You are on page 1of 9

This article was downloaded by: [Central Michigan University]

On: 10 October 2014, At: 09:59


Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Journal of Projective
Techniques
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/hzpt20

The Color-Shading Response


and Suicide
a a
Stephen A. Appelbaum & Philip S. Holzman
a
The Menninger Foundation , USA
Published online: 16 Nov 2010.

To cite this article: Stephen A. Appelbaum & Philip S. Holzman (1962) The Color-
Shading Response and Suicide, Journal of Projective Techniques, 26:2, 155-161, DOI:
10.1080/08853126.1962.10381091

To link to this article: http://dx.doi.org/10.1080/08853126.1962.10381091

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [Central Michigan University] at 09:59 10 October 2014
The Color-Shading Response and Suicide’
STEPHEN AND PHILIP
A. APPELBAUM S. HOLZMAN
The Menninger Foundation

INTRODUCTION shading within the colored area, and


Hunch or impression, bolstered by not to achromatic color, scored C’, or
persistent efforts to verify it, often the shading to delineate areas, scored
leads to fruitful advances in under- (C) on achromatic cards. It is analo-
standing. In the present study, we at- gous to shading in achromatic areas,
tempt to determine the validity of a scored Ch by Rapaport.
clinical hunch. Some examples of these responses
For many years, several of us work- are:
ing with psychological tests at T h e 1. Card 8, lower center half; “Shape
Downloaded by [Central Michigan University] at 09:59 10 October 2014

Menninger Foundation have had the and color of an iris; the velvety in-
impression that a particular combina- sides because of the shading.”
tion of shading and color determin- Scored FC (C) .
ants of responses on the Rorschach 2. Card 8, upper center blue: “Bluish
Test is associated with suicidal ten- crocheted material, nubby knit or
dencies in patients.2 None of us was boucle.” In inquiry, the texture
bold enough to base any pronounce- was attributed to the shading.
ments of suicidal tendencies on the Scored C (C) F.
appearance of this one sign because
we were averse to a blind sign ap- 3. Card 10, center pink detail: “Some
proach to tests. But, as we continued kind of tissue, skin; it’s burned at
to amass testing experience, the color- the edges.” Inquiry brought out
shading determinant appeared with both color and shading as deter-
insistent regularity in cases of patients minants. Scored C (C) .
who had attempted suicide, some- Over several years of unsystematic
times in the absence of any other observation, the impression grew that
alerting sign in a battery of tests. regardless of the degree of morbidity
Curiosity about this consistency led of the response’s content the color-
us to test this hunch by systematic shading determinant was the crucial
search. variable indicating suicidal tenden-
The responses under consideration cies.
combine color and shading as deter- METHOD
minants whether form is a primary or
a secondary determinant. In scoring If the color-shading determinant
the Rorschach Test, following Rapa- and suicidal proclivity are associated,
port’s notation (Rapaport, et al, the color-shading combination occurs
1946), the responses were denoted as more often among people who either
FC (C) , F/C ( C ) , C (C) F, C (C) /F, or attempted or committed suicide sub-
C (C) . As used here, (C) refers to sequent to testing, or who unsuccess-
~

fully attempted suicide prior to test-


We gratefully acknowledge the helpful criti- ing, than among people who made no
cism of several colleagues, among whom are
Robert R. Holt, Roy Schafer, and Seward suicidal attempts. To test this rela-
Hiltner. An abbreviated version of this paper tionship, Rorschachs were gathered
was read at the 1961 meetings of the Ameri- from seven groups of subjects: two ex-
can Psychological Association, New York, perimental and three control groups
New York.
This observation was originally called to our
matched on age and IQ, and two
attention in 1946 by Dr. Michael Dunn at control groups not matched on age
Winter V.A. Hospital, Topeka, Kansas. and IQ.
156 T h e Color-Shading Response and Suicide
Experimental Groups Control Groups
Suicides. Thirty Ss, seventeen fe- Non-suicidal Psychiatric Patients.
male, thirteen male, who subsequent Ninety-six patients, forty female, fifty-
to being tested, committed suicide. six male, who never made suicidal at-
Twenty-nine were atients in various tempts. These cases were randomly
f
psychiatric hospita s.3 Their psychia-
tric diagnoses included four neurotic
selected from the records of the C. F.
Menninger Memorial Hospital. T h e
depressions, seven character disorders, mean age was 35; the mean IQ was
one involutional depression, two ad- 117. Their diagnoses included thirty-
justment reactions of adolescence, m e seven character disorders, fourteen
obsessive-compulsive neurosis, and neurotic depressions, five psychotic
nine schizophrenic reactions. One depressions, twenty schizophrenic re-
Downloaded by [Central Michigan University] at 09:59 10 October 2014

