This article was downloaded by: [New York University]

On: 05 May 2015, At: 20:00

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 Personality Assessment

Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/hjpa20

A Psychoanalytic View of the Rorschach

Comprehensive System 'Special Scores'
J. Reid Meloy & Jacqueline Singer
Published online: 10 Jun 2010.

To cite this article: J. Reid Meloy & Jacqueline Singer (1991) A Psychoanalytic View of the Rorschach
Comprehensive System 'Special Scores', Journal of Personality Assessment, 56:2, 202-217, DOI: 10.1207/
s15327752jpa5602_2

To link to this article: http://dx.doi.org/10.1207/s15327752jpa5602_2

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
 JOURNAL OF PERSONALITY ASSESSMENT, 1991, 56(2), 202-217
Copyright o 1991, Lawrence Erlbaum Associates, Inc.

A Psychoanalytic View of the

Rorschach Comprehensive System
"Special Scores"

J. Reid Meloy
University of California, San Diego
Jacqueline Singer
California School of Professional Psychology
Downloaded by [New York University] at 20:00 05 May 2015

We analyzed the contemporary empirical and theoretical literature concerning the

two predominant approaches for scoring formal thought disorder on the
Rorschach, the Comprehensive System special scores, and the methodology of
Rapaport, Gill, and Schafer (1946/1968). The psychoanalytic research related to
selected special scores is reviewed, and some linkages to psychoanalytic develop-
mental theory and psychopathology are made. Recommendations are presented to
bridge the gap between these two important avenues of Rorschach research, with
an emphasis on empirical rigor and intrapsychic contextual meaning.

Rapaport agreed with the criticisms that his original work o n diagnostic
psychological testing had "shortcomings as a piece of controlled quantitative
researchn (Rapaport et al., 1946/1968, p. 1). Although these shortcomings have
been addressed by Exner's (1986a) empirically based approach to the Rorschach,
the theoretical rationale for formal thought disorder has not been set forth in his
Comprehensive System. Weiner (1986) contended that a n empirical approach
to the Rorschach should be enhanced by a conceptual approach, which
questions "why" a certain response occurs. Through a n examination of the
Comprehensive System (Exner, 1986a) special scores, a n empirically reliable
measure of formal thought disorder, we explored the notion that the validation
of these test findings is found in their linkage to external behavior, empirical
data, and a developmental theory of personality and psychopathology.
We are committed, nonetheless, t o both the empirical rigor of Exner (1986a)
and his colleagues and to the search for intrapsychic contextual meaning of the
psychoanalytic Rorschach clinicians and researchers. It is our intent to review
 PSYCHOANALYSIS OF SPECIAL SCORES 203

the contemporary areas of convergence and divergence of these two traditions as

applied to the understanding of formal thought disorder and the Rorschach. We
hope that through such an empirical and theoretical ancalysiswe can bridge the
unnecessary gap between these approaches and make useful recommendations
for future research.
The psychoanalytic approach to the Rorschach assumes that data collected,
including indices of formal thought disorder, are both motivated and meaning-
ful. Freud (1900/1953) asserted that form may represent concealed subject
matter in dreams, and we hypothesize that formal thought disorder encompasses
more than "problems of disordered thinking and inaccurate perception" (Exner,
1986a, p. 414).
To date, psychoanalytic Rorschach researchers have empirically investigated
the lielationship between formal thought disorder and boundary disturbance
(Blatt 6rRitzler, 1974;Blatt &Wild, 1976; Lerner, Sugarman, & Barbour, 1985;
Downloaded by [New York University] at 20:00 05 May 2015

Quinlan, Harrow, Tucker, & Carlson, 1972; Quinlan & Harrow, 1974; Wilson,
1985) and have inferred, but not empirically measured, a relationship to
defensive operations (Lerner, Sugarman, & Gaughran, 1981) and object rela-
tions (Sugarman, 1986). Athey (1974, 1986) probed the individual differences
among patients regarding their thinking and object representations at different
levels of regression or fixation and was critical of the nomothetic work of others
for ignoring idiographic subtleties. These psychoanalytic researchers primarily
used the Rapaport et al. (1946/1968) system of administration, scoring, and
interpretation, making direct comparison with the Comprehensive System
(Exner, 1986a) difficult, but not impossible.
The Comprehensive System was developed as an empirically defensible
system to be used by all Rorschach clinicians and researchers (Exner, 1986a).
Special scores were categorized to quantify structural aspects of the Rorschach
response that previously had been qualitatively interpreted. Exner, Weiner, and
Schuyler (1976) originally proposed five special scores for the Comprehensive
System, derived from the work of Rapaport et al. (1946/1968), Schafer (1954),
and Weiner (1966). There are currently 12 special scores in the Comprehensive
System (Exner, 1986a). We chose 7 for this analysis, because they appear to have
the closest link to the Rorschach psychoanalytic research concerning formal
thought disorder. The Comprehensive System divides these unusual verbaliza-
tions into three groups: deviant verbalizations, inappropriate combinations,
and inappropriate logic. The specific categories are the deviant verbalization
(DV), deviant response (DR), incongruous combination (INCOM), fabulized
combination (FABCOM), contamination (CONTAM), and inappropriate logic
(ALOG). We added the seventh, confabulation (CONFAB), score to our
analysis, because a similar category (DW) occurs in the Rapaport et al.
(1946/1968) scoring system.
The Comprehensive System (Exner, 1986a) views these unusual verbaliza-
tions as a form of "cognitive slippage" and limits the use of these scores to
 204 MELOY AND SINGER

