Chapter 25

Accuracy of Defense Interpretation

in Three Character Types

J. Christopher Perry, Jonathan Petraglia, Trevor R. Olson, Michelle D. Presniak,

and Jesse A. Metzger

Keywords Defense mechanisms • Interpretation of defenses • Personality disorders • Psychother-

apy outcome • Psychotherapy process

Defense mechanisms are one of the most durable constructs in psychoanalysis and dynamic psychia-
try/psychology, spanning theoretical, clinical, and research approaches. While the construct origi-
nated with Freud’s 1894 [1] publication, The Neuro-Psychoses of Defence, the first seven decades of
psychoanalytic writing largely advanced the theoretical understanding and clinical approaches to
defense mechanisms, while the research did not begin in earnest until about the last 40 years, accel-
erating somewhat more recently. Much of this research has understandably concentrated first on
issues of how to assess defenses [2, 3], second, on the relationship of defenses to clinical disorders,
such as depression [4] and personality disorders [5, 6], and, third, on change in defenses over time
and long-term development [7]. In recent years, this latter avenue has expanded to include treatment
outcome studies indicating that defenses and defensive functioning improve with treatment [4, 8–10].
To date, these have been naturalistic observational studies of patients in treatment and follow-up, but
they have also begun to examine the role of defenses in the processes of change with psychotherapy.
Kramer et al. [11] found that change in distress was mediated by prior improvement during psycho-
therapy of defensive functioning, but not of conscious coping. Perry and Bond [12] reported that
change in defense mechanisms at 2.5 years of long-term dynamic psychotherapy predicted change in
multiple measures of symptoms and functioning at 5 years. While we await additional research to
establish that change in defenses mediates improvement in symptoms and functioning, it is important

J.C. Perry, M.P.H., M.D. (*)

McGill University at SMBD Jewish General Hospital, Montreal, Quebec, Canada
e-mail: jchristopher.perry@mcgill.ca
J. Petraglia, M.A.
Department of Counseling Psychology, McGill University, Montreal, Quebec, Canada
e-mail: jonathan.petraglia@mail.mcgill.ca
T.R. Olson, Ph.D. • M.D. Presniak, Ph.D.
FIT for Active Living, Saskatoon City Hospital, Saskatoon, Saskatchewan, Canada
e-mail: Trevor.Olson@usask.ca; Michelle.presniak@saskatoonhealthregion.ca
J.A. Metzger, Ph.D.
Utting Fellow for Studies in Depression, McGill University at the Institute of Community and
Family Psychiatry, SMBD Jewish General Hospital, Montreal, Quebec, Canada
e-mail: jesse.metzger@mail.mcgill.ca

R.A. Levy et al. (eds.), Psychodynamic Psychotherapy Research: Evidence-Based Practice 417
and Practice-Based Evidence, Current Clinical Psychiatry, DOI 10.1007/978-1-60761-792-1_25,
© Springer Science+Business Media, LLC 2012
418 J.C. Perry et al.

to explore and delineate therapeutic processes that lead to change in defenses. This chapter, then, is
an effort to examine some early hypotheses and approaches to determining how therapeutic interven-
tions lead to change in defensive functioning within and across psychotherapy sessions.

Background

In the previous volume in this series, our research group reviewed the theoretical and clinical char-
acteristics of defenses, and the rationale and methods for studying defense mechanisms in patients
undergoing psychotherapy [13]. We briefly summarize these points here. Defenses are automatic
mechanisms that deal with internal and external stress and conflict [14]. They occur partly or wholly
out of awareness preceded by signal anxiety, that is, a momentary sense of distress. Everyone has a
repertoire of defenses that they habitually use; hence, some defenses appear trait-like. However,
they are dynamic so that different stressors, conflicts, and states of mind may occasion the use of
different defenses. Defenses are attempts to adapt inner and outer realities, and they underlie symp-
tom formation and character traits. There is no basis for a necessary and sufficient list of defenses,
say, the way there is for the elements in the periodic table. Rather, we choose which defenses to study
based on criteria such as having good definitions, differentiation from other defenses, and evidence
of validity. Each defense is associated with a usual level of adaptation; hence, the defenses can
be arranged hierarchically by this level (Table 25.1). Nevertheless, every defense is adaptive in some
circumstances, which accounts for the persistence of some defenses that are only occasionally adap-
tive. There may be defense sequences in which individuals shift from using lower to higher adaptive
defenses, both in development and normal maturation, across treatment and even across the sequence
of dealing with severe stressors over time. One sequence often seen by clinicians is that of acting out
shifting to reaction formation, which later shifts to self-assertion or altruism. These potential
sequences require empirical delineation, and, if correctly delineated, would help clinicians identify
positive developments in defensive functioning whenever they occur. Defense and coping mecha-
nisms conceptually overlap, but the latter are sometimes differentiated from defenses as conscious

Table 25.1 DMRS hierarchy of defense categories, levels, and individual defense mechanisms
I. Mature
7 High Adaptive Level (Mature): Affiliation, altruism, anticipation, humor, self-assertion, self-observation,
sublimation, suppression
II. Neurotic
6 Obsessional Level: Intellectualization, isolation of affect, undoing
5 Other Neurotic Level: (a) Repression, dissociation, and (b) reaction formation, displacement
III. Immature
4 Minor Image-distorting Level (Narcissistic): Devaluation of self or object images, idealization of self or object
images, omnipotence
3 Disavowal Level: Denial, projection, rationalization. Although not a disavowal defense, autistic fantasy is scored
at this level
2 Major Image-distorting Level (Borderline): Splitting of other’s images, splitting of self-images, projective
identification
1 Action Level: Acting out, hypochondriasis, passive-aggression
IV. Psychotic
0 Defensive dysregulation Level (psychotic): Distortion, psychotic denial, delusional projection, psychotic dissociation,
concretization
Overall Defense Maturity (ODF): 0–7 scale summarizes defensive functioning by taking the mean of all the defense
scores, each weighted by the above 0–7 scheme
25 Accuracy of Defense Interpretation in Three Character Types 419

and intentional [15], whereas defenses protect the individual from awareness of threat, anxiety, and
conflict, often at the expense of constricting awareness, reducing flexibility of response, and optimi-
zation of adaptation. Finally, we noted that specific defenses may be associated with specific
disorders and problems in treatment, study of which may inform management and treatment of indi-
viduals. The use of individual defenses also serves as a marker, alerting the clinician that a core issue
is activated and informing on how it is being handled. Considering the defenses in use then offers the
clinician opportunities to intervene at specific times and with specific interventions. This issue is the
focus of the current report.
In the remainder of the chapter in the previous volume [13], we developed four hypotheses about
characteristic ways defenses change over the course of psychotherapy and then examined them on
four cases, with follow-up varying from 1 to 14 years. Understanding patterns of defensive change
across sessions and time is an important precursor to examining how to influence change within a
session. We briefly recap those hypotheses and results (Fig. 25.1).
The first hypothesis was that as individuals change, they increase their overall level of defensive
functioning, while at the same time, variability in defensive functioning tends to decrease, indicating
increased resilience to stress. The evidence indicated that this is true over the long term (years) [13].
However, the time frame over which this is true may depend on patient and treatment characteristics
as well as time. For instance, we recently reported a study of long-term psychotherapy in which we
found that change in defenses by two-and-a-half years was, on average, not yet associated with a
decrease in variability [12]. This appears to stem from the fact that, as a group, the patients had not
yet attained the neurotic level of functioning. Thus, future tests of this hypothesis may need to deter-
mine the degree to which variability is a function of the usual level of defensive functioning.
Decreased variability may not occur until improvement in overall defensive functioning (ODF) has
reached a certain level, which is yet to be determined.
The second hypothesis was that change in defense levels occurs in a stepwise fashion in which
individuals trade off defenses lower on the hierarchy for those in the middle and only later develop
those at the top of the hierarchy. Specifically, as lower level (principally immature) defenses decrease,
mid-level (principally neurotic level) defenses increase, and then as improvement continues, mature
defenses increase. This hypothesis was generally found to be true in the three cases with enough data
to examine it. However, in one intensively treated case, high adaptive (mature) level defenses began
to increase early on. This latter phenomenon was also reported in a subsequent study [12], in which
improvement was larger for the high adaptive defenses than for the mid-level defenses. As a result,
the hypothesis of stepwise improvement requires refinement, as mid-level defenses may not change
in unison in some individuals. For instance, over the course of therapy with most individuals, repres-
sion may decrease, whereas with severe personality disorders in which splitting is initially prevalent,
repression may paradoxically increase in early years of treatment before decreasing much later.
The third hypothesis was that over a given period of time, individuals and classes of individuals
(e.g., a diagnostic group) have their own rates of change which may vary across naturalistic and dif-
ferent treatment conditions. For example, in episodic disorders, such as major depression, state
changes (depressed to remitted) may be associated with initially large changes that then decelerate.
By contrast, individuals with stable traits, such as some personality disorders, may have long initial
periods of induction into the therapeutic process (“priming”) before change commences. Thereafter,
the underlying trend of improvement may become more or less linear. Treatments that increase this
rate of change are likely to be seen as more effective. While we found evidence consistent with this
in four cases [13], a full test requires multiple assessments across time for a variety of disorders and
treatment types. Furthermore, controlled trials would be required to determine whether specific
treatments alter the naturalistic rate of change in defensive functioning.
Our fourth hypothesis, in line with most of the research to date, was that as defensive functioning
improves, symptoms decrease and other aspects of functioning improve. Of the four hypotheses, this
420 J.C. Perry et al.

