- tuc-v'l-.n‘fi . V

. 4h.-

 

  
 
    

 
 

’ LIBRAR Y '

. Michigan Stat:
University

we... lull“HIMlullllllljllllzllllllfllfll

3 1293

This is to certify that the
thesis entitled
ADAPTIVE REGRESSION AND ARTISTIC PREFERENCES
AMONG VISUAL ARTISTS

presented by

Martha Orrick

has been accepted towards fulfillment
of the requirements for

Master of Arts degree in P§ychology

[liar/fut

Major professor

 

 

 

Date July 6, 1979

 

0-7639

 

 

OVERDUE FINES ARE 25¢ PER DAY
PER ITEM

Return to book drop to remove
this checkout from your record.

 

 

ADAPTIVE REGRESSION AND ARTISTIC PREFERENCES

AMONG VISUAL ARTISTS

BY

Martha Orrick

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of'

MASTER OF ARTS

Department of Psychology
1979

ABSTRACT

ADAPTIVE REGRESSION AND ARTISTIC PREFERENCES
AMONG VISUAL ARTISTS
BY

Martha Orrick

Kris's theory of regression in the service of the
ego identifies two stages in artistic creation: emergence
into consciousness of drive-dominated material, followed
by its rational evaluation. It was hypothesized that art-
ists showing this positive response to internal disorder
will also respond positively to external disorder in the
form of complex or asymmetrical designs. Forty painters
and potters were administered the Revised Barron-Welsh Art
Scale and the Holtzman Inkblot Technique. Holt's system
for scoring primary process in Rorschach responses was
adapted to the HIT. Contrary to predictions, scores on
the art scale did not correlate significantly with adaptive
regression scores on the HIT. An interesting finding was
that there were no differences on these measures between
male and female artists. Speculations focused on the homo-
geneity of the sample and changes in artistic tastes since
the construction of the art scale. Suggestions for future

research were discussed.

For Joyce

ii

ACKNOWLEDGMENTS

I was told when I embarked on this study that artists
were a hard group to involve in research, notoriously sus-
picious of psychologists and their intentions. They may
have good reason to be. But I would like to put it on ree
cord that I found the community of artists in and around
Lansing, where this study was conducted, to be very welcom-
ing. The percentage of acceptances to my letter soliciting
their participation (67%) testifies to that. Subjects in
the study traveled to the interview and gave two hours of
hard-working time, without compensation. They were good-
humoured, Open, and as curious about the creative process
as I was. I am in their debt for what they taught me and
I hOpe they found some rewards in the experience, too.

The person I have most to thank for my introduction
to this community is Eileen McCollough, a gifted artist in
her own right. It was her interest in the project that
first encouraged me to believe that it would be acceptable
to the pe0ple I wanted to do it with. Polly Freeman and
Beverly Rollins also shared their enthusiasm and experience
with me at a critical juncture.

I would especially like to thank Dr. Roger Funk for
providing me with a congenial space for interviewing.
Others who assisted along the way with various needs and
tasks included Andrew Foster, Roger Halley, Jane Harrington,

Frear ‘Simons, and Zoli Zlotogorski.

iii

I could not have completed the study without the help
of Dr. Robert R. Holt. It was very gratifying to discover
that the author of much of the work which had inspired me
was himself so generous with his time and critical attention.
His willingness to advise me on my use of his scoring system
and his offer to check over some of my material himself was
crucial. This does not necessarily imply, of course, that
he endorses the study.

I have been very fortunate in my relations with my
committee. Dr. Robert A. Zucker and Dr. Albert Aniskiewicz
are men for whom I have great liking and respect. Their
perceptive criticisms of my work have been offered in a
spirit which makes me look forward to doing further re-
search--something I have not always assumed I would feel at
this stage. The chairman, Dr. Norman Abeles, a man whom
one of his students called "a professional's professional,"
has been my mentor since I first came to Michigan State.

He knows, I think, how grateful I am for the steadfast way
in which he has shared his expertise with me, but it is a
pleasure to have the opportunity to thank him formally.

For years I have been reading Acknowledgments which
end with a tribute to the author's spouse. Now I know why.
But I find it hard to believe many women in training for a
new profession have the degree of undiluted support I have

from my husband, James Malcolm.

iv

TABLE OF CONTENTS
Page
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . vi

I. INTRODUCTION AND OVERVIEW: THE STUDY OF
CREATIVITY . . . . . . . . . . . . . . . . . . 1

II. PSYCHOANALYTIC THEORIES OF ARTISTIC
CREATIVITY . . . . . . . . . . . . . . . . . . 9

III. EMPIRICAL STUDIES OF VISUAL ARTISTS . . . . . . 20

Observational Studies . . . . . . . . . . . . . 20
Personality Studies . .-. . . . . . . . . . . . 21
Rorschach Studies . . . . . . . . . . . . . . . 23
Other Instruments . . . . . . . . . . . . . . . 30
Studies Of Adaptive Regression in Artistic
Creativity . . . . . . . . . . . . . . . . . 33
Comments . . . . . . . . . . . . . . . . . . . 37

IV. HYPOTHESES AND PREDICTIONS . . . . . . . . . . 41
V. METHODS AND PROCEDURES . . . . . . . . . . . . 44
Subjects . . . . . . . . . . . . . . . . . . . 44
Instruments . . . . . . . . . . . . . . . . . 46
Holtzman Inkblot Technique . . . . . . . 46
Revised Art Scale Of the Welsh Figure
Preference Test . . . . . . . . . . . . . . . 53
Procedures . . . . . . . . . . . . . . . . . . 58
VI. RESULTS . . . . . . . . . . . . . . . . . . . . 60
VII. DISCUSSION . . . . . . . . . . . . . . . . . . 74
VIII. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . 83
APPENDIX . . . . . . . . . . . . . . . . . . . . . . . 85

REFERENCES . . . . . . . . . . . . . . . . . . . . . . 87

Table

l.

10.

ll.

12.

l3.

14.

LIST OF TABLES

Mean Scores Of Artist Groups on the Barron-
Welsh Art Scale . . . . . . . . . . . . . . . .

Mean Scores Of Various Groups on the Barron-
Welsh Art Scale . . . . . . . . . . . . . . . .

Mean Scores on the Revised Art Scale, According
to Sex, Medium, or Sex and Medium . . . . . . .

An Analysis Of Variance for the Effect of Sex
or Medium on Scores on the Revised Art Scale .

Correlations Of Adaptive Regression Scores with
Scores on the Revised Art Scale, According to
Sex, Medium, or Sex and Medium . . . . . . . .

Mean Adaptive Regression Scores, According to
Sex, Medium, or Sex and Medium . . . . . . . .

An Analysis of Variance for the Effect of Sex
or Medium on Adaptive Regression Scores . . . .

Mean Percentage Of Responses Scorable for
Primary Process, According to Sex, Medium, or
sex and Medium 0 O O O O O O O O I O I O O O 0

An Analysis of Variance for the Effect of Sex
or Medium on the Percentage of Responses Scor-
able for Primary Process . . . . . . . . . .

Mean Scores on Defense Demand, According to
Sex, Medium, or Sex and Medium . . . . . . . .

An Analysis Of Variance for the Effect of Sex
or Medium on Defense Demand . . . . . . . . . .

Mean Scores on Defense Effectiveness, Accord-
ing to Sex, Medium, or Sex and Medium . . . . .

An Analysis Of Variance for the Effect Of Sex
or Medium on Defense Effectiveness . . . . . .

Pearson Correlations Of Scores on the Revised
Art Scale with Mean Scores for Defense Demand
and Defense Effectiveness and with Percentage
Of Primary Process, According to Sex, Medium,
or Sex and Medium . . . . . . . . . . . . . . .

vi

Page

61

62

64

64

66

66

67

68

69

70

70

71

71

72

I. INTRODUCTION AND OVERVIEW:

THE STUDY OF CREATIVITY

Fascination with the person of the artist and the
way in which a work of art comes into being must be almost
as Old as the activity Of art itself. The reason for this
fascination, no doubt, is that creativity is mysterious.»
Neither artists nor laymen have ever been able fully to
trace the sources of the artist's ideas, which seem to lie
outside ordinary awareness. The artist has therefore been
called either divinely inspired or insane, or, very fre-
quently, both. Plato, for example, wrote that a poet was
"a light and winged thing, and holy, and never able to com—
pose until he has become inspired, and is beside himself,
and reason is no longer in him." The purpose Of this, he
continues, is "in order that we listeners may know that it
is not the poets who utter these precious revelations while
their mind is not within them, but that it is the god him-
self who speaks, and through them becomes articulate tO
us" (The Ion, in Rothenberg & Hausman, 1976, p. 32).

Over the centuries since Plato this image Of the
artist has changed remarkably little. The 19th century
physician Lombroso agreed that the artist is not in his

right mind-~but he attributed it to congenital insanity

F4

rather than to the voice Of God (1891). The Romantic era
glorified the artist in his vivid relation to the irra-
tional. In popular culture the composer seeking inspira-
tion at the piano, or the painter attacking his canvas,
remains a familiar stock figure, as any number Of Holly-
wood film biographies can demonstrate. Perhaps one reason
this picture of the artist persists is that, distorted as
it may be, it matches at some points the artist's own ex-
perience. Anthologies Of creators' descriptions of the
creative process, such as those by Ghiselin (1952) or Rosner
and Abt (1970) yield abundant testimony to the feeling that
images or ideas often appear suddenly and unpredictably,
with the quality of having arrived from outside conscious-
ness.

A picture comes to me from far Off, who knows

how far, I divined it, I saw it, I made it, and

yet next day I myself don't see what I have done.

(Picasso, in Ghiselin, 1952,
p. 60)

With the advent of psychoanalysis, interest in the
artist was intensified, if not, in some ways, essentially
altered. The unconscious replaced the deity as the source
Of inspiration but the process remained ineffable. "Before
the problem of the creative artist," said Freud, "analysis
must, alas, lay down its arms" (1928/1961). In spite Of
this disclaimer, early analysts found the artist relevant
to their explorations in several ways. It was evident to

them that the productions of artists had much in common

with dreams and fantasies and could be studied in the same
way for clues to the Operations of elusive regions Of the
mind. As the complexities of human motivation unfolded
before them, they also recognized that great artists had
anticipated their insights. Freud, for example, drew

freely on literary works to supplement his clinical mater-
ial, as in his analysis Of Shakespeare's Macbeth and Ibsen's

Rosmersholm (1916/1957b). In other instances Freud treated

 

a work of art as a symptom, a reflection Of the infantile
conflicts Of its creator, as in his study of Leonardo da
Vinci's painting "Virgin and Child with St. Ann" (1910/
1957a).

Psychoanalytic and neo-psychoanalytic theories on
art, with their accompanying case studies, therapeutic con-
cerns, and critical essays, represent a continuous tradition
from Freud to the present. Major trends in this tradition
will be outlined in the following Chapter.

Gestalt psychology promised a different approach, de-
fining the creative experience in terms not of individual
psychodynamics but of cognitive processes (e.g., Wertheimer,
1945) which resolve a problem by reorganizing elements in
its field--a reorganization which takes place instantane-
ously, in a manner similar to the "flash" of inspiration or
insight. Arnheim's work has contributed more specifically
to the conceptualization Of the perceptual processes of the
Visual arts (1954, 1969). His theories, however, have

rarely been translated into experimental investigations.

Prior to 1950, psychoanalysis and Gestalt theory seem
to have accounted between them for the attempts Of psycho-
logists to understand and explicate the creative process
in the arts. Only three empirical studies of visual artists
appear before this date: Patrick, 1937; Prados, 1944; Roe,
1946. The only other source regularly cited in the liter-
ature is the work Of Wallas (1926), whose intuitively de-
rived phases Of the creative process were for some time
widely accepted as an accurate description.

By all accounts, 1950 was a watershed year in the
study Of creativity. Guilford's presidential address to
the American Psychological Association (1950) served to
announce, if not exclusively to inspire, the wave of invest-
igations which followed. Several explanations have been
Offered for this burgeoning Of interest in a previously
neglected area. Barron (1969, p. l) attributes it to the
destructiveness of World War II and the subsequent urgency
Of the need to understand and develop man's potential for
constructive, creative behavior. Stein (1969, p. 909)
identifies three sources: (1) the "cold war" and the avail-
ability of government funds for research leading to the
identification and development Of creativity in the
sciences; (2) concern over an ethos of conformity, as ex-
emplified by Whyte's "organization man"; (3) a general move-
ment in psychology away from the preoccupation with path-
ology and toward the study of health and superior function-

ing. This last has continued to expand, its major spokesman

being Maslow (1959) with his concept Of the self-actualizing
person.

A distinguishing characteristic Of Maslow's work and
of most of the post-World War II research is a broadening
Of the definition of creativity, leading to its separation
from the province Of the arts. Earlier speculations on the
creative person, such as Galton's work on inherited abil-
ities (1870), or even Lombroso's ideas on insanity and cre-
ativity (1891), may be subsumed under the more general head-
ing Of genius. Terman's monumental study of gifted individ-
uals (1925-1959) also belongs to this tradition, a tradition
which is consistent with the time-honored View Of the artist
as set apart. By contrast, in many studies Of the 50's and
60's which used artists as subjects, the tenor of the re-
search was tO regard them either as carriers Of certain
traits believed to be factors in all creativity, or as cri-
terion groups for the development of tests designed to
identify creative individuals. Even the investigators at
the Institute for Personality Assessment and Research at
Berkeley, who sought--and gained-~the participation of
highly successful and eminent creators, applied their find-
ings toward a generalized portrait of the creative person
rather than toward an analysis of unique achievement. A
democratic attitude toward creative abilities was clearly
articulated by Guilford in his presidential address, when
he cited as one Of the reasons for the neglect Of creativ-

ity by psychologists the difficulty of studying the

phenomenon if research subjects are restricted to superior
performers. He went on to say,

It is probably only a layman's idea that the crea-

tive person is peculiarly gifted with a certain

quality that ordinary people do not have. This con-
ception can be dismissed by psychologists, very likely
by common consent. The general psychological convic-
tion seems to be that all individuals possess to some
degree all abilities, except for the occurrence of
pathologies. Creative acts can therefore be expected,
no matter how feeble or infrequent, Of almost all in-
dividuals. The important consideration here is the
concept of continuity. Whatever the nature of crea-
tive talent may be,those persons who are recognized
as creative merely have more Of what all of us have.

It is this principle Of continuity which makes pos-

sible the investigation of creativity in people who

are not necessarily distinguished.
(1950, p. 446)

One outcome Of this change of focus from the excep-
tional to the less-than-distinguished creator was a wide-
spread series Of attempts to develOp tests for identifying
potential creativity in individuals who had not produced
anything Obviously creative. The major work in this field
has been done by Torrance (1962) on the identifying of
talent in school settings; by Getzels and Jackson (1962)
on the relation of creativity to intelligence; and by
Guilford himself (1959), using factor analysis to isolate
the cognitive abilities associated with creative thinking.