"identify events in which some difficulty occurred in cognitive processing"

(Exner, 1986a, p. 375). Psychoanalytic researchers have asserted, on the other
hand, that primary process thinking provides insight into levels of thought
organization (Blatt & Wild, 1976): "Various types of thinking disturbances can
be distinguished by the degree to which they require a separation of self from
nonself, of action from object, and of object from its mental representation" (p.
55). The resurrection of the concept of levels of personality organization
(Kernberg, 1984) and empirical reviews confirming the ubiquity of formal
thought disorder in certain psychopathological groups, such as borderline
personality disorder (Gartner, Hurt, & Gartner, 1989), lend further support to
this search for intraps~chiccontextual meaning in formal thought disorder.
Table 1 compares the 7 special scores from the Comprehensive System
(Exner, 1986a) and indices of formal thought disorder from Rapaport et al.
(1946/1968). The two systems for scoring formal thought disorder are not
Downloaded by [New York University] at 20:00 05 May 2015

entirely different. Rather, it appears that the meaning, or lack thereof, implied
by each system differs.

FABCOM

The Comprehensive System FABCOM response is essentially the same as the

psychoanalytic fabulized combination response and is defined as "an implausible
relationship . . . between two or more objects identified in the blot" (Exner,
1986a, p. 163). The psychoanalytic interpretation, however, has included one
object containing incongruous parts and two incongruous objects, identifying
the former as more pathological (Lerner et al., 1985).
Psychoanalytic Rorschach researchers have viewed the FABCOM as a less
severe form of thought disorder when compared to CONFABS and CONTAMs
(Johnston & Holzman, 1979; Rapaport et al., 1946/1968). Blatt and Ritzler
(1974) understood FABCOMs as representing a laxness or flexibility of bound-

TABLE 1
A Comparison of the Special Scores From the Comprehensive System and Rapaport,
Gill, and Schafer's (RGS) Formal Thought Disorder Categories
Desaiption of Response Comprehensive System RGS

Two percepts combined in an impossible way FABCOM FABCOM

One percept with impossible details INCOM FABCOM
Extensive arbitrary elaboration of a response DR CONFAB
Fusion of two percepts into one CONTAM CONTAM
Illogical cause-effect relation ALOG ALOG
Peculiar, non sequiturs DV Pecuhar, Queer
Inappropriate perceptual generalization CONFAB DW
 aries. A boundary is a construct that implies both structure and content; it refers
to a capacity to create cognitive and affective distinctions along a bipolar
coordinate where previously no difference was possible (Waelder, 1962; Wils~on,
1985). Blatt and Ritzler (1974) and, later, Lerner et al. (1985)found a theoretiical
place for boundary disturbance in the developmental object representational
work of Kernberg (1975,1976).The concept of boundary also appears to have an
affinity with the psychoanalytic literature concerning thle experience of the self
(Kohut, 1971). Boundary formation in object relations theory is critically
important to the genesis of the preoedipal self through the developmental
necessity of splitting defenses to demarcate internal representations and their
respective primordial pleasure-pain feeling states.
The various FABCOMs share two characteristics: ~erceptualclarity and
temporal or spatial discontinuity. In the FABCOM response, the spatial rjela-
tionship is taken to be the real-world relationship. FABCOMs irnplicate a
Downloaded by [New York University] at 20:00 05 May 2015

problem with object concepts, involving faulty juxtaposition, without a loss or

permeability of boundaries between those concepts; object percepts, however,
are largely unimpaired (Meloy, 1985, 1986).
The empirical literature concerning FABCOM responses has ~ieldetla
number of findings. FABCOMs are more common in borderline personality
disordered records than in normal records (Bloomgarden, 1980; Buttenheim,
Lohr, & Kerber, 1985; Singer & Larson, 1981). Thley occur with greater
frequency in borderline personality disordered records when compared to
published norms (Carr, Goldstein, Hunt, & Kernberg, 1979; Exner, 1986b;
Patrick & Wolfe, 1983). FABCOMs are more frequent in borderline personality
disordered than neurotic records found in the work of Singer and Larson (1981),
but Wilson (1985) found no significant mean difference between borderline
personality and neurotic records. Wilson's finding was replicated by Lerner et al.
(1985) who found no significant mean difference in FABCOMs across four
groups: outpatient neurotics, outpatient borderlines, inpatient borderlines, and
inpatient schizophrenics.
Singer and Larson (1981) and Exner (198613)found differences in the number
of FABCOMs in borderline personality disordered and schizophrenic subjects.
Unfortunately, the direction of proportionality in each study was opposite:
Exner (1986b) found more FABCOMs in the schizophrenic subjects, Singer and
Larson (1981) found more in the borderline personality disordered subjects.
Wilson (1985) validated Singer and Larson's (1981) findings when he found a
greater arithmetic and covariate mean of FABCOMs in borderline inpatients
when compared to schizophrenic inpatients. His definition of FABCOM-
benign equated with the combining of Exner's (1986a) FABCOM and INCOM
in this particular study. Wilson (1985) scored FABCQM-benign when two
contiguous parts of the blot of good form quality were combined in an
improbable fashion or when "two independent percepts are arbitrarily juxta-
posed" (p. 349).
 206 MELOY AND SINGER