Hypothesis Results: 4 cases with 1 to 13 years follow-up

1a. Improvement: The general level of Yes in 3 cases

Overall Defensive Functioning rises + (slight improvement) in 1 case

1b. Improvement: Variability in defensive Yes in 3 cases, (1 case insufficient data)

functioning decreases Variability decreased over time

2. Improvement follows the hierarchy of Yes in 4 cases

defenses Increased (2), little
High Adaptive change (2)
Neurotic Decreased in 4 cases
Immature Decreased greatly in 4
cases

3. Individuals and classes of individuals Depression:

have unique rates of change in a given Defenses return to
condition: Natural history or specific neurotic levels by 6-18
treatment months (2 cases)
Personality Disorders:
Clear improvement by
2-5 years (2 cases)
Development of
healthy functioning by
13 years (1 case)

4. Improvement in defensive functioning Yes

correlates with improvement in Social functioning
symptoms, functioning improved (4/4)
Decreased depression
(3/3)
Improved coping (1/1)
Developed healthy
functioning (1/1)

Overall summary Defenses are a robust measure of how

personality structure functions at any
time
Measured over time,
they reveal whether
that structure is
changing (improving)
Therapy appears to
promote improvement,
including developing
healthy functioning
Improved defensive
functioning is
associated with other
improvements in
symptoms and
functioning

People improve at
different rates

Fig. 25.1 Results of four defense hypotheses examined in four cases (Data from [13])
25 Accuracy of Defense Interpretation in Three Character Types 421

has garnered the most support in recent studies [8, 11, 12], including evidence suggestive that
improvement in defenses acts as a mediator of change in symptoms and functioning.
While each of the above hypotheses will benefit from additional study, we believe that the find-
ings to date have firmly established that defense mechanisms are clinically meaningful, playing a
central role in adaptation. However, there is a question as to how therapists directly intervene with
defenses, say, by using interpretation. A recent review of psychotherapy studies – all of which identi-
fied therapist interventions using the Psychodynamic Intervention Rating Scales (PIRS) – found
consistent evidence that dynamic therapists often directly address defensive functioning [16]. In
reviewing four studies of short and longer term dynamic psychotherapy and psychoanalysis, the
authors found that defense interpretations were used more commonly than transference interpreta-
tions. Furthermore, defense interpretations increased from early to later sessions and became some-
what deeper.
Several reports indicate that addressing defenses has important effects. In a case series, Foreman
and Marmar [17] reported that at difficult therapeutic impasses, addressing the patient’s defenses,
such as denial, resulted in improved therapeutic alliance. Despars et al. [18] considered the effect of
interpretation on patient defensive functioning and suggested that interpreting defenses at the same
level as the patient was actually using should have an effect of improving the alliance. Following
this idea, Junod et al. [19] examined the accuracy of therapist interpretations in the middle two ses-
sions of a four-session Brief Psychodynamic Investigation. They divided patients into a high or low
alliance group based on the mean alliance of the two sessions. In the high alliance group, they found
that therapists tended to interpret the patient’s most commonly used defense level (43% of cases) or
a level just slightly above it (43% of cases), whereas in the low alliance group therapists tended to
interpret below the patients’ most commonly used defense level (75% of cases). In a study of short-
term psychotherapy, Winston et al. [20], using a different methodology, found that therapists address-
ing defenses (TAD) was associated with improvement in neurotic defenses. While the heterogeneity
of design and methods precludes specific conclusions, we can safely conclude that examining how
therapists address defenses is promising.
A review of all of the possible mediators of in-session change is beyond the scope of this report,
but focusing on defenses, the above research leads us to link the following. Defenses are clinically
meaningful. Therapists recognize defenses and interpret or otherwise address or manage them with
some frequency. Attempts to modify defenses should be associated with change in defensive func-
tioning and subsequently with change in measures of symptoms and functioning. The question we
will explore is how can we empirically determine that some specific aspect of interpreting defenses
is associated with change in defensive functioning in the process of psychotherapy. By contrast with
our previous chapter [13], which dealt with whether defenses change and in what patterns, the cur-
rent chapter deals more directly with how they change.

Addressing and/or Interpreting Defenses

Our research group recently conducted a review of the general theoretical and clinical literature
related to the addressing defenses in psychotherapy [21] in which we enumerated 74 separate hypoth-
eses related to the process of improving defensive functioning. Among the most promising to exam-
ine was theme 14, that identifying specific individual defenses can be necessary for successful
interpretations, previously noted by Rangell ([22, p. 168]) and ourselves ([23, pp. 532, 538]). We will
explore two conceptually related hypotheses regarding change at the level of the individual psycho-
therapy session and across sessions. We will examine the feasibility of this approach, applying it to
several cases reflecting borderline, histrionic (hysterical), or narcissistic personality disorder types.
422 J.C. Perry et al.

Hypothesis 1: Defensive changes within sessions predicts overall defensive

change across sessions

In line with the findings that defensive functioning improves over time, we hypothesize that defen-
sive functioning will change over the course of a session, and that this general trend within sessions
will relate to the overall change across sessions. Operationally, we will examine the defenses across
each individual session selected to determine the trend in the defense level score (1–7). A positive
trend indicates that ODF is improving, while a negative trend indicates that ODF is regressing.

Corollary

A clinical implication of this hypothesis is that large moves in defensive functioning within a session
would indicate that something of particular interest has occurred, which could reflect the patient’s
response to a stressor, the patient’s response to the therapist, or the patient–therapist interaction.
These large moves might be “hot spots” of good or poor therapeutic activity, warranting particular
clinical attention.

Hypothesis 2: Accuracy of defense interpretation predicts change

in defensive functioning within and across sessions

This can be examined in two parts: 2a. On average within a session, the accuracy of interpretation
will be reflected in the direction and amount of change in patient defensive functioning; 2b. Across
sessions, on average, the accuracy of defense interpretation will relate to the rate of overall change
in defensive functioning. Operationally, we will examine the defenses prior to and immediately
following therapist defense interpretations to look for the direction and amount of change in ODF.
The level of accuracy of interpretation (defense adjustment) within and across sessions should then
mirror change in ODF within and across sessions.