Nicholls (1972) suggests that there are two major
approaches to the study Of creativity in normal popula-
tions: Guilford's approach, in which a trait (or traits)

Of creativity is assumed and its distribution is examined,

and a product-based approach, in which individuals ranked

according to criteria of creative achievement are examined
on a number of psychological dimensions, without making the
assumption that these dimensions in themselves constitute
creativity. An example Of the product-based approach is the
long-term project undertaken at the Institute Of Personality
Assessment and Research (IPAR) on the campus Of the Univer-
sity of California at Berkeley. The project lasted a dozen
years, and over the course of it hundreds Of people be-
lieved to be creative were tested, interviewed, and Ob-
served, many Of them in the course of weekends spent "living-
in" with the research staff in a house set aside for this
purpose. Among the groups studied were architects, writers,
mathematicians, industrial researchers, physical scientists,
and engineers. The results of these assessments have been
extensively reported in the literature (e.g., Barron, 1969;
Crutchfield, 1962; Gough, 1961; Helson, 1961, 1967; Mac—
Kinnon,1961, 1962). Although the IPAR investigators started
from no theoretical premise regarding the creative person,
Barron's portrait summarizing their findings gives support
to the psychoanalytic view, and returns us to the image of
the artist as one who Operates outside the bounds Of conven-
tional reality:

The creative individual, in his generalized
preference for apparent disorder, turns to the dimly
realized life of the unconscious, and is likely to
have more than the usual amount Of respect for the
forces Of the irrational in himself and in others.

This respect consists in a faith that the irra-
tional itself will generate some ordering principle

if it is permitted expression and admitted to con-
scious scrutiny. TO put the matter more strongly,

I believe that the creative individual not only re-
spects the irrational in himself, but courts it as
the most promising source of novelty in his own
thought. He rejects the demands Of society that he
should shun in himself the primitive, the uncultured,
the naive, the magical, the nonsensical; that he
must be a "civilized" member Of the community.
Creative individuals reject this because they want
to own themselves totally, and because they perceive
a shortsightedness in the claim Of society that all
its members should adapt themselves to a norm for a
given time and place.

When an individual thinks in ways which are cus-
tomarily tabooed, his fellows may regard him as men-
tally unbalanced. In my view this kind Of imbalance
is more likely to be healthy than unhealthy. The
truly creative individual stands ready to abandon
Old classifications and to acknowledge that life,
particularly his own unique life, is rich with new
possibilities. TO him, disorder Offers the poten-
tiality Of order.

(1958, pp. 163-164)

Couched in the psychological terminology of the
twentieth century, Barron's description seems an echo Of
Plato's belief that the poet composes by forsaking reason
and that although the state of inspiration may be abnormal
through it can come revelations with value for the entire

community.

 

This brief overview can do no more than set a con-
text for a study of artists. For more Comprehensive re—
views and bibliographies Of creativity research, the reader
is referred to Golann, 1963; Mackler and Shontz, 1965;
Rothenberg and Greenberg, 1974(a), 1974(b); Rothenberg and
Hausman, 1976; Stein, 1969; Stein and Heinze, 1960.

II. PSYCHOANALYTIC THEORIES OF

ARTISTIC CREATIVITY

Freud himself articulated no comprehensive general
theory Of art or the creative process. There are, however,
two sources in his works for adumbrating such a theory:
remarks on art scattered through his theoretical essays,
and papers analyzing individual artists and their works.
Summaries Of Freud's writings on art are Offered by FraiQ-
berg (1956), Jones (1957), and Sterba (1940). Sterba points
to the paradox that it is Freud's work on neurosis which
carries the ingredients of an art theory, while the studies
on art actually function to illustrate more general psych-
analytic ideas. For example, his psychosexual analysis of
an early memory of Leonardo da Vinci (1910/1957a) traces
the etiology Of sublimated homosexuality, not the creative
process, while his theoretical works on primary and second-
ary process are more applicable to the production Of a work
Of art.

According to the orthodox psychoanalytic View, art,
like the neurotic symptom, is born of conflict. In Formu—
lations on the Two Principles Of Mental Functioning (1911/
1958), Freud outlines the course Of development. The

earliest way of functioning, the "primary" process, follows

10

the pleasure principle. The infant turns toward pleasure
and away from pain. When pleasure is not forthcoming, the
infant is presumed to attempt to regain it by hallucinating
satisfaction, a tactic whose insufficiency compels the for-
mation Of a concept Of the real world and the adaptive de-
velopment of an ability to interact with it. This is the
"secondary" process, which follows the reality principle.
Motor activity, which previously served to discharge tension,
is now restrained and converted to thought. One type of
thought is split Off from this sequence and remains subject
to the pleasure principle: fantasy. The essence Of
Freud's idea Of art is that it reconciles the two principles
Of reality and pleasure. The artist is one who has resisted
the necessity of renouncing gratification in favor Of real-
ity and who continues to satisfy his desires in his fan-
tasies. Without resting there, however, the artist

finds a way back from this world of phantasy by

making use of special gifts to mould his phantasies

into truths Of a new kind, which are valued by men

as precious reflections of reality.

(1911/1958, p. 224)

Yet the truth is that the appeal of art is in its reflection
not of real life but Of the forbidden fantasy life, which
the artist's audience has not wished to renounce any more
than he.

In Creative Writers and Daydreaming (1908/1959),

Freud explores this idea further. Popular fiction makes

the origin in fantasy plain through its Obvious similarity

11

to the daydream. The popular writer allows us to enjoy his
daydreams--which are also ours--without shame. Sachs
(1942) elaborates on this "community Of daydreams" in his
work on the creative unconscious.

There are problems with these suggestions, as Jones
(1957) makes clear. First, the motivations Freud attributes
to artists are insufficient, especially when he reduces them
to "honour, power, and the love Of women." Second, his
speculations are prompted by the art Of fiction and are
less applicable tO the visual arts. In this connection,
Jones quotes Fry's critique Of Freud's neglect Of the
aesthetic factor.

With the exception of Sachs, psychoanalytic theor-
ists following Freud have not pursued the problem of aesthe-
tics, but have built on one or the other Of his original
ideas: art as a solution to conflict, namely the binding Of
sexual energy, and art as a reflection of the psychosexual
history of the individual artist.

Rank (1932), although he elaborates his portrait of
the artist considerably with the addition Of such concepts
as that of will, and although he contrasts rather than
identifies the artist with the neurotic, Still agrees with
Freud that sexual impulses are the root of conflict and that
art represents sexual energy transformed. Sachs (1942)
identified guilt over forbidden impulses as the artist's
motivation to create and the relief Of guilt as the reward

of creating. This theme of guilt persists in the theories

12

of other analysts such as Lee (1940) or Fairbairn (1938),
who repudiate the idea of sublimation.altogether and inter-
pret the making of a work of art as an act of restitution.

Havelka (1968, p. 67) ascribes the somewhat negative
or reductionistic trend of psychoanalytic explanations to
a dichotomy in Freud's work similar to that noted by Sterba
(1940): the distinction between Freud's psychopathology,
(which Havelka calls ”a moral position on mental health")
and his metapsychology, or psychology of mental functions.
It is the psychopathological theory, he says, which has
led to oversimplifications regarding art and to an unfort-
unate tradition of "detective work" on artists' neuroses,
on the assumption that these are the determining factors in
their work--in spite of the fact that the making of a work
of art requires a level of both spontaneity and control not
usually available to the neurotic. Freud's metapsychology,
on the other hand, provides the basis for a general theory
Of artistic creativity through its treatment of imagination
and the articulation of symbols, processes which, in Havel-
ka's phrase regarding the unconscious, assume "the property
of what might be termed an operational control function
rather than that of a mysterious, hidden repertory of vio-
lent but unknown mental events" (1968, p. 37).

Freud, While he initiated the ideas of the conscious,
unconscious, and preconscious systems, never fully pursued
their implications. This development has taken place in

recent years through the work of the ego psychologists,

13

especially Kris, whose concept of regression in the ser-
vice of the ego appears to have fulfilled some of the impli-
cations of the original theory. Kris restates Freud's
assertion regarding fantasy, while placing it in the context
of neutralized and non—neutralized energy at the disposal

of the ego:

Fantastic, freely wandering thought processes tend
to discharge more libido and aggression and less
neutralized energy; purposeful reflection and solv-
ing problems, more neutralized energy. In fantasy,
the processes of the ego are largely in the service
of the id . . . In reflective thinking the contrary
is likely. Reflective thinking, according to Freud
(problem solving as we would prefer to say) serves to
a higher degree the autonomous ego interests. Dis-
charge of libido and aggression is therefore likely
to be minimized, and that of neutralized ego energy
to be of greater relevance.

(1952, pp. 311-312)
Fantasy and the creative process are not synonymous for

Kris.

Topographically, ego regression (primitization of
ego functions) occurs not only when the ego is weak-—
in sleep, in falling asleep, in fantasy, in intoxi-
cation, and in the psychoses--but also during many
types of creative processes. This suggested to me
years ago that the ego may use the primary process
and not only be overwhelmed by it. This idea was
rooted in Freud's explanation of wit (1905) . . .
according to which a preconscious thought is "en—
trusted for a moment to unconscious elaboration,"
and seemed to account for a variety of creative or
other inventive processes.

(Ibid, p. 312)
The experience of creative inspiration takes place during
the phase of creation in which the ego withdraws its count-
ercathexis of id impulses and allows them, or their deriva-

tives, to enter preconscious thought. The sense Of

l4

revelation attested to by so many creative people is, in
Kris's View, linked with the passivity of this phase.

"The feeling of triumph and release from tension remind
the individual of a phase in his development in which
passivity was a precondition of total gratification, and
in which the hallucinated wish fulfillment became reality:
the period of nursing" (p. 317). The preconscious thought
is sexualized at the level of oral gratification.

The creative act, however, does not end with inspir-
ation. A phase follows which Kris calls "elaboration."
Countercathexis is re-established and energy is directed
to the ego functions of reality-testing--or secondary pro-
cess--by which the artistic work is formulated, criticized,
and prepared for communication to others. These two phases
are not simply sequential but are repeated in an oscilla-
tion which may recur repeatedly over a varying length of
time.

Schafer (1954, 1959) has further explicated the con-
cept Of regression in the service of the ego and has re-
lated it to performance on projective tests. Ambiguous
stimuli require what may be called a creative act. A
Rorschach or TAT card is the raw material which must be
shaped into an image or story. Regression is facilitated,
in that subjects are encouraged to take a non-critical,
playful attitude toward the task and to say whatever occurs
to them. The situation is passive; the percept is exper-

ienced as outside, existing in the stimulus, and

15

responsibility for it need not be fully assumed. Responses
will Show to what extent regression occurs or is prevented
from occurring, to what extent primary process material
arises and is expressed, and to what extent these primi-
tive Operations are brought, or not brought, under control.
Schafer discusses in general terms how this regression may
be analyzed, and Holt has provided a detailed scoring
system for measuring it in the Rorschach (Holt, 1970; Holt
& Havel, 1960).

Perhaps the earliest to reject the major premise of
Freud's theory was Jung, who drew a distinction between
the personal and the collective unconscious. The personal
unconscious he defined as the repository Of psychic pro-
cesses and contents repressed by reason of their unaccept-
ability. While acknowledging that this level may exert
influence over some artists, he called the products of its
influence "symptomatic art" and considered it an appropri-
ate object of reductive Freudian analysis (1928). The more
significant works, which he called "symbolic art," he be-
lieved had their source in a collective unconsCious, a
stratum of the personality common to all human beings.
Symbols formed at this level are no longer merely personal
in their genesis or their import; it is their universality
which accounts both for the power of great artists and the
depth of the response which their work awakens in others.

Other contemporary writers also find the hypothesis

of regression unnecessary or inappropriate, but for

16

different reasons than Jung. Schachtel, for example,
taking a more existential view, proposes the idea of auto-
centric, or subjectéoriented, and allocentric, or object-
centered, perception as a more parsimonious explanation
for creative experience (1959). The motivation for art,
in his View, is not the discharge of tensions generated by
libidinal drives but the free play of attention, thought,
feeling, and perception in an "open encounter" with the
world. His version of the stages of creating is that a
period of immersion in the various aspects of the object
is followed by a period in which the resulting insights or
perceptions are articulated, assimilated and given form

in the work. He compares the attitude of the artist with
the intensity of interest and the spontaneity of vision of
the child.

An especially outspoken critic of the concept of
regression in the service of the ego is Kubie (1958). In
his view, conscious and unconscious processes alike tend
toward rigidity: the conscious processes by virtue of
their literal and convention-bound relation to reality;
the unconscious processes because they are unconscious and
are therefore by their nature unmodifiable and unrelated
to the external world. Only preconscious processes are
flexible enough to serve as the basis of creative activity.

In between [the conscious and unconscious] come

the preconscious functions with their automatic

and subtle recordings of multiple perceptions,

their automatic recall, their multiple and over-
lapping linkages, and their direct connections

17

to the automatic processes which underlie affective

states. The rich play of preconscious operations

occurs freely in states of abstraction, in sleep,

in dreams, and as we write, paint, or allow our

thoughts to flow in the nonselected paths of free

aSSOCIatIon.

(Kubie, 1958, p. 44)

Creativity, in fact, depends on free association, which
Kubie calls analogous to the creative process-—"an analogy
which is so close as to be almost an identity" (ibid,
p. 54). In order to promote free association he agrees
that controls must be relaxed, but in the act of relaxing
them the creative process is exposed to the influence of
the unconscious by which it can only be deformed and dis—
torted.

What Jung, Schachtel and Kubie have in common seems
to be the desire to divorce the creativity of the artist
altogether from its Freudian connection with sexuality,
aggression and conflict. Art is more than primitive wishes
or sublimated drives, they are saying, it is a superior
form of functioning in its own right, in which any personal
neuroses of the artist can only be a hindrance, never an
intrinsic factor. Others have sought a similar formulation.
Arieti (1976), for instance, although he retains the con-
cepts of primary and secondary process (redefined without
dependence on the structural concepts of id, ego and super-
ego) does not consider them adequate to describe creativity,
and proposes the term "tertiary process" to suggest the

unique quality of the activity by which archaic and obso-

lete processes combine with others which are normal and

18

logical in what he calls "the magic synthesis" (1976, pp.
12-13). Arieti's emphasis is on cognition rather than on
libido. In this sense, it is similar to Kubie's and
Schachtel's definitions of creativity as the making of
associative connections.

Schafer himself speaks of a "progression in the
service of the ego," suggesting that a

. . . progression Of psychic functioning to a

level more advanced than the individual's usual

level may also be involved in the creative process.

