair, consisting of the most and the least preferred colours, is also available. Each interpretation consists of brief (2 to 10 15 lines) 'personality descriptions', constructed for the most part with psychological jargon, and in some instances, with physiological termi-nology. For example, the following description can be found in the tables describing an individual who preferred violet seventh and orange-red eighth: Physiological interpretation: Stress resulting from frustration in his attempts to achieve security and understanding. Nervous strength can become seriously depleted and there is a possibility of cardiac trouble. Psychological interpretation: Is responsive to out-side stimuli and wants to experience everything in-tensely, but is finding the existing situation ex-tremely frustrating. Needs sympathetic understanding and a sense of security. Distressed by his apparent powerlessness to achieve his goals. (Scott edition, 1969, p. 147). In scoring preferences Luscher regards the second t r i a l as more valid, "especially in doubtful cases'. He therefore advocates that i t be used more frequently than the first t r i a l for making inferences about personality. Scoring procedures are also available for measuring the intensity of psychological and physiological problems resulting from stresses, anxieties, conflicts, and compensations. Basically, the rule of thumb is as follows: the more disliked the four primary colours are, the greater the intensity, and similarly, the more liked the four auxiliary colours are, the greater the intensity. A score of 1 is allotted; if one of the four primary colours is in the sixth position in the order of preference, a score of 2 i f a primary is in the seventh position, and 3 i f one is in the eighth or last 11 position. In addition, a score of 1 is also given i f either black, grey, or brown are third most preferred, a score of 2 if one. of these colours is second most preferred, and 3 i f one of them is most preferred. According to Luscher, a normal, well adjusted, stable, healthy individual is one who prefers a l l four primaries in at least the first five positions. Preference of the primaries beyond fifth position is indicative of some degree of abnormality. Scores can range from 0 to 12. The following preferential order of colours would be given a score of 0: blue, green, red, violet, yellow, black, grey, brown, while a score of 12 would result i f the colours were liked in the reverse order. In this paper the scoring procedures for measuring what Luscher refers to as the "intensity of deficits" (degree of anxiety, severity of conflicts, intensity of compensations) are used as a means of judging the 'degree of psychopathy'in the subject. Thus a score of 0 equals no psychopathy, while a score of 12 would indicate serious psychopathy. Personality interpretation in the Full LCT follows similar scoring procedures as in the abridged edition, but is more complicated because there are more selections to be made and because the methods of selection do not a l l involve a straightforward rank ordering of the colours. Brief personality descriptions are given for the results of colour preference on each of the pages in the Full LCT. Two points should be noted regarding scoring and interpretation of test results. The first is that the method of scoring conflicts with the response pattern requested of the subject. In performing the task, subjects are asked to avoid reflecting on colour combinations and rather to select 12 colours o n e a t a t i m e . However, when the results are scored the colours are grouped in p a i r s and interpreted as such. Luscher does not give reasons as to why this particular method of sorting the data and scoring was chosen. Secondly, there are 56 personality descriptions offered in each of the four functions (the +, x, =, and - functions), yet they are not a l l different. Preferring violet first and green second for example, yields the same personality description as does preferring green first and violet second. Similarly, blue followed by black in the fifth and sixth positions 'means the same' as does black followed by blue in those two positions. Inspection of the interpretation tables reveals many other examples. These findings plainly illustrate that in certain in-stances the exact order of colour preference does not matter at a l l . There is no internal logic in the interpretation tables which explains this, and Luscher offers no explanation for these obvious inconsistencies. /3 II. COLOUR PREFERENCE The Luscher test shares some similarities with the Colour Pyramid Test (Pfister, 1950) in that they both use colour preference data to make inferences about personality. Pfister, however, uses this data in con-junction with other test results and so does not rely solely upon colour preference data as the basis for his conclusions. Similarly, colour responses on the Rorschach Inkblot Test (1942) are interpreted in the light of other test responses. Luscher's test, then, is unique in that personality assessment is based exclusively upon colour preference. Some of the assumptions which underly Liischer's test however, have been the subject matter of much investigation: that colours have an intrinsic meaning, for example, or that colour preference can be interpreted and used as a diagnostic tool. In the following sections, the Luscher test is examined in relation to research in this area and comparisons between his assumptions and the findings of other researchers will be drawn such that his work may be considered in some perspective. Investigations of Colour Preference Colour preference-studies have a long history in experimental psychology. Answers have been sought to questions such as whether or not there exists a universal order of colour preference, and, i f such an order indeed exists, whether or not i t is dependent upon age and/or sex variables. Attempts have also been made to determine what factors affect or effect colour preference. The first studies,(Conn, 1894, 1900; Major, 1895; Preyer, 1897; 13 14 Jastrow, 1897) produced conflicting results concerning an order of preference. They also disagreed as to whether or not hue, saturation, or brightness was the predominant factor determining preference. In .subsequent years, many similar studies were conducted which produced results which were also not altogether consistent with one another. In an attempt to establish once and for a l l whether there was a consistent preferential order of colours, Eysenck (1941) examined the results of 29 studies which reflected a total of 21,060 subjects' responses. By obtaining an average ranking for a colour from each study, the most preferred was determined to be blue, followed by red, green, and violet. Similar preferences were found for males and females, but the former preferred orange over yellow, the latter yellow over orange. Furthermore, he determined that there were no r a c i a l differences in colour preference. His findings must be viewed cautiously however, because the studies he examined used different coloured stimuli (wool, velvets, coloured lights, coloured crayons, coloured papers), and because he focused only upon the six most highly saturated colours that a l l the investigators had used. Even from a simple mathematical point of view, Eysenck's findings can be shown to be misleading, as was illustrated in an earlier study (Dorcus, 1926). Sex and Age Factors Eysenck's results revealed a sex difference in colour preference. Sex differences have been reviewed by Norman and Scott (1952) who report 15 that with the exception of Jastrow (1897) and Warner (1949) , most investigators found sex to be of l i t t l e importance as a preference determinant. Since their review however, Choungarian (1967, 1968), Child et a l . ? (1968), and Spiegel and Keith-Spiegel (1971) have a l l reported sex differences in colour preference, and Guilford (1959) has suggested that there are differences i n the relative strengths of preference between men and women. The Warner study (1949) is interesting i n relation to the Luscher test because i t dealt with preferences of psychotic populations. Warner found significant sex differences for most hues, saturations and bright-nesses. On the Luscher test however, sex is not considered as a variable. Colour preferences are interpreted the same for a l l individuals. As Luscher puts i t the only limitation which can be placed on the general applicability of the test is the neces-sity to communicate to the person being tested; i f he can understand what is required of him... and state his preferences, then the test applies to him (1969, p. 51). If there are indeed sex differences in colour preference, as the more recent studies are indicating, the LUscher test would then be selectively distorting personality inferences. The Luscher test also does not take into consideration the age variable. There i s evidence to show that preferences do change with age, though the reported changes are not consistent from study to study (Beebe-Center, 1932; Normal and Scott, 1952; Burnham et a l . , 1963, Child et a l . , 1968). Since there i s no mention in the Luscher manual of whether 16 the test should be used with adults only, one must assume that children can also be administered the test. The quotation above implies just. that. However, i f there are age differences in the preference of Luscher colours, this further complicates the interpretation of personality. Inspection of the personality interpretations does suggest that they are better suited to a description of adults rather than to children, as can be seen for example, when a testee places brown and yellow in the f i f t h and sixth position respectively, and discovers that he i s : very exacting in the standards he applies to his choice of a partner and seeking a rather unreal-i s t i c perfection i n his sex l i f e (1969, p. 124). Hue, Saturation, and Brightness Many of the studies i n colour preference have been c r i t i c i z e d because of poor s t a t i s t i c a l treatment of data, and, because hue, satura-tion, and brightness variables were not given proper consideration. These failings have been redeemed most notably by Guilford (1934, 1940, 1959) and by Granger (1955) who have very systematically analyzed the importance of the three component variables of colour in determining preferences. Guilford has established a system of isohedonic charts where the affective value of a colour (equated to relative preference in Guilford's system) can be predicted on the basis of known variable values. Granger's findings show that preferential order for any one variable remains invariant with changes in the other two variables i n the Munsell colour solid. The Luscher test does not provide any information about the psychological colour variables. The manual states that "each of the 17 eight colours has been carefully chosen because of its particular psychological meaning" (p. 51). Klar (1961) states that the colours on the Full LCT were deliberately chosen "according to psychological criteria from more than 4,500 (colours) when he (Luscher) developed his test" (p. 19). There is no evidence supporting these statements however", no criteria are given. Specification of hue, saturation, and brightness values will be provided later in the paper. Methods of Determining Preferences Studies of colour preference have employed varying experimental methods. Subjects have been asked to indicate preferences by rating colours on numerical scales, by choosing a colour in a paired comparison task, by choosing a colour in painting, by rank ordering a set of colours, and by presenting an example of a colour selected for apparel, to mention only some of the methods. What part the method may play in determining preference has not yet been properly established (Hass, 1963). There is a considerable dif-ference in task situations however, between rank ordering a set of colours and making judgements about single colours one at a time. The former must take into consideration the interaction among the colours, and is con-ceivably more affected by the Gestalt of the situation than is the latter. The interaction of colours in combination as well as the preference of colour pairs had been studied quite extensively prior to the publication 18 of the Liischer test in 1948 (Chevreul, 1890; Bullough, 1910; Geissler, 1917; Washburn et a l . , 1921; Guilford, 1931; Allen and Guilford, 1936; Moon and Spencer, 1944) but the findings of these studies are not given any consideration in the design of the Luscher test. A more recent study (Haas, 1963) has shown that preferences did not change s i g n i f i -cantly when three different methods were used — coloured t i l e s pre-ference, free painting technique, and questionnaire — but this finding reflects the responses of only 20 undergraduate volunteers, and therefore needs repetition before the results can be generalized. In the Luscher test a subject is exposed to a l l eight colours at once and is asked to point to the one he likes most. This is then removed from his visual f i e l d , and he continues to indicate preferences u n t i l a rank order is established. There is no possibility for tied ranks i n this method, despite the scoring procedures previously men-tioned which suggest tied ranks (see page 11). LUscher does not specify why this particular method of rank ordering was chosen. He seems to assume that once the cards are dis-played before the person to be analyzed, that person w i l l view each card independent of the surround. However, whatever the arrangement of the cards, the person cannot help but see them as a Gestalt, with each colour being perceived and apperceived i n relation to a l l the other colours, and hence modified by whatever interrelation the person selects from this Gestalt. The very order of arrangement of the colour cards in the visual f i e l d — which changes each time the test i s ad-ministered — becomes a disturbing factor not taken into account 19 experimentally. In discussing the method used to establish colour preference in the Luscher test, a study by Rickers-Ovsiankina et al. (1963) seems relevant. In this study, performance on the Rorschach was compared with performance on,the Knapp Tartan Test (Knapp, 1962). (On the Tartan Test subjects rank order by preference 30 photographic reproductions of Scottish tartans). On both tests, responses were grouped according to whether they f e l l into the 'warm1 or the 'cool' colour categories. The results showed that subjects who frequently responded to warm colours on the Rorschach tended to reject the warm colours on the Tartan Test. Sim-ilarly, subjects who avoided warm colours on the Rorschach preferred these colours on the Tartan Test. The findings were interpreted according to the different task situations required on the test. On the Tartan Test a subject is given as much time as he wishes to think about his choices. The test essen-tially imposes a situation involving calibrated judgements which re-quire "careful deliberation and reflection". The Rorschach does not involve 'judgements' of the same kind. A subject presumably does not have time to reflect'on his responses to colour. They occur.(at least are supposed to occur) more or less automatically, or else they are scored as colour shock responses i f they do not. Any colour preferences which may come out of a Rorschach protocol are not the result of careful deliberation. Rickers-Ovsiankina's interpretation of the differences in response to the warm colours is as follows: the person who knows he has reacted 20 spontaneously to the impact of red when i t 'hits him by surprise' in the Rorschach test, will most probably defend himself against that same color (placing i t last) when the task situation allows him enough time to think about his reaction. The subject thus retains on both tasks his basic receptivity towards the environment but modifies the behavioral manifestations of this passivity in accordance with the forces prevailing in the overall (test), situation (p. 465). The point to note here regarding the Liischer test is that the nature of the task situation can decidedly influence the effect that a colour can have. Colour will n o t be more evocative as a pref-erence stimulus in one method than i t will be in another, but its response-evoking potential will change depending on the nature of the task. The method for establishing colour preference on the Liischer test is quite similar to that of the Tartan Test, in that subjects are given as much time as they wish to reflect on the colours before making their selections. Thus the test allows for controlled, thoughtful responses, although i t does not insist upon such responses. Rickers-Ovsiankina would argue that the Luscher test does not really allow for the fu l l affective impact of colour to occur, that i t encourages an 'intellectual' mode of response rather than an 'emotional' one. Liischer would counterargue that spontaneous 'Rorschach-type' reactions to colour are not necessary for purposes of personality description and diagnosis, that such response would in fact tend to lose vital information. In any event, the importance of the method used to e s t a b l i s h lour preferences requires more research. I I I . DETERMINANTS OF COLOUR PREFERENCE The Meaning of Colours Most researchers hold that hue, saturation, and brightness are the important factors in determining the preference of a colour. Luscher maintains that the subjective reaction towards a colour, as measured relatively by its preferential position, is determined by the objective meaning — the Structure — of that colour. Various colour meanings are provided in the manual. An inspection of these shows that Luscher has collected whatever has been said about the effects of colours, and people's responses to them, in addition to including his own intuitive feelings. He does not adhere to any pres-entational method; his presentation is unsystematic. In this sense, his presentation is naive, and remains so throughout. Table 1 presents the meanings and associations of the eight colours as given by Luscher. No attempt has been made to reorganize his materials. Note that only the four primary colours have been assigned a Basic Repres-entation (equal to the Structure). Perusal of the manual indicates that the following might be offered as basic representations for the auxil-iary colours, although they ate not stated as such; violet represents 'identification', brown 'sensation' or 'sensuousness', black 'negation', and grey 'neutrality'. The entries in the third column are labelled by Luscher as Affective Aspects and in some instances (the underlined words) as 'emotional content'. Some of the entries, nobably 'self-assertion' for blue-green and 'originality' for bright yellow, are not usually included by researchers 22 TABLE 1 MEANINGS AND ASSOCIATIONS OF THE EIGHT COLOURS IN THE LUSCHER TEST to Colour Basic Other Affective Biological Symbolic Sensory Other Representation Representations Aspects Effects/ • Perception Meanings Organs Dark- Depth of Passive Tenderness Reduces Loyalty, Calm Sweetness Balance, Truth blue Feeling Incorporative Tranquility blood water Trust, Love (#1) Environment- Contentment pressure, Quiet Dedication. effected Love and pulse, res- temperament Surrender Sensitive affection piration. Femininity Devotion Perceptive Pacified CNS Illumination Belonging Unifying Induces in