Methods

Study Design

We selected cases from our previously published naturalistic study of long-term dynamic psycho-
therapy. This sample was particularly apt for our purpose because we had previously found that
defensive functioning improved using the self-report Defense Style Questionnaire and that this
change correlated with change in other measures of symptoms and functioning [24]. In addition, we
had session audio recordings and transcripts which had been rated for defenses using the DMRS [12]
on a subsample (n = 21) of the study participants. As in the report on the DSQ results, we found that
the majority of individuals demonstrated improvement in defensive functioning, suggesting that an
examination of the process of defense change would be fruitful.
Study participants were referred from the outpatient psychiatric department of a university
teaching hospital. The design, inclusion and exclusion criteria, and subjects have been described
previously [25–28]. Briefly, the overall aim of the study was to examine the course and outcome of
long-term dynamic psychotherapy for subjects whom clinicians deemed that previous, usually
short-term, treatments had been inadequate. Selection criteria included having a depressive, anxiety,
25 Accuracy of Defense Interpretation in Three Character Types 423

and/or personality disorder, expressing a desire for psychotherapy, and agreeing to participate in the
research component. Participants gave written informed consent after the study was explained to
them and their questions addressed.
Twenty-two experienced practitioners of long-term dynamic therapy participated, with a mean of
13.1 years of post-doctoral experience. Twenty were psychoanalysts. Therapists treated a median of
three subjects each.
Dynamic psychotherapy was offered once or twice weekly at the discretion of the subject and
therapist at no cost to the participants. The design was naturalistic and observational, intended to
reflect long-term dynamic therapy as locally practiced. Neither specific therapy manuals nor super-
vision groups were used. While participants were offered a minimum of 3 years, they could terminate
at will, or try other therapies such as pharmacotherapy. The median duration of therapy was 3 years
or 110 sessions [95% CI: 52–141; range 4–339] [28].

Measures

Defense Mechanism Rating Scales

We identify defenses using the quantitative directions for the Defense Mechanism Rating Scales,
fifth edition (DMRS) [29]. The DMRS is a quantitative, observer-rated method [3] which is almost
identical in content to the qualitative Provisional Defense Axis in Appendix B of DSM-IV [14, 30].
Each defense from the list of 30 defenses is identified in sequence as it occurs in the session. This
method differs from other observer-rated methods that are qualitative or semi-quantitative ratings,
which yield global ratings for the whole interview (see review in [3]), missing moment-by-moment
defensive activity.
Once a session has been rated, three levels of scoring the whole session are used, all of which
yield continuous, ratio scales for the whole session.
• Individual defense score. A proportional or percentage score is calculated by dividing the number
of times each defense was identified by the total instances of all defenses for the session.
• Defense level score. The defenses are arranged into seven defense levels hierarchically arranged
by their general level of adaptiveness (Table 25.1). Each defense level has a proportional or per-
centage score calculated.
• Overall Defensive Functioning. The ODF score is obtained by taking the average of each defense
level score, weighted by its order in the hierarchy, yielding a number between 1 (lowest) and 7
(highest).
In addition, the defense level scores can be divided into several super-ordinate categories: mature,
neurotic, immature, and psychotic, described by Vaillant [6], although in most publications using the
DMRS, the fourth is not included. For the purpose of psychotherapy process research, an immediate
ODF can also be calculated from one or several defenses at any point in a session, allowing a
moment-to-moment representation of the level of defensive functioning across the session.
The PIRS [31, 32] is a systematic observer-based method for identifying therapist’s interven-
tions from therapy transcripts. The PIRS consists of a manual of definitions with examples of ten
types of interventions characteristic of psychodynamic therapies. The interventions include (1)
acknowledgments (Ack), (2) work-enhancing strategies (WES), (3) contractual arrangements (CA),
(4) questions (Q), (5) associations (Assoc), (6) support strategies (SS), (7) reflections (Rf), (8)
clarifications (Cl), (9) defense interpretations (D), and (10) transference interpretations (T).
These are sometimes further grouped into three broad functional categories: therapy-defining (2, 3),
424 J.C. Perry et al.

supportive (4 through 8), and interpretive (9 and 10). Banon et al. [16] found that in early therapy
sessions, the mean proportion of interpretive interventions varied from about 10% to 20%
across four studies. Each interpretation is given an additional rating based on a five-point scale
anchored by definitions, reflecting the depth and linkage of each. Briefly, for defense interpreta-
tions these are:
1 = The therapist specifies the methods used to diminish affect or diffuse meaning, or points out an
affect.
2 = The therapist specifies both the method used to diminish affect or diffuse meaning and also points
out an affect.
3 = The therapist alludes to methods used to diminish affect or diffuse meaning and inquires about a
possible motive (without specifying what the motive is).
4 = The therapist makes a remark which alludes to both the process of avoiding or mitigating affect,
and the motive as to why the affect is being avoided or mitigated.
5 = The therapist specifies the defensive process, the motive, and makes a link to past relationships.
In applying the PIRS, the rater first identifies the beginning and end of an intervention, and identi-
fies its type. Interpretations are then given the additional depth rating [1–5] above. Raw counts were
expressed as a proportion of total interventions for that session. For whole session scores, the indi-
vidual interventions are summed by category and divided by the total number of interventions to
yield a percentage score. The interrater reliabilities of the PIRS categories varied from k values of
0.83–0.99 [32].

Procedures

Participants had an initial Guided Clinical Interview (GCI) with a psychiatrist who made DSM-IV
Axes I through V diagnoses and obtained a personal lifetime history [24]. At baseline and every
6–12 months, research assistants interviewed subjects using the Longitudinal Interval Follow-up
Evaluation [33] – Adapted for the Study of Personality (LIFE-ASP) [29]. All psychotherapy sessions
were audiotaped.
Transcripts were made of sessions 3, 5, 7, three sessions at 6 months, and two at 2.5 years of treat-
ment for those still in treatment. Audiotapes and transcripts of sessions were disguised as to session
number and rated in random order for defenses. These data were then entered into computer files for
analysis of longitudinal change. A separate file was made to examine change in defenses within ses-
sions, in which each defense scored was entered in the order that it was scored, allowing analysis of
defense and defense level as the session progressed. Defenses were rated on 21 of the 49 participants
on whom we collected session and follow-up data. We selected patients who had completed at least
6 months of therapy, but preferably 3 years as the design allowed. Due to funding limitations, tran-
scribing and rating stopped at 21.
A rater blind to defense data rated the session transcripts above, identifying therapist interven-
tions using the PIRS. The same or a different rater then examined the transcripts with PIRS ratings,
selecting defense interpretations and then identifying the specific defense levels and/or individual
defenses that the therapist was interpreting. A research assistant then combined the information for
computer analysis as follows. For each session the defense interpretations were entered in sequence
along with the three immediately preceding and three immediately following defenses rated for the
patient, along with the defense levels and individual defenses interpreted by the therapist. From
these data, we calculated the patient’s ODF prior to interpretation (prior-ODF), the patient’s ODF
following the interpretation (post-ODF), and the prior–post difference (dif-ODF), which reflected
25 Accuracy of Defense Interpretation in Three Character Types 425

the direction and magnitude of the change in defensive functioning. The ODF of the therapist’s
interpretation was also calculated (interpretation-ODF). Taking advantage of the 1–7 hierarchy of
defense levels, we devised a defense adjustment score to reflect the accuracy of defense interpreta-
tion. A ratio was calculated dividing the therapist’s interpretation-ODF by the patient’s prior-ODF
in which a score of 1 = perfect agreement, <1 = interpreting below the patient’s mean prior-ODF, and
>1 = interpreting a defense level higher than the patient’s mean prior-ODF. Defense adjustment then
represents accuracy by the direction and magnitude of any difference from the mean of the patient’s
actual three defenses immediately preceding the interpretation.