The progression of psychic functioning will be

evident in the critical, elaborative, analytic and

synthetic manipulations of artistic images, scien-
tific concepts, or any creative material demanding
intense, hyperacute concentration, rationality or
sensitivity. We do rise above ourselves at times
tOO .
(1954, p. 81)

Whatever the theorist's orientation, there appears
to be little disagreement on the requirement for creative
work of a temporary loosening Of controls, the suspension
of criticism, and the ability to permit a fluid connecting
and disconnecting of associations, similarities, analogies,
and the like, without premature closure into familiar or
conventional configurations. Whether this process is
given the psychoanalytic title "preconscious," or is de—
scribed in terms of cognition or perception, its qualities
appear to be the same. Where disagreements occur Is in
regard to what is probably the major unresolved problem for

a general theory of artistic creativity: the question Of

whether primitive sexual and aggressive drives are an

19

inextricable or incidental component of the process.
Whether, to put it another way, the processes of problem
solving are a sufficient description of what occurs in
the production of a work of art, or whether additional
motivational or genetic considerations must be taken into

account.

III. EMPIRICAL STUDIES OF VISUAL ARTISTS

Observational Studies

 

Only three attempts have been made to observe art-
ists' behavior directly while they are engaged in creat-
ing. In the earliest, Patrick (1937) visited 50 artists
in their studios and asked each to draw a picture about,
or suggested by, Milton's poem "L'Allegro." She found
that the sequence of work generally followed Wallas'
(1926) four stages of creative thought: preparation, in-
cubation, illumination, and verification. Eindhoven and
Vinacke (1952) questioned Patrick's use of a priori cate-
gories and conducted their own observation of 13 artists
(comparing them to l4-non-artists). After watching the
artists develop drawings on a set theme over a period of
weeks, following a self-chosen schedule, these experiment-
ers concluded that the so-called stages are "not stages
at all, but processes which occur during creation" in
patterns which are complex and which may vary according
to individual differences (p. 161). More recently, Get-
zels and Csikszentmihalyi (1964), as part of a series of
studies Of students at the Art Institute of Chicago, have
made detailed observations Of 31 of these young artists

as they selected, arranged, and made drawings of objects

20

21

from an array provided by the experimenters. Photographs
were taken of the drawings as they developed and final
products were judged for aesthetic value, originality, and
craftsmanship. One of the authors' premises was that the
first step in creativity is the formulation of the problem
itself. The results of their study supported this: the
drawings rated most original and valuable were those by

the artists who had Spent the most time in exploring the
objects available before making their selections and begin-
ning to draw. The conclusion: "here, in what we have
labelled 'discovery-oriented behavior,‘ seems to lie a

key to the creative process." This behavior seems con-
sistent with the theories on preconscious processes de-
scribed above, in that flexibility and fluidity of associa-
tion appears to have been maintained by the artist. It
also illustrates Schachtel's "open encounter with the
world," with its resistance to quick or conventional solu-

tions.

Personality Studies

 

In contrast to this observational approach, are two
studies by Drevdahl (Drevdahl, 1956; Drevdahl & Cattell,
1958) in which the attempt to understand the creative pro-
cess is made by way of the personality and intelligence of
artists. In the 1956 study, 64 graduate and undergraduate
art and science students, divided into creative and non-

creative subgroups on the basis of faculty ratings, were

22

given Cattell's l6 Personality Factor Questionnaire,
Thurstone's Primary Mental Abilities Test, and tests from
Guilford's Factor Analytic Study of Creative Thinking.
According to the results, the more creative students (in
both art and science) scored higher on verbal facility,
fluency, flexibility, and originality. They were also more
withdrawn and quiescent. Creative art students were more
radical and self-sufficient, and both creative and non-
creative art students were more sensitive emotionally
and more bohemian.

In the 1958 study, Drevdahl and Cattell initially
contacted 356 Iactive and productive" artists and writers

(as judged by such evidence as listings in Who's Who and

 

ratings by a committee of experts). Of those contacted,
153 eventually completed and returned the Cattell 16 PF
Test. Their responses were compared to normative data

for the general population. The artists and writers
scored significantly higher on intelligence, ego strength,
dominance, adventurousness, emotional sensitivity, and
radical self-sufficiency, and were less subject to group
standards and control. The authors suggest that the find-
ings that creative people are neither conformist nor well-
adjusted should have implications for academic selection

criteria and for the goals of therapy.

23

Rorschach Studies

 

The majority of studies of visual artists have used
the Rorschach, many Of them in the hope of identifying
responses which might later serve as measures of artistic,
or other, creativity. Rorschach's own hypothesis was that
the protocols of creative individuals should show (1) a
high number of human movement responses, (2) a high number
of good form responses, (3) a high number of organized
whole responses, (4) a high number of original responses,
and (5) a low number of animal responses. The human move-
ment, or M response, in particular, he identified as an
indicator of imagination and the potential for creative
productivity (1942). Without attempting to review the lit-
erature on these response variables, Rorschach studies
of artists will be summarized, with attention to the above
hypotheses.

The earliest Rorschach study of artists is that of
Prados (1944). His subjects were 15 male and 5 female
painters, professionals at varying levels of success,
working in a variety of styles. In his summary of results,
he shows generally high totals of M responses, whole re-
sponses, and good form responses--all agreeing with Ror-
schach's hypotheses. He found "nothing striking," how-
ever, in the number of original responses. His conclu-
sions were that artists do share personality traits which
give common characteristics to their Rorschach records.

He suggests that the individual study of outstanding artists

24

may give a more detailed understanding of artistic crea-
tiveness.

Roe's study (1946) of 20 successful male painters
is frequently cited in the literature, both as a pioneer-
ing attempt to analyze the personality structure Of crea-
tive artists and as proof that the Rorschach cannot
identify creative ability. Summarizing her results, she
gives as the most consistent finding a high number of
whole responses. Form responses were not unusually high,
and in 7 of the 20 artists' protocols the form responses
were poor or vague in quality. She also found in many of
the protocols a low number Of human movement responses and
a high number of animal responses--the reverse Of Ror-
schach's expectations. Her conclusion was as follows:

For a number Of these subjects there are in their

Rorschachs no indications Of creative ability, as

this has usually been estimated. In view Of the

fact that these men are all at the top Of a crea-
tive profession this is a very striking finding.

Some Of the protocols, of course, abound in ele-

ments which have been interpreted as indicating

creative ability, but so many of them had few or

none Of these that it seemed important to have an-

other Opinion than my own.

(p. 403)

She therefore sent the protocols to Dr. Bruno Klopfer for
blind analysis. He found evidence Of creative ability in
only 5 of the 20 records and in no case did he identify
the subject as a successful artist.

Roe's study has been criticized by Dudek (1968) on

the basis Of the Special nature of her subjects, who were

25

taking part in a study of creativity and alcohol at Yale
and were all chronic alcoholics, with a mean age Of 51.
Dudek cites her own study Of a younger, non-alcoholic
group of writers and painters (1968, p. 539) in which she
Obtained a high number Of M responses. Cohen (1960), in

a later study Of adaptive regression in artists, finds a
different kind Of significance in Roe's work. To him "it
appears that the very material which so disturbed her
(presence Of primary process, 'Overproduction Of responses
with sexual content,’ etc.), combined with the adequacy of
popular responses and 'unusually great W' are some of the
same variables which present theory holds are determining
factors for creative productivity" (pp. 25+26). We will
return to this point later.

As part of a study Of the use Of the Rorschach in
industry, Steiner (1947) administered the Rorschach to 10
commercial artists. Due to the small number Of subjects,
she does not generalize her findings from this group but
she notes, among other tendencies, a high number Of M re-
sponses (average: 7).

Andersen and Munroe (1948) compared the Rorschach
responses of three groups Of students at Sarah Lawrence
College: 20 students Of creative painting, 20 students Of
composition and design, and 40 students in liberal arts
courses. They then compared the responses of the painting
students with the results Of the Roe (1946) and Prados

(1944) studies. The protocols of both groups of art

26

students were also examined according to Rorschach's or-
iginal assumptions regarding creative potential. On the
variables Of interest to this discussion, they found a
higher percentage Of human movement responses among the
painting group and a lower number Of animal responses.
The design group showed a higher percentage Of form re-
sponses characterized as constricted. NO significant dif-
ferences were found in the number of popular or original
responses. These findings contradict those Of the Roe
study but are similar to those Of the Prados study. They
also appear to bear out Rorschach's assumptions concern-
ing the proportion of human movement and animal responses.
Eiduson (1958) compared artists and non-artists to
test the hypotheses (a) that artists are significantly dif-
ferent from non-artists in characteristics of thinking and
perception, personality, and motivation; (b) that artists
who have sought help for psychological problems are not
significantly different from artists who have not sought
help. Her 65 subjects were divided into three groups:
(a) the experimental group, composed Of a mixed group Of
20 professional artists who had been engaged for at least
five years in literature, music, theatre, or visual and
graphic arts, all of whom had been studied as outpatients
at a psychiatric clinic; (b) control group I, 25 non-
artists admitted to the same clinic, selected for their
similarity to the experimental group on the variables of

age, education, intelligence, sex, diagnostic classification,

27

and stability Of vocation; (c) control group II, 15 art-
ists drawn from the same fields as those in the experimental
group and similar to them in stability of vocation, who

had never sought psychiatric help. All subjects were given
the Rorschach and TAT. Responses were rated globally and
scores are not included in the report. Eiduson gives
summary data showing that on variables Of thinking and per-
ception, emotion, and motivation, both groups of artists
differed significantly from non-artists but not from each
other. In addition, She concludes that the characteris-
tics common tO the artist groups do not fit the common
stereotype of artists as neurotic. Among these are the
search for originality, responsivity to sense data, tol-
erance fOr ambiguity, and the ability to "relax their
thinking without accompanying personality disorganization"
(p. 25). Eiduson's study is one Of the first in which this
qUestion of the pathological sources of artistic activity
is directly addressed.

Hersch (1962) presents a developmental analysis Of
the cognitive functioning of creative artists. He com-
pares Rorschach data from three groups: a "creative"
group of eminent artists, a "normal" group Of individuals
not distinguished for creativity in any form, and a
schizophrenic group of hospitalized patients. Basing his
study on Werner's comparative developmental theory, he
hypothesized that the artists would show a greater avail-

ability Of both "mature" and "primitive" Operations than

28

non-creative normals, and that the artists would also Show
more mature responses than the schizophrenics. No hypoth-
esis was stated regarding a comparison of primitive re-
sponses in artists and schizophrenics. Among the charac—
teristics defined as mature are: a high number of Movement
(i.e., human movement) responses, and a higher number of
Form Dominant responses. Primitive characteristics in-
cluded Form Subordinate responses and a category Of "Prim-
itive Thought" (including contaminations, confabulations,
and the like). The major hypothesis Of the study--that
artists will Show greater availability of both mature and
primitive operations-—is generally supported by the results.
The Hersch study is of special interest for two rea-
sons. First, it foreshadows the more recent "adaptive re-
gression" studies which have proven fruitful in the study
of artistic creativity. Although Hersch did not use the
Holt system for scoring primary and secondary process in
the Rorschach (Holt & Havel, 1960), his "primitive and
mature Operations" are conceptually similar tO Holt's
categories and are clearly related to the idea of regres-
sion in the service Of the ego. A second feature Of inter-
est in this study is the fact that for his creative group
Hersch utilized the data from the controversial Roe study
(1946). In the same artists' responses in which Roe
failed to find corroboration of creative ability according
to the traditional assumptions Of Rorschach scoring, Hersch

is now able, using a somewhat different set of assumptions,

29

to identify some of the cognitive processes which distin-
guish creative artists from non-creative normals. This
contrast in interpretations serves to illustrate the pro-
gress made in the years from 1946 to 1962 in the under-
standing Of how creative processes function in artists,
and supports Cohen's speculation (1960, pp. 25-26) that
the very material which disturbed Roe might, in the light
Of later theory, be considered evidence in favor Of crea-
tivity.

Two later studies of artists which follow this ap-
proach of looking for creativity through isolated categor-
ies of Rorschach responses are those by Rawls and Boone
(1967) and Rawls and Slack (1968). Citing earlier work
to show that the production of whole responses held up
better than any other variable as an indicator of artistic
creativity, the first study was designed (1) to verify
this statistically, and (2) to evaluate the effect Of in-
structions designed tO increase the number Of whole re-
sponses on their ability to distinguish artist from non-
artist groups. The artist group consisted Of 16 male and
8 female art teachers; the non-artist group was selected
from a pool of graduate students. Discrimination on the
basis of whole responses was statistically verified, as
predicted. In regard to the second hypothesis, it was
found that an instructional set to produce whole responses
did not improve discriminatory power. The artist group

increased their production only slightly, while such

30

responses on the part of the non-artists almost doubled.
The authors interpret this to mean that artists have a
predisposition to produce whole responses whether instruc-
ted to do so or not.

The second of these studies (Rawls & Slack, 1968)
was intended to clarify some of the inconsistencies in the
literature implied by the use of different scoring systems.
Scoring the protocols of an artist group (painting and
sculpture teachers) and a non-artist group (other teach-
ers), they found that both the Beck and Klopfer systems
discriminated the groups on the basis of variables asso-
ciated with creativity, including whole and human movement

responses.

Other Instruments

 

Very few studies Of artists have employed instruments
other than the Rorschach. Munsterberg and Mussen (1953)
used the Thematic Apperception Test for a study of the
personality structure of art students. Deriving a set Of
hypotheses from analytic writings, they matched them with
variables on the TAT. Using this method, they found sup-
port for beliefs that artists are more introverted, more
aggressive, suffer more from guilt feelings and are more
concerned with the acceptance Of their work than with per-
sonal acceptance or recognition. Although Munsterberg
and Mussen appear to have found it satisfactory, there are

indications elsewhere that the TAT may be ineffective with

31

professional artists. Roe's sample Of painters, for ex-
ample, found it difficult to respond to, due to what they
considered the poor artistic quality of the pictures (1946).
Writers, as well as visual artists, have made the same
complaint (Rexroth, 1959; Wilson, 1958).

The Barron-Welsh Figure Preference Test,or its deriv-
ative the Barron-Welsh Art Scale, has also been used in
studies Of creativity. This test, which figured promin-
ently in the IPAR studies Of creative people, is presumed
to measure the preference for complexity and asymmetry over
simplicity and symmetry, a preference which has been found
to be more common among artists than non-artists (Barron &
Welsh, 1952).