the manu- Fulfilment relaxation script Tenderness recuperation, Meditation Harmony tranquility reunion with The Skin* Gaea, Earth Mother, Time-lessness of Eternity, Represents lasting values, Perpetuates past Blue- Elasticity Passive Pride Represents Majestic Astringence Perseverance green of Will Defensive Persistence elastic sequoia Tenacity (#2) Environment- Self- tension Austere and Firmness. effecting assertion Smooth autocratic Constancy Retentive Obstinacy involuntary temperament . Resistance Possessive Self-esteem muscles* Tension in to change Immutable bow-string Table 1 continued on next page Colour Basic Other Affective Representation Representations Aspects Orange-red (#3) - Force of will Bright yellow (#4) Spontaneity Active Offensive-aggressive Environment-effecting Locomotor Competitive Operative Active Projective Environment-effected Expansive Aspiring Investigatory Desire Excitability Domination Sexuality Hopeful volatility Variability Expectancy Originality Exhilaration Biological Effects/ Organs Expenditure of energy Increases blood pres-sure, pulse, respiration Expresses nervous/ glandular activity Striated (voluntary) muscles* Sympathetic nervous system* Organs of reproduction Increases blood pres-sure, pulse, respiration, but less stable than red Induces re-laxation, dilation Sympathetic and parasym-pathetic nervous system* Symbolic Blood of conquest Pentacostal flame Sanguine temperament Masculinity Welcoming warmth of sunlight Halo /around Holy Grail Cheerful Sensory Perception Appetite Piquancy Other Meanings Impulse towards a l l forms of vitality, Power Sport, Struggle Competition Eroticism Enterprising productivity Lightness Cheerfulness Change Release from burdens and harassment to JN Table 1 continued on next page Colour Basic Other Affective Representation Representations Aspects Violet * Idenfication' C#5) Brown 'Sensation'/ (#6) 'Sensuousness' Black 'Negation' (#7) Grey (#8) 'Neutrality' Biological Effects/ Organs Passively receptive and sensory vitality Symbolic Mystic union Fusion between subject and object Roots-the hearth, the home Nothingness Extinction Renunciation Ultimate surrender Relinquishment Neither subject nor object Demilitarized zone Berlin Wall Iron Curtain Sensory Other Perception Meanings Enchantment Magical state Bodily sensations Non-involvement 26 as being 'affective 1 in nature, but rather as connotative. It is not known therefore what Luscher means by affect, what i t includes and what it does not include. Entries in the Biological Effects/Organs column are of two kinds. One consists of the effects of colour on bodily systems, blue reducing blood pressure for example, and the other (marked with *) has to do with specific systems that Luscher associates to colours, green with the smooth involuntary muscles for example. No evidence of experimentation is cited to substantiate either kind of association. The symbolic meanings assigned to the eight colours have various associations: political (Berlin Wall - grey), religious (Pentecost -red), mythological (Gaea - blue), linguistic (Sanskrit - blue), vegetative (sequoia - green), and intuitive (masculinity - red). It is not known what Luscher means by Sensory Perception, nor how he arrived at the meanings in this column. Luscher also gives various other meanings to some of the colours, and these have been listed and appropriately labelled in an additional column in Table 1. He does not specify whether these are affective, physiological, symbolic or connota-tive in nature. They are presented merely as qualitative meanings of a colour. For example, in reference to meanings of the colour blue, Luscher states that It is truth and trust, love and dedication, sur-render and devotion. Blue is the timelessness of eternity, representing tradition and lasting values, and so tends to perpetuate the past (1960, p. 56). Table 1 has been constructed to allow for easy comparisons between Liischer colour meanings and those that have been assigned from other studies. 27 It should be noted that the term 'meaning' is used here in a generalized sense, and is meant to include both the attributes and associations given to a colour, as well as the effects of and responses to a colour. In some cases i t is difficult to ascertain whether the meaning assigned to a colour refers to a stimulus attribute or to a response parameter, or to both. Red, for example, is assigned a meaning of 'offensive-aggressive' yet i t is not clear whether this is meant to refer to a property of the stimulus red, or whether i t refers to the response that is induced when perceiving the colour red. This is not exclusive to the Luscher test however, and emerges frequently in this kind of colour research. Synaesthetic Attributes One dimension of meaning commonly given to colours has to do with the non-chromatic sensations produced by colours — sensations of temperature, weight, size, and distance — which are referred to as synaesthesias. The American Inter-Society Council (Hanes, 1961) in-vestigated the phenomenon of synaesthesias and has published the "rela-tively unquestioned facts" as follows: Colours can convey the impres-sion of temperature (red, orange,yellow are the warm colours; green, blue,violet the cool), with hue the determining variable, as opposed to brightness or saturation. Maximum warmth was attributed to orange, but coolness was found to be distributed through the greens and blues. Apparent heaviness is predominantly determined by brightness, with dark colours judged heavier than lighter colours. Wright's review 28 (1962) of apparent weight and temperature confirms these findings. For apparent distance, red-appearing surfaces were found to 'advance', blue-appearing surfaces to 'recede', with saturation the major determinant of distance, although brightness was also found to play a role. There is conflicting evidence concerning apparent distance effects of colour (Payne, 1964; Ball, 1965). Finally, concerning apparent size, lighter objects appear to be larger than darker objects of the same dimensions. A high positive correlation between apparent size and the luminance factor was obtained. It can be concluded reasonably safely, that synaesthesic reactions do occur with some coloured stimuli, and that some of the sensations appear to follow regularities in relation to hue, saturation, and bright-ness values. It is to be noted that any, or a l l of such synaesthesias in combination, can influence colour preference. It is surprising that Luscher does not specifically refer to any synaesthesic attributes when he assigns meaning to colours. Apparent temperature is implied however, when 'calm water' is given as a symbolic meaning for blue, and 'Pentecostal flame' to red. - Connotative Meaning Studies dealing with the connotative meanings or attributes of colour generally employ the semantic differential approach. Subjects judge coloured stimuli on a number of bipolar scales reasoned to cor-respond to possible dimensions of meaning. These are then factor analyzed to determine the principle dimensions of meanings (Osgood et al., 1957). An assumption underlying these studies is that colour is considered as 29 a form of communication, where the message is the colour itself. The response to a colour will therefore depend upon the connotative meaning that the message holds for that individual. Wright and Rainwater (1962) have reported six principle dimensions of connotative meanings for colour: calmness, forcefulness, happiness, warmth, elegance, and showiness. It is of interest that the first three of these correspond quite closely to some of the meanings given to the Luscher colours blue, red, and yellow, respectively (see Table 1). The authors state however, that saturation contributes most to the connota-tions and that hue has almost no effect. In a study by Kansaku (1963), brightness was found to be the important factor determining connotative 2 meaning. On the Luscher test differential effects of these variables are not even considered. Some studies have investigated the connotative meanings of single colours. Williams (1964), using colour names as stimuli, showed sig-nificant differences in connotative meaning. The word white was found to be good, active and weak, for example, the word black to be bad, passive, and strong, the word red to be good, active and strong. In a later study (Williams and Foley, 1968), i t was reported that connotative meanings for the names of colours (the 'signs') were not different from the meanings given the actual colours (the 'significates') which repre-sented them. The advantage to experimenting with the semantic differential ap-proach is that colours are judged on more than one scale, thus giving a comprehensive and varied view of possible meanings (which are connotative 30 in nature). The procedure can be carried too far however, for instance when a colour has to be judged on as many as 24 different scales, which requires considerable concentration on the part of the one judging, and tends to overshadow any 'spontaneous' response to the colours in the process. Furthermore, after a prolonged period of judging, scales would tend to lose their individual meaning and significance to the one judging. Moreover, the nomenclature in certain semantic differential scales em-ployed — smart-boorish, chic-rusty, clear-muddy — might appear com-pletely foreign and/or incomprehensible, thus serving only to complicate and confuse the assignment of meaning. If colour is to be regarded as a form of communication, as the transmission of a message which is decoded into, among other things, connotative meaning, then i t might be advisable to have a subject dis-cuss such meanings in his own language at his own level of understandings rather than assign him a previously devised set of meanings (scales) to which he must apply his personal experience with colour. Some of the Swedish researchers have already taken steps in this direction, in par-ticular the phenomenological approach to colour responding advocated by Hesselgren (1965). ~ Luscher takes an extreme view of his delegation of meaning to colours. There is no evidence that any subjects ever made judgements about any colours. This is understandable within the context of his assumptions about colour, namely that experimentation is not necessary because each colour already has an intrinsic meaning. As he puts i t This meaning is of universal significance and is the same the world over, to young and old alike, 31 to men and women, to the educated and the backward, to the 'civilized' and the 'uncivilized' (Scott, 1969, p. 51). The research into the notion of connotative colour meaning could be viewed as part of a larger approach to the concept of colour responding, namely, that colour affects an overall cognitive response process. Cog-nitive operations are reasoned to be the main characteristics of colour responses on the ground that colour is almost always associated with familiar, and therefore cognitively meaningful objects (Committee on Colorimetry, 1953; Norman and Scott, 1952). There is no argument that colour has object-associations, but whether'familiar associations reflect a 'cognitive' type of response any more than they reflect an 'affective' type of response is irresol-vable. The argument is partly semantic in nature, anyway. Furthermore, no one has attempted to differentiate the two types of response exper-imentally. As was mentioned previously, some of the entries in the Affective Aspects column of Table 1 (e.g., Self-assertion, Originality) do not reflect what is usually referred to as affect. Similarly, some of the entries in the Other Meanings column — balance, constancy, impulse, for example — : "cannot properly be classed as either affective or cognitive in nature. The argument really centers on the issue of whether colour has meaning, and most of the research directed towards this issue has dealt with what is called affective meaning. Affective Meanings and Values The 'affective meaning* of colour is a term frequently employed, seldom defined, having variable meanings, and measured by different 32 procedures. Inspection of the studies dealing with colour and affect lead one to suggest the following as an understanding of the term: the affective meaning o f a c o l o u r refers to the feeling that is elicited i n t h e s u b j e c t as he perceives that colour (in an experimental setting). - If enough people respond to a given colour in a similar manner, by reporting their feelings, then the nature of the response is said to characterize the affective meaning of that colour, which in accordance with the usual meaning of the term affect, is more emotional in nature than anything else-. The term affective meaning is frequently interchanged with affective value, both of which then can be likened to the 'elicited motive capabilities of a colour'. For purposes of clarification, studies dealing with colour and affect can be differentiated into two kinds. One kind seeks an answer to the question of whether or not there is an o v e r a l l relationship between colour and an affective response. Such studies have generally chosen the Rorschach Inkblot Test as a focal point for investigation. The second kind seeks s p e c i f i c relationships between a colour and its particular affective meaning. Rorschach Research on Colour and Affect Rorschach initiall y postulated a relationship between colour and the dynamics of affectivity. Specifically, responses determined solely by colour and those determined by colour and form, with form of less importance, were said to reflect a more egocentric, uncontrolled, and often impulsive affectivity. Responses determined by form and colour, with colour of less importance, were said to reflect controlled, socialized, 33 and adaptive affectivity. The phenomenon of colour shock, represented by a prolonged reaction time to the coloured cards and/or a diminished number of responses to these cards, expressed according to Rorschach, neurotic repression of emotion. It is not known for certain whether Luscher had any contact either with Rorschach or with his test, but i t is li k e l y that he was at least familiar with the latter, i n view of the fact that Luscher was a European psychologist, a colour consultant, and that the Rorschach Ink-blot Test was published six years prior to the^Tull LCT. It is possible that the Liischer test i s an offshoot from some of Rorschach's ideas, a l -though there i s no evidence to substantiate this. Generally speaking, those who use the Rorschach test in c l i n i c a l settings have tended to accept Rorschach's interpretive meaning of the colour response. The colour-affect relationship has become common know-ledge and appears in a l l Rorschach textbooks as one of the basic hypo-theses. Klopfer and Davidson (1962), for example, assert that the five colored cards are emotionally challenging, and the subject's reactions to them provide a great deal of information as to how this emotional chal-lenge is met (p. 137). The most notable proponents of the Rorschach colour-affect relation-ship are Schachtel (1943), Rickers-Ovsiankina (1943, 1960), and Shapiro (1960). Their defense of the relationship l i e s mainly on theorteical grounds, and is supported by ' c l i n i c a l - i n t u i t i v e ' evidence. Experimental investigations have shed doubt on this relationship however. Some have shown that the presence of colour on the cards is not an important stimulus dimension, and that when i t is removed, the 34 phenomenon of colour shock s t i l l occurs (Lazarus, 1949; Sappenfield and Buker, 1949; Dubrovner et al., 1950; Allen, 1951; Wallen, 1948). A review of the problem by Cerbus and Nichols (1963) showed that there is almost no evidence supporting an association between colour responsi-veness and the personality variables of impulsivity, emotionality, and anxiety. Can the affective value of colour on the Rorschach be examined experimentally? Norman and Scott (1952), taking the side of the Rorschach proponents, argue that in order to treat the data statistically, the ex-perimentalists "have substituted measurable elements for responses which the test assumes to be in gestalt". The present author agrees. Taking colour 'out' of the Rorschach cards and holding constant other variables denies a l l of the interactional effects that the dimension of colour has with the other dimensions of the test, notably the form dimension. Can the affective value of colour on the Rorschach be validated by a clinical-theoretical approach? From his experience the clinician argues that a colour-determined response gives more information about the affective state of the individual than does for example, a form-deter-mined response. Similarly, a form-determined response is more inform-ative as to the cognitive, organizing, integrating state of the indivi-dual. What the clinician does is to continually compare the respective meanings of the two types of response. The significance of colour will depend upon the significance of form, and vice versa, and this will vary from individual to individual. When there is a 'healthy combination' of form- and colour-determined responses, this is taken as an indication of normality. 