Results

We selected the cases for this report to reflect several personality disorder types, with somewhat dif-
ferent responses to treatment and long-term outcome. Figure 25.2 shows the data from all the indi-
vidual session ratings, with the linear regression trend lines reflecting change in ODF for the three
cases. While all the cases show change in a positive direction, they differ in the rate of change, with
the highest rate 22 times greater than the lowest. There was some suggestion that initial ODF moder-
ated the rate of change in ODF in these cases, in that Case E with the lowest initial ODF (4.15) had
the slowest rate of improvement. This individual had borderline personality disorder (BPD), reflected
by a low ODF. Secondly, as we posited and found in our previous chapter in this series [13], the
session-to-session variability, which was initially large, also decreased over time, in these cases at
about 2.5 years. Thus, these cases represent a good opportunity to see whether the intra-session
response to interpretation is reflected in the overall rate of change.

Fig. 25.2 Change in overall defensive functioning over 2.5 years of therapy: individual session data and linear regres-
sion trend lines for three cases
426 J.C. Perry et al.

Case E (2035)

Miss E was a 22-year-old single woman working in the erotic services field, studying for one of the
broad helping professions, and doing an internship, when she was referred to the psychotherapy
research study. She was recently discharged from the psychiatry in-patient service after a highly
lethal suicide attempt with pills that required treatment in the intensive care unit. She had one of the
highest persistent levels of suicidal ideation in the study, with suicidal ideation most days of the
week. She showed clear signs of emotional instability, impulsive, self-destructive behavior and
intense, unstable relationships. On the Borderline Personality Disorder Scale [34] her score of 40.7
(28 is the cutoff for BPD) was in the extreme upper range of BPD individuals. She also met full
criteria for dependent and depressive PDs, and had significant self-defeating and antisocial traits.
At intake, she had five current Axis I disorders, including major depressive and dysthymic disorders,
generalized anxiety and post-traumatic stress disorders, and substance use disorder. Her GAF at
intake was 48 while 53 was her best level of functioning in the prior year.
She had a history of being molested by a male second degree relative. From school age onward,
her father would punish her by first ordering her to undress, then beating her. Her mother never
intervened, and both parents were emotionally neglectful except in the earliest years. In her late teen
years, she became addicted to heroin and cocaine by her boyfriend who also pimped her for financial
gain. In general, she had often been abused and abandoned by men.
She began therapy but requested and received a change of therapists early in the course of treat-
ment, remaining in therapy for a combined total of 189 sessions over about 4 years. She was usually
seen weekly, occasionally twice weekly. Her therapist was a male psychoanalyst. We present two
sessions with her second therapist.

Session 6

In this session, the patient had 22 defenses scored. Figure 25.3 shows the progression of the defense
level scores over the session. In the initial third of the session, she displayed largely level 5 neurotic
(especially repression) and some level 6 obsessional and level 7 mature defenses, while in the latter
two-thirds, she vacillated among neurotic (e.g., repression, displacement), disavowal (e.g., rational-
ization), and minor image distorting (e.g., devaluation), that is, levels 3 through 5. Her final defense
was an action level 1 defense evident in a story she told. The regression line in Fig. 25.3 indicates
that with each subsequent use of a defense, her ODF decreases by .1 of a point, which is a substantial
rate of change. Thus, this session would be characterized as one that challenged the patient’s initial,
neurotic level of defensive functioning, leading her to recount and explore stories highlighting her
lower defensive functioning from mid-session onward.
The therapist was highly active in this session, making a total of 37 interventions, a high propor-
tion of which were interpretive (40.5%). Figure 25.4 shows the individual data and linear regression
line for the adjustment level of interpretations (range 0.38–1.50) and the associated difference in
ODF from immediately before and after each interpretation (range 0.67−4.33). Both regression lines
trend negatively and in parallel, as the session progressed. In fact, for the 11 paired observations with
complete ratings, the correlation of defense adjustment and dif-ODF was quite high (rs = 0.80,
p = .003). At the outset, the therapist interpreted at the patient’s average level of defensive function-
ing (e.g., interpreting repression) or slightly higher. However, by mid-session as the patient began to
open up, she showed a wider range of defenses, and the therapist increasingly interpreted the lower
level defenses in her repertoire; thus, defense adjustment decreased, but only slightly. As this pro-
ceeded, the patient tended to reveal vignettes with more lower level defenses. The three following
selections represent this.
25 Accuracy of Defense Interpretation in Three Character Types 427

Fig. 25.3 Ms. E Session 6: Evolution of defense level scores across the session

Fig. 25.4 Ms. E Session 6: Parallel evolution of defense adjustment and dif-ODF scores across the session

This interchange began with the patient inquiring about who listened to the audiotapes of the session,
and whether that included the therapist. In the following interchange, the therapist is trying to explore
whether she has any fantasies about this. He interprets whether she might be defending against a fantasy
that might challenge her experience as to how well the therapy is going, and encourages exploring this.
The therapist’s interventions are noted on the leftmost margin, while the patient’s defenses are demar-
cated by their onset and offset. Interpretations also carry their depth rating [1–5] as a suffix.
428 J.C. Perry et al.

Selection E-1

WES I: …I’m interested in exploring what that thought would be in the back of your mind, what
I might say about things that would disappoint or upset you. Because you did think that.
S: [REPRESSION] Yeah, I don’t – I don’t know specifically just… [pause] I don’t know, I’m
just – I’m very naive and I don’t know if – I don’t know what you would say that would
upset me, but if there was something it would – you know what I mean?
Like there – there – the reason might be very simple and – and straightforward, but
I – I don’t know. [pause] [sighs] [repression ends].
WES I: What I’m trying to encourage you to do here is what we call free association. I realize
we’re not having a discussion that’s that much grounded in the official and realistic
answers. It’s more we’re having a discussion to explore what fantasies you might have in
your mind, because those might help us in the therapy.
T-3 I: [T-3 begins] So you did have a thought somewhere that if you were to ask me about
the tape, my answer might upset or disappoint you. So I would imagine that hiding behind
that thought are some specific possibilities of what I might say that would disappoint or
upset you.
It’s maybe hard for you to let them sort of come up to the surface and see that they’re
sort of underneath the surface themselves. They have to be there, because you wouldn’t
have the thought that you shouldn’t rock the boat, everything’s going well so far and that
if you do rock the boat, I’ll tell you it’s none of your business and say something that
would upset you or disappoint you.
So there’s something in there somewhere, but I don’t know if you could identify or tell
me. [end T-3]
S: I don’t know.
T-1 I: [T-1 begins] I mean, I realize it’s in the general ball park of whether or not you will be
disappointed and let down again and whether or not therapy really will be for you. What
you’ve said is that so far it’s going well…
S: Uh-huh.
I: …and you don’t want to rock the boat, as if you don’t quite believe that it’s going well,
that there’s got to be a rat somewhere. [T-1 ends]
S: Uh-huh.
WES I: So maybe we should explore that a little bit.
S: [pause] [speaking quietly] [inaudible] [pause] [SELF-ASSERTION] I just have to say that,
um, sometimes you use the word fantasize and fantasies and it just makes me very uncom-
fortable [self-assertion ends]. [RATIONALIZATION] Like I don’t – I’m very grounded,
I’m very like down to earth [rationalization ends]. [DENIAL] I don’t have fantasies [denial
ends]. Just, it makes me really uncomfortable.

Comment

In the above, his transference interpretation questions an affect and a defense, and the defense adjust-
ment is 1.03, indicating that he is interpreting at the level of defenses that she is using, the most
recent being level 5 repression. The repression is supported by three additional defenses. This selec-
tion also demonstrates that defense interpretations are contained within transference interpretations.
After several more minutes in like exchange, the patient opens up further.
25 Accuracy of Defense Interpretation in Three Character Types 429

Selection E-2

S: [REPRESSION] Just everything you’re saying, it’s like you’re verbalizing what I’m thinking,
what I’m feeling. [pause] I don’t know, sometimes it’s [sighs] I’m – I – I’m thinking things
and I don’t even realize till you say it, like [pause] I don’t know, [repression ends] I just
[pause] I just have so many thoughts and feelings. [PROJECTION] Like I’m afraid you’re
going to tell me that I’m too screwed up, you can’t help me or that I’m – I don’t have any
problems so you can’t help me or, I don’t know, you don’t want to. [pause] [projection
ends]
[RATIONALIZATION] There’s a part of me that wonders if you respect me and I think
about that and I’m like well, it doesn’t matter ‘cause you’re my doctor and you’re not here
to judge me, but it must affect the way you talk to me. [long pause] [rationalization ends]
D1 I: You stopped. Where did your thoughts go there?
S: I just wonder what you think about me.