For the most part, visual artists have served as
criterion groups in the development of the Barron-Welsh
scale, rather than as a focus in themselves. Some of the
correlates Of high scores on this instrument, however,
seem relevant to the processes of adaptive regression. The
Berkeley researchers found, for example, that a high de-
gree Of preference for complex and asymmetrical designs
was negatively associated with rigidity, conformity, and
constricted interaction with the environment (Barron,
1953). Barron's description Of the modal high scorer on
the art scale might serve as well for a description Of
high scorers on a measure of adaptive regression:

The decision in favor Of complexity, at its best,

makes for originality and creativeness, a greater
tolerance for unusual ideas and formulations. The

32

sometimes disordered and unstable world has its
counterpart in the person's inner discord, but the
crucial ameliorative factor is a constant effort to
integrate the inner and outer complexity in a higher
order synthesis. The goal is to achieve the psycho-
logical analogue Of mathematical elegance: to allow
into the perceptual system the greatest possible
ridhness of experience, while yet finding in this
complexity some over-all pattern. Such a person is
not immobilized by anxiety in the face Of great un-
certainty, but is at once perturbed and challenged.
(1952, p. 400)
Barron's conclusion is clearly that a positive response
to complex and asymmetrical designs is a marker for the
ability to regress constructively. This assumption and
the instrument itself are discussed further below, in the
Methods section.
Finally, Holtzman, Swartz and Thorpe (1971) tested
a carefully selected group of advanced art students on the
Holtzman Inkblot Technique and compared them with advanced
students in architecture and engineering drawing. They
found the artist group to be less concerned with details
of form and structural organization, more open to percepts
showing sex and anatomy, and more apt to produce deviant
or bizarre verbalizations. As with the Barron-Welsh stu-
dies of artists, this investigation was conducted primar—
ily for purposes of validating the test, rather than to

explore hypotheses about artists. This instrument is

also discussed below in the Methods section.

33

Studies Of Adaptive Regression
in Artistic CreatiVity

 

Kris' ideas on regression in the service of the ego
(1952) have provided the conceptual base for a number of
the empirical studies of artistic creativity. Tests Of
the validity Of the construct have been greatly facilitated
by the development of a method for scoring the operations
of primary and secondary process in the Rorschach by Holt
(Holt & Havel, 1960). Not all investigations of adaptive
regression have depended on this system, however.

Myden (1959) made a study of the personality charac—
teristics of creative people which pre-dates publication
Of the Holt system but which draws on Schafer's suggestions
concerning use Of primary and secondary process and shifts
Of level in Rorschach responses (Schafer, 1954). Myden's
creative sample consisted of 20 subjects (16 male and 4
femals) described as "well known" and "from the top ranks
Of diverse fields in the arts" (p. 142). Painters were
represented in this group, as well as theatrical and liter-
ary artists. These were compared with 20 equally success-
ful men and women from business or non-artistic profes-
sions such as medicine or law. Myden gave each Of his
subjects a test battery consisting of the Rorschach, TAT,
Bender-Gestalt, Human Figure Drawings, and Vigotsky Con-
cept Formation Test. Rorschach Responses were scored
quantitatively by one Of the traditional scoring systems.

They were also analyzed qualitatively for evidence Of the

34

sequence Of levels suggested by Schafer: the appearance
Of regressive, primitive material followed by progressive,
critical, and reality-oriented content. Both quantitative
and qualitative analysis supported Myden's hypotheses that
creative individuals would make greater use Of primary pro-
cess and, at the same time, show less evidence of repres-
sion. In regard to a third hypothesis, that the creative
group would show a lesser degree of psycho-sexual adjust-
ment, Myden found that there were evidences Of homosexual
experience in both groups but that its emergence in the
non-creative group was accompanied by increases in anxiety
not seen in the creative group.

A study by Hersch, based on research for a doctoral
dissertation completed in 1957 and published in an article
in 1962, has already been referred to. While Hersch does
cite Kris in his discussion, the framework for his investi—
gation Of cognitive functioning in the creative person is
based on Werner' comparative developmental theory, a
theory which shows Similarities to Kris' ideas. Hersch
quotes Werner:

Now, creativity, in its most general meaning is

an essential feature of emergent evolution, and

this, in turn, implies progression through reor-

ganization. Since we assume that such progress
through reorganization cannot be achieved without

"starting anew," that is, without regression, it

follows that a person's capacity for creativity

presupposes mobility in terms of regression and
progression.

(Hersch, 1962, p. 193)

Hersch found support for his hypothesis that artists'

35

Rorschach responses would show a greater use Of both
primitive and mature thought Operations than those Of
non-creative normals. He also discovered that the art-
ists gave an unusually high number Of a type Of response
designated as Physiognomic. Hersch's definition:
This category includes those responses in which the
blots themselves, and not something seen on the
blots, are considered as having affective or sym-
bolic qualities, dynamic properties, the attributes
Of living things and the like. The usual interpre-
tation of the blot as "looking like" a certain Ob-
ject or thing is minimal, and the blot itself is
described as being "gay," "vicious," "troubled,"
and so on . . . Such responses reflect a syncretic
relationship between the self and the external
world, and are therefore considered primitive.
(p. 195)
Although a small number Of such responses were also given
by schizophrenic subjects, in their case the percepts were
reacted to emotionally as Objective reality. In the art-
ists' protocols Hersch finds that although the subject
attributes characteristics directly to the inkblot--such
as saying that it is "quivering"--he then proceeds to re-
flect upon the perception. Hersch describes this as a
biphasic process consistent with the regression-progres-
sion sequence presumed to be essential for creative activ-
ity, and he concludes that "it is this response, more than
any other, which distinguishes the Rorschach protocols of
creators from those Of the other two groups," (p. 197).
Hersch's comparison of the artists to schizophrenics as

well as normals is particularly useful since it will be

remembered that these artists are the ones whose responses

36

were evaluated as non-creative by Roe (1946) and as path-
ological by Dubek (1968).

Cohen (1960) applied the Holt system Of analysis to
two groups of 20 advanced art students rated by their
teachers as high or low on creative ability. He found
that there were significant differences between the two
groups, not on the amount of primary process material pro-
duced but in the degree of effective control exercised over
such material. The creative group was also found to have
a lower score on the Rokeach Dogmatism Scale. Cohen's
study is Of interest in that it represents one Of the
early confirmations of the usefulness of the Holt system
as an Operational measure of adaptive regression (or re-
gression in the service of the ego).

Dudek (1968) compared the Rorschach records Of art-
ists defined as either successful or unsuccessful. Her
"successful" group included 22 writers, 3 sculptors, and
16 painters, who met the criterion of "acknowledged sta-
ture." There were 19 artists in the "unsuccessful" groups
(not specified as to media). She also compared both Of
these groups to a group of non-artists who had given a
high proportion Of human movement responses to Rorschach
stimuli. Responses were scored for regressive content
by categorizing them according to criteria adapted from
Holt: responses Showing primitive expressions of drive,
confabulatory responses, and responses showing autistic

elaboration. The protocols were also scored by a

37

traditional method. Neither the quantity nor quality Of
M responses was found to discriminate between groups.
Dudek did find evidence that the successful artists pro-
jected more regressive content and appeared to Shift
easily from regressive to normal levels. Although the un-
successful artists showed less Of this fluidity and less
regressive content, when this type of material did emerge
they were able to use it constructively. By contrast, the
non-artist group was not able to integrate primary process
material into their responses or to re-establish defenses

following its appearance.

Comments

 

With the exception Of a few observational studies,
the major investigations of artistic creativity have em-
ployed projective techniques, primarily the Rorschach.

The initial approach taken by investigators was to attempt
to substantiate assumptions made by Hermann Rorschach as
to the type, quality, and frequency of particular re-
sponses associated with creative ability. Emphasis was
strongest on the human movement, or M, response, with some
attention also being paid to Whole responses and to the
level and quality Of Form responses. Although some sup-
port has been found for Rorschach's assertions, the sum
total of results has been equivocal at best (see Rabin,
1968, pp. 590-594). Beyond this, it must be said that

this atomistic approach has advanced our conceptual

38

understanding Of the creative process only to a very lim-
ited degree.

Psychoanalytic insights into the regressive aspect
of the creative process remain intuitively convincing and
are consistent with artists' introspections. The diffi-
culty, as has always been the case with psychoanalytic
theory, has been in devising measures by which the propo-
sitions can be objectively tested. Kris's theory of re-
gression in the service Of the ego, by specifying more
closely the functional processes involved in the creative
act, has provided a firmer base for testing than earlier
theories which concentrated on the personal history of the
artist. Holt's application Of this theory to a quantifi-
able scoring system for primary process in the Rorschach
has made it possible to begin to accumulate findings with
greater confidence that we are Observing the substrates
of creativity rather than merely personal qualities of the
artist.

A flaw in some studies of creative artists is the
pooling of subjects from a variety of fields, on the ap-
parent assumption that all artistic activity is equival-
ent. An example of this practice is seen in the study by
Eiduson (1958), who included in her experimental group
writers, painters, musicians, actors, and directors.
While it may be true that the creative process is similar
in all the arts, at least at the level at which we are

presently able to measure it, this has not been clearly

39

demonstrated, nor has the influence Of the chosen medium
on the artist's internal processes been explored. It may
be that some Of our conclusions about artists are in fact
more typical of the samples most commonly tested (such as
painters and writers) than of artists in general or art-
ists working in other media. The present study attempts
to address this question by including both painters, a
group on whom a limited but cohesive body Of data has
been accumulated, and potters, who have not so far been
the subject Of published research.

A second set Of distinctions which have not always
been attended to are those among famous artists, lesser-
known artists, art students, and art teachers. Rawls and
Boone (1967), for example, appear to make the assumption--
like the assumption of homogeneity across fields—-that an
art instructor is psychologically equivalent tO a working
artist. Once again, this has not been demonstrated.

Other investigators, in search Of what they take to be the
purest examples of the phenomena of interest, seek out for
their sample artists who are particularly reknowned (cf.
Drevdahl & Cattell, 1958). While the assumption that these
creators have more of the creativity trait may be well
founded, the possibility needs to be considered that
celebrity may have its own correlates, which function in-
dependently Of talent in gaining fame. Studies such as

the one by Dudek (1968), which makes distinctions in lev-

els of success, help to clarify these questions. For the

40

present study, subjects were enlisted from the ranks Of
working artists at an intermediate level Of attainment--
past the student level but not, at this point in their
careers, nationally famous.

Finally, it is rare to find a consideration Of pos-
sible sex differences in artists' approaches to the task
Of creating. Many reports fail even to specify the sex Of
the subjects used in the study (e.g. Eiduson, 1958;
Eindhoven & Vinacke, 1952). The most typical research
strategy is probably that used by Munsterberg and Mussen
(1953) and Myden (1959): both males and females, in un-
equal numbers, are included in the artist group, which is
then compared to a control group matched on sex as on other
variables. NO analysis is Offered which differentiates
results according to sex. Only Getzels and Csikszentmi-
halyi (1964) report reasonably equivalent numbers Of males
and females, and their comparisons do reveal some differ-
ences between males and females within fields Of special-
ization. In this study, therefore, while no predictions
were made as to their potential differences, equal numbers
Of male and female artists were included so as to make

post hoc comparisons possible.

 

IV. HYPOTHESES AND PREDICTIONS

The stance Of the creative artist has been described
as one of Openness to complexities and irregularities in
experience. On the aesthetic plane, this is seen in a pref-
erence for designs which are complex and asymmetrical
over those which are simple and symmetrical. Perhaps this
is because complex, irregular stimuli do not present a
ready-made synthesis but require the viewer to perceive
or to construct some pattern of relationships which brings
the elements into a satisfying equilibrium.

On the plane Of inner experience, a parallel may be
drawn. When confronted with the ambiguous, provocative
stimulus Of an inkblot, the creative artist will not try
to suppress or deny awareness Of bizarre projected images,
nor become unduly disturbed by them. Ideas or feelings
which are highly charged with sexual or aggressive themes,
or which appear in a fantastic or illogical form, can be
organized into a coherent construct. What is deviant or
eccentric can be included, bound, in a new synthesis.

The challenge of disequilibrium is accepted and met.

Hypothesis 1: Artists generally, as compared with

 

non-artists, will tend to prefer visual stimuli which

encourage them to resolve the problem Of a disequilibrium.

41

42

Prediction 1: This sample Of artists will score

 

significantly higher than non-artists, using the norms
derived for the Barron-Welsh Art Scale.

If the appearance of primary process material may be
considered to create a temporary state Of disequilibrium,
then bringing it under control by giving it a coherent
form may be considered to be the restoration of balance.
In response to inkblot stimuli, the combination of sexual
or aggressive content and/or primitive or illogical think-
ing with good formal organization of percepts is consid-
ered to constitute an Operational definition of adaptive
regression. The second hypothesis relates the ability to
regress adaptively to the aesthetic preferences stated in
Hypothesis 1.

Hypothesis 2: A high degree Of adaptive regression

 

will be associated with a high degree of preference for
designs which are complex and asymmetrical.

Prediction 2: Results will show a significant posi-

 

tive correlation between high scores on the Revised Art
Scale and high adaptive regression scores on the Holtz-
man Inkblot Technique.

Hypothesis 3 is the corollary Of Hypothesis 2.

Hypothesis 3: A low degree of adaptive regression

 

will be associated with a low degree of preference for
designs which are complex and asymmetrical.

Prediction 3: Results will Show a Significant posi-

 

tive correlation between low scores on the Revised Art

43

Scale and low adaptive regression scores on the Holtzman
Inkblot Technique.

NO hypotheses are stated and no predictions are made
concerning potential differences between painters and pot-
ters. Since no studies of potters appear in the litera-
ture, their inclusion must be considered exploratory. If
significant differences do appear, then we may speculate
on distinctions in the nature Of creativity or the crea-
tive process required by the two fields or in the person-
ality patterns of those who are drawn to specialize in one

field or another.

V. METHODS AND PROCEDURES

Subjects

 

Previous studies Of artistic creativity have gener-
ally drawn their subjects from the ranks of either eminent
artists or art students. Although these populations may
tell us something about artists as compared to non-artists
at a similar level Of achievement, neither may be typical
of artists in general--people who are motivated to work at
their craft over a period of years without the rewards of
celebrity or the incentives of academic grades and dead-
lines. Subjects for this study were drawn fronithis latter
population. They may be characterized as "working artists,"
serious professionals who have not, for the most part,
achieved national reputations but who are committed to
their work and for whom there is some evidence Of local
or regional recognition.

Potential subjects were identified through three
channels: newspaper announcements of current shows,
gallery records, and survey data gathered by a regional
artists' association. Once a list of names of artists
meeting the criteria had been compiled, a letter was mailed
to each by the experimenter, followed by a phone call.

None Of the subjects in the study was a volunteer. The

44

45

rate of acceptance was 66.6%; 60 artists were contacted,
Of whom 40 agreed to participate.

Subjects were divided equally between two areas Of
specialization: painting and pottery. The intention here
was to avoid the uncertainties inherent in studies in which
creators from different fields are lumped together under
the single name "artist" (e.g. Eiduson, 1958). Such
specification also provided the opportunity to explore the
possibility that the creative process might vary with the
medium. Half the artists in each group were male and half
female. Criteria for inclusion in the study were (1) that
the artist should have been working for at least three
years in the chosen field, not including time spent in
training, and (2) that there should be evidence Ofaipro-
fessional level Of achievement recognized by Others, such
as having exhibited works in at least two juried shows or
gallery exhibitions. Virtually all the artists in the
study exceeded these criteria by a wide margin. It was
not required that the artists be self-supporting through
the sale of their work. A number of those who took part
derive or supplement their main income through teaching.
While teaching did not disqualify an artISt whose work
appeared in exhibitions, no artist was considered elig-
ible on the basis of teaching alone. NO attempt was made
to evaluate the artists' work for excellence.