was not a psychiatric patient, but a actions, eleven anxiety reactions, six
subject in another study. Five sub- obsessive-compulsive neuroses, two
jects were patients in another hospital adjustment reactions of adolescence,
and their diagnoses were not avail- and one chronic brain syndrome.
able. T h e mean age was 33; the mean Patrol. Fifty-three members of the
I Q was 120.4 Kansas Highway Patrol, all male, the
Suicide Attempts. Thirty-nine Ss, “normal Ss” used in the diagnostic
thirty-two female, seven male, who testing studies by Rapaport, et a1
made at least one suicidal attempt (1946) . These Ss are of a lower socio-
prior to being tested. All were pa- economic status than the other Ss in
tients in psychiatric hospitals. Their our study. At the present time, none
diagnoses included twenty character of the Patrol has committed or at-
disorders, five neurotic depressions, tempted suicide. T h e mean age was
four psychotic depressions, and nine 35; the mean IQ was 116.
schizophrenic reactions; one subject’s Thyroid Ss. Fifty-two Ss, all fe-
diagnosis was not available to us. Be- male, in a study investigating thyroid
cause physicians differ in what they gland dysfunction in women between
are willing to call an attempt, a his- the ages of 19 and 44. Twenty-four
trionic display, or a gesture, our cri- of these fifty-two were chosen from a
terion was simply that the patient midwestern city of 100,000 popula-
had made a motoric act of a suicidal tion by a stratified sampling techni-
nature. Consequently, the suicidal at- que. All twenty-four were without
tempts of the patients in this group demonstrable medical or surgical
varied considerably i n their appear- pathology. Thirteen others were
ance of seriousness. I t is a safe as- volunteer Ss for the thyroid study.
sumption, however, that all patients These thirteen were also without
in this group knew that their actions significant medical or surgical patho-
may well have been lethal. We did logy. T h e remaining fifteen Ss were
not include in this group patients referred to the thyroid investigation
who thought about suicide, whether by physicians who suspected thyroid
fleetingly or profoundly, if they did dysfunction. All subjects were given
not act upon these thoughts. T h e a battery of psychological tests and in-
mean age was 32; the mean IQ was terviewed by a senior psychiatrist at
119. T h e Menninger Foundation. T h e
wish to thank Dr. Arthur Kobler of mean age was 34; the mean IQ was
Seattle, Washington for allowing us to study 115.
the Rorschach protocols he obtained in a Psychiatric Residents. Fifty physi-
West Coast Hospital from patients who sub-
sequently committed suicide and from pa- cians, two female and forty-eight
tients who had made a suicidal attempt. male, who had been accepted in the
‘All IQ’s were determined from the Wechs- Menninger School of Psychiatry and
ler-Bellevue Intelligence Scale. who were tested as part of the school’s
AND PHILIPS. HOLZMAN
STEPHENA. APPELBAUM 157
selection procedure. T h e mean age of miss). In both suicidal groups, more-
29 was significantly younger than all over, the sign occurs with greater
except the woman college student than chance frequency, while in the
group. The mean IQ was 129 which control groups it occurs with less than
is significantly higher than all other chance frequency. A Chi-square Test
groups. indicates that the difference between
College Students. Seventeen stu- the combined experimental groups
dents, all female, between 19 and 21 and the combined control groups
years of age, youngest of all the could occur by chance less than once
groups. They were originally Ss in an in a thousand times (Chi-square=
investigation of cognitive function- 111.76) . These results obtain in spite
ing. Wechsler-Bellevue Intelligence of the as yet unknown effect of the
scores were not available for these high IQ and lower age of the resid-
Downloaded by [Central Michigan University] at 09:59 10 October 2014