These contradictory findings are especially interesting because Exner's (1986a)

definition of FABCOM is more restrictive than the analytic definition; the
Exner data (1986a, 1986b)indicate that schizophrenicpopulations have more of
both FABCOMs and INCOMs than borderline personality disordered subjects.
It may be that the schizophrenic's impairment in boundaries and disturbance in
thinking is not specific to any particular stimulus pull, whereas the borderline
disorder is manifested by disturbances in thinking around specific conflictual
areas.
Table 2 indicates the proportional frequency of FABCOMs and INCOMs in
the Exner (1986a, 1986b, 1989) clinical and normative samples. Several trends
are apparent: FABCOMs increase with severity of psychopathology when
subjects are selected according to descriptive diagnostic criteria, such as the
Diagnostic and Statistical Manual of Mental Disorders (3rd ed. [DSM-Ill]; Amer-
ican Psychiatric Association, 1980); and FABCOMs are twice as likely to appear
Downloaded by [New York University] at 20:00 05 May 2015

in the records of schizophrenics than in the records of borderline personality

disordered patients when defined more restrictively as two or more incongruous
objects in relationship.

INCOM

The INCOM is defined as the condensation of blot details or images "that are
inappropriately merged into a single object" by the Comprehensive System
(Exner, 1986a, p. 163). The INCOM response has been parcelled out of the
analytic FABCOM literature by Exner, so it is hidden within the empirical
analytic research concerning the FABCOM. Direct analysis of empirical psy-
choanalytic findings concerning the INCOM response is generally impossible
because there are no data, unless a category that can be equated with the
INCOM (e.g., FABCOM-serious) has been established for a particular study
(see, e.g., Lerner et al., 1985).
Psychoanalytic Rorschach researchers, however, have theorized that the
INCOM is a more pathological condensation of the FABCOM due to the

TABLE 2
INCOMs and FABCOMs in Exner Normative and Clinical Samples: Frequencies,
Means, and Standard Deviations
Study Subjects n INCOM M SD FABCOM M SD
Exner (1989) Normals 700 46% .52 .65 16% .17 .41
Exner (1986a) Normals 600 43% .54 .79 12% .18 .56
Exner (198613) Borderline 84 56% 1.11 1.34 36% .62 1.06
Exner (1986b) Schizophrenic 80 77% 1.76 1.66 73% 1.58 1.60
Exner (1986a) Schizophrenic 320 66% 1.51 1.71 78% 1.59 1.63
 PSYCHOANALYSIS OF SPECIAL SCORES 3-07

merger into one object instead of two (Blatt & Ritzler, 1974; Lerner et al., 1985;
Sugarman, 1986). This is not supported by Exner's (1986a) research despite the
analytic assumption that spatial or temporal incongruity between two objects is
less regressed than one object. Although INCOMs and FABCOMs necessitate
temporal or spatial incongruity, FABCOMs usually include an additional
movement incongruity: for example, a "four-legged chicken" (Exner, 1986a, p.
163) is an INCOM with spatial incongruity; "two chickens holding basketballs"
(Exner, 1986a, p. 163) is a FABCOM with both spatial and movement incon-
gruity. We think that inherent in the FABCOM is a greater violation of the
constraints of real-world objects and, consequently, secondary process thinking,
due to this additional movement incongruity. Such a marked departure impli-
cates more severe disturbance in thinking, which is supported by the empirical
literature.
Lerner et al. (1985) indirectly tested this hypothesis when their
Downloaded by [New York University] at 20:00 05 May 2015

FABCOM-serious, which they viewed as a self-other boundary problem, was

explicitly defined using the same criteria as Exner's (1986a)INCOM. They found
no significant mean differences between INCOMs across their four diagnostic
groups, but they did occur most frequently in their borderline inpatient group (n
= 21). They noted that the INCOM, which they called FABCOM-serious,
may not be as related to CONTAMs as they thought.
Table 2 illustrates the high frequency of INCOMs in both normal and
psychopathological records, with INCOMs being more likely to appear in more
psychopathological samples. The trend is in the same direction as was noted for
the FABCOMs and may represent a positive correlation.
Exner's introduction (1986~) of Level 1 and Level 2 scoring criteria for certain
measures of formal thought disorder will certainly lead to a fuller understanding
of the INCOM and its relationship to FABCOM. Exner (1987) developed these
distinctions to more effectively discriminate mild from severe forms of thought
disorder. Level 1 is scored when there is mild or moderate disturbance; Level 2
is assigned when a severe disturbance in thinking is evident. Discriminated on
the basis of clearly less bizarre features, the Level 1 INCOMs and FABCOMs
occur at a frequency of 46% and 16%,respectively, in the nonpatient group (n =
700). Level 2 INCOMs and FABCOMs, however, marked by their identifiable
bizarreness, occur at a frequency of only .4% and 1.7'%, respectively, in the
nonpatient group (Exner, 1989). This is a notable finding when empirically
understood to mean that FABCOMs are four times more likely to appear than
INCOMs at Level 2 in normals, although both are a rare occurrence, and the
FABCOM is considered indicative of more psychopathology in the Compre-
hensive System. Level 1 and Level 2 distinctions in normative and clinical
samples are presented in Table 3. This Level 1-Level 2 dichotomy may eventu-
ally confirm the original psychoanalytic hypothesis that spatial or temporal
incongruity is less regressed in two objects than in one object in certain clinical
populations.
 208 MELOY AND SINGER