Ages Of subjects ranged from 26 to 61. The mean age

for the total group was 40, with the mean for the females

46

(42) somewhat higher than for the males (38). Twenty-nine
Of the artists were married (3 were currently divorced)

and 26 had children.

Instruments

 

Two instruments were used as means Of Observing the
artists' responses to disorder: the Holtzman Inkblot
Technique and the Revised Art Scale of the Welsh Figure

Preference Test.

Holtzman Inkblot Technique (HIT)

Background. The HIT was developed by Wayne Holtz-

 

man and his associates at the University of Texas at Aus-
tin to meet the need for a research instrument incorporat-
ing the principles of the Rorschach while improving on its
quantifiability. In place of the ten inkblots Of the
Rorschach, to which the subject may give an unlimited num-
ber Of responses, the HIT provides a series Of 45 cards
(plus two practice cards) to which the subject gives one
response apiece. The examiner's inquiry follows the re-
sponse to each card and is limited to a few standard
questions. These limitations provide for equivalence in
the numbers of responses across subjects; lengthening the
test improves the reliability of scores. For purposes of
re-testing, the HIT is available in two forms, drawn from
the original pool of 135 items and matched on criteria of
"goodness." Form A alone was used in the present study.

For a more detailed description Of the development

47

of this measure, see Holtzman, Thorpe, Swartz and Herron
(1961).

Reliability. Standardization of the HIT was initi-

 

ally carried out in a program involving 15 sample groups
of school children, normal adults, and institutionalized
patients, for a total of 1334 subjects. Holtzman at El:
(1961, p. 104) report intra-scorer consistencies on 9 Of
the test Variables ranging from .89 to .97 and inter-
scorer consistencies averaging from .73 to .89. Split-
half reliabilities are also reported for each of the 22
variables in the standard scoring system. Among those
given, the most relevant for the scoring system used in
this study are those pertaining to the variables of Form
Definiteness and Form Appropriateness. Using 15 samples,
Holtzman reports split-half reliability coefficients for
Form Definiteness ranging from .81 to .96 with a median of
.88 (1968, p. 149). For Form Appropriateness, the average
intra-scorer consistency is given as .92 and the average
inter-scorer consistency as .73 (ibid, p. 104). These
figures indicate that the test is acceptably reliable.
Holtzman also cites support for his contention that situa-
tional influences such as examiner differences or tone Of
administration have less effect on HIT responses than they
have been shown to have with the Rorschach (1968, p. 148).
The HIT and the study of artists. Two pairs of re-
searchers have used the HIT to Obtain measures of crea-

tivity: Lake and Telford (1970) and Bowers and van der

48

Meulen (1970). In both cases, the creativity index was
based on Rorschach's assumptions that Human Movement (M)
responses represent the capacity for inner creation. A
high score on this variable on the HIT was presumed to in-
dicate a high level of creativity.

Only one reported study of artists to data has used
the HIT. Holtzman, Swartz and Thorpe (1971) gave the HIT
to three groups Of advanced students in art, architecture,
and engineering as one of a series of studies aimed at ex-
tending the construct validity Of standard scores for the
test. While primary process is not a variable in the
standard scoring system, it was found that the artists
produced significantly higher scores on variables Of Sex,
Anatomy, and Pathognomic Verbalization. This was inter—
preted by Holtzman, 93 31. to indicate greater comfort
with irrational feelings. The artists did not differ
from other groups on Form Definiteness, indicating that
admission of more highly charged material involved no loss
Of clarity in their percepts. The finding that artists'
inkblot responses show both a high level Of primitive con-
tent and good form has implications for the scoring sys-
tem used in this study (see Scoring method, below).

Advantages of the HIT. The value of inkblot stimuli

 

in the testing of artists has been demonstrated (see Em-
pirical Studies, above). By their ambiguity, they require
what may be considered a creative act in the organizing of

visual data into a coherent percept. An increasing number

49

Of studies also show that responses to inkblots can be
analyzed for Operational evidence Of the theoretically
postulated primary and secondary processes. That is,
primitive sexual and aggressive content can be identified
and its constructive utilization evaluated (see Cohen,
1960; Dudek, 1968; Hersch, 1962; Myden, 1959). All such
studies, however, have used the Rorschach.

In spite of the neglect of the HIT as an instrument
for the testing Of artists, it would appear to have sev-
eral potential advantages over the Rorschach. First, it
was designed with serious attention to its aesthetic qual-
ity and should have a high degree of acceptability to
artists as a visual stimulus. This proved to be the case
with the subjects in this study. Aesthetic acceptability
can be a significant factor, as attempts to test artists
with the TAT have demonstrated (see Roe, 1946). Second,
control over the total number of responses reduces the
effect of a variable which may inflate or skew other socres.
Cronbach (1949, pp. 409-410) gives examples Of how produc-
tivity alone may account for differences between groups
on Rorschach scores.

Scoring method. The HIT responses were scored accord-

 

ing to the procedures outlined in Holt's Manual for the
Scoring Of Primary Process Manifestations in Rorschach

Responses (1970). Although this system was not developed

 

for the HIT, there is no inherent incompatibility between

Holt's scoring method and the Holtzman stimuli, which

50

closely follow the Rorschach model as a projective test.

Indeed, Holt specifically states that his system is

applicable to a variety Of verbal productions:
Drawing on Freud's general descriptions Of uncon-
scious processes and on Rapaport's "verbalization
scores" for the Rorschach, we tried to frame scor-
ing categories that would fit the unique situation
presented by this test. Yet when my colleagues...
and I have turned to other types of data and have
attempted to see how far we could go in applying
the manual to samples Of free speech, to dream
texts, and to TAT stories, we have found that rel-
atively few changes had to be made-~most1y a few
additions, in order to cover problems posed by a
discursive or narrative format, which Rorschach
responses do not have...

(1970, p. v)
Gray (1969) has applied Holt's Operational definitions of
primary process to the HIT in a study Of the effects of productiv-

ity on creativity scores. _The complete scoring system was not

used; he simply summed the number of responses giving
evidence Of primary process.

In transferring the Holt ratings to the Holtzman
blots, only one of the four categories contributing to
an Adaptive Regression Score required special adjustments.
This is the Form Level score, which draws on Rorschach
tables Of F+ and Popular responses. The Holt manual in-
cludes, however, a 7 point scale, developed by Mayman,
for rating the clarity and appropriateness of the per-
cepts. Mayman's criteria were used in rating the responses
to the HIT. The Holtzman manual provides categories for
judging Form Appropriateness and these were incorporated

into the Mayman ratings. The HIT Form Appropriateness

51

scores are restricted to three points, which would reduce
variability in deriving an Adaptive Regression Score.

For the complete rating scale used in this study, see

the Appendix.

Following Holt's system, each response was rated on
three dimensions in addition to Form Level: Content,
Formal, and Control and Defense. Content scores rate the
expression Of libidinal or aggressive drives through the
activities or objects referred to in the response. Formal
variables are those related tO the fluidity or disorder
in thinking which can be inferred from the subject's ver-
balizations. Control and Defense ratings take into account
the extent to which drive expression is bound by making it
appear reasonable or socially acceptable. The first two
categories, Content and Formal, can be thought of as those
involving primary process, while secondary process, or ego
functions, are at work in the Operations Of Control and
Defense. Within the Content and Formal categories distinc-
tions are also drawn between Level 1 ("crude, direct and
primitive" expressions) and Level 2 ("indirect, controlled
and socialized"). For example, a simple reference to a
crocodile, without the mention Of teeth, would get a
Level 2 Oral Aggressive score of 1; if teeth were men-
tioned, the percept would be given a Level 2 score of 2;
if the crocodile were seen tearing a person or other animal
apart with its teeth, this would receive a Level 1 score

of 4. Level 1 scores are generally seen only in psychotic

52

records; they were rare in the responses Of the present
sample.
Content and Formal subscores are then used to Obtain
a rating of Defense Demand (DD). Holt Offers two Options.
In the first, a global DD rating is made Of the response
as a whole, using a 6 point scale ranging from a score of
1, indicating "virtually no apparent need for defense,"
to 6, "greatest need for defense." In the second Option,
the DD values in the Content and Formal categories are
simply added together. The second alternative was selected
for the present study. In a system which relies so heavily
on judgments, it seemed desirable wherever possible to
use automatic methods of arriving at a score. Moreover,
the summation method provides what is essentially a weight-
ed score, taking into account the number as well as the
level of primary process elements appearing in each response.
Form Level ratings establish a tentative score for
Defense Effectiveness (DE), which is then adjusted accord-
ing to the apparent Operations of Control and Defense. A
summary score rating the degree of adaptive or maladaptive
regression implied in each response is figured as the pro-
duct of these two scores (DD X DE). An overall score for
each subject is then computed on the formula ARS =
2(DD x DE)/R, where ARS = Adaptive Regression Score,

DD = Defense Demand, DE = Defense Effectiveness, and

R the total number Of responses given. In the case of

the HIT, R is constant at 45.

53

TO sum up the meaning of the final score, Adaptive
Regression is considered to be represented by an averag-
ing of the demand for defense generated by the percepts
which the subject sees in the inkblots multiplied by the
effectiveness with which the subject handles this poten-
tially threatening material.

Validity and reliability Of the scoring method.
Since no manual or Other published guide exists for scoring
the Holtzman inkblots by the Holt method, a representative
protocol was submitted to Dr. Holt for his comments and
critical evaluation. The author selected this protocol on
the basis Of two criteria: (1) it fell in the upper range
Of the sample with regard to the percentage of responses
showing primary process (76% for this subject vs. a mean
of 60% for the total sample); (2) the variety of scoring
subcategories represented was wide in comparison with
other protocols. The protocol was rescored by Dr. Holt.
The product moment correlation was then computed for the
34 responses showing scorable primary process. For the
34 pairs of scores, r = .89. There is therefore reason
to believe that the scoring method used in this study is
an acceptable application of the Holt system.

The Revised Art Scale of the Welsh
Figure Preference Test

Background. The Welsh Figure Preference Test (WFPT)

 

grew out of an attempt to construct a picture test for

detecting psychiatric abnormality which could be easily

54

and validly responded to by patients who did not speak
English, merely by indicating "Like" or "Dislike" to a
series of black and white designs. Factor analysis of
subjects' preferences on the initial version of this test
revealed a bipolar factor of simplicity-symmetry and com-
plexity-asymmetry (Barron and Welsh, 1952). Qualities
which appeared to characterize high and low scorers on
these poles suggested that these preferences were associ-
ated with artistic taste and possibly also with related
personality variables. The 400 items from the original
test were therefore administered to a sample Of 37 art
students and a larger group Of 137 "people in general."
The scale derived from an item analysis of the preferences
exhibited by subjects in this study, which is known as the
Barron-Welsh Art Scale, consisted of 40 items disliked
more Often by artists than non-artists and 25 items liked
significantly more often by artists. The items more often
disliked by the artists were found to be Of the simple-
symmetrical type, while those more Often liked were Of the
complex-asymmetrical type. However, since the uneven num-
bers Of items in each category allowed a negative re-
sponse set to be confused with a complexity-asymmetry pref-
erence, the test was re-edited. The latest version, the
Revised Art Scale (RA), consists of 60 items, 30 keyed

in the "Like" and 30 in the "Dislike" direction. It is
this revised scale which was used in the present study.

The RA scale is empirically derived and, in this

55

study, no assumptions were made as to personality variables
the scale may draw upon, beyond the original preference

for complexity and asymmetry Over simplicity and symmetry.

This was taken to be an Operational measure Of the artist's
attitude toward disorder: visual stimuli are preferred in

which balance is not Obvious or conventional but has to be

discovered as it were in the eye of the viewer.

The RA scale and the study Of artists. Aiken (1967)

 

offers a review of the literature on the BWAS and the RA
scale, with a table of mean scores for a number Of profes-
sional groups, including artists, architects, writers,
research scientists, and mathematicians (p. 7). He also
summarizes research showing significant correlations be-
tween scores on the BWAS and ratings of creativity.

Rosen (1955) administered the BWAS to 44 art students
(22 beginning students, 22 advanced), 8 artist faculty mem-
bers from the same school, and a matched group Of non-
artist faculty members. His findings were that (a) scores
on the BWAS did not increase with training; (b) while the
mean scores Of the three artist groups were not signifi-
cantly different from one another, there was a 19 point
difference between the mean scores Of the artist and non-
artist groups (p < .01); (c) the scores of the art stu-
dents correlated significantly and positively with faculty
ratings of the originality Of their work (r = .40, p < .02)
and, to a lesser degree, with their grade point averages

(r = .34, p < .05).

56

Getzels and Csikszentmihalyi (1964) used the BWAS
in their very thorough study Of students at the Chicago
Art Institute. Comparisons by sex and field of specializa-
tion showed no significant differences in mean scores, nor
did they find significant correlations with scores on an
intelligence test.

The Holtzman, Swartz and Thorpe study Of student
artists, architects, and engineers (1971), which is cited
above in the discussion of the HIT, also made use Of the
original 400 item WPFT. Of nine scales contained in this
version of the test, only one failed to show at least one
significant correlation with a variable on the HIT. While
the authors note that the relationship Of these scales to
inkblot responses seems to warrant further investigation,
they do not attempt to interpret their findings.

Studies Of adaptive regression. Three reports show

 

correlations of BWAS scores with scores on adaptive re-
gression. As support for the Holt scoring system, Holt
and Havel (1960) display a table of "measures of person-
ality that correlate with Rorschach measures of adaptive
vs. maladaptive regression" (p. 313). The BWAS shows a
positive correlation of .74 (p < .01). NO reference is
given for the study from which this figure is Obtained.
The authors acknowledge that the comparisons summarized in
their table are suggestive rather than conclusive. They
also add, as a word of caution, that preliminary results

from another study (also unnamed) involving female college

57

students do not show the same pattern.

It seems likely that the studies referred tO are in-
vestigations Of creativity and primary process undertaken
by Pine and Holt (1960) and Pine (1962). For male student
subjects in the earlier investigation, Pine and Holt found
significant correlations of adaptive regression scores with
five Of the ten creativity measures used. For the female
students in the same sample, however, adaptive regression
scores were not significantly related to scores on any Of
the creativity tests. Significant correlations EEES
found for the female subjects (a) between control of pri-
mary process and the overall creativity score and (b) be-
tween amount Of primary process expressed and the quality
Of created productions (including TAT stories). Both these
results are contrary to predictions.