35 This leads one to suggest that the role of colour on the test i s only important when i t is either completely denied as a stimulus pro-perty, or i s overemphasized to the exclusion of other stimulus properties. In either case, whether colour responses have to do exclusively with the dynamics of affectiv i t y (as Rorschach postulated), or whether they pre-dominate in different kinds of impairments, affective disorders being only a special case (as Shapiro postulated) i s s t i l l open to question. A point to note from the Rorschach studies is that the interactional effects of colour and form must also be considered in colour preference testing. Colour i s a l w a y s combined with form. On the Luscher test, colour appears on curved-corner, rectangular shaped cards and i t remains to be seen whether preferences are independent of the shape of the cards. Another point to note i s that the Rorschach colours not only inter-act with form but also interact among themselves, presumably reinforcing (either positively or negatively) the affective meanings of each other. Such interaction i s also important in colour preference testing, and i t remains to be seen whether preferences on the Luscher test are independent of the interactional relationship between the colours when they are f i r s t presented to a subject i n a given order. Affective Meanings of Individual Colours The affective meanings of individual colours have been considered in various review (Norman and Scott, 1952; Schaie and Heiss, 1964; Ba l l , 1965; Zausznica, 1969). Tables 2 and 3 repeat the summary tabulations for the Schaie and Heiss and the Zausznica reviews, respectively. TABLE 2 AFFECTIVE VALUES FOUND TO BE ASCRIBED TO VARIOUS COLOURS IJO O N Study Red Orange Yellow Hevner happiness (1935) restlessness, agitation Lewinski stimulating, stimulating, stimulating, (1938) hot Karkowski exciting (1938) Alschuler affection (1943) and love, or aggression and hate Schachtel striking, (1943) exciting, explosive Bricks (1944) Kouwer (1949) Napoli (1951) Wexner (1954) h o s t i l i t y , aggression active, Intense exciting, stimulating hot, un-pleasant exciting tempered emotions warmth, delight merry most un-pleasant exciting Green Blue Purple dignity, sadness tenderness most pleas- depressing and, cool leisurely leisurely vigorous controlled drives to-emotion- ward control a l i t y Black White Brown serene, cheerfulness, envy h o s t i l i t y , aggression sad solemn sad intense, anx-iet i e s , fears disturbed, cheerful, distressed, j o v i a l , upset joyful youthful social controlled security, emotion- drive a l i t y secure, comfort-able, tender, soothing depression disagree- sad, vague, pure spir- dis-able, sad disagree- ited agreeable deep but evasion, optimistic, fear, depression depression dignified, stately powerful, strong, masterful Table 2 continued on next page Study Red Orange Yellow Green Blue Purple Black White Brown Murray (1957 exciting, stimulating, defiant, contrary, hostile, powerful, strong, masterful Hofstatter stong, (1958) Schaie (1961) active, f u l l protective, exciting, powerful, strong, masterful, defending, stimulating cheerful, jovial, joyful happy exciting, stimulating, exciting, stimulating, cheerful, j ovial, joyful, pleasant secure, comfort-able, calm, peaceful, serene young, i l l , fresh tender, soothing strong, f u l l , great, deep pleasant, secure, comfort-able, tender, soothing despondent, dejected, melancholy, unhappy f u l l dignified, stately deep, strong, great, old, sad • distressed, disturbed, upset, defiant, contrary, hostile, dignified, stately, powerful, strong, masterful, despondent, dejected, melancholy, unhappy empty f u l l tender, soothing secure, comfortable (from Schaie and Heiss, 1964) TABLE 3.. THE DIFFERENT VALUES FOUND TO BE ASCRIBED TO VARIOUS COLOURS ASSOCIATION COLOUR FEELING OBJECT RED ORANGE YELLOW love crime glowing forth warmth lig h t daylight joy/gladness GREEN l i f e BLUE spatiousness t r a v e l l i n g VIOLET mourning dignity [PURPLE secrecy ceremonious IWHITE BLACK brightness c l a r i t y marriage hygiene h e l l mystery f i r s blood f i r e setting sun an orange l i g h t sunshine nature vegetation heaven ocean water flowers bishop amethyst flowers flowers snow hospital night death PSYCHOLOGICAL warmth dynamicism i r r i t a t i o n exciting/provoking s h i n i n g / b r i l l i a n t passionate/ardent joy/cheerfulness . l i v e l y peace/quiet. freshness restfulness s p i r i t u a l balance brightness freshness lightness transparency peace/quiet beloved freshness peace/quiet freshness delicacy/softness brightness/clarity cleanliness/purity grief/sorrow EFFECT PHYSIOLOGICAL thought-provoking warmth-penetrating exciting/provoking emotional favours digestion v i s u a l l y exciting pacifying calmness/quiet soothing/relieving calming/reassuring bringing r e l i e f calming/reassuring (provoking/ exciting calming/reassuring (provoking/ exciting restfulness PHYSICAL very eye-catching same as above eye-catching sane as above not at a l l obtrusive weakly eye-catching d i f f i c u l t to qualify v i s u a l l y AFFECTIVE CHARACTERISTIC fullness of l i v i n g trouble disturbance SYTEOLISii RELIGIOUS SECULAR joy gaiety patience/ forbearance trouble/anxiety quiet/ peacefulness grief-sorrow sadness melancholy grace love excitement 'of the doctrines' power . truth f a i t h regeneration j udgeinent int e l l i g e n c e immortalitv repentance hope appraisal/esteen dignity sa t i s f a c t i o n / holiness pleasure brightness/purity neatness cleanliness darkness imagination/ fancy (From Zausznica,. 1969; translated by P. Melhuish) piety innonence/ v i r g i n i t y fate/destiny.' pity/charity eternally r e s t f u l / imrrtality penance/ atonement love excitement DEGP.EE OP PREFERENCE +5 treason corruption hope ( e v i l eye) wisdom science +45 (maximum) repentance dignity innocence cleanliness health +30 +5 +30 death -20 (minimum) 39 A look at the research dealing with affective meanings reveals that there have been at least eight methods employed to investigate such meanings. A suggested classification of methods appears below; i t is not intended to be exhaustive. 1) Method of Verbal Description Subjects are confronted with a colour and asked to give verbal judgements about the colour (Bullough, 1908), or, to describe their feelings experienced with the colour (Lewinski, 1938). 2) Method of Adjective Checklist Subjects are confronted with a colour and asked to check adjectives which they consider to be an appropriate description of colour (Hevner, 1935). Also included in this method should be those studies i n which subjects check paired adjectives representing 'mood-tones' that are said to 'go with' certain colours (Wexner, 1954; Murray and Deabler, 1957; Schaie, 1966). 3) Method of Scale Ratings Subjects are confronted with a colour and asked to rate i t s relative pleasantness on a scale from 1 to 7, representing the two extremes of pleasantness and unpleasantness (Guilford, 1934, 1940, 1959). The semantic dif f e r e n t i a l studies of colour meaning should also be included with this method, although they are really, a combination of 2) and 3). 4) Method of Mood Associations via Music Subjects associate colours to various selections of music, which have previously been related to various moods. The colours and moods are then related (Karwoski and Odbert, 1938; Odbert et a l . , 1942). 40 5) Painting Analysis The colours that children use in paintings are examined in relation to the children's stages in development and to their personality conflicts (Alschuler and Hattiwick, 1943; Bricks, 1944). The colours are then as-sociated with related feelings, moods, emotional states, etc. 6) Association to Emotional Situations Subjects are either confronted with stimuli (stories) representing situations that have obvious emotional overtones and are asked to asso-ciate a colour to the situation (Lawler and Lawler, 1965), or, they are confronted with colours (differently coloured figures of men) and asked to choose which colour is associated with a particular situation (Crane and Levy, 1962). 7) Method of Posthypnotic Suggestion A subject is hypnotized and instructed while under hypnosis that everything he sees is tinted with a given colour. He is then asked to describe his feelings, and these posthypnotic reports are then related to the colour (Aaronson, 1964). 8) Method of Cli n i c a l - h i s t o r i c a l - i n t u i t i v e Association Colours are given- affective meanings on the basis of c l i n i c a l experience, on the basis of experience with colour and hence intuition, on the basis of hi s t o r i c a l surveys of how others like Goethe (1810), E l l i s (1900), Chevreul (1890) and famous artists view colour, and on the basis of standard accepted symbolic associations. Schachtel (1943) and Birren (1950) use such methods. Luscher (1948) also takes this approach. 41 The different methods employed to study affective meanings make comparisons between studies difficult. Associating colours to music utilizes quite different criteria of meaning than does the analysis of finger painting. Differences in the coloured stimuli used in these studies also make comparisons difficult. Names of colours, coloured lights, coloured cards (of Munsell, Ostwald, Milton-Bradley variety), coloured Tartans, and coloured designs have been employed as stimuli. In spite of the differences in method and in stimuli, i t is interesting to note the similarities in meaning for some of the colours. An examination of the red, green, and blue columns of-Table 2 shows that these three colours are the most consistent in affective meaning. Zausznica's tabulations (Table 3) correspond closely to Schaie and Heiss' (Table 2) for red and green, but differ for blue. Black, when i t is included in studies, similarly tends to be associated consistently with distinct feelings. Luscher's meanings correspond quite closely in places to those in Tables 2 and 3 for red, blue, and yellow, but differ significantly for green, violet, and brown. An overall view of affective meanings shows that reds, oranges, and yellows (the warm colours) tend to be associated with a stronger and more exciting stimulation than do the greens, blues, and purples (the cooler colours) which seem to have a lower, more peaceful stimulation value. Such a generalization is commonly adhered to}both in the l i t -erature of colour psychology, and in applied settings such as advertising and industry. 42 The Physiological Meaning of Colour The perception of colour is said to produce physiological reactions apart from those in the visual system per se. The 'physiol-ogical value' of a colour then, reflects the capacity of that colour to affect bodily functions. Schaie (1966) speaks of a colour's physiol-ogical value as its excitatory potential, or its arousal value. It is not known at present however, whether a l l colours can produce dis-tinct bodily reactions. Evidence from animal studies shows that certain colours have a biological cue function (Tinbergen, 1942). However, if innate biolo-gical cue functions of this type s t i l l exist in man, they have not yet been detected. Luscher does not use the term biological cue function, but he does maintain that there are innate physiological significances of colour. These, he suggests, were established very early in time, when the behavior of primitive man was dictated by night and day, by darkness and light. As he puts i t Day brought an environment in which action was possible, so he set forth once more to replenish his store and forage or hunt for his food. Night brought passivity, quiescence and a general slowing down of metabolic and glandular activity; day brought with i t the possibility of action, an increase in the metabolic rate and greater gland-ular secretion, thus providing him with both energy and incentive. The colors associated with these environments are the dark-blue of the night sky and the bright yellow of daylight (Scott, 1969, p. 11). Several studies have attempted to establish physiological reactions to colour without relying on intuitive and fanciful generalizations. The 4 3 work of Goldstein (1939, 1942) is best known in this area. Basically, Goldstein found differential effects of coloured lights in his experi-mental work. For example, red was found to be disagreeable, upsetting, disruptive, forceful, and nausea-producing. Green was found to produce the opposite effects — peaceful and quietening feelings. In his theoretical assumptions, extracted from his data, Goldstein states that red has an expansive effect on the body and mind, making one more susceptible to the outside world whereas green has a contractive ef-fect, making one draw inside oneself, away from the outside world. He argues that every colour stimulation is accompanied by a specific response of the•organism. His evidence comes from a small unrepresentative group of brain-damaged patients however, and only a few coloured lights were used for the stimulation of subjects. His generalizations have been pointed out by almost every author reviewing his work, nevertheless he is continually cited because of the thoroughness of his research. It is not known whether Luscher had any contact with Goldstein, or with his work, but i t is likely that he was familiar with the latter. Luscher places great emphasis upon the physiological effects of colour, and Goldstein's findings are the only ones which appear prior to the publication of Luscher's test, with the exception of Pressey's (1921). Luscher, like Goldstein, also relies upon ideas about colour meanings expressed by Kandinsky and Goethe. Gerard's research (1958) lends some support to Goldstein regarding the differential effects of coloured lights, though not necessarily to the substance of his theory. Gerard found that blood pressure, respiration, and frequency of eyeblinks increased under red light, decreased under 44 blue light. Palmar conductance increased under both blue and red light, but after a period of time the arousal remained higher for red. Cortical activity was affected by red, blue, and green light, but remained consistently greater for red than for blue. No differences were found for heart rate. Nakshian (1964) tested Goldstein's theory on a normal population, and with the exception of performance on a few psychomotor tasks, no support was found for the differential effects of colour. Thorpe (1961) similarly has found no differential effects of coloured lights on chronic schizophrenics when a number of measures were taken, including GSR. On the other hand, Wilson (1966) found significant differences in the arousal properties of red and green when GSR was measured. He proposed a U-shaped function for the effects of colour, namely, that the hues at the ends of the spectrum were more arousing than those located towards the middle. Such differences, he felt, were due to the fact that the end of spectrum hues often occur "in association with the potentially dangerous wavelengths just outside" (the visible spectrum). This function was later verified, at least in part by Nourse and Welch (1971) who found violet to be more arousing (via GSR) than green. They con-cluded that more research is needed. The inclusion of Wohlfarth's publications (1956, 1957) in the Luscher test bibliography suggests that these may be sources which Luscher used to delineate the physiological effects of colour. Klar (1961), fore-most proponent of the Ltischer test, similarly mentions the work of Wohlfarth in his review of the test. Wohlfarth's findings are similar to Gerard's 45 concerning the differential effects of the warm (red, orange, yellow) and the cool (black, blue, green) colours on pulse, respiration, and blood pressure. They were published considerably later than the f i r s t appearance of the Luscher test (1948) however, which suggests that they could only have been additions to later editions of the test, notably to the Scott edition (1969). Luscher's original version of the test must either have placed less emphasis upon the physiological effects of colour, or have relied upon the work of Goldstein exclusively. In addition to physiological effects of colour, Luscher associates colour with various organs of the body. The organs associated with red, for example, are the striated voluntary muscles, the sympathetic nervous system, and the reproductive organs. The source of such associations, or any experimental evidence to substantiate them, i s not given. Symbolic Values The symbolic value of a colour is similar to the connotative meaning that a colour holds. It can be likened to a learned association, but may be unconscious or conscious, although generally held to be the latter. Most researchers in colour psychology pay only l i p service to symbolic meanings of colour, emphasizing that because they are un-conscious, they are therefore d i f f i c u l t to measure. Those who have given them some thought (Goethe, 1890; E l l i s , 1900a and b; Schachtel, 1943; Luscher, 1948; Yazmajian, 1964, 1968; Woltmann, 1965; Suinn, 1966, 1967; Zausznica, 1969; Trevor-Roper, 1970) have drawn upon mythology, folklore, cultural tradition, natural objects, free association, oneirics, 46 parts of the body, and the like, as source material. For any one colour, the symbolic values are rarely constant and often tend to contradict each other (see Tables 1 and 3). Red, for example, is the colour of blood and fire, i t may signify the Red Cross, the Pentecostal flame, or a barber, shop, and i t is said to symbolize love, excitement, anger, impulse, and danger. Which of these meanings, i f any, plays a part in a given individual's response to red cannot readily be ascertained. While i t is generally accepted that symbolic meanings are often closely related to affective meanings, and that the symbolism of a colour may play a role in determining the emotional reaction to that colour, the exact nature of these relationships is unknown, and likely will remain so. Summary of Colour Meanings The purpose of research into the meanings of colour is to delineate some of the stimulus dimensions of preference, whether such dimensions can be so neatly differentiated into synaesthetic, connotative, affective, physiological, or symbolic is open to question. The above presentation of meanings has been organized in this fashion, because this is princi-pally the way in which Luscher has differentiated colour meanings, and secondly, because various studies have tended to concentrate on only one dimension at a time. Presumably, the meanings that colours have are factors to be reckoned with in determining colour preference. Luscher holds that this is so. It is not unreasonable to assume that a man may like a given red 47 best because i t reminds him of love (symbolic), because i t makes him feel warm when he perceives i t (synaesthetic - physiological), and because he associates aggression to i t (connotative - affective). He may also like that red however, because that is the colour of his car, because he dreamt of i t the night before, or because i t appeals to him when i t is placed next to green. Any one of these factors, or a l l of them in combination, may play a role in determining colour preference PART TWO EXPERIMENTAL INVESTIGATIONS WITH THE LUSCHER TEST IV. OBJECTIVE ANALYSIS OF THE TEST COLOURS Research conducted with personality tests that use coloured stimuli has been mainly concerned with the responses to the colours: how they are scored, how they can be interpreted, whether they can be related to other data. Very l i t t l e information concerning the colours themselves is available. In Rorschach research for example, much is known about the meaning of the Colour Response, but nothing is known about the characteristics of the colour or colours to which i t refers. The common reference to 'Rorschach red' in the literature for example, shows that the individual colours on the cards are not being given proper attention. Such a term may mean a great deal to the clinician and personality researcher, but i t holds l i t t l e meaning to those interested in the specifics of colour perception and colour measurement. As a general statement about these tests then, more is known about the colour r e s p o n s e s of the tests than is known about the coloured s t i m u l i which eli c i t those responses. Specification and/or measurement of these stimuli should be basic procedure to a l l psychological research involving the response to colour, but up to this point i t has been neglected. On the Luscher test, the only information given about the colours is their names — dark-blue, orange-red, bright yellow, for example. There has been no application of available colour measurement techniques to the stimuli on any of the three versions. If personality differences 49 50 can be assessed on the basis of colour preference alone, as the Luscher test claims to do, then a first step towards assessing the test would be to delineate the characteristics of the test colours in question. Visual inspection alone shows noticeable differences between the three versions in the printing of colours where in fact they should be the same.Colour measurement can provide a more accurate picture of these differences. In addition, colorimetric specification will determine i f the test can be used with colour deficient subjects, a question which was brought up by Pickford (1971) in his review of the test, and which must depend upon exact colour specification for an answer. The concern of this study then, is not with the presumed diagnostic usefulness of the test responses, but rather with the colours that are used for that purpose. Reflectance characteristics