TABLE 3
Level 1and Level 2 INCOMs and FABCOMs in Exner (1990) Normative and Clinical
Samples: Frequencies, Means, and Standard Deviationsa
Subjects
Normalsb Character Disordered' lnpatient Schizophrenicsd
INCOM 1
Frequency 46% 51% 71%
M .52 1.06 1.53
SD .65 1.36 1.44
INCOM 2
Frequency .4% 17% 49%
M .OO .36 1.17
SD .07 .68 1.68
FABCOM 1
Frequency 16% 39% 44%
Downloaded by [New York University] at 20:00 05 May 2015

M .17 .63 .72

SD .4 1 .57 1.07
FABCOM 2
Frequency 1.7% 12% 63%
M .02 .36 1.83
SD .13 .79 2.04
"Standard deviations are probably unreliable and misleading. They should not be used to estimate
expected ranges, nor should they be included in most parametric analyses (see Exner, 1990). bn =
700. 'n = 180. dn = 320.

The INCOM response appears to be neither sensitive nor specific to psycho-

pathology. It may be that Exner's (1986~)Level 2 scoring criteria will further
delineate these pathological features. This research has yet to be done.

The Comprehensive System DR is virtually synonymous with the analytic

literature's CONFAB response. The DR is slightly more expansive, with the
inclusion of both circumstantial responses and inappropriate phrases (Exner,
1986a). The latter are usually denoted in the analytic research as "confabulation
tendency" (Wilson, 1985).
Psychoanalytic theorists have historically understood CONFAB as repre-
senting a more severe form of thought disorder than FABCOMs, but less severe
than CONTAMs (Johnston & Holzman, 1979; Rapaport et al., 1946/1968;
Watkins & Stauffacher, 1952). Lerner et al. (1985) noted that CONFAB was a
Rorschach manifestation of displacement, a genotypic form of primary process .
(Gill, 1967; Meloy, 1986).It is a disturbance in the association of ideas that may
 PSYCHOANALYSIS OF SPECIAL SCORES 209

be expressed through tangential, circumstantial, rambling, or flighty verbaliza-

tions.
Blatt and Ritzler (1974) theorized that CONFAB represented a loss of
inner-outer boundary from a developmental object relations perspectnve.
Lerner et al. (1985) agreed and theorized that CONFABs suggested a develop-
mental defect during the rapprochement subphase of the separation-
individuation process (Mahler, Pine, & Bergrnan, 1975), when there was a
predominance of illusion and the need for transitional objects (Winnicott,
1953).
We think, however, that the theoretical proposition that CONFABs repre-
sent a loss of inner-outer boundary is somewhat simplistic and misleading
because it suggests psychosis or a loss of reality testing. We offer the alternative
hypothesis that the inner-outer boundary is maintained, but the patient's
ability to identify interoceptive and exteroceptive stimuli correctly is confused:
Downloaded by [New York University] at 20:00 05 May 2015

which is within and which is without? Bion's (1977) analogy of the container and
the contained sheds light on our distinction between the maintenance of
boundaries (the container), yet the confusion concerns the origin of psycholog-
ical content (the contained). This alternative manner of understanding
CONFABs is allied with the projective-introjective defensive cycling of psycho-
logical content found in borderline personality organization (Kernberg, 1975;
Meloy, 1988):The patient can distinguish between internal and external stimuli,
but is not sure which is which.
Wilson (1985) found that CONFAB-tendency means were significantly
greater in inpatient borderlines than inpatient neurotics or psychotics. His
CONFAB-tendency definition was virtually identical to Exner's (1986a) IDR
definition. Lerner et al. (1985) also found significantly greater mean CONFABS
in inpatient borderline personality disordered patients when compared to
inpatient schizophrenics and outpatient borderlines and neurotics.
Exner (1986b) found that CONFABs did not significantly differentiate be-
tween borderline personality disordered subjects and schizotypal subjects. The
proportional frequency of DRs in the normative sample (n = 700; Exner, 1989)
was 15% at Level 1; in the borderline personality disordered sample (n = 84;
Exner, 1986b),the frequency was 23%; in the schizophrenicnormative sample (n
= 320; Exner, 1986a), the frequency was 46%; and in another schizophrenic
sample (n = 80; Exner, 1986b),the frequency was 60%. Again, the trend when
comparing frequency of occurrence in a large sample is that a CONFAB is more
likely to appear in the records of more severely disturbed patients. Yet the
comparison of means of CONFABs among various psychiatric groups appears
to yield a consistent finding of more CONFABs in Rorschach protocols from
borderline personality disordered (DSM-111; American Psychiatric Association,
1980) patients, a descriptive diagnostic group of patients who can be reasonably
assumed to be organized at a borderline personality level (Kernberg, 1984).
 210 MELOY AND SINGER