In order to pursue the question of generalizability,
Pine repeated the study with a sampleaof 50 male actors
(1962). Although the BWAS was included among the measures
used in both the 1960 and 1962 studies, scores are given
only in the later report, where comparisons are made among
all three groups. While the male students show high correl-
ations of BWAS scores with both adaptive regression (r =
.74, p < .01) and control of primary process (r = .64,

p < .05), correlations for the other groups--male actors
and female students--are low and not significant. Pine's
conclusion is that sample differences make a difference in

the relation of creativity and adaptive regression.

58

The present study was designed to gather further in-
formation on the influence of sample variations. Since
this study included both males and females in an artist
sample, it was expected to shed light on whether the sex
differences found by Pine and Holt were peculiar to their
student sample.

Reliability Of the measure. The reliability of the

 

Barron-Welsh Art Scale is reported by Barron (1965, p. 21)
as follows: odd-even reliability, 196; test-retest reli-
ability after six months, .91. The manual does not report
reliabilities for the Revised version Of the test, but
correlation of the RA scale with the BWAS is reported as

.85 (Welsh, 1959, p. 9).

Procedures

Subjects were told that the study was concerned with
the creative process and how it might vary among artists
in different media, and that the experimenter was inter-
ested in artistic preferences and the organization of vis-
ual perceptions. Each artist was interviewed and tested
individually by the experimenter, in a single two hour
session. The art scale was administered first in all
cases, followed by Form A of the Holtzman Inkblot Tech-
nique. Each artist was then asked to sort the HIT cards
into two groups on the basis of "Like" and "Dislike," as
he or she had also been asked to do with the art scale

designs. In the time remaining, the artist and experimenter

59

discussed informally the artist's own insights into the
creative process, with emphasis on preferred working con-
ditions. This material was not formally analyzed but pro-
vided background material for interpretation of the main
hypothesis.

An effort was made to standardize testing conditions
and provide an environment congenial to the subjects. All
but five of the artists were interviewed in a small art
gallery maintained by the university art department for the
exhibition of student work. In the five cases in which the
subject did not live or work near this facility, equival-
ent space was Obtained from the nearest college or univer-
sity art department. In no case was the subject inter-
viewed either in his or her own studio or in a setting
within the psychology department.

Responses to the HIT cards were tape recorded and
scoring was done from a verbatim transcript, as recom-
mended by HOlt (1970, p. x). Scoring Of the HIT protocols
was done by the experimenter. In order to reduce poten-
tial bias, responses were identified by code only and then
scrambled, so that the initial scoring was performed on
reconstructed composite protocols.

Scoring of the RA scale is Objective, using a key

provided in the manual.

VI. RESULTS

Hypothesis 1

 

The first hypothesis predicts that this sample of
40 artists will show a mean score on the art scale similar
to mean scores reported for other artist groups in the
literature. This hypothesis is not supported.

Mean scores are compared in Table l, which shows that
the mean score Of 35.2 for this artist sample is signifi-
cantly lower than the mean score of 40 derived from earl-
ier studies. This score does prove to be similar to those
reported for some art student groups and for other "crea-
tive" groups, especially as compared to non-artist groups
such as mathematicians. These figures are listed in
Table 2 .

There are certain difficulties in comparing scores
on the art scale, as a result of the changes which have
been made in the instrument over the years. The original
Barron-Welsh Art Scale (BWAS) included 65 items, of which
40 were to be checked "Like" and 25 "Dislike" to gain the
maximum score. The current manual (Welsh, 1959) gives
scoring keys for two later versions: a 62 item "Barron-
Welsh Art Scale" (24 Like, 38 Dislike) and a 60 item "Re-

vised Art Scale" (30 Like, 30 Dislike). Both scales can

60

61

Table 1. Mean Scores of Artist Groups on the Barron-
Welsh Art Scale

 

Source and Sample Number Of Mean Standard
Subjects Score DeVIatIon

 

Barron & Welsh
(from Barron, 1965)

standardization 80 40.3 12.9
sample

cross-validation 30 39.1 13.8
sample

Rosen (1955)

"established artists" 8 41.1 (not reported)

"advanced students" 22 40.7 (not reported)
Orrick (1979) *t (39)

"working artists" 40 35.2 8.3 = 3 68**

 

*A score Of 40 was used as the population mean

** p < .001, two tailed

62

Table 2. Mean Scores of Various Groups on the Barron-
Welsh Art Scale

 

 

Number Of Mean Standard
Source and Sample Subjects Score Deviation
Getzels &
Csikszentmihalyi
(1964)
Male fine arts 35 31.7 (not reported)
students
Female fine arts 44 34.7 (not reported)
stUdents
Barron (1965)
"Creative" architects 40 37.1 9.8
"Representative"
architects 41 26.1 12.1
"Creative" writers 19 32.9 11.1
"Creative" women
mathematicians 16 28.1 12.5
"Representative"
women mathematicians 28 26.9 15.4

Unselected adult males 343 13.9 11.2

 

63

be derived from the test booklet shown to the subjects in
this stddy, which contains a total of 86 designs.

The 62 item Barron-Welsh scale in the manual corres-
ponds to the earlier BWAS, with two duplicate designs and
one all black design eliminated. Scores on this scale were
therefore used for purposes of comparing the artists in
this study with those in earlier reports. For all other
analyses, however, the artists' protocols were rescored
according tO the later 60 item Revised Art Scale (RA).

The earlier version with its preponderance of "Dislike"
items was considered to be more vulnerable to response set.
Correlation between the two sets of scores is high:

r = .85 (p < .001)

On the Revised Art Scale, the mean score of the total
sample is 33.73, with a range from 17 to 49 and a standard
deviation Of 8.04. Mean scores for the various subgroups
in the sample are reported in Table 3. An analysis of
variance shows no significant main effects or interactions
according to either the sex or medium of the artists in

this sample (see Table 4).

Additional analyses

In an attempt to assess the influence Of response set
on their ratings Of the art scale designs, subjects were
asked at the conclusion of testing to sort the 47 Holtzman
cards according to whether they liked or disliked them.

Of the 16 artists who liked more than half of the 86

64

 

 

Table 3. Mean Scores on the Revised Art Scale, According
to Sex, Medium, or Sex and Medium. (N = 40)
Number Of . Standard

Group Subjects Mean Variance Deviation
Sex

Males 20 33.30 59.91 7.74
Females 20 34.15 72.45 8.51
Medium

Painters 20 33.05 57.95 7.61
Potters 20 34.40 73.83 8.59
Sex and medium

Male painters 10 34.20 78.84 8.88
Female painters 10 31.90 40.54 6.37
Male potters 10 32.40 45.84 6.77
Female potters 10 36.40 101.16 10.06

 

 

 

Table 4. An Analysis of Variance for the Effect of Sex
or Medium on Scores on the Revised Art Scale.
(N = 40)
. . Sum Of Mean Signif-
Source of VarIation Squares DF Squares icance
Main effects 25.45 2 12.73 .19 n.s.
Sex _7.23 l 7.23 .11 n.s.
Medium 18.23 1 18.23 .27 n.s.
Two-way interactions 99.23 1 99.23 1.49 n.s.
Sex X medium 99.23 1 99.23 1.49 n.s.
Explained 124.68 3 41.56 .62 n.s
Residual 2397.30 36 66.59
Total 2521.98 39 64.67

 

65

designs in the art scale test booklet, 10 also liked more
than half of the HIT cards. Of the 24 disliked more than
half the art scale designs, 18 also disliked more than

half Of the HIT cards. Of the artists tested in this study,
therefore, 70% showed a tendency to rate in the same dir-
ection on both sets Of designs. The correlation of the two

"Like" totals is .42 (p < .01).

Hypotheses 2 and 3

 

The second and third hypotheses predict that there
will be a significant positive correlation between soores
on the Revised Art Scale and Adaptive Regression Scores
Obtained from responses to the Holtzman Inkblot Technique,
that high scores on the RA will be associated with high
ARS and low scores on the RA will be associated with low
ARS. These hypotheses are not sustained by the results.

For the sample as a whole, the correlation of RA
scores with ARS is negative and not significantly different
from zero: r = -.06, p = .36. These correlations do not
vary significantly according to sex or medium (see Table 5).
Mean Adaptive Regression Scores for each subgroup are given
in Table 6. An analysis of variance shows no significant

effects for either sex or medium (Table 7).

Additional Analyses
For a discussion Of processes of adaptive regres-
sion, other categories in the Holt scoring system are of

interest, apart from the summary Adaptive Regression Scores.

66

 

 

Table 5. Correlations Of Adaptive Regression Scores with
Scores on the Revised Art Scale, According to
Sex, Medium, or Sex and Medium. (N = 40)
Number of . . .
Group Subjects Pearson r SIgnIfIcance
Sex
Males 20 -.06 n.s.
Females 20 -.07 n.s.
Medium
Painters 20 -.19 n.s.
Potters 20 .19 n.s.
Sex and medium*
Male painters 10 -.05 n.s.
Female painters 10 -.39 n.s.
Male potters 10 .01 n.s.
Female potters 10 .48 n.s.

 

* A x2 test for homogeneity, with 3 df, shows the

differences among these groups to be nonsignificant.

 

 

Table 6. Mean Adaptive Regression Scores, According to
Sex, Medium, or Sex and Medium. (N = 40)
Number of . Standard
Group Subjects Mean VarIance Deviation
Sex
Males 20 1.51 1.47 1.21
Females- 20 1.53 1.14 1.07
Medium
Painters 20 1.66 2.04 1.43
Potters 20 1.38 .53 .73
Sex and medium
Male painters 10 1.33 2.77 1.67
Female painters 10 1.98 1.30 1.14
Male potters 10 1.70 .26 .51
Female potters 10 1.07 .64 .80

 

67

Table 7. An Analysis of Variance for the Effect Of Sex
or Medium on Adaptive Regression Scores.

 

 

(N = 40)
Source Of Sum of D Mean F Signifi-
Variation Squares Square cance
Main effects .75 2 .38 .30 n.s.
Sex .00 1 .00 .00 n.s.
Medium .75 l .75 .61 n.s.
Two-way interactions 4.11 l 4.11 3.31 n.s.
Sex X medium 4.11 1 4.11 3.31 n.s.
Explained 4.86 3 1.62 1.31
Residual 44.68 36 1.24'
Total 49.55 39

 

These are (l) the percentage Of responses scorable for
primary process (PPR) and (2) the scoring categories re-
lated to Defense Demand (DD) and Defense Effectiveness
(DE). PPR indicates the degree to which drive dominated
material has entered into the subject's responses. The
mean score on Defense Demand can be thought of as a mea-
sure Of the intensity or potential threat of this material.
And the mean DE score provides an estimate of how effec-
tively the subject has been able to neutralize aggressive
or libidinal content by using it imaginatively or by
making it appear more socially acceptable.

For the sample as a whole, the mean PPR is .60, with
a range from .29 to .89 and a standard deviation of .16.
The median is .62. Table 8 gives the mean scores on this

variable for the various subgroups. An analysis of variance

68

Table 8. Mean Percentage of Responses Scorable for Pri-
mary Process, According to Sex, Medium, or Sex
and Medium. (N = 40)

 

 

Number of . Standard

Group Subjects Mean VarIance Deviation
Sex

Males 20 .63 .03 .16
Females 20 .58 .02 .15
Medium .

Painters 20 .63 .03 .18
Potters 20 .58 .01 .12
Sex and medium

Male painters 10 .66 .04 .20
Female painters 10 .60 .03 .17
Male potters 10 .60 .01 .10
Female potters 10 .56 .02 .13

 

shows no significant effects for either sex or medium
(see Table 9).

The overall mean score on Defense Demand is 3.47,
indicating that the average response earned between three
and four points for the level of potential threat imposed
by the primary process implicit in the percept. Since
the DD scores were summed in the method used for this
study, this constitutes a rather modest level of threat.

The mean for Defense Effectiveness is .87. To in-
terpret this figure, it should be noted that Holt's rat-
ings extend from -3 (seen only in psychotic records) to
+2 (a completely successful response). A score of .87
falls between "fairly successful" and "successful" control

and defense.

69

Table 9. An Analysis Of Variance for the Effect of Sex
or Medium on the Percentage of Responses Scor-
able for Primary Process. (N = 40)

 

 

Source of Sum Of DF Mean Signifi-
Variation Squares Square canoe
Main effects .05 2 .03 1.07 n.s.
Sex .02 1 .02 .92 n.s.
Medium .03 l .03 1.22 n.s.
Two-way interactions .00 1 .00 .05 n.s.
Sex x medium .00 l .00 .05 n.s.
Explained .05 3 .02 .73 n.s.
Residual .88 36 .02
Total .93 39 .02

 

Mean scores for the subgroups on Defense Demand and

Defense Effectiveness are shown in Tables 10 and 12.
Once again the homogeneity of this sample Of artists is
evident. An analysis of variance in each case shows no
significant main effects or interactions (Tables 11 and
13).

While Adaptive Regression Scores do not correlate
significantly with scores on the Revised Art Scale as pre-
dicted, certain of the subgroups within the sample do show
significant correlations with RA for some of the other var-
iables. These are shown in Table 14. For the male artists
as a group there is a significant positive correlation of
RA with PPR and-a significant negative correlation of RA
with mean DE. This latter correlation is also signifi-

cantly different from that of the female artists as a group.

j

.m

7O

 

 

Table 10. Mean Scores on Defense Demand, According to
Sex, Medium, or Sex and Medium. (N = 40)
Number Of . Standard

Group Subjects Mean VarIance Deviation
Sex

Males 20 3.59 .75 .87
Females 20 3.35 1.04 1.02
Medium

Painters 20 3.52 1.12 1.06
Potters 20 3.42 .69 .83
Sex and medium

Male painters 10 3.66 1.30 1.14
Female painters 10 3.38 1.03 1.01
Male potters 10 3.52 .27 .53
Female potters 10 3.33 1.17 1.08

 

 

 

Table 11. An Analysis of Variance for the Effect of Sex
or Medium on Defense Demand. (N = 40)

Source of Sum of DE Mean F Signifi-

Variation Squares Square canoe

Main effects .65 2 .33 .35 n.s.
Sex .56 1 .56 .59 n.s.
Medium .09 1 .09 .10 n.s

Two-way interactions .02 l .02 .02 n.s.
Sex X Medium .02 1 .02 .02 n.s.

Explained .67 3 .22 .24

Residual 33.89 36 .94

Total 34.56 39 .89

 

71

Table 12. Mean Scores on Defense Effectiveness, According

 

 

to Sex, Medium, or Sex and Medium. = 40)
Number of . Standard

Group Subjects Mean VarIance Deviation
Sex

Males 20 .86 .10 .32
Females 20 .87 .08 .28
Medium

Painters 20 .88 .10 .32
Potters 20 .85 .07 .27
Sex and medium

Male painters 10 .80 .16 .40
Female painters 10 .96 .04 .21
Male potters 10 .92 .04 .21
Female potters 10 .78 .10 .32

 

Table 13. An Analysis of Variance for the Effect of Sex

 

 

or Medium on Defense Effectiveness. (N = 40)

Source of Sum of DF Mean Signifi-
Variation Squares Square canoe
Main effects .01 2 .01 .06 n.s.