Comment

The above defense interpretation points out that she is using a defense, but addresses the highest
level of her previous three defenses, level 5, repression – “You stopped. Where did your thoughts
go?” – so defense adjustment is quite high (1.50). The therapist explored and interpreted her con-
cerns about respect and self-evaluation, leading to a transference interpretation (not shown), follow-
ing which she opened up further about her belief that she trusts too easily and that others will just
take from her and she ends up devastated. This is a pivotal point following which she is somewhat
stressed by the line of inquiry and responded both adaptively (level 7 self-assertion), and somewhat
more defensively (level 3 rationalization and denial). The patient then associated to her general
experience of people taking things from her and devastating her. Both the above example and the
material that followed are consistent with the importance of addressing neurotic level defenses,
whenever surrounding lower level defenses are evidently active in protecting the inhibitions. It is
tempting to interpret the lower level defenses, but because they support the central role of repres-
sion, the interpretation of repression was warranted. Hence, the high defense adjustment score was
associated with an opening up of exploration.

Selection E-3

S: [REPRESSION] Sometimes I don’t even realize that I am being mistreated till it’s too late. I just
think it’s normal to feel that – feel certain ways and then I’ll find out, no, it’s wrong, so it’s –
I just, I blame myself for not knowing or for putting myself in that. [pause].
And if I don’t – I don’t know, [repression ends] [REACTION FORMATION] sometimes
I just smile or I laugh, ‘cause it’s just [laughs] I don’t know. Sometimes, if I don’t – if I don’t
laugh, I’ll cry [reaction formation ends]. [DEVALUATION-SELF] Sometimes it’s just so – it’s
so sick, it’s ridiculous. It’s funny [long pause] [devaluation ends].
[[PASSIVE-AGGRESSION]] Sometimes I’ll recognize that I’m doing something wrong and
I just – I can’t help it, I just – I see myself repeating patterns. Just it’s really confusing, ‘cause
then I hate myself for doing it and I hate whoever I’m with, ‘cause if you loved me, then you
wouldn’t do it to me, but it’s my own fault for letting you. [long pause] [passive-aggression
ends]
430 J.C. Perry et al.

D-4 I: [D-4 begins] I think that what you’re saying and what we’re talking about here is really a
crucial issue for your well-being and I think it’s a large reason why you’re here in the first
place. That you’re in this rut of repeated, lousy relationships where you get abused and you
have an awareness and an insight that you’re part of the pattern, because you allow it to
happen.
And you’re quite mixed up as to who to blame. And you probably fluctuate between
blaming the other person and hating them to pieces, and hating yourself, because you clearly
do. You’re suicidal and you want to cut yourself.
S: Uh-huh.
I: You want to be hurt. So we have to really try to see if we can get a handle on this really core
problem for you.
S: Like I…
I: Because there’s a simultaneous, a very intense hatred of yourself and the other person. [D-4
ends]
Assoc Sorry I interrupted you.
S: [PASSIVE-AGGRESSION] [pause] Um, I – I’ve been in therapy for enough years to know
myself a little bit and I just recognize that I’m just – this – these past couple of weeks I’ve
just been regressing, getting worse, just like I used to be.
Um, when Jacques and I first started going out, I had a big fight with him, I was angry
with him and I slept with his best friend. And later he found out and I denied it and to this
day I think deep down he knows it, but I won’t admit it and he – we just will never talk
about it…
The session continued exploring this story, ending with a D-4 interpretation by the therapist.

Comment

Overall, these selections demonstrate two related common phenomena. First, the patient, who made
an initially neurotic, inhibited presentation, responded to well-adjusted interpretation – in our tech-
nical meaning – and opened up. Second, because the patient has BPD, the new material that fol-
lowed included vignettes evidencing lower level defenses. As the therapist began to interpret the
lower level defenses, the patient vacillated between immature and neurotic levels. In this instance,
the therapist interpreted “aggressively,” as evidenced by two things. First, the level of defense adjust-
ment decreased slightly as the therapist tended to pick up on the lower levels of the patient’s defenses.
Second, the pre–post difference in ODF after each interpretation tended to be negative from mid-
session onward, and to become more negative. However, despite her diagnosis of BPD, the patient
did not regress in the malignant sense of using major image-distorting and action defenses toward
the therapist. Hence, the therapy promoted exploration at the price of some regression, but “con-
tained” the patient well enough, thereby avoiding enactments (e.g., projective identification), which
would be evidence of more severe regression.

Session 20

In this session, the patient used neurotic level defenses early and late on but generally exhibited her
lower defensive functioning in the middle. In the section that follows, the patient discussed her
ambivalent feelings and actions toward her boyfriend, including an odd feeling of getting pleasure
from fighting with him. In this selection, she emphasized her wish to help him, although letting him
keep drugs for sale in her house made her nervous. The selection includes the sixth through eighth
of 11 interpretations.
25 Accuracy of Defense Interpretation in Three Character Types 431

Selection E-4

S: [REPRESSION] Well, what am I doing? I still don’t understand [repression ends].

WES I: Good question. I don’t know. Let’s look into it. You must have your own reasons for being
in this situation. It’s not just, you know, something that just dropped you in this situation.
S: [REACTION FORMATION] I just – I want to help him out. [reaction formation ends] I
don’t know. He kept the stuff at his house till he was arrested. But the more I think about
it, the more nervous and upset I get about it.
SS I: Of course; you could go to jail.
S: See, but if I were to say something, I feel like I’m being unfaithful, you know what I
mean?
SS I: Yes, I know what you mean. [inaudible]
S: What do I do, tell him to take it out?
WES I: It’s the same – look, the answer to that is the same as the answer [inaudible]. You want to
know, you know, why – you’re saying to me, why am I in this nightmare? Well, we have
to maybe take our time and look why you’re in that nightmare, figure it out.
D-1 [D-1 begins] You’re giving – and again, I come back to the whole business of your
laughing. It’s like you yourself have an official version of what goes on. The laugh tells us
that something else is true underneath [inaudible]. [end D-1]
S: [laughs] [pause] I don’t know, for a while, when we first started going out, I felt very
guilty, uh, [PASSIVE-AGGRESSION] because from day one I wasn’t – I wasn’t faithful
and he had some suspicion that – obviously I never told him [passive-aggression ends].
[REPRESSION] [laughs] I don’t know why I’m laughing [repression ends].
D-1 I: [D-1 begins] Well, because it was your method, from the very beginning, of stabbing him in
the back, because you allow him to stab you in the back all the time and you have various
ways of stabbing him back.
S: I was just afraid that he…
I: So in that sense it’s a [inaudible]. [end D-1]
S: He just seemed too good to be true. Like I was – I was going out with this Caribbean guy,
I got pregnant and that’s when I met François who convinced me, “No, don’t have a baby.
You’re Quebecoise, you’re white, whatever, you’re not going to stay with him,” na, na, na.
So I told François that I broke up with the other guy, had the abortion and I think I was
still with the other guy. Then I broke up with him and continued with François and François
found out, ‘cause he was snooping or whatever, he was looking at my books.
And then he – [DEVALUATION-OTHER] maybe he was an asshole from the begin-
ning [devaluation ends]. I don’t know, [PASSIVE-AGGRESSION] but I mean, 2, 3 months
after we were going out, I slept with his best friend and I remember it was to get back at
him. I was just furious with him [passive-aggression ends]. And that turned out to be the
biggest mistake because later he found out and he was – I wouldn’t admit it and I felt from
then on everything changed, [same DEVALUATION-OTHER as above] that from then on
he was an asshole [Devaluation ends].
D-4 I: [D-4 begins] Let’s say you’re going to two-time a guy and you’re going to sleep with
somebody else as a method of getting back at him, of getting something nice for yourself
[inaudible] or something, okay? You do it in such a way, though, as to get caught and get
shit. You got caught with the Caribbean guy and you got caught with his best friend. It’s
like you didn’t use your head, and if you’re going to do that, why not do it in such a way
that you’d never get caught? I would suspect that the reason is, is that there’s a part of you
that is quite masochistic and every time you give shit you have to be paid back and receive
shit. [end D-4]
S: I hate how I feel now, like I’m so uncomfortable; I want to change so much, but I can’t.
432 J.C. Perry et al.