The CONTAM response is defined almost identically by the Comprehensive

System (Exner, 1986a) and psychoanalytic research. Both involve "two or more
impressions that have been fused into a single response in a manner that clearly
violates reality" (Exner, 1986a, p. 163). CONTAM on the Rorschach is phe-
nomenologically rooted in condensation, a genotypic mechanism of primary
process (Gill, 1967), which Freud (1950/1954) originally defined as "latent
elements which have something in common being combined and fused into a
single unity in the manifest dream" (p. 171).
Schwartz and Lazar (1984) described three categories of CONTAMs: simul-
taneous or overlapping perceptual images interrelated by the perception of
action, fused or combined images without action, and perceptual change
through influence of an extraneous perception. They reported that 42% of their
Downloaded by [New York University] at 20:00 05 May 2015

simultaneous CONTAMs contained an incorporative or merger fantasy, and

50% of all their CONTAMs contained human content (Lazar & Schwartz,
1982). Rather than the contiguity in space or time seen in FABCOMs,
CONTAMs involve simultaneity in either space or time. Schwartz and Lazar
(1984) developed two dynamic hypotheses from their study of a large sample of
CONTAMs: (a) a CONTAM indicates a change in state which reduces
reflective self-awareness, and (b) it signals both a wish for merger and a
counterwish to maintain contact with reality.
Blatt and Ritzler (1974) theorized that CONTAMs indicated a loss of
boundaries between self and others, an object relations developmental defect at
the symbiotic stage (Mahler et al., 1975). CONTAMs appear to involve a severe,
if transient, impairment in both concept and percept formation (Meloy, 1985).
Conceptual representations are condensed and perceptual distinctiveness is lost
(Card IV: "It's seaweed, it's a monster, it's a seaweed monstern).
CONTAWS,in contrast to CONFABS,represent a loss of boundary wherein
interoceptive and exteroceptive stimuli merge and are no longer demarcated.
Rather than contrasting loss of internal-external boundaries (the CONFAB
hypothesis) with loss of self-other boundaries (the CONTAM hypothesis) as
Blatt and Ritzler (1974) did, we argue that CONTAMs suggest a temporary loss
of boundary at a variety pf levels: the sensory barrier between internal and
external stimuli; internal representations of objects, both self and others; and
internal representations of percepts, both self and others (Meloy, 1985).
CONTAMs, although rarely occurring in any psychiatric sample, signal the
momentary, but catastrophic, dissolution of all boundary (Grotstein, 1986;
Meloy, 1984). Psychological content no longer oscillates between inner and
outer boundaries, thus leading to CONFAB at a borderline level of personality
organization; but the dissolution of boundary, or containment (Bion, 1977),
leads to CONTAM at a psychotic level of personality organization (Kernberg,
1984).
 PSYCHOANALYSIS OF SPECIAL SCORES 211

CONTAMs appear to be pathognomonic of schizophrenia (Rapaport et al.,

1946/1968; Johnston & Holzman, 1979),but they rarely occur. They are specific
to schizophrenia but are not sensitive (Edell, 1987). Wilson (1985) found that
CON TAM^ discriminated schizophrenic from both borderline personality
disorder and neurotic records. Lerner et al. (1985) reported that CONTAMs
had significantly higher means in inpatient schizophrenics than inpatient or
outpatient borderlines and outpatient neurotics. Rapaport et al. (1946/1968)
reported that schizophrenics gave approximately four CONTAMs per 100
responses.
The CONTAM response has yielded the most consistent empirical findings
in both the psychoanalytic and the Comprehensive System research. Table 4
illustrates the proportional frequencies of CONTAM responses in five different
studies using different samples of schizophrenics. The data repeatedly show that
a CONTAM will occur in only one out of six or seven schizophrenic Rorschach
Downloaded by [New York University] at 20:00 05 May 2015

protocols. CONTAMs appear to be quite specific, but not sensitive, when

diagnosing schizophrenia, with the expectation of few false positives. Rapaport
et al. (1946/1968) found that their borderline schizophrenic sample (n = 33)
produced CONTAMs in only 6% of the records. Their neurotic and control
groups combined (n = 149) produced CONTAMs in less than 1% of the
protocols. Exner (1986b) reported that a borderline personality disordered
sample produced CONTAMs in 4% of the records (n == 84), whereas the new
Exner (1989) normative sample (n = 700) produced no CONTAMs.
A CONTAM response should be viewed as a pathognomonic suggestion of
schizophrenia, but its lack of sensitivity to schizophrenia should be kept in
mind, given the likelihood of a false negative rate of 80%-85%.

ALOG

An ALOG is scored by the Comprehensive System "whenever the subject,

without prompting, uses strained reasoning to justify his or her answer" (Exner,
1986a, p. 164). It is virtually identical to the psychoanalytic literature references

TABLE 4
CONTAM Frequencies in Schizophrenicsin Five Studies
-
S ~ Y Subjects n Frequency
Rapaport et al. (1946/1968) Combined Schizophrenics 75 17%
Johnston and
Holzman (1979) Acute and chronic schizophrenics 69 13%
Exner (1986a) Inpatient schizophrenics 320 18%
Exner (198613) Inpatient schizophrenics 80 15%
Edell (1987) "Early" schizophrenics subchronic course 30 16.7%
 212 MELOY AND SINGER

to autistic logic in the Rorschach (Rapaport et al., 1946/1968), and finds its roots
in the paralogic or paleologic of the primary process (Meloy, 1986). Most
CONTAMs occurring on the Rorschach contain ALOG, but ALOG responses
usually appear in the absence of a CONTAM response. ALOG usually violates
one of the four principles of Aristotelean logic (Meloy, 1986),including identity,
contradiction, excluded middle, or sufficient reason.
Autistic logic in the Rapaport et al. (1946/1968) data was present in only the
combined schizophrenicsample (n = 75) with a frequency of 20%.Johnston and
Holzman (1979) reported a frequency of 33% in their acute and chronic
schizophrenic sample (n = 69 combined). Exner (1986a, 198613, 1989) reported
ALOG frequencies of 4% in his normative adult sample (n = 700), 54% in his
normative schizophrenic sample (n = 320), 13% in his borderline personality
disorder sample (n = 84), and 26% in his schizotypal personality disorder sample
(n = 76). ALOG, or autistic logic, responses to the Rorschach appear to be
Downloaded by [New York University] at 20:00 05 May 2015

specific and are more sensitive than CONTAM responses to schizophrenic, or

so-called "schizophrenic-spectrum," disorders.