Sex .00 1 .00 .02 n.s.
Medium .01 l .01 .11 n. .
Two-way interactions .23 l .23 2.72 n.s.

Sex X medium .23 1 .23 2.72 n.s.
Explained .25 3 .08 .95 n.s
Residual 3.10 36 .09

Total 3.34 39 .09

 

72

Table 14. Pearson Correlations of Scores on the Revised
Art Scale with Mean Scores for Defense Demand
and Defense Effectiveness and with Percentage
of Primary Process, According to Sex, Medium,
or Sex and Medium. (N = 40)

 

Group Number Of Mean DD Mean DE PPR With

 

 

Subjects with RA with RA RA
Sex
Males 20 .26 -.42* .49** } *
Females 20 -.23 .20 -.15
Medium
Painters 20 .02 -.40* .20
Potters 20 -.06 .19 .15
Sex and medium
Male painters 10 .29 -.42 -.05
Female painters 10 -.45 -.27 -.33
Male potters 10 .18 -.39 .01
Female potters 10 -.ll .57* .48
* p < .05
** p < .025

¢ The difference between these two correlation co-

efficients is also significant (x2 with 1 df, p < .05).
If the sample is divided into painters and potters, we see
the same negative correlation of DE with RA for the paint-
ers. Only the female potters show a significant positive
correlation.

Of the four groups it is the female potters who come
closest to performing as predicted. Their mean score on
the art scale is the highest of the four subgroups: at
38.30 on the BWAS it is not significantly different from

the mean of 40 predicted for artist groups. They also

73

show a positive trend in the correlation of RA scores
with Adaptive Regression Scores (r = .48), although with
an n of 10 this does not reach the level of significance
(p = .08). An inspection of their scores shows that this
correlation accompanies low rather than high levels of
adaptive regression (Hypothesis 3). NO subject in the
female potter group had an ARS as high as half a standard
deviation above the mean, while six Of the ten had soores
more than half a standard deviation beigy_the mean.

The female painters show negative correlations of RA
with all four scoring categories, generally Opposite to
the trend of the female potters. The male painters and
potters, on the other hand, Show scores which are closely
parallel with one another: positive correlations with
percentage of primary process and Defense Demand, but
negative correlations with Defense Effectiveness. It
appears that in this sample Of artists the choice of med-
ium may express greater differences for the females than

for the males.

VII . DISCUSSION

The major hypothesis of this study is drawn from
Kris's theory that the creative process involves an adap-
tive stance toward the products of unconscious or precon-
scious thought, a type of functioning which he calls
"regression in the service of the ego" (1952). As Opera-
tionalized in Holt's scoring system for Rorschach responses,
this type of "voluntary" regression can be seen in an un-
threatened willingness to entertain bizarre, unconven-
tional, violent, or openly sexual images--images which do
not flood conscious processes or provoke self-protective
reactions but which are received as the raw material for
an original synthesis. This has been described as the
creative response to disorder, and it has been suggested
that an artist finds apparent disorder stimulating since
it leads to synthesis-seeking, an activity which is ex-
perienced as enlivening and inherently rewarding. If
this can be accepted as one way of describing a creative
response to the ambiguous visual elements of an inkblot,
then,the hypothesis continues, the same mechanism should
be apparent in the response to other visual stimuli. If
an artist is given the choice between designs which are

simple, symmetrical, and orderly and those which are

74

75

complex, asymmetrical, and less Obviously ordered, he or
she will tend to prefer the latter. Like the inkblots,
they serve to initiate the seeking Of a new and more
subtle type of order.

The art scale developed by Barron and Welsh (1952)
has been widely assumed to tap into this process. A num-
ber of artists or other "creative" groups have shown con-
sistent preferences for the complex and asymmetrical de-
signs among those which the test presents. In this study
it was predicted that a sample of working artists--paint-
ers and potters--would share these preferences, and that,
in addition, these preferences would prove to be correlates,
or predictors, Of the more complex processes of adaptive
regression. Neither of these expectations has been con-
firmed. The mean score of these artists on the Revised
Art Scale is significantly lower than what has come to be
considered typical for artist groups. In addition, their
art scale soores bear no statistically significant rela-
tion to the Operations of adaptive regression, insOfar as
these can be observed in responses to the Holtzman ink-
blots.

Reasons for the failure of the group as a whole to
match the mean scores of the standardization and cross-
validation samples on the art scale are difficult to pin
down. It may simply be due to sample differences: the
tastes of a 1978 group of midwestern artists may simply

be different from those of the groups tested by Barron and

76

Welsh in the early fifties. To assume this would be to
dismiss the question too quickly, however, since these
artists' scores do resemble those reported by Getzels and
Csikszentmihalyi in 1946 for a talented art student sample.
They are also clearly more similar to scores given in the
literature for other creative groups than to scores of
non-artists. See for example the mean scores which Barron
(1965) lists for "creative" architects and the scores he
cites for non-artists and less creative groups. It may

be that this group is more typical of what Barron calls
"representative" artists--those who fall somewhere between
the untalented and the famous. Indeed, the artists in
this sample were selected precisely because they could be
characterized as "working artists," committed but not un-
usually successful professionals who might be more repre-
sentative of the artist population as a whole. To test
the possibility that this accounts for their lower scores
would require more detailed information on the make-up of
earlier samples, plus a means of rating creativity or ex-
cellence. It cannot be assumed that the art scale in it-
self is a creativity measure, although this is frequently
done in the literature because Of its generally high cor-
relations with other measures Of artistic ability. Nor
does this proposition explain why these artists' prefer-
ence scores should in some cases be exceeded by those re-
ported for pre-professional students (Rosen, 1955).

The question of changes in artistic preferences over

77

time needs to be considered. There may be a conventional
view within the art field regarding symmetry, for example,
which differs appreciably from that of non-artists but
which varies in pOpularity according to the problems which
stimulate artists in any given era. Since the scale was
devised, there has been a renewed interest on the part of
some artists in the potentialities of classically simple
forms. Several of the artists spontaneously commented
while taking the test that they supposed they were meant
to lite the more geometrical designs because these were
"in." There is some support for this interpretation in
the ratings which some of the subjects gave to designs
classified by the authors of the test as "ruled, simple."
These include such simple symmetrical figures as squares,
rectangles, and the like. While the bulk of these artists
checked "dislike" for these designs, as expected, a sub-
stantial minority (13 out of 40) "liked" more than half

of them.

The artists' informal comments on the test were en-
lightening in other ways. A number of them complained
about the workmanship in the instrument itself, pointing
out the uneven line quality or the crowded format of the
test booklet, which they found distracting. "You can
tell it was done by a psychologist," said one, who added
that he might have liked more of the designs if they had
been better executed. The influence of response set was

clear, confirming the early finding of Barron and Welsh

78

(1952) that one of the two major factors influencing
scores was a predisposition to accept or reject. Some
subjects said that they considered themselves critical
and hard to please. A couple remarked that they enjoyed
evaluating almost anything they looked at.

In regard to the hypothesis that the ability to re-
gress adaptively will tend to correlate with preferences
for complex and asymmetrical designs, this proves not to
be the case for this sample. Indeed, evidence from the
results suggests that for many of these artists the oppo-
site is true. At least in the case of the males as a
group, the more primary process produced the more likely
they are to prefer the predicted types of designs but the
less likely they are to show effective defenses. This is
entirely a matter of correlations--in their mean scores
on Defense Effectiveness they do not differ from the fe-
males. Why their reactions to the Barron-Welsh should
Show this negative association with the ability to neu-
tralize primary process is not at all clear. The same
trend holds to some extent for painters as a group, but
not for potters.

Differences in medium appear to be more salient for
the women than for the men. The male painters and potters
more nearly resembled one another in their responses than
the female painters and potters, who showed opposite
trends. One can only speculate as to reasons for this.

For example, it may represent a social rather than a

79

psychological phenomenon. Pottery is a popular and soci-
ally recognized artistic activity for a woman who has

been occupied in raising a family. While a number of
these women are gifted, and recognized as such, some Of
them are just beginning to emerge into full productivity
after years of giving priority to husband and children.
For a woman to paint at a level of involvement which qual-
ifies her for professional level exhibitions a more radi-
cal choice may be implied.

The female potters are the only group for whom the
relation between design preferences and effectiveness of
defense tends to correlate as predicted. As noted above,
however, this correlation accompanies low rather than high
levels of adaptive regression. Is it possible that the more
adaptive the response to primary process the more idiosyn-
cratic and the less predictable artistic taste will be,
at least in an artist sample? In interviewing these art-
ists, certainly, the idiosyncracies Of individual approaches
was evident. As a conversation starter, each artist was
asked, "What helps you begin the process? What kind of
conditions do you like to set up for yourself before you
set to work?" Their responses often seemed to illustrate
the seeking of a personally satisfying balance between
orderly and disorderly processes and environments. An
artist whose responses were rich with uninhibited primary
process described a worktable as meticulously laid out as

a surgeon's tray. Another, whose responses were

80

impressionistic, fanciful and loose in form, said that
work was impossible until the house was clean. Still an-
other, whose 1ife tends to be heavily structured by ex-
ternal demands, looked forward to time in the studio as
an experience of release. "It's very therapeutic," said
one potter, "to be up to your neck in mud." This antici-
pation of release seemed especially poignant in the women
artists, much of whose time has to be spent in organizing
family life. There has been no attempt to quantify the
interview material, but the interviewer's subjective im-
pression was that each artist seemed to be working toward
a personal equation balancing internal and external free-
doms and constraints. Occasionally the balance was sur-
prising. Some of the least repressed material came from
artists functioning in highly organized settings. Some
Of the most controlled--or most uneasy--came from those
whose style of life and work more nearly matched the
stereotype of artistic unconventionality.

When all is said and done, perhaps the most striking
outcome of this study is the very paucity of significant
differences between groups--the notable homogeneity of
the scores on the various dimensions of adaptive regression.
Whatever the specific differences in the technical problems
imposed by the different media, the psychological process
of creating, at the level defined by Kris's theory and
measured by Holt's scoring system, appears to be common to

both these types of visual artists.

81

A previous study by Pine (1962) may cast some light
on these similarities. Pine also correlated scores on the
Barron-Welsh with a measure Of adaptive regression. In
a sample of 27 students, he found that the males' adaptive
regression scores correlated highly with the Barron-Welsh
scores (r. = .74, p < .001), while the correlation for
the females was low and nonsignificant (r = -.10). For
a group of 50 male actors also tested by Pine the correl-
ation was again low and nonsignificant (r = .15). Indeed,
neither the female students nor the male actors showed
significant correlations between any of Pine's creativity
measures and their adaptive regression scores on the Ror-
schach.

In this discussion, Pine surmises that the differ-
ences are in part related to the degree of control which
.the subjects exert over primary process expression. He
notes that both the female students and the male actors
were less controlled in their responses than the male stu-
dents. Referring to the male actors, he suggests the pos-
sibility that

. . . in a group that as a whole can so readily

express primary process content, minor variations

in this are not significant predictdrs of varia-

tions in the quality of imaginative productions.

This would be in contrast to the sample of male

undergraduates who gave fewer and less extreme pri-

mary process responses than either the actors or

the female undergraduates and who, by personal im-

pression as well, were a more controlled group.

In such a group, the differential capacity to "open

up" expressively (with adequate control) appears

to be a significant predictor of quality of created

productions.
(1962, p. 510)

82

Perhaps the same applies to subjects in this study. With
few exceptions, they were highly expressive and entered
readily into the imaginative task. In an all-artist sam-
ple, minor variations in the components Of adaptive regres-
sion are not significant predictors Of artistic preferences
for order or disorder.

In his introduction, Pine suggests that Kris's dis-
cussion of regression in the service of the ego and its
relation to creativity should be viewed as a broad approx-
imation, which might be subject to variation according to
the nature of the sample under investigation. The present
study has attempted to specify two groups of artists care-
fully according to medium and has found that for visual
artists responding to visual stimuli the creative process
appears to be, psychologically at least, markedly similar.
To test this proposition further, it would be desirable
to extend it into the specializations and subspecializa-
tions of other, nonvisual artists, such as writers,
musicians, or performers.

The similarity in scores between male and female
artiSts is especially provocative, given the rarity Of
such comparisons in this literature. Certainly, this

avenue of research ought to be pursued.

VIII. SUMMARY AND CONCLUSIONS

Kris's theory of regression in the service of the
ego has been widely endorsed as a description of the dy-
namics of artistic creation. This is conceived of as a
two-stage process. In the first stage, primitive illog-
ical thinking is allowed to occur, or ideas with sexual
and aggressive content are allowed to enter awareness, in
a temporary relaxing of normal controls. In the second
stage, controls are reinstituted and the artist's rational
and critical faculties are applied to the products of pre-
conscious thought. This can be called a positive approach
toward the experience of disorder, which has within it for
the artist the potential for discovery or creation of a
new order. Such an attitude has an analogue in the pref-
erence for complex and asymmetrical designs found by
Barron and Welsh to characterize artist groups as compared
to non-artists.

It was hypothesized that these two styles of approach
--to intrapsychic experience and to external stimuli--
should be positively correlated for visual artists. A
group of 40 working artists was administered the revised
form of the Barron-Welsh Art Scale and the Holtzman Ink-

blot Technique. Scoring for the HIT followed the system

83

84

developed by Holt as a measure of adaptive vs. maladap-
tive regression in Rorschach responses. Findings‘were
that this group scored significantly below the norm for
creative artists on the Art Scale and that their scores

on the scale did not correlate positively and significantly
with their Adaptive Regression Scores. Both these find-
ings are contrary to predictions. Additional findings
were that there were no significant differences on these
measures between male and female artists or between art-
ists in the two media represented, painting and pottery.
Indeed, the sample proved to be markedly homogeneous on
all variables of interest, including the percentage of
primary process responses, the level of defense demand
generated by these responses, and the effectiveness of the
defenses used to deal with potentially disturbing percepts.
These findings were discussed in the light of Pine's spec-
ulation that within a group which already tends to be open
and expressive fine variations on the dimension of adap-
tive regression do not have a significant influence on the
quality of created productions. It may be that while
artists at a similar level Of professional achievement
show similar internal processes of creativity, their
tastes remain idiosyncratic and unpredictable. In this
study, at least, artistic preferences were found to be

uncorrelated with processes of adaptive regression.

APPENDIX

APPENDIX

Form Level Rating Scale

 

(From Holt, 1970, p. 91, after Mayman)

Rating
7
6

Sharp, convincing forms, easily seen by E.
Popular and near-popular forms.

Reasonably plausible, but not terribly convinc-
ing forms; takes a little stretching to see.

Forms that bear only a light resemblance to the
blot area; not very plausible, or based on only
one point of resemblance.

Arbitrary forms, very little or no resemblance.