WES I: I think the stuff I’m referring to in your life is something [inaudible] and I don’t think you
can change until we can slowly but surely trace out what it is, why it’s happening, how it
got to be that way, what it’s all about, what’s going on here. And until that we’re pretty
much in the dark [inaudible] this that and the other thing.
S: [pause] Like how – how do I tell François I don’t want the shit [referring to drugs] in my
house? [inaudible]. [pause] [RATIONALIZATION] My rationalization has always been
that I have his lawyer’s card and it’s a great lawyer and if the charges couldn’t ever be
dropped, God forbid, then François would take the rap [rationalization ends].
SS I: That’s a very pretty rationalization, but the one who may end up in jail is you, the one who
may end up pregnant is you and the one who may get killed because of the guns, is you.
S: [DENIAL] I don’t care about that, I just care about going to jail [denial ends].
While the above began with neurotic level defenses, the therapist interpreted the lowest level
defenses after they began to appear in the vignette. As the story unfolded, the therapist’s next inter-
pretations focused on repression, which then elicited passive-aggression. The therapist then
responded further by interpreting how the passive-aggression functions when she is uncomfortable
over her wish for something good for herself. This deep, challenging interpretation led to disavowal
of some concerns accompanied by more material. This series of interpretations leading up to a
deeper interpretation reflects how interpretation that targets the lowest of the defenses used – that is,
show low defense-adjustment levels – can lead to increased exploration but also to some regression
in defensive functioning. Although the patient’s responses were volatile, Table 25.2 indicates that on
average the patient had an increasingly positive response to interpretation across session 20, reflected
in the positive slope of dif-ODF.

Discussion Case E

For the four sessions rated on all measures (Table 25.2), the patient’s mean ODF across all four ses-
sions was 4.36, SD = .59, ranging from 3.77 to 5.16, indicating an extreme range of functioning from
low borderline to neurotic level functioning. This variability is consistent with Stern’s [35] seminal
description of “the borderline neuroses,” that individuals with BPD may sometimes appear neurotic
but readily regress in treatment. The rates of change in ODF within the sessions varied from −0.11,
to +0.021. While the patient regressed over the course of half the sessions, in one instance, session
6, the magnitude of the regression was substantial.
Her therapist was aggressively interpretive, often including motive, object relationships and
transference in addition to affect and defense. His mean adjustment to defense score was .80, range
.59 to 1.00, indicating that on average he tended to address the defenses somewhat below the moving
average level of her defensive functioning. Furthermore, as most sessions evolved, the trend of his
defense adjustment score was negative, indicating that as she began to reveal more lower level
defenses, the therapist preferentially tended to interpret the lower level defenses. Related to this, the
mean difference in ODF before and after interpretation was positive in only one (25%) session and
the trend within the sessions was positive in only one (25%). The overall result of the case was that
at 2.5 years, she evidenced a very slow rate of improvement in ODF, with raw change (Δ) = +.05,
about one tenth of an effect size. She was still within the range of defensive functioning, consistent
with BPD. Left for consideration is whether greater improvement would have followed on average
interpretations with higher defense adjustment scores on average.
25 Accuracy of Defense Interpretation in Three Character Types 433

Table 25.2 Comparing defense and defense-adjustment changes within and across sessions
Within session Across 2.5 years
ODF dif-ODF Defense-adjustment ODFa
Subject Session Mean Slope Mean Slope Mean Slope By model Slope per session
Ms. E 4 4.38 −0.013 0.0 −0.0090 0.80 −0.0029
6 4.14 −0.11 −0.43 −0.011 1.00 −0.011
20 3.77 +0.021 +0.20 +0.0078 0.79 −0.0071
21 5.16 +0.0027 −0.10 −0.0076 0.59 −0.0057
Mean 4.36 −0.025 −0.083 −0.0050 0.80 −0.0067 4.15|4.20 +0.00034
SD 0.59 0.051 0.23 0.075 0.15 0.0029 Δ = +0.05
Positive 50% 25% 25% 0%
Mr. F 5 4.90 −0.013 0.0 +0.18 1.08 +0.033
24 5.35 −0.0024 −0.92 −0.071 0.89 +0.0043
26 4.50 −0.0018 +0.26 +0.053 0.93 +0.0047
125 5.24 +0.0083 +0.39 +0.037 1.16 −0.0043
Mean 5.00 −0.0022 −0.068 +0.050 1.02 +0.0094 4.96|5.23 +0.0021
SD 0.38 0.0075 0.51 0.089 0.11 0.014 Δ = +0.27
Positive 25% 50% 75% 75%
Ms. G 4 4.78 −0.038 +0.17 +0.015 0.97 +0.00046
5 6.04 +0.028 +0.42 −0.0020 0.87 −0.00063
21 3.98 +0.035 +0.13 −0.0056 1.17 −0.015
22 4.73 −0.0027 +0.20 +0.020 0.93 −0.0057
23 4.94 +0.040 +0.99 +0.032 1.43 −0.010
66 5.46 +0.010 −0.38 −0.0022 0.91 −0.0039
Mean 4.98 +0.012 +0.26 +0.0095 1.05 −0.0058 4.86|5.37 +0.0076
SD 0.70 0.027 0.41 0.014 0.20 0.0054 Δ = +0.51
Positive 67% 83% 50% 17%
a
First | second ODF scores from the modeled data refer to predicted values at outset and at 2.5 years of therapy; Δ = raw
ODF change

Case F (2015)

Mr. F was a man in his mid-20s who was referred to long-term therapy after completing a short-term
therapy which had been precipitated by the end of his relationship with a girlfriend. He felt he was
still “a basket case.” He had a long history of substance dependence on cannabis beginning at age 13
but had been abstinent for over a year. He had no other Axis I disorders, except a history of child-
hood conduct disorder. On Axis II, he had definite histrionic and narcissistic personality disorders,
with the former predominating clinically, along with some significant antisocial, self-defeating, and
borderline traits.
The patient felt loved by his parents in his early years, although his mother was strict, not show-
ing her emotions readily, but unconditionally loving and understanding. He lost an eye at 5 due to
illness and remembered the event as suffused with caring. Grammar school went well; there were no
academic problems, and he had friends. The parents argued a lot and the father was physically abu-
sive to an older brother, who in turn from mid-childhood on became verbally and physically abusive
to the patient. The children could tell that their parents were heading for divorce. While at summer
camp at age 12, his mother was hospitalized, allegedly for anorexia but in fact had made a suicide
attempt. After discharge, she went to live with relatives. After the divorce, the children remained
living with father. Father was preoccupied with a new girlfriend, and exercised no oversight. From
14 onward, the patient felt very alone and became hungry for attention. He started lying to build up
his self-image, and did anything he could to be popular. He began smoking, having lots of sexual
434 J.C. Perry et al.

encounters, and stealing, first from his father, and later elsewhere, such as at a part-time job. There
was no direction, caring, or understanding at home. In later teenage years, he began using cannabis
regularly and worked out an arrangement where he bought the drug for his father who gave him a
share in return. He repeated a grade of high school and went through college in a desultory fashion
while taking a series of jobs that he quit or left before being fired. He got into financial trouble with
credit cards. He had intense relationships with girlfriends, desperate to connect with them. He was
purposefully exhibitionistic, and women found him entertaining, even captivating, but ultimately
needy. He had concerns about trust and fidelity and became excessively angry when disappointed by
them. At the outset of the project therapy, he hoped to develop a new career in drama or finance,
make a lot of money, and be seen as important.
He saw a male therapist once weekly for 125 sessions, terminating at about 2.5 years due to a
move across country for economic reasons. Follow-along continued by phone and mailed question-
naires for 7 years from intake. We examine one of six sessions rated.