and luminance values will be determined for the Random House version, and CIE, USC, and Munsell specifications will be provided for the colours in a l l three versions. Excitation purity, dominant wavelength, and a more exact colour naming identification are also included. The Random House abridged version will be referred to as 'LCTj', the Pocket Books abridged version as 'LCl^' , and Full LCT will identify the unabridged edition. Methods Illumination Al l instrumental and visual luminance measurements were done under a laboratory-built illuminant, which uses Corning Glass (Judd, 1962) to give CIE type 'C' illumination with a colour temperature of 6500°K. LABORATORY ILLUMINANT V 0 © © 0 e a o o e © © e O a e e o © 500 600 WAVELENGTH (nm) Energy distribution of laboratory-built Illuminant 52 Fig. 1 shows the relative energy distribution of this source, normalized with respect to wavelength of 555 nm. The CIE chromaticity coordinates are, for x = 0.305, for j_ - 0.315. Thus the characteristics of this laboratory source do not differ essentially from the CIE illuminant 'C'. Sample Specifications Reflectance characteristics were measured for the LCT^ only, as this was the version used in other experimental work. Reflectance measurements were obtained on a System 2400T Telespectroradiometer (Gamma Scientific Inc., 1971) with high efficiency monochromator assembly (1350 lines per nm), using 45/90 geometry. The resultant curves were cor-rected to equal energy source. The luminance of standard white, BaSO^ , supplied in powder form by Zeiss and pressed into a plaque on their Powder Press 60, was the reference standard. Its luminance was 90.5 nits. Luminance was measured only for the colours on the LCT 1 } and measurements were obtained with a Pritchard Photometer (Kollmorgen, 1970), which utilizes the response characteristics of the relative luminosity function (VA) of the CIE Standard Observer. The geometry of the readings was 45/90. Four sets of readings were taken under the laboratory-built Illuminant 'C. Colorimetric Measurements Colorimetric data was obtained on a Lovibond Schofield Tintometer, Type 1A (Schofield, 1939) by the author, age 26. The suitability of the author's colour vision for making the measurements was determined as fo l -lows: no colour confusion (Ishihara 10th edition and Dvorine Pseudo-53 isochromatic plates), perfect colour discrimination (Farnsworth-Munsell 100 Hue Test), acceptance only of mixture ratios on a l l three colour mixture equations at the means for the equivalent age group (Pickford-Nicolson anamaloscope, modified, Lakowski, 1973). Efficiency on the tintometer was determined by matching five plastic samples previously used in'a study on visual colorimetry (Lakowski, 1967). The matches were in agreement with the means obtained from a group of industrial colorists. The tintometer was modified to Illuminant 'C' viewing conditions. All measurements were made in a darkened room with an Illuminant 'C' lamp providing the only source of adaptation. BaSO^  was used as standard white, prepared in the Zeiss Powder Press 60. For the colours of the Full LCT, a 2" x 2" BaSOi^  standard was prepared in a thin card-board container and taped behind the appropriate port of the tintometer. This was necessary because the Full LCT colours are not on separate card but rather are printed on pages of a book, which had to be placed flush to the tintometer port. The Powder Press 60 was too thick to allow for this. Three matches were made on each of the eight colour cards in the LCTi and L C T 2 . Two matches were made on each of 29 colours in the Full LCT (5 neutrals, the 8-colour panel, the 4 'blues', 4 'greens', 4 'reds' and 4 'yellows'). Thus a l l numerical values presented in the Tables and Figures represent the means of a number of observations. Values in the RYB Tintometer system were first transformed into CIE coordinates x_ and v_. Dominant wavelength (Ad) , excitation purity 54 (%pe), and brightness (Y%) were also determined from RYB values, x and y_ coordinates were then converted into u and v_ values of the CIE (1960) UCS space and plotted accordingly. Munsell values were obtained by plotting x and y_ coordinates on the appropriate conversion charts (Newhall, et al., 1943). Finally, with the use of exact Munsell values, each colour was correctly identified by name, which were obtained by the ISCC-NBS method of designating colours (Kelly and Judd, 1955). Results  Figs. 2 and 3 give the objective characteristics in the form of reflectance curves, for the four primary and the four auxiliary colours of LCT}, respectively. The curves for the six chromatic stimuli are a l l broad band, and are very similar to those one would obtain from printers' inks, with reflectance values ranging from about 4% to 80%. None of the colours are metamerically related. Note that the two achromatic stimuli, grey and black, exhibit an even energy distribution, and thus should be considered as good neutral stimuli in a colour preference situation, un-affected by variations in the energy distribution of illuminants. Table 4 gives the actual luminance values in nits for the eight LCTi colours. Measured under Illuminant 'C', these values serve here as the basic photometric measurements, representing the values that would be obtained by the CIE Standard Observer. The yellow sample has the highest luminance, black the lowest. Also included in Table 4 for com-parative purposes are the Y% values for the colours, obtained from visual colorimetry. Ln 100-UJ o z < I— o UJ LU Blu< Green-blue » o o © © 600 W A V E L E N G T H nm Fig. 2. Reflectance characteristics of the four primary colours in LCTj 100 UJ o z < r-u LU LU OH 80 -60 -40 -20 -0 Grey — — Vio le t o o © o Block Brown o © » 0 .• o o o e T T 400 500 600 W A V E L E N G T H nm ON Fig. 3. Reflectance characteristics of the four auxiliary colours in LCTi 57 TABLE 4 Luminance Values in Nits and Derived Y% Values for the Eight Colours of LCT^ (Luminance Value of ; BaS0i+ under Ilium. 'C = 90.2 nits) PRIMARY COLOURS AUXILIARY COLOURS Colour Nits Y% Colour Nits Y% Dark blue 4.2 4.65 Violet 7.9 8.81 Blue-green 13.2 14.63 Brown 13.9 15.41 Orange-red 17.0 18.84 Black 2.7 2.99 Bright yellow 61.5 68.18 Grey 16.9 18.73 The results of a l l colorimetric measurements and conversions are presented in Tables 5 and 6, and in Figs. 4 and 5 in the form of two sets of data, one dealing with the eight colours which appear in a l l three versions, and the other with the additional panels of the Full LCT. Table 5 and Fig. 4 thus allow for comparisons between the tests, while Table 6 and Fig. 5 do so within a test. The names given to the eight colours (Table 5, Column 1) are those of Luscher (translated by Scott, 1969, p. 24), but since some colours differ from version to version, more specific nomenclature is included in the Table (Column 2). Similarly, the ISCC-NBS designated names of Full LCT colours (Table 6) are included so that the colours may be referred to more accurately. Note the large discrepancies in nomenclature between the Luscher and the NBS methods. TABLE 5. Colour Specifications for the Three Versions of the Eight Colour Panel LCT1 Colour lift IDS ISCC-H3S Nomenclature V. [unsell Specification C L E . Coordinates' Excitation X y AD Purity (%pe) Dark blue dark, blue 5.5 PB 3/7 .209 .188 7.24 474 38.40 Blue f,reen strong bluish ^rcen 10.0 BG 5/7 .203 .251 17.18 48S 40.90 Orange red viv i d reddish orange 7.0 II 5/13 .542 .339 20.42 606 o4. ex Bright yellow vi v i d greenish yellow S.O Y 3/12 .433 .434 66.07 574 79.33 Violet strong purplish red 5.0 .id? 4/12 .393 .223 11.46 503 40.73 Brown Cioderkte. reddish orange 1.5 YR 5/7 .474 .354 15.67 596 c 56.52 Black dark greyish blue. 4.0 PB 2/1 .283 .237 3.67 477 8. So Grey medium grey H 5/0 .306 .318 21.38 - 1.72 LCT2 Dark blue dark blue 6.5 PB 2/61 .213 .135 4.26. 4701 37.77 Blue greea strong yellowish greea 2.0 0 5/9 .265 .466 17.30 532 48.31 Orange red strong reddish orange 9.0 TI 5/11 .561 .349 19.05 604 73.51 Bright yellow vi v i d greenish yellov 7.0 T 8/12 .450 .490 63.10 575 87.40 Violet v i v i d reddish purple .2.0 RP 4/13 .364 .206 13.18 518 40.95 Brown s crony brovn. 7.5 YE 4/7 .478 .408 15.85 537° 70.49 Blacfc bluis . ' i black 5.0 PB 1/1 .294 .236 2.04 472 5.00 Grey medium &rey N 5/0 .308 .315 16.60 — 0.00 FELL LCT Dark blue moderate purplish blue 8.0 PB 4/7 .239 .209 10.7 466 25.89 Blue fjreen strong bluish green 4 .^0 BG 6/11 .194 .343 26.3 474 45.57 Orange red vivid reddish orange 7.5 R 6/14 .535 .340 28.5 606 63.83 Bright yellow vi v i d greenish yellow 7.5 T 9/12 .432 .478 33.1 574 30.85 Violet viv i d reddish purple 3.0 HP 5/14 .373 .223 19.9 509 37.36 Brown greyish reddish oraag,e 2.5 YR 5/7 .443 .363 23.9 592 c 50.96 Black dark bluish grey 4.0 PB 3/1 .285 .290 5.5 476 8.26 Grey light bluish Krey 6.0 3 7/1 .302 .312 39.8 482 2.70 Fig. 4. Loci of the eight colours in the three versions of the Luscher Colour Test 60 Conversions of CIE x and coordinates into Munsell specification was carried out to the first decimal place, however, the values listed in the Tables are given to the nearest half step of Hue, and to the nearest f u l l step of Chroma and Value. CIE coordinates take the standard presentation. The most significant difference between the three versions of the eight-colour panel can be found in the colour blue-green (Table 5). In terms of Munsell specification, blue-green changes a f u l l 18 steps of Hue from 10.0BG5/7 in LCTj to 2.0 G 5/9 in LCT2. The change in ISCC-NBS nomenclature also shows the extent of the difference. The other colour with a large difference is brown, which involves a 6.5 step difference in Hue between LC^ and LCT2. The remaining colours do not change above 3 steps of Hue from version to version. Chroma and Value are more consistent between versions than is Hue, with the exception of Chroma for blue-green (compare LCTX with the Full LCT). A further means of showing the differences in 'blue-green' can be seen in the change of dominant wavelength. Blue-green in the LCTj has a Xd of 488 nm, while in the LCT2, Ad is 532 nm. The Ad change between these two versions for violet is 15 nm, and for brown is 9 nm. Ad also changes for dark-blue (8 nm), but this difference is between the LCTj and the Full LCT. Other findings emerge from similar comparisons. Inspection of brightness values shows that Y%'s are higher for each colour in the Full LCT than they are for corresponding colours in the other two versions. There are also considerable differences in Y% between individual colours. TABLE 6 Colour Specifications for Additional Panels i n the FULL LCT ACHROMATIC PANEL ISCC-NBS Nomenclature Munsell Specification C L E . Coordinates XD Excitation X y y% Purity (%ps> Bluish grey 10.0 B 5/1 .294 .304 17.4 480 5.24 Dark bluish grey 5.0 PB 3/1 .286 .290 6.68 477 7.98 Light grey N 8/0 .302 .310 61.00 - 2.77 White N 10/0 .304 .311 97.70 - 1.92 Light Bluish grey , , 6.0 B 7/1 .306 .314 39.80 478 1.28 BLUE PANEL Moderate purplish blue 8.0 PB 4/7 .241 .211 12.9 466 25.17 B r i l l i a n t blue 3.0 PB 6/11 .200 .207 29.5 477 42.37 Vivid purplish blue 8.0 PB 5/16 .206 .154 19.0 466 42.65 lioderate blue 9.3 B 5/7 .215 .245 18.6 481 35.81 GREEN PANEL Moderate yellowish green 8.5 GY 6/7 .324 .446 26.3 557 50.00 B r i l l i a n t yellowish green 9.0 GY 9/10 .325 .470 69.2 557 58.09 B r i l l i a n t green 7.5 G 6/8 .242 .394 33.9 505 33.64 B r i l l i a n t bluish green 8-0 BG 7/11 .200 .312 38.0 490 42.34 RED PANEL Moderate reddish orange 8.5 R 5/9 .478 .344 17.9 603 49.66 Vivid yellowish pink 10.0 R 7/13 .504 .371 46.2 596 66.95 Vivid reddish orange 7.5 R 6/15 .540 .344 30.2 605 66.99 Moderate Ked 3.0 R 5/9 .447 .308 17.0 635 30.00 YELLOW PANEL Moderate orange yellow 10.0 YR 8/9 .446 .419 52.5 583 68.18 Vivid greenish yellow 9.0 Y 9/12 .428 .483 81.2 574 78.90 -B r i l l i a n t orange 6.5 YR 8/11 .469 .406 64.6 586 69.87 B r i l l i a n t yellow green 5.5 GY 8/9 .372 .468 66.1 567 66.26 Y e I l o w s .2 .3 Fig. 5. Loci of the colours in the additional panels of the Full LCT 63 Grey and yellow, for example, change 23.2% and 20.0% respectively, from LCT2 to the Full LCT. Blue and black are the most consistent in Y% from version to version. In terms of excitation purity, brown changes the most with a 19.53% difference between the LCT2 and the Full LCT. For a l l other colours the range of change in pe between versions is less than 10%. The two colours which remain the most consistent between versions of the test are red and black. Fig. 4 shows the eight colours of the three versions plotted in uniform colour space, and this visually conf irms what Table 5 gives with more precision. By inspection, it can easily be seen that the differences between versions are greatest for the blue-green stimuli, and least for the two nonchromatic, the yellow and the red stimuli. Table 6 gives colorimetric specifications for the colours in the additional panels of the Full LCT. Each chromatic panel consists of four colours, one of which is intended to be a primary colour in the eight-colour panel. Compare for example, 'moderate purplish blue' in the blue panel (Table 6) with 'dark blue' in the eight-colour panel (Table 5). The given values show that the two colours are almost equivalent colori-metrically, and are therefore perceptually the same. Y% differs only 2.2%, pe differs only 0.72%, Munsell specification and Xd are equal, and the J C , y_ and _u, y_ coordinates differ only in the third decimal place (the largest difference is .002 for x and . Similar colorimetric equi-valence can be shown by comparing specifications of orange-red and bright yellow in the eight-colour panel with 'vivid reddish orange' and the 64 'vivid greenish yellow' in the red and yellow panels,respectively. The situation is quite different for the greens however. 'Brilliant bluish green' in the Full LCT green panel should be equivalent to the primary blue-green in the Full LCT eight-colour panel, but i t is not. Colorimetrically, Y% differs almost 12%, Ad changes 4 nm, pe changes more tfran 3%, and some of the x» y_ and _u, v coordinates are significantly different in the second decimal place. The primary colour blue-green thus does not repeat itself in its respective chromatic panel, as do the other three primaries in their panels. The difference is noticeable per-ceptually. Table 6 shows that there is no apparent colorimetric relationship between the colours in any one panel. In terms of Munsell specification for example, a l l of the colours in the red panel are red Hues, but the colours in the yellow panel consist of two yellow-reds, a yellow, and a green Hue. Fig. 5 depicts the relationships in UCS space and shows that with the exception of the achromatic panel, the stimuli in each panel are not perceptually equidistant from each other. Furthermore, it can be seen that the panels themselves do not follow any apparent perceptual pattern. The green and-yellow panels suggest a circular pattern equi-distant from neutral grey, but this does not apply to the choice of the reds and blues. Discussion Among tests which use colour for personality assessment, the Luscher test has rapidly gained popularity. This may be due in part to the fact that i t has been published in three versions, which makes i t 65 easily accessible to the public. Yet those who use the test know nothing about the colours which comprise i t . The preceding data is in-tended to provide this information. •Luscher claims that each colour on the test has been carefully chosen, and he mentions that the choice involved selecting eight out of a total* of 4500. This number incidentally is approximately the number of colour samples available in the Munsell collection in the late 1940s. Were the colours really chosen carefully, or was the selection based on other criteria? Regarding the two achromatic colours in the eight-colour panel (LCTj for convenience), both black and grey show even reflectances across the visible range of the spectrum, which suggests that they are then 'good choices', if i t is neutrality of stimulus that Luscher desired. Grey, in particular, is very similar to Munsell N5. The choice of the four primaries however is not as easy to reconcile. Beginning with dark-blue and bright yellow, these two primaries are al-most perfectly complementary to each other. Their respective coordinates lie on a line which passes almost directly through Illuminant 'C' in the colour space. They also represent the two extremes of luminance as re-flected in nits and in Y% (disregarding black). In addition, they represent the two extremes in excitation purity. Yellow has the highest and blue the lowest pe% of the chromatic cards. Thus, i f i f was a com-plementary relationship that Luscher desired, and if i t was bipolarity in the brightness and saturation variables that he had in mind, then these two colours are also 'good choices' for primaries. 