Vague forms that fit the blot quite well, and
non-definitive form combined with appropriate use
of color or shading ("fire," "flowers," "dark
clouds," "splashing water," "inkstain, running
down a water color").

Vague forms with no other determinant, or forced
use thereof: "clouds," "islands," "cave mouth,"
"piece of dough."

Amorphous responses, in which form plays ng_role
(and could not, by the nature of the concept).
"Sky," "water," "night," "spring" (and other ab-
stract concepts), "urine," (but "wine stain" seen
as having some sort of shape).

Spoiled form responses, to be used when the sub--
ject gives what is basically a familiar and good
response but introduces some specification that
has the effect of markedly lowering the accep-
tability of the response as a whole.

These definitions were used in rating the form

85

86

level of responses to the Holtzman inkblots, and were com-
bined with the criteria in the Holtzman manual for "Form
Appropriateness." The Holtzman "Good" category was con-
sidered equivalent to a rating of 7, "Fair" to a rating of
5, and "Poor" to a rating of 4 or below. Holtzman does
not provide a separate category for popular responses;
they may appear as either "Good" or "Fair." The rule
followed was to assign the higher rating only if the per-
cept was well elaborated. For example, on Card 12 a
popular response involves figures dancing around a fire.
Holtzman rates this both "Good" and "Popular." If the
subject merely referred to "people dancing around a fire"
this was scored 6, as a Popular; if the subject specified
"girls dancing around a fire, with clouds in the sky" it

was given the full score of 7.

REFERENCES

REFERENCES

Aiken, L. R. A review of research on the Welsh Figure
Preference Test. Greensboro, North Carolina:
Creativity Research Institute Of the Richardson
Foundation, June, 1967.

Andersen, B. S. and Munroe, R. Personality factors in-
volved in student concentration on creative paint-
ing and commercial art. Rorschach Research Exchange,
1948, 13(3), 141-154.

Arieti, S. Creativity: The magic synthesis. New York:
Basic BoOks, Inc., I976.

Arnheim, R. Art and visual perception: A psychology of
the creaEIVe e e. Berkeley: UnIverSIty O CalI -
ornIa Press, 1954.

Arnheim, R. Visual thinking. Berkeley: University of
California Press, 1969.

 

 

Barron, F. Personality style and perceptual choice.
Journal of Personality, 1952, 20, 385-401.

 

Barron, F. Complexity - simplicity as a personality di-
mension. Journal of Abnormal and Social Psychology,
1953, 48, 163-172.

Barron, F. The psychology of imagination. Scientific
American, 1958, 199, 150-166.

 

 

Barron, F. The psychology of creativity. In F. Barron,
et al. New directions in psychology II, New York:
Holt, Rinehart & Winston, 1965}

 

Barron, F. Creativg_persons and creative_process. New
York: H61t, Rinehart and Winston, 1969.

Barron, F. and Welsh, G. S. Artistic perception as a pos-
sible factor in personality style: Its measurement
by a figure preference test. Journal of Psychology,
1952, 33, 199-263.

87

88

Bowers, K. and van der Meulen, S. J. Effect of hypnotic
susceptibility on creativity test performance.
Journal of Personality and Social Psychology, 1970,
£4(3), 247-256}

 

Cohen, I. H. Adaptive regression, dogmatism, and crea-
tivity. UnpuBlished doctoral diSsertatiOn, MiChigan
State University, 1960.

 

Cronbach, L. J. Statistical methods applied to the Ror-
schach: A review. Psychological Bulletin, 1949,
46, 393-429.

 

Crutchfield, R. S. Conformity and creative thinking. In
H. E. Gruber, G. Terrell & M. Wertheimer (Eds.),
Contemporary approaches to creative thinking. New
York: Atherton, I962.

 

Drevdahl, J. E. Factors of importance for creativity.
Journal of Clincial Psychology, 1956, £2, 21-26.

 

Drevdahl, J. E. and Cattell, R. B. Personality and crea-
tivity in artists and writers. Journal Of Clinical
Psychology, 1958, ii, 107-111.

 

 

Dudek, S. 2. Regression and creativity. A comparison of
the Rorschach records Of successful vs. unsuccessful
painters and writers. Journal of Nervous and Mental
Disease, 1968, l41(6), 5353546.

 

Eiduson, B. T. Artist and nonartist: A comparative study.
Journal of Personality, 1958, 26, 13-28.

 

Eindhoven, J. E. and Vinacke, W. E. Creative processes
in thinking. Journal Of General Psychology, 1952,
g, 139-1640

 

Fairbairn, W. R. D: Prolegomena to a study of art.
British Journal of PsycholOgy, 1938, 28, 288-303.

 

Fraiberg, L. Freud's writings on art. International Jour-
nal of Psycho-analysis. 1956, 31, 82-96.

 

 

Freud, S. [Leonardo da Vinci and a memory of his child-
hood]. In J. Strachey (Ed. and trans.), The stand-
ard edition of the complete psychological works of
SigmundIFreud (VOIT'IIYI London: Hogarth, I957(a).
(Originally published, 1910).

 

 

Freud, S. [Some character types met with in psychoanal-
ytic work.] In J. Strachey (Ed. and trans.), 222
standard edition of the complete psychological
works of Sigmund Freud (Vol. 14f. London: Hogarth,
1957(b). (OriginaIIy published, 1916.)

 

89

Freud, S. [Formulations on the two principles of mental
functioning.] In J. Strachey (Ed. and trans.), The
standard edition of the complete psychological WEEKS
of Sigmund Freud (Vol. 12). London: Hogarth, 1958.
(Originally published, 1911).

 

Freud, S. [Creative writers and daydreaming.] In J.
Strachey (Ed. and trans.), The standard edition of
the complete psychological worEs of SigmundIFreud
(Vol. 9). Lofidon: Hogarth, 1959. (Originally
published, 1908.)

Freud, S. [Dostoevsky and parricide]. In J. Strachey
(Ed. and trans.), The standard edition of the com-
plete psychological works of Sigmund Freud (Vol. 21).
London: Hogarth, I961. (OriginaIly puBlished,
1928.)

Galton, F. Hereditary_genius. New York: D. Appleton,
1870.

 

Getzels, J. W. and Csikszentmihalyi, M. Creative think-
ing in art students: An exploratory study. Cooper-
ative research project No. E-008. Chicago: Univer-
sity of Chicago, 1964. (ERIC NO. ED 003 377).

Getzels, J. W. and Jackson, P. W. Creativity and intel-
ligence. Explorations with gifted students. New
York: Wiley, 1962.

Ghiselin, B. (Ed.) The creative process: A symposium.
New York: New American Library, 1952.

Golann, S. E. Psychological study of creativity. Psycho-
logical Bulletin, 1963, 33(6), 548-565.

 

Gray, J. J. The effect of productivity on primary pro-
cess and creativity. Journal of Projgctive Tech-
niques and Personality Assessment, 1969, 33(3), 213-
2R.

Gough, H. G. Techniques for identifying the creative re-
search scientist. In Proceedings, The gteative
person. A conference presented at the Tahoe Alumni
Center, October 13-17, 1961. Berkeley: University
of California, Institute of Personality Assessment
and Research, 1961.

Guilford, J. P. Creativity. American Psychologist, 1950,
3, 444-454.

Guilford, J. P. Traits of creativity. In H. H. Anderson
(Ed.), Creativity and its cultivation. New York:
Harper, 1959.

9O

Havelka, J. The nature of ths_creative_process in art:
A psychological stfidy. The Hague: MarEinus Nijhoff,
1968.

 

Helson, R. Creativity, sex, and mathematics. In Proceed-
ings, the creative person. A conference presented
at the Tahoe Alumni Center, October 13-17, 1961.
Berkeley: University of California, Institute of
Personality Assessment and Research, 1961.

 

Helson, R. Personality of women with imaginative and
artistic interests: The role of masculinity, orig-
inality and other characteristics in their creativ-
ity. Journal of Personality, 1967, 33, 214-233.

Hersch, C. The cognitive functioning of the creative per-
son: A development analysis. Journal Of Projective
Techniques, 1962, 33, 193-200.

 

Holt, R. R. Manual for the scoring of primary process
manifestations in Rorschach responses. 10th draft.
Research Center for Mental Health, New York Univer-
sity, March, 1970.

Holt, R. R. and Havel, J. A method for assessing pri-
mary and secondary process in the Rorschach. In
M. A. Rickers-OvSiankina (Ed.), Rorschach psychology.
New York: Wiley, 1960.

Holtzman, W. H. Holtzman inkblot technique. In A. I.
Rabin (Ed.), Projective techniques in personality
assessment. New York: Spring, 1968.

 

Holtzman, W. H., Swartz, J. D., and Thorpe, J. S. Artists,
architects, and engineers--Three contrasting modes
of visual experience and their psychological cor-
relates. Journal Of Personality, 1971, 33(3), 432-
449.

Holtzman, W. H., Thorpe, J. S., Swartz, J. D. and Herron,
E. W. Inkblot_perception andtpersonality. Austin:
University of Texas Press, 1961.

Jones, E. A. Art. In The life and work Of Sigmund Freud
Vol. 3, Chapter 15. New York: Basic Books, 1957.

Jung, C. G. [On the relation of analytical psychology to
poetic art.] In Contributions to analytical psy-
chology, H. G. and C. F. Baynes (trans.). New York:
Harcourt, Brace & CO., 1928, 225-249.

Kris, E. Psychoanalytic explorations in art. New York:
InternatiOnal Universities Press,’1952.

91

Kubie, L. S. Neurotic distortion of the creative process.
Lawrence, Kansas: UniVersity of Kansas Press, 1958.

 

Lake, A. W. III and Telford, W. H. Jr. Influence of crea-
tivity on formation of subjective units. Journal of
General Psychology, 1970, 33, 227-237.

 

 

Lee, H. B. A theory concerning free creation in the inven-
tive arts. PSYChiatEXI 1940, 3.

 

Lombroso, C. The man of genius. London: W. Scott, 1891.

 

MacKinnon, D. W. The study of creativity. Creativity in
architects. In Proceedings. The creative person.
A conference presented at the Tahoe Alumni Center,
October 13-17, 1961. Berkeley: University of Cal-
ifornia, Institute of Personality Assessment and
Research, 1961.

 

MacKinnon, D. W. The nature and nurture of creative tal-
ent. American Psychologist, 1962, 33, 484-495.

 

Mackler, B. and Shontz, F. C. Creativity: Theoretical
and methodological considerations. Psychological

 

Maslow, A. H. Creativity in self-actualizing people. In
H. Anderson (Ed.), Creativity and its cultivation.
New York: Harper, 1959.

 

Munsterberg, E. and Mussen, P. H. The personality struc-
tures of art students. Journal of Personality, 1953,
33, 457-466.

Myden, W. Interpretation and evaluation of certain person-
ality characteristics involved in creative production.
Perceptual and Motor Skills, 1959, 3, 139-158.

Nicholls, J. G. Creativity in the person who will never
produce anything original and useful: The concept
of creativity as a normally distributed trait.
American Psychologist, 1972, 33(2), 717-727.

 

Patrick, C. Creative thought in artists. Journal of
Psychology, 1937, 3, 35-73.

 

 

Pine, F. Creativity and primary process: Sample varia-
tions. Journal of Nervous and Mental Disease, 1962,
134, 506-511.

Pine, F. and Holt, R. R. Creativity and primary process:
A study of adaptive regression. Journal of Abnormal
and Social Psychology, 1960, 33(3), 370-379.

 

92

Prados, M. Rorschach studies on artists - painters. Ror-
schach Research Exchange, 1944, 3, 178-183.

 

Rabin, A. I. (Ed.) Projective techniques in personality
assessment. New York: Sprihger,—1968.

 

 

Rank, 0. Art and artist. New York: A. A. Knopf, 1932.

 

Rawls, J. R. and Boone, J. N. Artistic creativity and
Rorschach whole responses. Journal of Projective
Techniques and Personality Assessment, 1967, 31(3),
18-23. "

 

 

Rawls, J. R. and Slack, G. K. Artists vs. nonartists:
Rorschach determinants and artistic creativity.
Journal of Projective Techniques and Personalipy
Assessment, 1968, 33, 233-237.

 

Rexroth, K. My head gets tooken apart. In Bird in the
bush: Obvious essay_. New York: New Directions,
1959.

 

 

Roe, A. The personality of artists. Educational and
Psychological Measurement, 1946, 3, 40l-4UB.

 

Rorschach, H. [Psychodiagnostics.] P. Lemken & B. Kron-
berg (trans.) New York: Grune & Stratton, 1942.

Rosen, J. C. The Barron-Welsh Art Scale as a predictor of
originality and level of ability among artists.
Journal of Applied Psychology, 1955, 33(5), 366-367.

 

Rosner, S. and Abt, L. B. (Eds.) The creative experience.
New York: Grossman, 1970.

Rothenberg, A. and Greenberg, B. The index of scientific
writipgs on creativity.3§eneral: 1566-1974. Ham-
den, Connecticut: Archon Books, 1974(aYT

Rothenberg, A. and Greenberg, B. The index of scientific
writings on creativity. Creative men a women. Ham-
den, Connecticut: Archon Books, 1974(bTT

 

Rothenberg, A. and Hausman, C. R. (Eds.) The creativity _
question. Durham, North Carolina: Duke University
Press, 1976.

 

Sachs, H. The creative unconscious: Studies in the_psy-
cholanalysis of art. Cambridge, Massachusetts:
Sci:Art Publishers, 1942.

 

Schachtel, E. Metamorphosis. New York: Basic Books, 1959.

 

93

Schafer,Rn Psychoanalytic interpretation in Rorschach
testing: Theory_and application. New York: Grune
& StrattOn, 1954.

 

Schafer, R. Regression in the service of the ego: The
relevance of a psychoanalytic concept for personal-
ity assessment. In G. Lindzey (Ed.), Assessment of
Human Motives. New York: Rinehart, 1959f

 

 

Stein, M. I. Creativity. In Borgatta & Lambert (Eds.),
Handbook of_personality theory and research. Chap-
ter 16. Chicago: Rand McNally, 1969.

Stein, M. I. and Heinze, S. Creativ3ty: Summaries of
selected literature. Chicago: The Free Press, 1960.

 

 

Steiner, M. E. The use of the Rorschach method in indus-
try. Rorschach Research Exchange, 1947, 33, 46-52.

Sterba, R. The problem of art in Freud's writings. Psy-
choanalytic Quarterly, 1940, 3, 256-268.

 

Terman, L. M. (Ed.) Genetic studies of genius (5 vols.).
Stanford: Stanford University Press, 1925-1959.

Torrance, E. P. Guiding cpeative talent. Englewood
Cliffs, New Jersey: Prentice-Hall, 1962.

Wallas, G. The art of thought. New York: Harcourt,
Brace & Co., 1926.

 

Wilson, R. N. The poet and the projective test. Journal
of Aesthetics and Art Criticism, 1958, 33, 319-327.

Wertheimer, M. Productive thinking._ New York: Harper,
1945.