subjects, but we presumed that they ents, and lower age of the college
were at least of bright normal intelli- students.
gence. While one might have anticipated
The Rorschach Test responses of that the experimental groups would
all subjects in all experimental and show fewer color responses with form
control groups were read and scored as the primary determinant (FC)
than the control groups, all groups
separately by each author. T h e au-
thors then compared their judgments used C (C) F proportionately more
and reconciled their differences in often than FC (C) and C (C) , with no
the few cases of disagreement. One significant difference between the
author was unaware of what groups combined experimental and com-
bined control groups (Chi-square=
the test protocols were drawn from
while he was scoring. T h e color-shad- 2.417, p<.30).
ing determinant was scored when the T h e incidence of color-shading re-
subject specifically mentioned the sponses with “morbid” content (e.g.,
shading as well as the color as a de- “decaying flesh”) is greater in the
terminant, or when the scorers were combined experimental groups
(seven out of 69 color-shading re-
satisfied from the content that shad-
ing had been used, but the subject sponses) than in the combined con-
had not been specific in verbalizing it trol groups, including non-suicidal
or inquiry was not adequate to elicit patients (two out of 51 color-shading
an unequivocal statement by the sub- responses). T h e low frequency in all
groups of such “morbid” responses,
ject (“soiled wallpaper” in the mot- however, indicates that the content of
tled green area of Card IX) .
the color-shading responses is not a
RESULTS
Multiple appearance of the de- TABLE I-Incidence of the Color-
terminant was no more often associat- Shading Determinant in Suicidal
and Non-Suicidal Subjects
ed with suicide than single occurrence
Color-Shading
(Chi-square=0.224) . Therefore, the Present Absent
single occurrence of the sign was Experi- Suicides 27 3
given the same weight as multiple mental
occurrence. Table I shows the in- Groups Attempted Suicides - 32 7
-
cidence of the color-shading determin- Sum 59 lo
ant in the suicide, attempted suicide Non-Suicidal Patients 18 78
Thvroid
and control groups. There is a greater Control Study Subjects 9 43
incidence of the sign among the suc- Highway
Groups Patrol Subjects 5 48
cessful suicidal patients (9 hits for
every miss), than among the suicide
- Psychiatric Residents 12
College Students 7
38
13
attempt groups (4.5 hits for every - -
Sum 51 220
158 The Color-Shading Response a n d Suicide
good distinguishing measure between twenty-eight per cent as the base rate
the experimental and control groups. of suicidal actions in our hospital pa-
For both experimental and control tients. Table I indicates that the test
groups the color-shading responses sign correctly identifies eighty-six per
appeared in descending order on cent of suicidal patients, a significant
VIII, IX, X, I1 and 111, perhaps on increase in predictive power beyond
the basis of the structure of the card. the twenty-eight per cent base rate.
Use of the sign incorrectly labels
Thus, the significant link is be- nineteen per cent of the control popu-
tween suicide or attempted suicide lation as potential suicides. For our
and the simple presence of the color- hospital population the test sign is
shading determinant, irrespective of clearly an alerting signal of suicidal
the number of instances in an in- behavior. Absence of the sign in our
Downloaded by [Central Michigan University] at 09:59 10 October 2014

dividual protocol, how this determin- hospitalized patients, however, is no


ant is combined with form, the mor- valid indication that there is no dan-
bidity of content, or on what card it ger of suicidal behavior. T h e four-
appears. teen per cent false negatives (10 out
Is the sign related to other deter- of 69) are evidence that the sign is
minants such as C and Ch or the not an invariable predictor, and that
number of responses ( R ) ? Positive its presence in a record is a surer alert
correlations were found in all groups to suicidal action than its absence is
between the color-shading determin- a guarantee of no such action. The
ant and the total number of responses incidence of false positives in the con-
per protocol (rp,bis= .31, p < .001,) trol groups (nineteen per cent) how-
presence of achromatic shading re- ever, exceeds our estimate of the base
sponses (rtet= .38, p < .001), and rate of attempts and suicides in the
number of color responses independ- general population (approximately
ent of those combined with shading one percent) 6, indicating that the
(r,,bis=.58, p<.OOl) . Would these sign is not an efficient predictor of
other scores, then, also successfully suicide in the general population.
predict suicide? Forty-nine out of
ninety-six nonsuicidal hospitalized DISCUSSION
patients gave at least one Ch response, T h e color-shading determinant oc-
and eighty-four of ninety-six gave at curs with significantly greater fre-
least one C response. T h e high rate quency in groups of suicides and at-
of false positives invalidates their use tempts than in five control groups.
as a sign of suicidal actions. For our population, at least, predic-
T h e effectiveness of the color-shad- tions of suicide of hospitalized psy-
ing determinant as a predictor of chiatric patients made from the use of
suicide depends upon the sign's in- the color-shading determinant are
creased p r e d i c t i v e power beyond considerably more effective than pre-
what one would expect from a knowl- dictions made from the base rate of
edge only of the base rate of suicide suicide in that population. But the
in the population for which one is determinant is also used by some
predicting. T o estimate the base rate patients (nineteen per cent) who do
of suicide in our population we select- not attempt suicide, although it is
ed a random sample of 252 other likely that a few of these nineteen
patients from admissions to the C. F. per cent false positives would make a
Menninger Memorial Hospital dur- suicidal attempt were they not hos-
ing 1958, 1959 and 1960. From the
records of these patients we deter- There are no reliable studies of the rate of
mined that twenty-eight per cent had suicidal attempts. It seems to us unlikely
that the rate of attempts in the general
at some time attempted or committed population would approach nineteen per-
suicide. We took this figure of cent.
AND PHILIPS. HOLZMAN
STEPHENA. AI~PELBAUM 159
pitalizetl. The clinical records of red and the shading indicates it is en-
several of these patients show reports crusted.” Rather than “blue pillows”
of serious ruminations about suicide, he sees “blue satin pillows,” the shad-
but so do the records of those who do ing indicating the sheen of satin.
not give the determinant. It would Thus, there is a searching, going-be-
be interesting to be able to see which yond-the-givens quality; a spade is
patients subse uently make suicidal more than just a spade. But this ap-
7
attempts; whet ier those patients who
translate their suicidal rumination
proach is also used in giving shading
responses to achromatic areas, and
into suicidal action come from a our data have shown that Ch is not a
population of “susceptibles” that may useful predictor of suicide. Two pos-
be identified by the color-shading de- sibilities present themselves: (1)
terminants. Since more people use achromatic-
Downloaded by [Central Michigan University] at 09:59 10 October 2014