66 Orange-red and blue-green on the other hand, are not complementary to each other. They are similar in Y%, but quite dissimilar in pe% (al-though they are the second most saturated and the second least saturated of the chromatic cards). The perceptual relationship of one to the other is not really clear, however. Blue-green is perceptually equivalent in saturation to violet, but orange-red is considerably different in satura-tion from brown, the other auxiliary colour. No relationship is apparent regarding the brightness variable. The perceptual relationship between these two primaries and the other colours then,is also not clear. If there is a logic to the choice of blue-green and orange-red as primaries, i t is not obvious. From a perceptual point of view the choice seems somewhat more random than i t is deliberate. Any other explanation must refer back to Luscher's rationale that each colour was carefully chosen because of its deep psychological/symbolic meaning, but this is neither a comforting nor a satisfactory explanation. Violet and brown also do not seem to be related perceptually or colorimetrically to the other colours, or to each other. Liischer mentions that the auxiliary colours are included in the test only to increase the scope of choice available for preferences, which in itself seems a con-tradiction to the 'deep psychological meaning' hypothesis. The question arises: would any two colours suffice as chromatic auxiliaries to the test? Explaining the choice of colours which comprise the Full LCT is even more problematic. The fact that the blue-green primary in the eight-colour panel is not repeated in the green panel, as the other primaries 67 are, is particularly confusing, and one would like to insist that this is a printing error. The arrangement of colours in UCS space reveals an interesting pattern, suggestive of a circle around neutral grey. The circular arrangement does not complete itself however, and the pattern is broken. This does not necessarily rule out the possibility that 'a circle' was what Luscher had in mind when selecting the colours, but i t does make the choice of colours more suspect. In the publication of the Luscher test, or of any test which claims a clinical usefulness, i t is important that the printing of colours be exact, presumably in accordance with the wishes of the designer. It is just as important that the colours do not vary from one printing of a test to another. Test colours m u s t be standardized properly if responses are to be interpreted meaningfully. This requirement has clearly not been met with the Luscher test. The publications by Random House, Pocket Books, and Test-Verlag differ significantly in their printing of the standard eight colours. Certain colours in particular vary greatly from one version to the next, notably the blue-green and the brown. In fact the blue-green is no longer a 'blue-green' in the Pocket Books version. These findings have a number of implications. One of them concerns the basic assumptions on which the test is based, and that is that each colour has a unique psychological and/or symbolic meaning intrinsic to i t — the Structure of that colour. The question then arises, i f the colour changes from one version of the test to the next, what then happens to its psychological meaning? Should i t somehow change accordingly? The 68 test manuals accompanying the abridged editions are not different from one another, yet the colours that they are describing are different. Thus, the psychological/symbolic meaning of colours remains the same between versions, but the colours themselves change depending upon which version is purchased. A second assumption involves colour preference per se. The rel-ative preference of a colour is of course dependent upon its hue, sat-uration, and brightness, as has been systematically demonstrated by the work of Guilford (1934, 1940, 1959) and of Granger (1955. (According to Luscher, preference is dependent upon meaning, but this shall be dis-regarded for the time being.) In the eight-colour panels of the Luscher test subjects are required to rank order the colours by preference, yet those who take one version of the test are not confronted with the same coloured stimuli as those who take another version of the test. The colours are different, the task has therefore changed, and in-tertest comparisons of results become difficult. Whether the changes in hue, saturation, and brightness are significant enough to effect changes in the rank ordering of the eight colours is irresolvable at present. Data on each version will need to be collected so that normative respon-ses can be compared for any differences in the average rank order. Yet another assumption about the interpretation of colour preference is this. If colour preferences are affected by the differences in print-ing, i t follows that the assessment of personality will also be affected. For example, if a subject prefers the brown of one version in fourth pos-ition, and the brown of another version in eighth position, the resulting two personality descriptions will be considerably different, Th should cause some concern to those who take the test seriously. 7o V. COLOUR VISION AND THE LUSCHER TEST The practicability of projective tests which use colour has not been examined in relation to colour-deficient persons. The Luscher test which involves colour preference i s no exception. Quite apparently, i f test stimuli are not perceptually distinct to a subject, i t i s of l i t t l e value to ask him to choose one over the other. Those who administer such projective tests do not usually inquire about the state of a testee's colour vision. In Rorschach protocol, for example,no consideration is given to the matter. In view of the psychol-ogical significance placed on the Colour Response i n the Rorschach test this could be a serious omission of relevant information. If, for exampl colour-deficient persons produce fewer colour responses than do normals, they would be misjudged when such responses are analyzed for emotional content (in accordance with scoring procedures). Eight percent of a l l men and one percent of women are congenitally colour-deficient, and up to 10% of the population have acquired dyschroma topsias. Are these persons at a disadvantage on such projective tests, or i s their deficiency of no importance? Scott (1969), quoting from Steinke's research (1960), states that in regard to the Liischer test "colour vision need not be a l l " (p. 16). According to Pickford (1971), however, who reviewed the Steinke research, this state-ment i s an overgeneralization, as Steinke did not examine the responses of protanopes i n his research. Pickford also stated that the Luscher test colours were not l i k e l y 70 71 to cause much confusion to most red-green dichromats, but that the orange-red and brown "might be almost indistinguishable both in hue and brightness to some protanopes, and violet and grey to some deutera-nopes" (p. 153). He concluded that further research was desirable. The present author agrees with Pickford and the investigations below are intended as i n i t i a l steps in this line of research. Their overall purpose is to investigate the use of the test with colour-deficient subjects. Major consideration is given to the question of whether dichromats can differentiate test colours. This can be answered by plotting the colours in CIE space to determine i f any of them f a l l on known confusion lines. Pickford's statement about certain colours being indistinguishable can then be verified. It should be noted however, that Pickford was referring only to one version of the test, and i t is possible that some colour-deficient persons may have more trouble with one version than they would with another. Further consideration is given to brightness perception. Photo-metric measurements made by colour-deficient subjects will be compared with measurements made by subjects with normal colour vision. The question of whether or not colour preferences on the test are affected by differences in perception will be considered, and comparisons will be made accordingly. Finally, an attempt will be made to discover whether or not the colour preferences of deficient persons are determined or influenced by the brightness factor. The relationship between preference and two methods of brightness measurement will be considered. 72 Methods Regarding the part of the experiment not involving subjects, the x and y_ coordinates converted from tintometer readings (from Tables 5 and 6) were plotted i n CIE space. The origins of confusion lines for the four types of dichromats were plotted (Wyszecki and Stiles, 1967), however only those lines which connected two or more colours were ex-tended f u l l y . It should be noted that while there are many colours in the test, a subject is only confronted with a maximum of eight at any one time. A certain red in the red panel may" be confused with a green in the green panel, but since the two are never seen together, i t is not relevant that they f a l l onto a confusion line. Thus a confusion line was extended from i t s origin only i f i t connected two or more colours in the s a m e , p a n e 1 , or in the s a m e v e r s i o n of the eight-colour panel. Eighteen red-green male deficient subjects participated in the subjective part of the experiment. They were selected from approxima-tely 350 undergraduate students who were individually screened with pseudoisochromatic plates at the rear of a university classroom. Sub-jects who were tentatively classed as deficient were later administered the 10th edition of the Ishihara as well as the Dvorine pseudoisochro-matic plates, the Farnsworth-Munsell 100-Hue Test, and were tested on a modified Pickford-Nicolson anomaloscope (Lakowski and Tansley, 1973) in the Visual Laboratory under the previously described Illuminant 'C'. A l l testing took place with the subject seated in front of a cubicle which was painted in a f l a t neutral grey. For complete information on 73 tests and testing procedures see Lakowski (1969). Of the 18 subjects, seven were dichromats, four protanopes (P) and three deuteranopes (D). Eleven were anomalous trichromats, of whom two were protan-type, one extreme protanomalous (EPA) and one simple protanomalous (PA). The other nine were of deutan-type — three extreme deuteranomalous (EDA), and six simple deuteranomalous (DA). The c l a s s i -fication of defects i s according to Pickford (1951), and symbols denote the accepted genetic nomenclature. Each subject was administered the Random House version of the Ltischer test twice, one 30 seconds after the other. The cards were presented to the subject in a horizontal row and in random order. After the subject indicated each preference, the card was removed from sight, and the remaining cards moved to form a continuous row without gaps. Each subject also measured the luminance of the cards. Luminance measurements were made with a Salford E l e c t r i c a l Instruments (S.E.I.) exposure meter (Ilford, 1972) adapted to Illuminant 'C' viewing condi-tions, and held in place on a retort stand, with the eyepiece 15 inches from the cards. A l l measurements were taken using a 50/90 geometry with the cards directly under the Illuminant 'C' source. A subject was f i r s t introduced to the instrument by way of a demonstration using an N5 colour patch from a page in the Munsell Colour Atlas (Munsell, 1929). The subject looked through the viewing piece while the brightness of the dot i n relation to the N5 background was i n -creased or decreased by the experimenter. The subject was then asked to adjust the dot so that i t was equal in brightness to the background. 74 Following this, the page was moved so that the dot centered over N7, and the subject readjusted the brightness accordingly. N9 was then put into view and the procedure repeated, and this was followed by an exposure to N3." When the subject was familiar with adjusting brightnesses for achro-matic stimuli, (N3, N5, N7, N9) , the procedure was repeated with a Munsell colour. The colour so chosen was that which the subject had previously indicated to be his 'favorite colour', and since 12 out of 18 indicated blue, the 5.0 B page of the atlas was most frequently chosen (5.0 B 3/6, 5.0 B 5/6, 5.0 B 7/6 for example). The subject was then told he was to make a reading on each of the eight cards he had previously seen in the colour preference situation. He was instructed not to hurry and to make certain that for each card the dot was equal in brightness to the background. The choice of eye was left to the.subject. A second reading was then made on each card, so that an average could be taken and so that a measure of his consis-tency was available. If two readings for a given card differed more than 0.06 log foot lamberts, the subject was asked to make a third reading on that card. He was then shown whichever of the first two readings did not match his third, and asked i f he would accept i t as an equal match. If he did not, and this was the case with one exception, he was asked to make a fourth reading. The reading which was not acceptable was excluded from the average. For most subjects however, two readings were sufficient for most of the eight colours. 75 The subject was then introduced in more detail to the concept of a brightness variable in colour, with samples from various pages of the Munsell Atlas (Munsell Color Company, 1950). He was shown how to judge" which of two colours was brighter when they differed in hue and saturation. The eight cards of the Luscher test were then arranged in front of the subject in two rows of four, so that each was uniformly exposed to the illumination from the source. The subject was told that this part of the experiment required him to make brightness judge-ments 'with the naked eye', and he was asked to point to the brightest colour. The indicated card was removed, and the procedure repeated until a subjective rank order was established. Two subjects with no colour confusion, perfect colour discrimi-nation, and perfect colour matching functions made comparative photo-metric measurements with the S.E.I, photometer. Eighteen subjects with normal colour vision were selected to comprise a control group to be ad-ministered the Liischer test, and ten of these were asked to rank sub-jectively the colours by brightness. A l l subjects were tested over a number of sessions. Results Colour Confusions Confusion lines relevant to the colours of the Luscher Test are shown in Figs. 6 to 9. On Fig. 6 for example, i t can be seen that the red and brown of the Random House version (LCTX) are indistinguishable in hue to protanopes. A protanopic confusion line also connects the 76 green and black of the LCT^, however the difference i n brightness between these two colours is greater than i t is between the red and brown (see Table 5). Deuteranopes w i l l then see the LCT^ violet and black as the same hue, but w i l l have no problem differentiating the violet and grey. A l l LCT} colours w i l l be distinguishable to tritans and tetartans, and the lines which have been extended from these origins are only to show which colours w i l l be seen as 'very similar' in hue. An interesting picture emerges with the Pocket Books (LCT2) colours (Fig. 7). For each type of colour deficiency, there are two colours which w i l l be the same in hue: red and brown for the deutera-nope, violet and blue for the protanope, blue and brown for the tetartan, and blue and green for the tritan. (The top two protanope lines are again included only to show hue similarities.) Note that with the LCT 1 } red and brown l i e on a protanopic confusion line, while i n the LCT2 they l i e on a deuteranopic line. For the eight-colour panel of the F u l l LCT (Fig. 8) only three confusions are apparent, red and brown, and blue and violet for the protanope, and black and green for the deuteranope. Fig. 9 shows the colours in the chromatic panels of the Full LCT which w i l l be confused in hue by dichromats. Note particularly that three of the four colours i n the yellow panel l i e on a protanopic confusion line. These three colours dif f e r among each other only 3%% in excitation purity, and 13%% in Y% (see Table 6). Note also that every panel has at least two colours which w i l l be confused by one of the four types of dichromats. 77 F i g . 7 . Confusion lines relevant to the eight colours i n the Pocket Books version of the Liischer Test Fig. 8 . Confusion lines relevant to the eight-colour panel i n the F u l l LCT Fig. 1 0 . Comparison of luminance values for six colour-deficient subjects vs normal on the four primary colours i n LCTj 82 Luminance Measurements Figs. 10 and 11 compare the relative visual photometric values of six types of colour deficiencies with those of normal subjects, for the primary and auxiliary colours of the LCTj respectively. The mean of measurements for the normal group is indicated for each colour as a 100% reference line, while the points above and below this line represent percentage differences in brightness values for the six deficient subjects. Each point represents the value obtained from the mean of a number of actual photometric measurements (representing nits). These six indivi-duals were chosen out of the pool of 18 as they show most clearly the relative luminance differences between the various types of colour defectives. There are of course individual differences between observers within each type of deficiency. The resultant luminance values in Fig. 10 could have been predicted from the well-known differences found among dichromats. It can be seen that the deuteranope, the EDA, and the DA see the orange-red sample as much brighter than do the protanope, the EPA, and the PA. The opposite effect can be seen for the blue and blue-green samples. Luminance values for yellow do not differ as much. The points plotted on Fig. 11 indicate that luminance values for violet follow the same pattern as those for red. Relative luminance values for brown, black, and grey however, vary considerably. This is an interesting but puzzling observation, as i t implies that values for the auxiliary colours could not have been predicted from the relative luminosity functions. 60 H Fig. 11. Comparison of luminance values for six colour-deficient ejects vs normal on the four auxiliary colours i n LCTj 84 a) Dark-blue 1 : Ordinal Position c) Orange-red 1 1 ;: ijS sr n 3 4 9-e I s P r e f S e 4-r 3 b) Blue-green 5f 1 -X 3 •* 5 Ordinal Position d) Bright yellow Ordinal Position A 3 Y 6 Ordinal Position Colour-Deficient D Colour-Normal Ordinal Position g) Black a 3 4 5 <» Ordinal Position P r e f e ? f) Brovfti A 3 4 5" Ordinal Position h) Grey IL I i I a. 3 «- t Ordinal Position Fig.12. Distribution of Luscher-preferences for colour-deficient and colour-normal subjects 85 Colour Preferences Fig. 12, a) to h) depicts the distribution of colour preferences for colour-deficient and colour-normal subjects for each of the eight possible ordinal positions. A histogram depiction was chosen because it quickly indicates obvious differences between the two groups in the ordinal positioning of colours. Very few differences exist. No single colour stands out as being clearly preferred or not by one group more than i t is by the other. The four primary colours in particular show close matches in ordinal positioning. Black and grey are consistently the two colours least liked by both groups. Note that neither group prefers grey in the first two positions or blue in the last two, and, that red is never the least preferred, brown never the most preferred. The histograms show l i t t l e difference between the two groups in colour preference per se. To determine if there are group differences when colour preference is analyzed as personality data, an analysis of variance was conducted. Each subject was scored according to the Luscher method of interpreting preferences (see page 11). The difference between the two groups is negligible. For the 18 deficient subjects x = 1.61, s = 2.08, for the colour normals x = 1.72, s = 1.94. The analysis showed no significant differences, with F (1, 34) = 0.02. The histogram depiction and the analysis of variance deal with the 18 subjects taken collectively, yet the group is composed of six dif-ferent classes of deficiencies. When the average rank orders are 86 determined for those classes with three or more subjects in each, some differences emerge (see Table 7). Note in particular that the protanopes prefer green much less than the others, and that the simple deuteranomalous prefer violet much more than the other types. TABLE 7 MEAN RANK ORDERING OF LUSCHER COLOURS BY PREFERENCE FOR COLOUR-DEFICIENT GROUPS Colour Protanope Deuteranope Extreme Simple deuteranomalous deuteranomalous ( n=4) (n=3) (n=3) (n=6) Dark-blue 1 1 1 4 Blue-green 6 1 3 4 Orange-red 2 3 1 3 Bright yellow 3 4 4 2 Violet 4 4 5 1 Brown 4 6 8 6 Black 8 7 7 8 Grey 7 8 5 7 Colour Preference and Luminance To determine if luminance is a factor in colour preference, order correlations were computed between luminance measurements and colour preference. For each type of deficiency with three or more sub-jects., the mean rank order of colours determined by preference was cor-related to the mean rank order of colours determined by photometric measurements. Correlation coefficients are given in Table 8. None 87 of them is significant. TABLE 8 RANK ORDER CORRELATIONS BETWEEN PREFERENCE AND PHOTOMETRIC MEASUREMENTS Type of Deficiency Number of Spearman's Significance Subjects p Protanope 4 +0.387 n.s. Deuteranope 3 +0.513 n.s. Extreme deuteranomalous 3 +0.406 n.s. Simple deuteranomalous 6 +0.257 n.s. On the other hand, a l l correlations between the mean subjective rank ordering of colours by brightness and the mean rank ordering of colours by preference are significant for the deficient groups (see Table 9). For the control group however, the correlation is not significant. TABLE 9 RANK ORDER CORRELATIONS