The Comprehensive System DVs include both neologisms and redundancies.

The comparable psychoanalytic research concerns only neologisms and is
therefore much more restrictive.
Rapaport et al. (1946/1968) reported that neologisms only occurred in their
schizophrenic sample (n = 75) with a 7% frequency. None occurred in their
borderline schizophrenic, depressive, neurotic, or control groups. Johnston and
Holzman (1979)found a frequency of 4% in their schizophrenic sample (n = 69).
Exner (1986b), using his definition of DR to include both neologisms and
redundancies, found that the frequency was significantly different between a
sample of borderline (40%) and a sample of schizotypal (76%) personality
disorders.
Exner (1989) nonpatient norms indicate that the frequency of Level 1 DV
responses was 53%, whereas the frequency of Level 2 DV responses was 1%. It
appears that the Level 2 DV would be quite similar to the neologism research in
the analytic literature due to its more delineated bizarreness and oddity. Level 1
DVs probably have no interpretative significance; Level 2 DVs, although rarely
occurring, may have certain dynamic or motivational meaning if they represent
condensation in the form of a neologism.

CONFAB

The Comprehensive System CONFAB response occurs when a "subject attends

only to a detail area of the blot, but generalizes a response from that detail to a
 PSYCHOANALYSIS OF SIPECIAL SCORES 213

.
larger area . . the overall response will be inappropriate for the total area
involved" (Exner, 1986a, p. 166). This is the Rorschach DW response in the
psychoanalytic literature (Rapaport et al., 1946/1968)1.Called transductive
thinking by Piaget (1954), it is the discovery of the reality significance of the
inkblot through the autistic logic of pars pro toto-the part is the whole (Meloy,
1986; Rapaport et al., 1946/1968).
The difficulty with the carefully scored CONFAB is that it rarely occurs. In
the Rapaport et al. (1946/1968) samples, the frequency in the combined
schizophrenic group (n = 75) was 27%; in the borderline schizophrenic group (n
= 33) 18%;and in the combined depressive, neurotic, and control groups (n =
145) 3%. Exner's samples (1986a, 198613, 1989), however, have yielded signifi-
cantly less: the schizophrenic normative group (n = 320) found a zero frequency
of CONFABs, as did the new adult normative group (n = 700).
The differencein frequency of CONFABs between the Rapaport samples and
Downloaded by [New York University] at 20:00 05 May 2015

the Exner samples appears to be in the scoring methodology. Although both

have defined CONFABs virtually alike, Rapaport et al. (1946/1968) cautioned
against too much inquiry, and Exner (1986a) encouraged a much more aggres-
sive inquiry to rule out "lazy or resistive articulationn (p. 166). The probable
reason for the increased frequency of CONFABs in the Rapaport sample was the
assumption of DW when it was not there, and integration failure had not truly
occurred.
CONFAB, due to its rare presence, is neither sensitive to nor specific of
psychopathology. It appears to have little use as a special score.

RECOMMENDATIONS

We offer the following recommendations to both psychoanalytic and Compre-

hensive System Rorschach researchers interested in formal thought disorder.

1. The psychoanalytic researchers should use the Comprehensive System

special scores reviewed in this article due to their high interjudge reliabilities
(93%-99% in two studies; Exner, 1986a) and their inclusiveness of most tradi-
tional psychoanalytic categories. This would make research by both Compre-
hensive System and psychoanalytic investigators more comparable and would
probably accelerate new findings in the field.
2. The Comprehensive System special scores need to be contextualized as to
both motivation and meaning. Such research could begin by studying the
clustering of special scores around certain content themes, such as the associa-
tion of special scores and color responses as indicative of affective disturbance or
the association between special scores and human responses in paranoid condi-
tions (Exner, 1986a). Other promising lines of research include the empirical
relationship between special scores and defensive operations (Cooper & Arnow,
 214 MELOY AND SINGER

1986; Lerner & Lerner, 1980), object representations (Blatt, Brenneis, Schimek,
& Glick, 1976), and interpersonal modes of relating (Kwawer, 1980).
3. The Comprehensive System should eliminate the CONFAB response, and
rename the DR CONFAB. This would historically link future Comprehensive
System research to previous psychoanalytic Rorschach research and provide
impetus to the already promising findings concerning confabulation and its
discrimination between neurotic, borderline, and psychotic groups.
4. The Comprehensive System should add the word because as a necessary
criteria for scoring ALOG and should define strained reasoning as private,
nonconsensual logic. We think the current definition does not sufficiently
explain the nature of an ALOG (Exner, 1986a).
5. The new Level 1 and Level 2 scores in the Comprehensive System appear
to be quite promising. Their retest reliabilities are high and range from .82 to .93
(Exner, 1989). Interjudge agreement was 80% in an initial 1986 study (Exner,
Downloaded by [New York University] at 20:00 05 May 2015