Session 26

Raters identified 70 defenses in this very active session. In descending order, his most prominent
defenses were minor image distorting (30%), obsessional (24%), disavowal (20%), other neurotic
(11%), and high adaptive (7%) levels. The very high proportion of minor image-distorting defenses
reflected that his narcissistic, and to some extent hysterical, character issues were salient in the ses-
sion, as was a tendency to compartmentalize his own affective reactions by use of obsessional
defenses.
The session largely concerned several events that had recently transpired at a restaurant where the
patient worked as a waiter. At some point in each story, the patient talked from the vantage point of
an expert about how things should be done, how to run the floor, how to sell to customers and so on,
in part to protect or boost his self-esteem or to deal with related conflicts while minimizing his
uncomfortable feelings. Thus, there was a swing between narcissistic and largely obsessional neu-
rotic defenses.

Selection F-1

The first vignette demonstrates one of two large shifts in defensive functioning in the session. In it,
he described problems that he has with the floor manager, relating some interchanges along with
comments on how he thinks things should be run.
S: And again, just to jump ahead a little bit with that Brad conversation, you know, he started
explaining to me that, you know, “[subject’s name], you know, don’t worry about it. But if she’s
in her face and you’re in her face, just walk away, man. Do whatever she says. Just say yes and
don’t worry about it.” You know.
And that’s true. I mean, I could – I mean, [DISPLACEMENT] this restaurant business is very
high pressure. I mean, there is a lot – I mean, it’s little things, right. You think like, oh my God,
where’s the lettuce for the hamburgers, right? Well, when you got an order to get out and there’s
no lettuce for the hamburgers, you start going crazy for a piece of stupid lettuce. I mean, it’s
really funny, but it’s a piece of… But it could be a milker, it could be a little spoon, it could be
a million things right? [laughs][displacement ends]. [INTELLECTUALIZATION] And you
would think why is this person upset over a little spoon or a piece of lettuce, but you’ve got a
table full of six plates to get at and you’re waiting for a piece of lettuce to get the six plates, you
go crazy. It’s the nature of the business, you know. [intellectualization ends]
25 Accuracy of Defense Interpretation in Three Character Types 435

[DISPLACEMENT] So by definition her job is that much more difficult because she’s got
the - she’s in an intermediary position. She’s got management, senior management on her
head to say, “We need this place to perform,” and then she’s got a staff of about fifteen to
make them perform. And she’s being told, “Get these guys working,” you know. And she’s
trying to do her best job of making it happen. And sometimes she’s playing the tough cop
right now. It’s like the good cop, bad cop thing. [displacement ends]
And, uh, and I got caught up with the whirlwind of that. And until I spoke to Brad yes-
terday and found out, you know, you’re cool, man. You’re what we want.
D-2 I: [D-2 begins] But you were taking it very personally…
S: Completely.
I: …and feeling unappreciated and not understood.[end D-2]
S: [RATIONALIZATION] Well, I had to source it and I came to the understanding the source
of my insecurities, just for myself, never mind myself, it was her. That was the relationship
that wasn’t working in the restaurant. And you know what, I came to that – I sat down and
I wrote it…[rationalization ends]
D-1 I: [D-1 begins] You weren’t getting the strokes that you needed to keep you going, eh, to reas-
sure you that you’re doing okay and that she knows what you’re doing. [end D-1]
S: Well, yeah. It’s like I recognize her position and, um, I recognize it’s a very key relationship
for me to maintain in that restaurant and it was frustrating for me that I couldn’t maintain it,
you know. [RATIONALIZATION] It was frustrating for me that the little things were getting
in, like a phone call or a 15 min break when I’m working a 15 h day is making this relation-
ship sour. It’s stuff like that.[rationalization ends] It’s really very difficult for me to accept
this because this is silly, you know, when you work 15 h days, consecutive days, you know,
not like one day here, but doing like 10 h, 15 h, 13 h. I mean, I work a weekend, out of 48 h,
I work twenty-eight of them, you know. [UNDOING] The staff works very hard and we’re
compensated very well and we’re treated well, but you know, this one relationship in the
restaurant with me, just was really a source of, like really, it was very scary [undoing
ends].

Comment

The patient initially distanced himself from his feelings and reactions by displacements and intel-
lectualization [mean pre-interpretation ODF = 5.33]. The therapist made interpretations, numbers 2
and 3 out of 10 this session. The first spoke to his taking the problem with the floor manager person-
ally, which is a turning against the self, rated as passive-aggression [not shown]. This was directed
toward a defense occurring much earlier in the story, but at this point in time, it scored as a very low
defense adjustment = 0.19, the lowest value of the session’s 10 interpretations. This was followed by
an immediate decrement in ODF [dif-ODF = −1.33], reflecting two rationalizations and one undoing
[mean post-interpretation ODF 4.00]. This interpretation, way below the level of the patient’s imme-
diately preceding defenses, resulted in downward shift in defensive functioning, reflecting some
disavowal of his own role in the problem. However, the second interpretation addressed affective
experience only (D-1), without reference to a specific defense, and so defense adjustment is not
calculable. However, it was followed by a move upward in post-interpretation ODF, indicating a
positive response.
436 J.C. Perry et al.

Selection F-2

This selection occurred late in the session. He related a vignette of which he is proud wherein he
gave away free coffee at the end of a meal to a group of wealthy customers. This raised an issue of
conflict of interest between the effect on the bill versus his tip.
S: [DEVALUATION-OTHER] It’s totally – no, I mean, and then the grade’s important too,
right? I’m going to be upset if I get like a shit grade, if I put a hard effort in, you know.
But I’m not going to worry about, like if I’m writing one test and I’m worrying about
that A after, you know, when I got to write ten tests in the semester and I get like a C or a B
minus and I’m like, “Oh, my God, I got to get that A,” that’s a crock of shit. You know, all
I got to do is do my best and it’ll take care of itself [devaluation ends]. [RATIONALIZATION]
And that’s why I don’t worry about it, because I look at the percentage. I calculate my tip
at the end of the night, I do my division there, my tips over my total sales. I get my points.
And if I’m in the range, my range from fourteen and a half percent to eighteen percent, hey,
man, I’m making scratch and I’m getting paid for it, you know, and that’s an average. And
that’s all I will get, you know. I don’t care if I make fifteen cents on a dollar, I make fifteen
percent. That’s all I think about. I’m not worried about anything else [rationalization ends].
And those are good points, man. Waiters make good points. You know, salespeople – most
sales jobs you’re making six, seven, eight percent and even that much, you know.
[DEVALUATION-OTHER] Telemarketing, scumbag business that is, [devaluation ends]
they pay you 20%, you know. So we’re pretty high in percentage points, you know.
D-4 I: [D-4 begins] That’s why when you got sort of the possible conflict of interest with the cof-
fee, because the likelihood is what you’ve done is you’ve knocked five dollars off the
bill…
S: Hoping to get the five dollars in my pocket.
I: Well, the chances are they will appreciate the service and [unclear]. [end D-4]
S: So, what would it be, it would be about seventy-five more cents for me, an extra dollar.
That’s what it would work out to. Actually, yeah, seventy-five cents. You’re right. [DENIAL
(or rationalization)] But you know what, when I was doing it, I mean, I know that.
I wasn’t even thinking about – I won’t say – I was thinking about let’s serve them, you
know, and if that meant, yeah, a better tip, it also meant a better name for the house. Okay,
it wasn’t just for me. I wasn’t selfish in the act. It was really a selfless act. You know, ‘cause
for me a dollar here, a dollar there is not going to change my life. [Denial (or rationaliza-
tion) ends]
D-2 I: [D-2 begins] Yeah, sure. No, it’s not that it doesn’t sound like it made good business sense,
but it sounds as if you were acting more on the basis of being a principal of the outfit, of the
restaurant, where you would have that option and that flexibility, where you wouldn’t have
to answer to somebody else.[end D-2]
S: But I am constrained.
Ack I: Sure.
S: [DISPLACEMENT] Yeah. [laughs] Again, it’s so funny, man. The little things that we argue
about in the restaurant, like you know, it’s funny, you know. I mean, it’s like I was telling
you about spoons, right? [laughs] It could be about anything. You could take shit for [raises
voice] “Why did you leave the mayonnaise in that little…” like it would be one little may-
onnaise and, “You can’t leave it like that.” [laughs] And I’m like, “Oh, my God.”
[Displacement ends]
25 Accuracy of Defense Interpretation in Three Character Types 437