BETWEEN PREFERENCE AND SUBJECTIVE BRIGHTNESS Type of Deficiency Number of Subjects Spearman's P Significance Protanope 4 +0.784 <.05 Deuteranope 3 +0.842 <.05 Extreme deuteranomalous 3 +0.850 <.01 Simple deuteranomalous 6 +0.793 <.01 Colour normal 10 +0.664 n.s. 88 Discussion It is clear that there are certain colours on the Luscher test which dichromats will find indistinguishable in hue. This applies to al l three eight-colour versions, as well as to the chromatic panels of the Full LCT. Concerning the latter, protanopes in particular should be expected to encounter considerable difficulty in making their preferences on the yellow panel. The findings confirm what Pickford stated about orange-red and brown being almost indistinguishable in hue to protanopes, and this holds true for the Random House version and the Test-Verlag edition. In the Pocket Books edition however, these two colours will be problematic for deuteranopes rather than protanopes. Pickford's sta-tement about deuteranopes confusing violet and grey does not hold true. These two colours in a l l versions do not lie on any confusion lines. The findings imply that dichromats must in certain cases make preferences on the basis of brightness cues alone. In most cases of colour confusion, the.brightness differences are large enough so that the two colours can be differentiated, however in a few instances the differences are minimal, thus rendering the colours perceptually equi-valent. In such instances i t makes l i t t l e sense to ask for preferences. That colour-deficient subjects see the Luscher colours differently than normal colour vision subjects was also shown when photometric mea-surements were compared. Yet the perceptual differences do not seem to affect colour preference when i t is scored and interpreted according to 89 Luscher. The analysis of variance was conducted between two groups of male subjects, a normal and a deficient colour vision group, however the latter is not homogeneous with respect to colour vision. The sub-jects f a l l into six different classes of deficiency, and include dichromats and anomalous trichromats. Further research is needed to determine i f colour preference on the Luscher test will differ with varying deficiencies. Two of the four protanopes tested referred to the green of the test as a grey, and to the grey as a green. The confusion of these two colours in a preference test causes problems where the interpretation of preference is concerned. Green is one of the primary colours and according to Luscher should be preferred in the first five positions to indicate 'normality'. The protanopes preferred green on the average in sixth position (grey was seventh), while the other classes of colour-deficient subjects did not prefer green past fourth position. Either the protanopes did not like the green, or they were not certain what colour i t was in relation to grey and therefore placed them both toge-ther towards the end of the preferential order. If colour preference is interpreted according to the Luscher method, such persons would thus be incorrectly scored. The relationship between the colour preference and subjective brightness is an interesting one. The rank order correlations for a l l types of deficiencies were found to be significant, which suggests that persons who knew that they have difficulty in discriminating hues will tend to rely on other perceptual and/or cognitive cues — subjective 90 brightness being the most obvious. This is not to imply that bright-ness is synonymous with preference in colour-deficient subjects, but rather that i t plays a significant role in the determining of preference. It does not play the same role with normal vision subjects, whose ordering of colours by preference was not related to the ordering of colours by subjective- brightness. The results of these i n i t i a l investigations do suggest that the Luscher test is poorly designed from the point of view of the perception of a dichromatic population. Despite the fact that such persons are at a disadvantage because of perceptual difficulties, their preference scores are comparable to those of normal subjects. In conclusion, i t should be noted that the study of colour pre-ferences of colour-deficient persons is a new area of research and must be investigated more fully before more specific conclusions can be drawn about the Liischer test. The correlations between subjective brightness and colour preference suggests that a more systematic study of the type conducted by Guilford or Granger is in order. 7/ VI. COLOUR PREFERENCE AND PSYCHIATRIC STATUS In the last twenty years or so various studies have investigated the relationship between the response to colour and personality variables. These studies have tended to take one of two approaches to the problem. One approach has been to relate colour preference data with scores on personality variables, both of which have been obtained from normal populations. The other approach has dealt with populations with known psychiatric status. Barrett and Eaton (1947) were the first to take the normal population approach, and found that subjects preferring Milton Bradley 'colours' were more extroverted, while those preferring Milton Bradley 'tints' were less so. Riffenburgh (1959), using the Woods Polychrome Index (Woods, 1956), found a highly significant relationship between preference and the Re-sponsibility scale of the MMPI. Choungarian (1967, 1972), the leading proponent of cross-cultural studies in this area, has found l i t t l e relationship between colour preference and introversion, extraversion, and neuroticism, with the exception that neurotics prefer red and purple more than extraverts, and extraverts prefer yellow-green more than neurotics. Spiegel and Keith-Spiegel (1971), using the Brentwood Colour Test (Spiegel and Acker, 1960), found a relationship between high manifest anxiety and dislike for the colour blue. Suffice i t to say that the results of these studies are a long way from establishing a strong rela-tionship between colour preference and personality variables in non-psychiatric populations. With the exception of Colour Pyramid Test research, to be discussed 91 92 below, only four studies have taken the abnormal population approach. The first study (Shikiba, 1927) showed that manic patients preferred red above a l l colours, while depressives rated red the least preferred. Katz (1931) found only that manic-depressives preferred red more strongly than did schizophrenics. In another study (Block and Caldwell, 1959), blue wars found to be- the most preferred colour by male paranoid schizophrenics.^ The most extensive and carefully controlled study of this type was conducted by Warner (1949), who tested the preferences of 120 catatonic schizophrenics, 60 anxiety neurotics, and 60 patients in both the manic and depressed phases of a manic-depressive psychosis with the paired comparison technique. Hue, saturation, and brightness were accounted for by the use of Munsell coloured papers, although possible colour vision deficiencies, length of hospitalization, and progress of the illness were not considered. Only the yellow-green pair of colours was found to dif-ferentiate the psychiatric groups, with the anxiety groups preferring green to yellow more strongly than did any other group. Neurotics were found to prefer the lighter colours. In effect however, very l i t t l e of significance emerged from Warner's study. The only studies which have continued to find relationships between colour preference and psychiatric status are those which use the Colour Pyramid Test (Pfister, 1950). These have been thoroughly reviewed by Schaie and Heiss (1964). It should be noted that on the Colour Pyramid Test a subject chooses colours which he judges will be pleasing when placed into the form of a pyramid, and i t is assumed that the colours so chosen 93 are the most preferred ones. This difference must be kept in mind when comparing Colour Pyramid findings with other preference studies. In general, research on the Colour Pyramid Test has shown i t to have quite good concurrent validity for different psychiatric groups. Wewetzer (1951) for example, first showed that manic depressives preferred purple,, brown, and grey, but not red, yellow, and orange, and Brengelmann (1957) found low preferences for yellow in depressive subjects. The col-lection of normative data on the test is also growing rapidly. There is, however, no comparative collection of data for the Ltfscher test. Since the Luscher test is a personality test, i t is reasonable to assume that it may be helpful as an aid to the diagnosis and/or classification of psychological illnesses. In the interest of assessing the test in this respect, i t was decided to use the test with a psychiatric population. The main purpose of the study is to deter-mine i f the test is sensitive enough to make a differentiation between a psychiatric and a nonpsychiatric population. Depressive patients were chosen to represent a psychiatric popula-tion, because they are the only group Luscher (1960) mentions in a paper dealing with colour psychology. Specifically, he states that depressives will "long for the lightest colour, yellow". This has not yet been substantiated empirically. Furthermore, i t seems a contradiction to the findings of the Colour Pyramid research, in which yellow is not a popular colour with depressives. Methods Sixteen subjects with a depressive illness were tested from the 94 wards of Crease Clinic at Riverview Hospital in Coquitlam, British Columbia. Selection procedures were handled by the Administration at Crease Clinic, who when informed of the nature of the project, selected an appropriately classified group of patients for testing. Five subjects were male, eleven female, with ages ranging from 21 to 67. The average age was 41 years, 6 months. These 17 patients com-prised the total depressive population at Crease Clinic at the time of testing, thus the unequal sex ratio of subjects. Length of hospitalization was recorded for each subject and ranged from less than one week up to one year, with an average of 16 weeks. The type of medication administered to the subjects varied considerably from patient to patient, although diazepan, a relaxant, was the most common. Three patients were receiving no medication at the time of testing. Also recorded were the number of previous hospitalizations, which ranged from none to seven, with an average of three. The severity of the depression at time of testing was often difficult to judge from individual records of patients, however seven were judged to be acute, seven moderate, and two light. At time of ad-mission the severity was generally at an acute stage. All subjects were -screened for colour vision deficiencies with the 10th edition of the Ishihara pseudoisochromatic plates and with the Farnsworth Munsell 100 Hue Test. , One male depressive, aged 62, was found to be of deutan-type, either a deuteranope, or an extreme deuteranomalous, and was therefore excluded from the study. Two groups of nonpsychiatric subjects were also tested. One group consisted of 16 nonhospitalized subjects, matched for age and sex with the 95 depressive group. The other group consisted of 16 hospitalized patients, each one matched for age, sex, and length of hospitalization with a mem-ber of the depressive sample. The second group was included as a check on the possibility that hospitalization itself would be a factor in the response to colour. A few of the nonhospitalized subjects were tested in the visual laboratory, the remainder in a darkened room in the basement of the apartment block in which they resided. Selection was handled by door to door knocking in apartment blocks in a residential area. A l l of the non-psychiatric hospitalized patients were tested in the Fairview Pavilion: of the Vancouver General Hospital. This particular ward was chosen because it housed serious orthopaedic cases, and i t was felt that the effect of hospitalization in this Pavilion might parallel the effect in the de-pressive wards of the psychiatric hospital. Most of the 16 subjects were amputees, either confined to bed or beginning to make the adjustment to wheelchair and/or a r t i f i c i a l limbs. The order of test presentation varied slightly from one group to the next. The nonhospitalized control group was administered the Ishihara plates followed by the 100 Hue Test, and then by the Luscher test. For the depressives, the Luscher test was given before the 100 Hue Test. This was done because the administration warned that some of the depressives might not have the energy to complete a l l three tests. As a result a l l patients completed the pseudoisochromatic plates and the Luscher test, but three were not able to finish the last two boxes of the 100 Hue Test. The nonpsychiatric hospitalized group were not administered the 100 96 Hue test. Many could not freely move about in bed, and since the 100 Hue requires some concentration and effort, as well as some degree of manual dexterity, i t was decided not to subject them to any unnecessary discomfort. They were thus screened for colour vision deficiencies with the Ishihara plates only. Test illumination was provided by a portable Macbeth lamp which gives Illuminant 'C' lighting conditi ons. Those few subjects who were tested in the visual laboratory were tested under the Illuminant discussed previously (see page 50) and were presented the Luscher cards in the grey cubicles. The remainder of subjects were tested in darkened rooms. In these rooms, the Luscher cards were presented on a piece of grey cardboard which functioned as a neutral background. For those subjects who could only raise their heads and shoulders however, the cardboard had to be raised at an angle so the cards could be seen. In these situations the Macbeth lamp was angled appropriately from a bedside table. Results Table 10 gives the mean rank ordering by preference of the eight Luscher colours for the three groups, based on the average ordinal position of each colour in each group. Note that yellow is preferred on the average in sixth position by the depressives compared to third by the two non-psychiatric groups. Green is most preferred by the two hospitalized groups, but is fifth in relative preference for the nonhospitalized group. All three groups have similar preferences for the black and the grey. The rank order correlation coefficients for the three possible two-group comparisons are given in Table 11. A significant correlation implies that there are similarities in the overall colour preference 97 TABLE 10 MEAN RANK ORDERING OF LUSCHER COLOURS BY PREFERENCE FOR THREE GROUPS Colour Dark-blue Blue-green Orange-red Bright yellow Violet Brown Black Grey Depressive 4 1 3 6 2 5 7 8 Nonpsychiatric Hospitalized 5 1 2 3 4 6 7 8 Nonpsychiatric Nonhospitalized 4 5 2 3 1 6 8 7 between two groups. Note that depressives have similar colour preferences to the two nonpsychiatric groups, and that i t is the two nonpsychiatric groups that dif f e r most in their respective colour preferences. TABLE 11 RANK ORDER CORRELATIONS BETWEEN PREFERENCES OF THE THREE GROUPS Group Comparisons Spearman's Significance P D and NH D and NN NH and NN +0.810 +0.666 p < .05 p < .05 p > .05 +0.642 D: Depressive; NH: Nonpsychiatric, hospitalized; NN: Nonpsychiatric, nonhospitalized 98 Two d i f f e r e n t methods were chosen to examine the data on i n d i v i d u a l preferences. A f i r s t method was to determine i f there were differences between the three groups when preferences were interpreted psychologi-cally-, according to the LUscher method. This involved giving each subject a score as was done with the c o l o u r - d e f i c i e n t versus colour-normal groups. Remember that the higher the score, the greater the 'degree of psycho-pathy' . The basic group dif f e r e n c e s can be seen i n Table 12. Analysis of variance showed these differences to be s i g n i f i c a n t with F (2,45) = 3.68; p < .05. An orthogonal comparison between the p s y c h i a t r i c and the two nonpsychiatric groups showed the differences to be s i g n i f i c a n t at the .01 l e v e l of s i g n i f i c a n c e . The remaining orthogonal comparison between the two nonpsychiatric groups (see Table 13) was not s i g n i f i c a n t . TABLE 12 GROUP DIFFERENCES IN COLOUR PREFERENCE SCORED BY LUSCHER METHOD Group x s Depressives 3.18 2.90 Nonpsychiatric, 1.18 1.38 h o s p i t a l i z e d Nonpsychiatric, 1.44 1.87 nonhospitalized A second method was chosen to determine i f there were dif f e r e n c e s between the groups when no psychological i n t e r p r e t a t i o n was given to the preferences. What was needed was a m u l t i v a r i a t e system of analysis which could handle o r d i n a l data without having to assign a score to each •r -!roo1 -$o -#<r ~fr>' -U> -h'o -Jo 100 TABLE 13 SUMMARY OF ANOVA Source SS df MS F P Total 256.82 47 • Group 36.13 2 18.065 3.68 <.05 Vl: D vs NH, NN 37.50 1 37.50 7.90 <.01 V2'> NH vs NN 0.5 0.50 <1 <.05 D: Depressive; NH: Nonpsychiatric, hospitalized; NN: Nonpsychiatric, nonhospitalized individual. In consultation with a Social Science computer specialist, the Guttman-Lingoes Scalogram Analysis was selected (Lingoes, 1968). Basically, the method is like factor analysis, and begins by computing correlations between individual sets of ordinal data (between each subject's rank order of the eight Luscher colours, notated by their identifying numbers). Each subject can then be plotted in a two-dimensional space. The collection of points is called a scalogram. The less the distance between any two points, the more similar are the colour preferences of those two persons. If colour preference on the Luscher test were a perfect means of discriminating between groups, the scalogram would show three dis-tinct clusters of points, with perhaps some overlap. If colour prefer-ence were not a good means of discriminating between groups, then the points would not f a l l into any apparent clusters, but rather would appear to be arranged in a random fashion in the scalogram. Fig. 13 is a reproduction of the scalogram. Each point.represents 101 one of the 48 subjects, and has been notated appropriately for group membership. The two axes do not in themselves mean anything. The units on the axes similarly are arbitrary. . Remember that distance on the scalogram is the means of gauging similarities in colour preference. The points numbered 1 and 2, for example, are separated by the most distance on the scalogram, while the points numbered 3 and 4 are very close to each other. The actual colour preferences of these four subjects are given in Table 14. TABLE 14 COLOUR PREFERENCES OF FOUR SUBJECTS SELECTED FROM SCALOGRAM 1 Subject 1 V Dissimilar 2 Gn 3 Gn Similar 4 Y BI: Blue; 'Gn: Green; Rd: Red; Y: Yellow; V: Violet; Br: Brown; Bk: Black; Gy: Grey Note that subjects 1 and 2 do not share any similarities in colour preference for the eight Luscher colours. No single colour is liked or disliked to the same degree by the two subjects, yet they are both depressives. Subjects 3 and 4 on the other hand, have the same relative preferences for red, brown, blue, grey and black. These two individuals are not members of the same group, one is a depressive, the other non-Ordinal Position 2 3 4 5 6 7 8 BI Bk Gy Gn Br Rd Y Y Br Rd V Gy BI Bk Rd Br V BI Y Gy Bk Rd Br Gn _B1 V Gy Bk 102 psychiatric, nonhospitalized, but