Among the control groups, how- shading than color-shading as re-


ever, the number of false positives far sponse determinants, it may be that
exceeds the death rate from suicide in insofar as the Rorschach Plates are
the general population (approxim- constructed, color-shading is more
ately one per cent) and our estimate difficult to use in a Rorschach test
of the number of suicidal attempts in response. All people who can make
the general population (less than one use of color-shading can use achro-
per cent). Using this determinant as matic-shading, but not all people who
a predictor of suicide is defensible, can use achromatic-shading may be
therefore, only with hospitalized psy- able to use color-shading. Thus, peo-
chiatric patients. These data support ple who use color-shading may possess
the proposition that the process un- the searching, penetrating character-
derlying the use of the color-shading istics hypothetically related to this re-
determinant leads to suicide only sponse to a greater extent than those
when the person’s total functioning who do not, or cannot, use this color-
becomes so impaired that he requires shading determinant. (2) The way
hospitalization. one responds to color on the Ror-
schach Test is assumed to reflect the
What might be the significance of person’s idiosyncratic affect organiza-
color-shading responses for adequate- tion. O n e could, therefore, i n f e r
ly functioning people and for those that the searching, highly differen-
whose personality functioning is im- tiated quality of the color-shading
paired? That the determinant is more response reflects a sensitivity to
common among the psychiatric resi- nuances of feeling. Rapaport, et al,
dents and college students than stated that an accumulation of these
among our other control groups sug- color-shading responses . . . suggests

gests that it may reflect a potential a person given to much mingling of
asset in well-functioning people of affect and anxiety, as is seen in com-
high intelligence, a speculation con- plex emotions like nostalgia (sweet-
sistent with the fact that suicide is sorrow) .” (Rapaport, et al, 1946)
positively related to occupational and Both of these explanations suggest
socio-economic levels. that persons who give even one color-
The color-shading d e t e r m i n a n t shading response have a keener eye
seems to reflect an unusual differen- for what is there. They are more
tiation in responsiveness to color in aware of nuances, not only of the
the Rorschach Plates. Rather than colors they see but perhaps of objects
seeing the colored area as a unity- in general, and can see novelty where
“blood because it is red”-the subject others see only the familiar. T h e
is sensitive to additional properties “work” done in using the color-shad-
which make the response more in- ing determinant consists in an articu-
dividual-“dried blood because it is lating, discovering and penetrating
160 The Color-Shading Response and Suicide
activity. In adequately functioning tion to his entire life situation. A
people of superior intelligence, such too close look at a tree obscures the
sensitivity and capacity for differen- recognition that it is but one tree in
tiated responsiveness may be available a forest. Not only does the person see
for adaptive uses, for example, in dis- only one aspect of himself, but there
criminations between ideas, deep in- is the implication that time too is re-
quiry into the uniqueness of objects, stricted to the all-encompassing pres-
empathy for the nuances of another’s ent. The past, when things were bet-
experience, and the appreciation of ter, seems blotted out as does the fu-
art. ture when things could be better.
However, patients excessively con- Gone with the sense of time are its
cerned with their difficulties and palliative and protective functions.
caught up in their anxieties about liv- With the loss of perspective of self
Downloaded by [Central Michigan University] at 09:59 10 October 2014