1987). This may increase both the sensitivity and specificity of DV, DR,
INCOM, and FABCOM.
6. Psychoanalytic researchers must be careful ta not equate borderline per-
sonality organization (Kernberg, 1984) with the descriptive diagnosis of border-
line personality disorder (American Psychiatric Association, 1980, 1987). None
of the psychoanalytic studies we reviewed drew this distinction, but rather
developed their theoretical inferences concerning borderline personality orga-
nization from empirical findings based on samples diagnosed as borderline
personality disordered. We appreciate the difficulties inherent in sample selec-
tion, but the Kernberg (1984) criteria for levels of personality organizaqion, if
valid, should be operationalized without the use of DSM-111 or DSM-111-R; or,
at least, other independent variables should be used in addition to descriptive
diagnostic criteria to define subject samples in a more valid psychoanalytic
context.
7. Both Comprehensive System and psychoanalytic Rorschach research
concerning formal thought disorder should report specific categories of formal
scores, including means, standard deviations, and frequencies of occurrence in
samples. This will allow for more rigorous comparison between studies for
validation purposes. Statistical analyses also must be carefully chosen for
appropriateness of parametric versus nonparametric tests and means versus
frequencies based on the psychometric nature and clinical meaning of the formal
thought disorder index being researched.

We hope that these recommendations will help build a bridge between the
empirical rigor of Exner's (1986a) Comprehensive System special scores and the
theoretical and conceptual depth of psychoanalytic psychology's formal thought
disorder research. The joining of definitions and the alliance of methodologies
should greatly enhance the reliability and validity of future research efforts in
this important area of research. Perhaps, as Rapaport (1951) wrote, "after the
 PSYCHOANALYSIS OF SPECIAL SCORES 215

many attempts to explain thought-disorders by a simple theory have proved

partial or irrelevant, this complex phenomenon will finally exact a complex
explanation" (p. 730).

REFERENCES

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd
ed.). Washington, DC: Author.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed.
rev.). Washington, DC: Author.
Athey, G. (1974). Schizophrenic thought organization, object relations and the Rorschach test.
Bulletin of the Menninger Clinic, 38, 406-429.
Athey, G. (1986). Rorschach thought organization and transference enactment in the
Downloaded by [New York University] at 20:00 05 May 2015

patient-examiner relationship. In M. Kissen (Ed.),Assessing object relations phenamena (pp. 19-50).

New York: International Universities Press.
Bion, W. (1977). Seven sewants. New York: Aronson.
Blatt, S., Brenneis, B., Schimek, J., & Glick, M. (1976). A developmental analysis of the concept of the
object on rhe Rorschach. Unpublished scoring manual.
Blatt, S., & Ritzler, B. (1974). Thought disorder and boundary disturbance in psychosis. Joumcrl of
Consulting and Clinical Psychology, 42, 370-381.
Blatt, S., & Wild, C. (1976). Schirophrenia: A developmental analysis. New York: Academic.
Bloomgarden, J. W. (1980). A n exploration of splitting and integration on the Rorschach test in borderline
and schizophrenic patients. Unpublished doctoral dissertation, New York University.
Buttenheim, M., Lohr, N., & Kerber, K. (1985, August). The impact of depression on ego functioning.
Paper presented at the meeting of the American Psychological Association, Los Angeles.
Carr, A., Goldstein, E., Hunt, H., & Kernberg, 0. (1979). Psychological tests and borderline
patients. Journal of Personaliy Assessment, 43, 582-90.
Cooper, S., & Arnow, D. (1986). An object relations view of the borderline defenses: A Rorschach
analysis. In M. Kissen (Ed.), Assessing object relations phenomena (pp. 143-171). New York:
International Universities Press.
Edell, W. (1987). Role of structure and disordered thinking in borderline and schizophrenic
disorders. Journal of Personality Assessment, 51, 23-41.
Exner, J. (1986a).The Rorschach: A comprehensivesystem: Foundations (Vo~l.1, rev.). New York: Wiley.
Exner, J. (198613). Some Rorschach data comparing schizophrenics with borderline and schizotypal
personality disorders. Journal of Personaliry Assessment, 50, 455-471.
Exner, J. (1986~).Rorschach Workshops Alumni Newsletter. (Available from Rorschach Workshops,
P.O. Box 9010, Asheville, NC 28815-9010)
Exner, J. (1987). Rorschach Workshops Alumni Newsletter. (Available from Rorschach Workshops,
P.O. Box 9010, Asheville, NC 28815-9010)
Exner, J. (1989). Rorschach Workshops Alumni Newsletter. (Available from Rorschach Workshops,
P.O. Box 9010, Asheville, NC 28815-9010)
Exner, 9. (1990). A Rorschach workbook for the comprehensive system. Asheville, NC: Rorschach
Workshops.
Exner, J., Weiner, I., & Schuyler, W. (1976). A Rorschach workbook for the comprehensive system.
Bayville, NY: Rorschach Workshops.
Freud, S. (1953). The interpretation of dreams. In J. Strachey (Ed. and Trans.), The standard edition
of the complete psychological works of Sigmund Freud (Vol. 4-5). London: Hogarth. (Original work
published 1900)
 216 MELOY AND SINGER