both are female and both approximately the . same are. On the scalogram there do appear to be clusters of points, the boundaries of which have been suggested by the dotted lines. These clusters do not however, correspond in any way to the three groups in question. Each apparent cluster contains points which represent subjects from each of the three groups. If the 48 points are divided into two groups — psychiatric versus nonpsychiatric •— i t is s t i l l not possible to make sense out of the arrangement of points in the scalogram. If they are looked at in terms of sex, or in terms of age, or where concerned in terms of length of hospitalization, there is no relationship between the arrangement of points and the variables in question. The arrange-seems random. Discussion This application of the Luscher test does not support Luscher's statement that depressives long for yellow - yellow was found to be the third least liked colour among the Crease Clinic patients. A number of psychiatric patients indicated that the test yellow was "too bright", but stated that they generally liked the colour yellow because i t signified sunshine and the environment outside the hospital. The blue-green of the test similarly reminded many of them of nature and the environment outside the hospital, but because the blue-green card was "not too bright", i t was more highly preferred than yellow. It seems then, that while a viewer may adjudge a symbolic 103 value to 'a Color' (Yellow) and find i t to be pleasant, the percep-tion of a particular 'representative colour' (the test yellow) may be found to be unpleasant. Two groups of hospitalized patients were involved in the study, one hospitalized for medical illnesses, the other for psychological i l l -ness. Both groups had significant similarities in their average rank ordering of the eight colours, yet when individual preferences were sub-mitted to scoring procedures, the psychologically i l l group was clearly different from the medically i l l group. The former certainly did 'worse' on the Luscher test than did the latter. However, since both groups were matched for length of hospitalization, patient by patient, i t is not likely that hospitalization is the cause of the preference differences. The most interesting finding to emerge from this study is the dif-ference between assigning meaning to preferences on the Liischer test and not assigning meaning. When meaning is imputed, that i s , when an in-dividual is scored for degree of psychopathy on the basis of where he places the primary and auxiliary colours, the test seems to be able to differentiate a psychiatric from a nonpsychiatric sample. However when there is no imputation of meaning onto preference, that is when preferences are examined as ordinal data only, then there is no rela-tionship between an individual's preferences and his psychiatric status. The test can only differentiate on a group basis; individual preferences (and hence scores) do not give sufficient information about the psychiatric status of the preferee. Four of the depressives for ex-ample, obtained a score of zero, indicating a stable,• normal personality, 104 On the other hand, 16 out of 32 nonpsychiatric subjects obtained scores of 1 or more, indicating a tendency away from a stable, normal person-ality, according to Luscher. This is not the kind of data which lends support to the concurrent validity of a test. These findings are not unexpected in light of the previously men-tioned studies dealing with colour and personality. The complexity of the relationship between the response to colour and personality variables has only been acknowledged, but is not really understood. The LUscher test was designed as a personality test more than 20 years ago, before any systematic research had really been started. It is therefore not surprising that i t can only distinguish between psychiatric status with the use of a built-in interpretive system. The unabridged edition may prove to be somewhat more proficient in this respect. \0b VII. STABILITY OF LUSCHER COLOUR PREFERENCES Investigations of colour preference have generally tended to test subjects only once, the result being that l i t t l e is known about the stability of colour preference over a period of time. A few studies have considered the problem of stability by including some form of the test-retest method in the experimental design. Warner (1949) for example, presented each of his colour pairs four times to psychiatric patients and found consistency from t r i a l to t r i a l . Barrett and Eaton (1947) retested subjects after various intervals (average 6 months) and found a high degree of agreement between the first and second sessions. Haas (1963) similarly found colour preferences to be reliable after a four week period. To the author's knowledge, these are the only studies that have incorporated retesting into the experimental design. Granger (1973) compared the results of an early preference study (1955) with preferences collected more recently (1970s) and found that many changes had occurred over the 15 years. He suggested the changes were due to cultural-societal values and attitudes, reflected in a dif-ferent colour consciousness than had previously existed. The problem of stability of colour preference becomes important when considering the Luscher test, where colour preference is the key to personality description. If the preferences are not the same over a period of time, i t follows that the resulting descriptions of personality will change accordingly. The following is a pilot study which addresses _ 105 106 itself to this problem, and is intended specifically to determine i f the Liischer test gives a personality portrait which is consistent over a number of weeks. Method Fifteen subjects, eight male and seven female, participated in a longitudinal colour preference experiment. Subjects ranged in age from 20 to 28, and included graduate and undergraduate students. A l l subjects were screened with the 10th edition of the Ishihara plates and with the 100 Hue Test. The Luscher test was administered individually to each subject once a week for six weeks. In some instances i t was not possible to contact a subject exactly one week after the previous session, however no inter-test interval was less than 6 days or more than 10. Each administration of the test included two presentations of the cards, with a 30-second interval between trials. The cards were presented in a different random order on each presentation. Subjects were asked to try to not consciously duplicate their responses from previous weeks, and were asked to try to respond to the cards spontaneously each time. All testing was done in the Visual Laboratory under Illuminant 'C' conditions, with the subjects seated in grey cubicles. Results Fifteen subjects tested twice a week yields 180 individual rank orders of eight colours, 90 of these representing colour preferences on a second t r i a l . For a l l subjects over six weeks, each colour changed 108 less than one position in the order of colours from the first to the second t r i a l . Most subjects stated that they could easily remember their preferences 30 seconds previously, however a few subjects did exhibit considerable changes from t r i a l to t r i a l . Changes were generally made because certain colours "looked better" when displayed in the new random 'order. No particular analysis was applied to the change in pref-erence which took place from minute to minute. For the changes which took place from week to week a scalogram analysis was conducted. The raw data consisted of 90 first t r i a l sets of ordinal data, each of which was correlated to each other, yielding 3.960 correlation coefficients. Only two of these were +1.00, meaning that only twice were two colour preferences the same, and these two were preferences of one individual on different weeks. The .correlations were factor analyzed and plotted in two-dimensional space. Fig. 14 is a reproduction of the resulting scalogram. Each sub-ject is represented by a number, which occurs six times, once for each week. The subscript following the number indicates which one of the six weeks preference was tested, ^ being the first week, ^the sixth. The scattering of points on the scalogram shows the extent of colour preference differences on the Luscher test. It also indicates that most individuals do not have consistent preferences from week to week. The points connected by lines (subjects #2 and #11) are examples of how different the preferences were from one week to the next. The points that have been circled (subjects #7 and #10) on the other hand, are examples of consistency in preferences. The actual colour preferences 109 for these four individuals are given in Table 15. Note that #7 and #10 show the same relative preferences for at least four of the eight colours each week, with few exceptions. Thus the personality descrip-tions do not vary considerably. Individuals #2 and #11 do not exhibit this consistency. TABLE 15 COLOUR PREFERENCES OF FOUR SUBJECTS OVER SIX WEEKS Ordinal Position Inconsistent W e e k l g t 2 n d 3 r d ^ ~^ t h g t h ? t h g t h Subiects #2 #11 A Rd BI V Y Bk Br Gn Gy B Gn BI Rd . Br V Y Gy Bk C Rd Y Bk Gn V BI Br Gy D Rd BI Br Gn V Bk Gy Y E Rd Y Br BI Gn V Gy Bk F Rd Bk BI V Br Gn Y Gy A BI V Rd Br Bk Gn Gy Y B Bk BI V Br Gn Gy Rd Y C Br BI Bk V Gy Y Rd Gn D V Bk Br BI Gy Rd Y Gn E Bk V BI Br Gn Gy Rd Y F . BI Br Bk V Gy Gn Rd Y Consistent Subjects #7 ' #10 A V Gn Rd Br BI Gy Y Bk B Gn Rd V Br BI Gy Y Bk C Gn V Br Rd BI Gy Y Bk D Rd Gn V Br BI Gy Y Bk' E Rd Br V Gn BI Gy Y Bk F Rd Gn Br V BI Y Gy Bk A BI Gn Y Rd Gy V Bk Br B BI Gn Rd Y Gy Br Bk V C BI Gn Gy Y Rd Br Bk V D BI Gn Y Rd Gy Br Bk V E BI Gn Rd Gy Y Br Bk V F BI Gn Gy Rd Y Br Bk V BI: Blue; Gn: Green; Rd: Red; Y: Yellow V: Violet; Br: Brown; Bk: Black; Gy: Green 110 Only the two subjects #7 and #10 have similar preferences from week to week; the remainder resemble the pattern of change exhibited by #2 and #11. Most subjects have at least two weeks in which preferences are similar, but these weeks do not necessarily occur next to each other. The sex variable does not help to explain the data. Discussion A pilot study with a small sample size always limits the conclu-sions that can be drawn from the results. Nevertheless, this study sug-gests that colour preference on the Luscher test is subject to consider-able change in a six-week period (with the exception of two subjects). These findings do not agree with Barrett and Eaton's (1947), or with Hass' (1963), who reported consistencies in preference over time. They do agree with the findings of Warner (1949) however, who reported an agreement of preference at the same sitting. The immediate implication of these results is that colour preference is not a stable phenomenon, which within the interpretive scheme of the Liischer test, implies in turn, that personality is not a stable phenom-enon. This is highly unlikely, and most theorists of personality would argue against such a conclusion. Certain subjects showed changes in pref-erence from the first to the second t r i a l , changes which would render a different description of personality when the preferences were interpreted. It has been said that personality is a dynamic process, yet few individuals would suggest i t could change that quickly from one minute to the next. Alternatively the results suggest that the test is not measuring personality at a l l , but something else — something which is more I l l transitory in nature, such as 'mood' or what one might refer to as a 'general feeling'. The changes in colour preference which take place from week to week would then reflect differences in mood, which is more subject to rapid change than is personality. One could argue that the design of the experiment itself supports a mood-hypothesis, that is when persons are asked to indicate preferences each week for six weeks, that they are then susceptible at the time of testing to such factors as boredom or impatience, factors which would thus be responsible for the changes in preferences. On the other hand, these same or similar mood factors are not just operative in longitudinal sessions of preference testing, but are likely to be operating to affect the arrangement of colours in any single testing session. The effect of mood on pref-erences is an area which requires some elucidation. A third alternative suggestion is that the test is measuring neither personality nor mood, but colour preference only. One must remember that the Luscher test is a special kind of preference test, that the colours have been chosen for particular reasons, and that the eight colours do not represent the f u l l range of colours normally included in colour preference studies. Thus Luscher-preferences do not necessarily reflect colour preferences in the wider sense. On the Luscher test subjects are confronted with eight colours at a time, arranged next to each other in a row. The following week, the colours are physically the same, but the arrangement is different, producing a different visual effect, and a different stimulus value to the task. Certainly the perception of the task is not the same each week. 112 Each time a different aesthetic situation is involved, where a colour is preferred not necessarily on its own merit, but on the basis of complex factors such as harmony of adjoining colours, ambiguity of ad-joining colours, colour contrast, or in short, by the Gestalt of the arrangement. Since the original stimulus situation changes from week to week, i t is not surprising that the resulting responses will change accordingly. One could conclude then, that for some persons Luscher-preferences are a fleeting phenomenon with no solid foundation, due in part to the effect of mood factors and in part to the effect of the Gestalt. For other persons, preferences are more stable, particularly for certain colours, and are strong enough to override the effects of mood and/or the Gestalt. These differences alone suggest there are personality factors involved, but not of the simplistic nature as those advocated by Luscher to be reflected in preferences. I'3 SUMMARY CONSIDERATIONS Max Luscher, in his discussions of colour psychology, pays tribute to Goe*the who was perhaps the first to consider that coloured stimuli can e l i c i t responses which are subject to a meaningful interpretation. In 1810 Goethe wrote For some of our early observations we can conclude that general impressions produced by single colours cannot be changed, that they act specifically and must produce definite, specific states in the living organ. They likewise produce a corresponding in-fluence on the mind. (1890, p. 305)-J-Since Goethe's treatise on colour i t is well established that individuals do respond differently to colours, that some will find certain colours exciting and stimulating, while others will find those same colours disturbing or depressing. S t i l l others do not seem to react at a l l . The reactions of most individuals are mixed, and will vary with the particular colour or colours under observation. Most importantly, these reactions will be determined to a large extent by the context of the ob-servation, that is to say, by the object which is 'being coloured'. A reading of the Luscher test shows i t to be a naive attempt to systematize meaningfulness into colour responding, an attempt which avoids the complexities of the problems at hand. In no instances "does the literature of colour psychology justify the use of colour preferences as a complete diagnostic aid for personality assessment. On this basis alone, one is forced to doubt the sincerity of the designer as well as his scientific training. Experimental investigations similarly cast doubt on the validity of the test and point to its weaknesses. 113 114 The study with the depressive sample of subjects showed that the internal interpretive structure of the test is suitable for making a discrimination between a psychiatric and two nonpsychiatric populations. But such discrimination does not help in the assessment of personality, which is what the test claims to do. The personality profiles which emerge from the colour preferences are varied and are not characteristic of depressive subjects. What the test did show were differences in colour preference between the groups, and this is a matter which will require further study. The longitudinal study suggested that i f Luscher-preferences are taken literally as the key to personality assessment and description, then personality seems to change from minute to minute, from week to week, and from one month to the next. More than likely however, the results are suggestive of reactions to a task which calls for an aesthetic response of a different nature than colour preference alone. The task produced different reactions. Some persons were affected by the combinations of colours while others remained unaffected and adhered to original preferences. This difference may well reflect personality variables more than does the actual choice of colours. The colorimetric study suggests that colorimetric analysis should be applied to every projective test which uses colour. The use of such analysis with the Luscher test has been valuable, and the extent of the differences in colours between versions is now known. If the three versions of the Luscher test are any indication of future tests which will be published, then the question of standardization for test stimuli 115 must be given more emphasis. Colorimetric analysis is also necessary to determine i f colour-deficient subjects can use a test. Certain colours on the Luscher test were found to be indistinguishable in hue to some dichromats, and i t is suggested that other existing tests, notably the Rorschach, and Colour Pyramid Test, be examined similarly. It is of interest that colour may be thought of both as a stimulus and as a response. The difference has to do with whether colour is con-ceived of as something external to the organism, or as an inner experience. If colour is conceived of as an external stimulus, i t is generally capable of being measured objectively, and is therefore the subject matter of colour measurement. Colour measurement however, can include not only measurement of the standard stimulus variables (hue, saturation, bright-ness) but also measurement of variables a t t r i b u t e d to colour (its relative warmth, its affective value, its symbolic meaning). Some variables of colour are therefore less susceptible to measurement than are others. Colour conceived of as an inner experience is the subject matter of perception in the wider, sense. The experience (the colour response) is of course related to the properties of the stimulus which produce i t (the subject matter of psychophysics), but the nature of this relation-ship is exceedingly complex, and at best only partially understood. Satisfactory measurement of colour responding then, is also difficult to accomplish. The inherent problems in assigning a psychological interpretation 116 to colour preference are firstly, that the stimulus in question is multidimensional, and secondly that people prefer colours for a mul-titude of different reasons. There can be no objective system of rules which will explain colour preferences, just as there can be no guarantee in a preference study that subjects will respond on a single dimension. What do laboratory