ing may turn the capacity to pene- and history, the momentary situation
trate beyond the obvious toward an of travail becomes everything. Vari-
inquiry into their own existence, and ous alternatives and possibilities of
the purpose of their own being. They achieving change for the better seem
may then ask what they want from unavailable. Data on suicide further
life, whether and how they can get it, support our belief that the experience
and whether the struggle is worth it. hypothesized here leads to a short-
They face the problem of Sisyphus, sighted perspective of one’s historical
who could endure the physical strug- continuity and to a feeling that one
gle of continuing to push the rock u p has run out of alternatives: suicide
the hill, but who was tortured by the rates decrease significantly during
recognition that it was all without times of national crisis, when there
purpose, for the rock immediately are socially imposed purposes and
tumbled down again. T h e price of goals, and the rates increase during
such near-sighted clarity is probably times of surplus leisure and personal
high for it seems to lead to a loss of idleness. These patients would be
perspective, a “centration,” in Piaget’s better served if they were buffered by
terms, upon his feelings in the im- the refuge of greater generality, less
mediate present. (Piaget, in press) involvement, and an increased ability
Note Freud’s insight into the height- to “let it go at that.”
ened awareness of people in misery as We do not suggest that there is
he describes the puzzling astuteness of only one kind of suicide, either from
melancholiac patients: . . . he has a the standpoint of dynamic motivation

keener eye for the truth . . . When in or descriptively. Shneidman and Far-
his heightened self-criticism he de- berow (1957) have detailed varieties
scribes himself as petty, egoistic, dis- of suicides, and a study by Dublin
honest, lacking in independence, one and Bunzel (1933) shows the multi-
whose sole aim has been to hide the ple roots of self-destruction. Presum-
weakness of his own nature, it may ably, suicides can occur as “pure”
be, so far as we know, that he has ritual acts or as a sensible alternative
come pretty near to understanding to intractable pain. Or, perhaps the
himself; we can only wonder why a experience of exquisite sensitivity re-
man has to be ill before he can be flected in the color-shading response
accessible to a truth of this kind.“ may be present in admixture even in
(Freud, 1917) these situations, for there remains the
question of why some people do not
What Freud knew, but merely sug- commit suicide under apparently the
gested in his faintly ironic, rhetorical same conditions obtaining when
question, is that this terrible clarity others do. T h e state of mind we have
of detail issues from a distortion of suggested here to explain our results
one’s perspective of himself in rela- is complementary to motivational
STEPHENA. APPELBAUM
AND PHILIP
S. HOLZMAN 161
theories of suicide (viz., Menninger, dicator of suicidal tendencies signifi-
1938). This state of mind provides cantly improved predictions of sui-
what might be a structural condition, cide or attempted suicide over those
which, regardless of the precipitating made from the base rate in that popu-
motivation or circumstances bringing lation. T h e color-shading determin-
about the distraught condition, leads ant occurred significantly more often
the person to view existence in such a among the suicide and attempt
way that suicide appears his only groups than among five control
available course of action. groups. We offered speculations about
We offer these speculations as psychological processes that may be
plausible maps to direct our search involved in the color-shading deter-
for the processes underlying the color- minant, and about some consequences
Downloaded by [Central Michigan University] at 09:59 10 October 2014

shading determinant and its relation- of the assumed processes in suicidal


ship to suicide. For further informa- and non-suicidal populations.
tion and greater degree of certainty,
it is necessary first to seek replication REFERENCES
of these findings from other psychia- Dublin, Louis I. & Bunzel, Bessie. T o be or
tric hospitals and then to find per- not to be. New York: Harrison Smith and
sonality attributes that are associated Robert Haas, 1933.
with the appearance of this determin- Freud, Sigmund. Mourning and melancholia
(1917) in Standard edition of the com-
ant. In the meantime, in populations plete psychological works of Sigmund
of patients such as the one presented Freud. Vol. 14. London: Hogarth Press,
here, it seems that the presence of this 1957. Pp. 237-260.
sign may take its place in the total Menninger, Karl. Man against himself. New
York: Harcourt, Brace and Company, 1938.
clinical examination as a significant
Piaget, Jean. Will and action. Bull. Mennin-
bit of evidence pointing toward sui- ger Clin.2 6 138-145, 1962.
cidal tendencies. Rapaport, David, et. al. Diagnostic psycholo-
gical testing. Vol. 2. Chicago: Yearbook
SUMMARY Publisher, 1946.
The use of shading in colored areas Shneidman, Edwin S. & Farberow, Norman L.
of the Rorschach Test Plates by hos- Clues to suicide. New York: Blakiston, 1957.
pitalized psychiatric patients as an in- Received October 25, 1961

You might also like