Freud, S. (1954). Project for a scientific psychology. In E. Kris (Ed.), The origins of psychoanalysis (pp.
140-174). New York: Basic Books. (Original work published 1950)
Gartner, J., Hurt, S., & Gartner, A. (1989). Psychological test signs of borderline personality
disorder: A review of the empirical Ilterature. Journal of Perscmalzty Assessment, 53, 423-441.
G111,M. (1967).The primary process. In R. Holt (Ed.), Psychoanalytzc essays in honor of David Rapaport
(pp. 260-298). New York: International Universities Press.
Grotstein, J. (1986). Schizophrenicpersonal~t~ disorder: ". . . and if I should die before I wake." In D.
Feinsilver (Ed.), Towards a comprehensivemodel for schizophrenicdisorders (pp. 29-71). Hillsdale, NJ:
The Analytic Press.
Johnston, M., & Holzman, P. (1979). Assesszng schzzophrenic thinkmng. San Francisco: Jossey-Bass.
Kernberg, 0. (1975). Borderline conditions and pathological narcissism. New York: Aronson.
Kernberg, 0. (1976). Object relations theory and clinzcal psychoanalysis.New York: Aronson.
Kernberg, 0 . (1984). Severe personallty disorders. New Haven: Yale University Press.
Kohut, H. (1971). Analysis of the self. New York: International Universities Press.
Kwawer, J. (1980). Primitive interpersonal modes, borderline phenomena, and Rorschach content.
In J. Kwawer, H. Lerner, P. Lerner, &A. Sugarman (Eds.), Borderlinephenomenaand the Rorschach
test (pp. 89-106). New York: International Universities Press.
Downloaded by [New York University] at 20:00 05 May 2015

Lazar, Z., & Schwartz, F. (1982). The contaminated Rorschach response: Formal features. Journal of
Clinical Psychology, 38, 415-419.
Lerner, P., & Lerner, H. (1980). Rorschach assessment of primitive defenses in borderline person-
ality structure. In J. Kwawer, H. Lerner, P. Lerner, & A. Sugarman (Eds.), Borderline phenomena
and the Rorschach test (pp. 257-274). New York: International Universities Press.
Lerner, H., Sugarman, A., & Barbour, C. (1985). Patterns of ego boundary disturbance in neurotic,
borderline and schizophrenic patients. Psychoanalytic Psychology, 2, 47-66.
Lerner, H., Sugarman, A., & Gaughran, J. (1981). Borderline and schizophrenic patients: A
comparative study of defensive structure. Journal of Nervous and Mental Disease, 169, 705-7 11.
Mahler, M., Pine, F., & Bergman, A. (1975). The psychologicalbirth of the human znfant. New York:
Basic Books.
Meloy, J. R. (1984). Thought organization and primary process in the parents of schizophrenics.
Bntish Journal of Medical Psychology, 57, 279-281.
Meloy, J. R. (1985). Concept and percept formation in object relations theory. Psychoanalytic
Psychology, 2, 35-45.
Meloy, J. R. (1986). O n the relationship between primary process and thought disorder. Journal of the
American Academy of Psychoanalysu,14, 47-56.
Meloy, J, R. (1988). The psychopathic mmd: origins, dynamics, and treatment. Northvale, NJ: Aronson.
Patrick, J., & Wolfe, B. (1983). Rorschach presentation of borderline personallty disorder: Primary
process manifestations. J o u d of Clinical Psychology, 39, 442-447.
Piaget, J. (1954). The consrrzution of realzty in the child (M. Cook, Trans.). New York: Bas~cBooks.
Qulnlan, D., &Harrow, M. (1974). Boundary disturbance in schizophrenia. Joumal of Abnormal
Psychology, 83, 533-541.
Quinlan, D., Harrow, M., Tucker, G., & Carlson, K. (1972).Varieties of "disordered" thinking on
the Rorschach. Journal of Abnormal Psychology, 79, 47-53.
Rapaport, D. (Ed.). (1951). Organization and pathology of thought. New York: Columbia University
Press.
Rapaport, D., Gill, M., & Schafer, R. (1968). Diagnostic psychological testing. New York: Interna-
tional Universities Press. (Original work publ~shed1946)
Schafer, R. (1954). Psychoanalyticinterpretation in Rorschach testing. New York: Orune & Stratton.
Schwartz, F., & Lazar, Z. (1984). Contaminated thinking: A specimen of the primary process.
Psychoanalytic Psychology, 1, 319-334.
Singer, M., & Larson, D. (1981). Borderline personality and the Rorschach test. Archives of General
Psychiatry, 38, 693-698.
 PSYCHOANALYSIS OF SPECIAL SCORES 217

Sugarman, A. (1986). Self-experienceand reality testing: Synthesis of an object relations and an ego
psychological model on the Rorschach. In M. Kissen (Ed.), Assessing object relations phenmama
(pp. 51-76). New York: International Universities Press.
Waelder, R. (1962). Psychoanalysis, scientific method, and philosophy. Journal of the American
PsychoanalyticAssociation, 10, 617-637.
Watkins, J., & Stauffacher,J. (1952). An index of pathological thinking in the Rorschach. Journal of
Projective Techniques, 16, 276-286.
Weiner, I. (1966). Psychodiagnosis in schizophrenia. New York: Wiley.
Weiner, I. (1986). Conceptual and empirical perspectives on the Rorschach assessment of psycho-
pathology. Journal of Personality Assessment, 50, 472-479.
Wilson, A. (1985). Boundary disturbance in borderline and psychotic states. Journal of Personulity
Assessment, 49, 346-355.
Winnicott, D. W. (1953). Transitional objects and transitional phenomena. international Journrrl of
Psychoanalysis, 34, 89-97.

J. Reid Meloy
Downloaded by [New York University] at 20:00 05 May 2015

7966 Caminito del Cid

La Jolla, C A 92037

Received December 28, 1989

Revised April 23, 1990