instructions such as 'make your preferences on the basis of the colour alone' really mean? Is i t possible to judge a colour on its own merit, without association or bias? Luscher argues that each colour, because of its intrinsic symbolic meaning, has the same effect on a l l mankind, but that each man in accordance with his personality, responds to a given colour stimulus in his own way. Hence the personal response of each man towards the effects of colour and their effects upon a l l men allows for an interpreta-tion of this man's personality, in accordance with the methods in the Luscher test. If i t is true that each man responds uniquely to each colour, and there is no argument with Liischer's statement on this point, then the question appears: how do we differentiate between the effects of a colour on a man, and his personal responses to such an effect? It is significant that Luscher does not formulate the problem as a scientific one. He lacks the methods of investigation and apparently follows his intuition. Intuition is valuable i f i t leads towards the formulation of problems to be investigated, and may even be convincing to a more or less degree. But i t cannot take the place of systematic research. Luscher also assumes that the test stimuli can be presented as 117 colours thought abstracted from outline, surface, texture, or mass of objects which man is familiar with in his environment. But the selected stimuli are presented on rectangular pieces of cardboard, and hence are presented in a specific form with a specific texture (also depending on which version of the test is used). Each colour will also be modified by whatever interrelation the person selects from the visual Gestalt. There seems to be no technique which will avoid the above compli-cations, and a l l Luscher can hope for is that this does not matter, as indeed he does, without supporting evidence from adequate experimentation. These factors are not trivial to the study of colour preferences. A mahogany brown on stained wood 'feels' very different from the same brown painted on a flat surface of concrete, or on a surface of iron. Similarly, two colours may be most attractive when they appear side by side on a flowering bush, and they may be unpleasant when they appear side by side on two cards of rectangular shape. It is known that combinations of colours-forms-masses-textures as met in everyday l i f e do affect a person's mood. When so formulated i t sounds rather foolish even to mention what is obvious. I am convinced that .these matters can be investigated, and even used eventually as diagnostic tools in clinical settings. At present however, the state of knowledge about these matters does not justify their inclusion into the realm of personality research. They must remain for the time being as the subject matter of aesthetics. FOOTNOTES 1. In addition to the Rorschach Inkblot Test, there is also the Colour Pyramid Test (Pfister, 1950), the House-Tree-Person Test (Hammer, 1955), the Polychrome Index (Woods, 1956), the Brentwood Color Test (Spiegel and Acker, 1960), and coloured versions of the Thematic Apperception Test. 2. The Kansaku study dealt with connotative meanings of colour pairs, while the Wright and Rainwater study involved colours presented singly. Procedural differences may explain why the findings regarding brightness and saturation are different. 3. Colour preference has also been studied in the retarded child (Gillibrand, 1970), and in a group of psychiatric aged patients (Mather et al., 1971), and while these groups do not represent a normal population, neither do they represent the degree of psychopathology found in the sub-jects of the other studies mentioned. The most preferred colours were found to be respectively, red and blue. 118 1/9 BIBLIOGRAPHY Aaronson, B.S. Hypnotic induction of colored environments. Perceptual and Motor Skills, 1964, 18, 30. Allen, R.M. The influence of color in the Rorschach Test on reaction time in a normal population. J. of Proj. Techniques, 1951, 15, 481-485. Allen, E.C. and Guilford, J.P. Factors determining the affective values of colour combinations. Amer. J. of Psychol., 1936, 48, 643-648. Alschuler, R.H. and Hattiwick, L.A. Easel paintings as an index of personality in preschool children. Amer. J. Orthopsychiat., 1943, 13, 616-625. Ball. V.K. The aesthetics of color: A review of fifty years of ex-perimentation. J. of Aesthetics and Art Criticism, 1965, 23, 441-452. Barrett, D.M. and Eaton, E.B. Preference for color or tint and some related personality data. J. of Personality, 1947, 15, 222-232. Beebe-Center, J.G. The Psychology of Pleasantness and Unpleasantness. New York: Van Nostrand, 1932. Birren, F. Color Psychology and Color Therapy: A Factual Study of the Influence of Color on Human Life. New York: McGraw-Hill Book Co. Inc., 1950. Block, D.S. and Caldwell, W.E. Relationship between color and affect in paranoid schizophrenics. J. of Gen. Psychol., 1959, 61, 231-235. 119 120 Brengelmann, J.C. Farbwahl und Variabilitaet der Farbwahl im Farbpyramidentest. (Colour choice and v a r i a b i l i t y of colour choice i n the Colour Pyramid Test.) Z. ges. NeuroL Psychiat., 1957, 196, 297-306. Bricks, M. Mental hygiene value of children's art work. Amer. J. Orthopsychiat., 1944, 14, 136-146. Bullough, E. The 'perceptive problem' in the aesthetic appreciation of single colours. Brit. J. of Psychol., 1908, 2, 406-463. Bullough, E. The 'perceptive problem' in the aesthetic appreciation of simple colour combinations. Brit. J. of Psychol., 1910, _3, 406-447. Burnham, R.W., Hanes, R.M., and Bartleson, C.J. Color: A Guide to  Basic Facts and Concepts. London: John Wiley & Sons, Inc., 1963. Cerbus, G. and Nichols, R.C. Personality variables and response to colour. Psychol. Bull., 1963, 60, 566-574. Chevreul, M.E. The Principles of Harmony and Contrast of Colours, and their Application to the Arts. London: George Bell and Sons, 1890. Child, I.L., Hansen, J.A., and Hornbeck, F.W. Age and sex differences in children's color preferences. Child Development, 1968, 39, 237-247. Choungarian, A. Introversion-extraversion and color preferences. J. of Proj. Techniques and Personality Assessment, 1967, 31, 92-94. 121 Choungarian, A. Color preferences and cultural variation. Perceptual  and Motor Skills, 1968, 26, 1203-1206. Choungarian, A. Extraversion, neuroticism, and color preferences. 'Perceptual and Motor Skills, 1972, 34, 724-726. Cohn, J. Experimentelle untersuchungen ueber die Gefuehlsbetonung der Farben Hellgheiten, und ihre Combinationem. (Experimental investigations of the feeling quality of colours and brightness, and their combinations.) Philos. Studien, 1894, JLO, 562-603. Cohn, J. Gefuhlston und Sattigung der Farben (Feeling qualities in the saturation of colours.) Philos. Studien, 1900, 15, 279-286. Committee on Colorimetry. The Science of Color. New York: Thomas Y. Crowell, 1953. Crane, R.R. and Levy, B.I. Color scales in responses to emotionally laden situations. J. of Consult. Psychol., 1962, 26_, 515-519. Dorcus, R.M. Color preferences and color associations. The Pedagogical Seminary and J. of Genetic Psychol., 1926, 33, 399-434. Dubrovner, R.J., Von Lackum, W.J., and Jost, H. A study of the effect of color on productivity and reaction time in the Rorschach Test. J. Clin. Psychol., 1950, 6, 331-336. El l i s , H. The psychology of red. Pop. Sci. Mon., 1900a, 57, 365-375. El l i s , H. The psychology of yellow. Pop. Sci. Mon., 1900b, 57, 517-526. Eysenck, H.J. A critical and experimental study of color preferences. Amer. J. of Psychol. , 1941, 5_4, 385-394. Geissler, L.R. The affective tone of colour combinations. Titchener Commemorative Volume, 1917, 150-174. 122 Gerard, R. Color and emotional arousal. Am. Psychologist, July, 1958. Gillibrand, J.A. Colour preference in the retarded child. South African Medical Journal, 1970, _70, 12-13. Goethe, J.W. von. Naturwissenschaftliche Schriften, Vol. I, Zur Farbenlehre; Didaktischer Theil. (Essays on the Natural Sciences, Vol. I. On the Theory of Colour; Didactic Part.) Weimar: Bohlau, 1890. Goldstein, K. The Organism. New York: American Book Company, 1939. Goldstein, K. Some experimental observations concerning the influence of colors on the function of the organism. Occup. Ther., 1942, 21, 147-151. Granger, G.W. An experimental study of colour preferences. J. of Gen. Psychol. , 1955, 52, 3-20. Granger, G.W. 'Colour harmony' in science and art: Paper read at the Second Congress of the International Colour Association, University of York, 2-6 July, 1973. Guilford, J.P. The prediction of affective values. Amer. J. of Psychol., 1931, 43, 469-478. Guilford, J.P. The affective value of color as a function of hue, tint, and chroma. J. Exper. Psychol. , 1934, JL7_, 342-370. Guilford, J.P. There is system in color preferences. J. of Optical Society of America, 1940, 30, 455-459. Guilford, J.P. and Smith, P.C. A system of color-preferences. Amer. J. of Psychology., 1959, 72, 487-502. Haas, W.A. Investigation of the stability of color preferences. J. of Consult. Psychol. , 1963, 27^, 537-539. 123 Hammer, E.F. The H-T-P Clinical Research Manual. Beverly Hills, Calif.: Western Psychological Services, 1955. Hanes, R.M. Fact and fancy concerning derived reactions to color. 'International Colour Meeting, Sektion 2. Gottingen: Musterschmidt-Verlag, 1961, 359-366. Hesselgren, S. Phenomenology in general and as a method of research in the field of colour. Proceedings of the International Color Meet ing, Luzerne, June 1-4, 1965. Hevner, J. Experimental studies of the affective value of colors and lines. J. Appl. Psychol., 1935, 19, 385-398. Jastrow, J. The popular aesthetics of color. Pop. Sci. Mon., 1897, 50, 361-368. Judd, D.B. Blue-glass filters to approximate the black body of 6500°K. Fabre, 1961, 10, 31-36. Kansaku, J. The analytic study of affective values in colour combina-tions; a study of colour pairs. In: Hogg, J. (Ed.) Psychology  and the Visual Arts. Middlesex: Penguin Books, 1969, 345-347. Karwoski, T.F. and Odbert, H.S. Color music. Psychol. Monogr., 1938, 50, Whole No. 2. ; Katz, S.E. Color preferences in the insane. J. Abnorm. Psychol., 1931, 26, 203-211. Kelly, K.L. and Judd, D.B. The ISCC-NBS Method of Designating Colors  and a Dictionary of Color Names. Washington, D.C.: National Bureau of Standards, Circular 553, 1955. Klar, H. The Luscher Colour Test - a highly reliable procedure in the 124 psychodiagnostics of functional disorders. Medico (Boehringer Mannheim), 1961, 4_, 17-28. Klopfer, B. and Davidson, H.H. The Rorschach Technique: An intro- ductory Manual. New York: Harcourt, Brace & World, Inc., 1962. Knapp, R.H. Attitudes toward time and aesthetic choice. J. Soc. Psychol., 1962, 56, 79-87. Lakowski, R. Small colour vision variations and their effect in visual colorimetry. In: Colour Measurement in Industry. London: The Colour Group, 1967, 167-193. Lakowski, R. Theory and practice of colour vision testing: A review. Part 2. Brit. J. Industr. Med., 1969, 2j6, 265-288. Lakowski, R. and Tansley, B.W. Energy modification of the Pickford-Nicolson anomaloscope. Paper read at the Second International Symposium on Recent Advances in Colour Vision Deficiencies, June 28-30, 1973. Lawler, CO. and Lawler, E.E. Color-mood associations in young children. J. of Gen. Psychol., 1965, 107, 29-32. Lazarus, R.S. The influence of color on the protocol of the Rorschach Test. J. Abnorm.•-& Soc. Psychol., 1949, 44, 506-516. Lewinski, R.H. An investigation of individual responses to chromatic illumination. J. Psychol. , 1938, 6_, 155-160. Lingoes, J.C The multivariate analysis of qualitative data. Mult. Behav. Res. , 1968, 3_, 61-94. Luckiesh, M. Color and Colours. New York: Van Nostrand Co., 1938. Luscher, M. Liischer-Test. Anleitung und Farbtafeln. Klinischer Test zur psychosomatischen Persdnlichkeits-diagnostik. (The Luscher 125 Test. Manual and Colour Plates. Clinical Test for the Psycho-somatic Diagnosis of Personality.) Basel: Test-Verlag, 1948, (1969 publication used) . Luscher, M. Package design and colour psychology. Palette, (Sandoz Limited, Basel), 1960, 5_, 2-13. Major, D.R. On the affective tone of simple sense impressions. Amer. J. of Psychol. , 1895, ]_, 57-77. Mather, J., Stare, C., and Breinin, S. Color preferences in a geriatric population. Gerontologist, 1971, 11_, 311-313. Mogensen, A. and Juel-Nielsen, N. Factors influencing preference rankings in a special picture test and in Liischer's Colour Test: A study of uniovular twins brought up apart. Acta Psychiatrica Scandinavica, 1962, 38_, 208-212. Moon, P. and Spencer, D.E. Geometric formulation of classical color harmony. J. of Optical Society of America, 1944, 34. > 46-59. Munsell, A.H. Munsell Book of Color. Baltimore, Maryland: Munsell Color Company Inc., 1929. Munsell Color Company. Munsell Book of Color: Neighboring Hues Edition. Baltimore, Maryland: Munsell Color Co. Inc., 1950. Murray, D.C. and Deabler, H.L. Colors and mood-tones. J. Appl. Psychol., 1957, 41, 279-283. Nakshian, J.S. The effects of red and green surroundings on behavior. J. of Gen. Psychol., 1964, 70, 143-161. Newhall, S.M., Nickerson, D., and Judd, D.B. Final report of the O.S.A. Subcommittee on the spacing of the Munsell colours. 126 J. of Optical Society of America, 1943, 33_, 385-418. Norman R.D. and Scott, W.A. Color and affect: A review and semantic evaluation. J. of Gen. Psychol., 46, 185-223. Nourse, J.C. and Welch, R.B. Emotional attributes of color: A compar-ison of violet and green. Perceptual and Motor Skills, 1971, 32, 403-406. Odbert, H.S., Karloski, T.F., and Ekerson, A.B. Musical and verbal associations of color and mood. J. of Gen. Psychol., 1942, 26, 153-173. Osgood, C.E., Suci, G,J., and Tannenbaum, P.H. The Measurement of Meaning. Urbana: University of Illinois Press, 1957. Payne, M.C. Color as an independent variable in perceptual research. Psychol. Bull., 1964, 61, 199-208. Pfister, M. Der Farbpyramidentest. (The Colour Pyramid Test.) Psychologische Rundsshan, 1950, 1, 192-194. Pickford, R.W. Individual Differences in Colour Vision. London: Routledge and Keegan Paul, 1951. Pickford, R.W. Review article: The Liischer Colour Test. Occup. Psychol., 1971, 45, 151-154. Pressey, S.L. The influence of color upon mental and motor efficiency. Amer. J. of Psychol., 1921, 32, 326-356. Preyer, W. Mental development in the child. Psychol. Rev., 1897, JL, Monog. Suppl. Rickers-Ovsiankina, M.A. Some theoretical considerations regarding the Rorschach method. Rorschach Res. Exch. , 1943, 7_, 41-53. Rickers-Ovsiankina, M.A. Synopsis of psychological premises underlying 127-the Rorschach. In: Rickers-Ovsiankina, M.A., (Ed.) Rorschach  Psychology. New York: John Wiley & Sons, Inc., 1960, 3-22. Rickers-Ovsiankina, M.A., Knapp, R.H., and Mclntyre, D.W. Factors affecting the psychodiagnostic significance of color perception. J. of Proj. Techniques and Personality Assessment, 1963, 27, 461-466. Riffenburgh, G.H. Responses to color combinations as indices of personality traits. J. of Gen. Psychol., 1959, 61, 317-322. Rorschach, H. Psychodiagnostics: A Diagnostic Test Based on Perception. New York: Grune & Stratton, 1942. Sappenfield, B.R. and Buker, S.L. Validity of the Rorschach VIII, IX, X per cent as an indicator of responsiveness to color. J. Consult. Psychol., 1949, 13, 268-271. Schaie, K.W. On the relation of color and personality. J. of Proj. Techniques and Personality Assessment, 1966, 30^ , 512-524. Schaie, K.W. and Heiss, R. Color and Personality. New York: Grune & Stratton, 1964. Schachtel, E.G. On color and affect: Contributions to an understanding of Rorschach's Test II. Psychiatry, 1943, 6_, 393-409. Schofield, R.K. The Lovibond Tintometer adopted by means of the Rothamsted device to measure colour on the C L E . system. J. Sci. Instr., 1939, 16, 74. Scott, LA. (Ed.). The Luscher Color Test. New York: Random House, 1969. Scott, LA. (Ed.). The Luscher Color Test. New York, Pocket Books, 1971. 128 Shapiro, D. A perceptual understanding of color response. In: Rickers-Ovsiankina, M.A. (Ed.) Rorschach Psychology. New York: John Wiley & Sons, Inc., 1960, 154-201. Shikiba, T. Color preferences of deranged persons and delinquent boys. Jap. J. Psychol., 1927, 2, 677-700. Spiegel, D. and Acker, C.W. The Brentwood Color Test. Los Angeles: Authors, 1960. Spiegel, D. and Keith-Spiegel, P. Manifest anxiety, color preferences and sensory minimizing in college men and women. J. of Clin. Psychol.. 1971, 27, 318-321. Steinke, L. Farbpsychologische untersuchungen mit dem Liischer-Test bei angeborenen Farbsinnstorungen. (Colour psychology investigations with the Liischer-Test concerning inherited moods of colour.) Dissertation, University Eye Clinic, Basel, 1960. Suinn, R.M. Jungian personality typology and color dreaming, Psychiat. Quart., 1966, 40, 659-666. Suinn, R.M. Anxiety and color dreaming. Mental Hygiene, 1967, 51, 27-29. Trevor-Roper, P. The psychopathology of colour. Transactions of the Ophthalmological Society of the United Kingdom, 1969, 89_, 251-257. Wallen, R. The nature of color shock. J. Abnorm. & Soc. Psychol., 1948, 43, 346-356. Warner, S.J. The color preferences of psychiatric groups. Psychol. Monog., 1949, 63, Whole No. 6. Washburn, M.F., Haight, D., and Regensburg, J. The relation of the 129 pleasantness of color combinations to that of the colors seen singly. Amer. J. of Psychol., 1921, 32, 145-146. Wewetzer, K.H. Befunde bei Psychosen. (Findings with Psychotics.) In: Heiss, R. and Hiltmaun, H. (Eds.) Der Farbpyramiden-Test. (The Colour Pyramid Test.) Bern,. Switzerland: Huber, 1951. Wexner, L.B. The degree to which colors (hues) are associated with mood-tones. J. Appl.. Psychol. , 1954-38, 432-435. Williams, J.E. Connotations of color names among Negroes and Caucasians. Perceptual and Motor Skills, 1964, 18, 721-731. Williams, J.E. and Foley, J.W. Connotative meanings of color names and color hues. Perceptual and Motor Skills, 1968, 26, 499-502. Wilson, G.D. Arousal properties of red versus green. Perceptual and  Motor Skills, 1966, 23, 947-949. Wohlfarth, H. Versuche zur Bestimmung eines eventuellen. Effektes von Farbstimuli auf das autonome Nervensystem. (Experiments to induce some lasting effects of coloured stimuli on the autonomic nervous system.). Psychotherapie, 1956, 3., 216. Wohlfarth, H. Psychophysische Auswertung der Versuche zur Bestimmung eines eventuellen Effektes von Farbstimuli auf das autonome Nervensystem. (Psychophysical results from the experiments to induce some lasting effects of coloured stimuli on the autonomic nervous system.) Psychotherapie, 1957, 2_, 86. Woltmann, A.G. A contribution to the symbolic use of color in dreams. Psychoanal. Rev., 1965, 52, 94-105. Woods, W. A polychrome index of social maturity. Paper read at the 130 Chicago Meeting of the American Psychological Association, Sept. 4, 1956. Wright, B. The influence of hue, lightness, and saturation on apparent warmth and weight. Amer. J. of Psychol., 1962, 75, 232-241. Wright, B. and Rainwater, L. The meanings of color. J. of Gen. Psychol 1962, 67, 89-99. Wyszecki, G. and Stiles, W.S. Color Science: Concepts and Methods, Quantitative Data and Formulas. New York: John Wiley and Sons, Inc., 1967. Yazmajian, R.V. Color in dreams. Psychoanal. Quart., 1964, 33^ , 176-193 Yazmajian, R.V. Dreams completely in color. J. Amer. Psychoanal.Assoc. 1968, 16, 32-47. Zausznica, A. Nauka o Barwe. (The Science of Colour.) Warsaw: Paristowe Wydawnictwo Nankowe, 1969. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items