Fig. 25.5 Mr. F Session 26: Parallel evolution of defense adjustment and dif-ODF scores across the session

[UNDOING] But everybody works so hard. I have to say it’s a tremendous, tremendous
staff and we’re being recognized for it, you know. We are being appreciated for it. Yeah, they
work us very hard because some of the people didn’t stick up front and the ones that stuck are
going to carry the load right now. And everybody’s putting in, on average, you know, any-
where between 10 and 15 h a day. [undoing ends]
Q I: What were you thinking of when you said you argued with the boss?

Comment

The action of giving away the coffee to the customers encompassed issues of his self-image and self
esteem – a salesman who knows the right things to do – versus his duty to his employer. His defenses
largely deflected the conflict in favor of boosting his self-regard. The two interpretations related his
actions to a wish to be appreciated, and later his wish to be powerful (“being a principal of the out-
fit”). Before the first interpretation, the patient’s pre-interpretation ODF was low (mean ODF 3.67)
followed by the therapist interpreting almost at the same level (defense adjustment = .82). In turn, the
first post-interpretation mean was higher (ODF = 4.67; dif-ODF = +1.00). The next pre-interpretation
ODF was the lowest of the session (mean ODF = 3.33), but the next interpretation had an even higher
defense adjustment (1.20). This led to a much higher post-interpretation mean ODF (5.33) and the
highest dif-ODF (+2.00) of the session. Because the two interpretations were separated by a single
defense, their effects are somewhat confounded. Alternately, their juxtaposition may have produced
some synergy, resulting in a large, positive dif-ODF. These examples also demonstrate that low-
level defense interpretations (D-1, D-2) can be supportive while still setting the stage for fuller
interpretation (e.g., about his conflict in doing his duty to the employer).
Figure 25.5 displays the evolution of both defense adjustment and associated dif-ODF across the
whole session 26. Both trended higher. In fact, for the eight paired observations with complete
ratings, the correlation of defense adjustment and dif-ODF was high (rs = 0.60, p = .11), albeit shy of
power to demonstrate statistical significance.
438 J.C. Perry et al.

Discussion Case F

For the four sessions rated, Mr. F had a mean ODF (5.00) on the border between immature and
neurotic levels of functioning with two sessions on either side. This is consistent with having a high
functioning personality disorder. Only one of four sessions (25%) showed a trend for ODF to rise,
principally due to an initial reliance on obsessional and other neurotic level defenses early in the
sessions while telling stories, with lower level defenses showing up later. However, when the inter-
pretive parts of the sessions were examined, we found that the mean dif-ODF following interpreta-
tions in each sessions was positive in 50% of sessions while the slope of dif-ODF within sessions
was positive in 75% of sessions. Defense adjustment varied slightly around a mean of 1.02, indicat-
ing, on average, high accuracy at interpreting at the level of the patient’s defensive functioning.
The overall pattern of evolution of defenses within each session is somewhat counter to the over-
all rate of change. Overall, his change in ODF was positive (+0.27) across the 125 sessions, moving
the patient up from the level of personality disorder to neurotic functioning. His change in ODF was
at about the median for the study. However, the data on dif-ODF were in line with the changes in
defenses at 2.5 years, suggestive that changes in defenses in the interpretive parts of the session were
the better predictor of overall change. Consistent with our hypothesis, defense adjustment was close
to 1, indicating that on average, the therapist interpreted at the level of the patient’s defensive func-
tioning. The session also demonstrated that sometimes deviations in defense adjustment can also be
associated with large swings in defensive functioning. Along with these metrics, those listening to
the sessions would describe the therapist as respectful, supportive, and well-attuned to the patient,
while taking an interpretive stance.

Case G (2006)

Ms. G was a 27-year-old married woman working in a joint small business with her husband, referred
for long-term psychotherapy when their couple’s therapist encouraged her to get individual help.
She had a history of recurrent major depressions of short duration with dysthymic disorder beginning
at age 10 but terminating 2 years before admission to the study. As a teenager, she had a 3-year
period of alcohol dependence and substance use disorder but had been largely abstinent for nearly a
decade. She had definite antisocial and narcissistic PD types, with significant borderline, histrionic,
dependent, and passive-aggressive traits. At intake, her current GAF was 62, as was her best level of
functioning in the past year.
She grew up in the suburbs where her father owned a small business. She was the eldest of four
children. Mother had a hard pregnancy and some post-partum depression and did not want to take
care of the patient for the first several months of life. She felt some affection from mother until the
birth of a brother at age 4. While mother always thought the patient was adorable from 4 onward, she
rejected showing physical affection toward the patient. In general, mother showed affection only
toward one brother and largely ignored the children conversationally. Mother was emotionally
neglectful, for instance offering no comfort if the patient hurt herself. At age nine, the patient began
to beat up the younger brother out of jealousy, and sibling rivalries were rampant. While father was
more attentive toward her, he too turned his attention more toward the brothers when they became
active in sports. She began to steal things such as cosmetics from 11 onward and at 17 began hanging
out with friends who stole a lot. While she liked primary school, she lost interest by 14 and began
skipping school. She engaged in some vandalism. At 15, she threatened to quit school if her
brother did not start treating her better, and when he did not, she quit and entered a jobs program.
She became disinterested in work, partied a lot with friends, would come late to work, and began
25 Accuracy of Defense Interpretation in Three Character Types 439

Fig. 25.6 Ms. G Session 23: Evolution of defense level scores across the session

drinking a lot and having sex. At 16, her family got fed up and after a big argument kicked her out
of the house with no notice. As she was crying and packing up, they sat down and watched TV. She
lived with various people on and off for 6 months with no fixed address, but got a food service job.
At 18, she followed a boyfriend into a recovery program and became abstinent. She worked as an
exotic dancer engaging in some prostitution as well. She was irresponsible with protection and had
a series of abortions. She met a boyfriend and became quite dependent upon him although he was
abusive, was frequently unfaithful, and even took her money. She was afraid to disagree with him,
but finally left after getting tired of being berated. She met her current husband while stripping. He
was a big spender and a show-off but a gentleman and was admiring of her. She was reluctant to get
involved, then, later after he proposed, reluctant to get married, but finally did in order not to lose
him. He brought her into his business as a partner. Soon after getting married, they began to fight
physically, and she would hit, scratch, or slap him, once receiving a black eye in return. Their recent
couple’s therapy ended the violence. She described herself as very much in love but not that happy
with her marriage, often complaining that he thinks he is perfect, and blames all problems on her.
She was jealous and possessive of him as well as competitive at work, although she frequently gave
in to him at work. Seeking therapy, she wished to improve herself, her self-image and confidence,
her relationship, as well as to develop her intense sense of ambition.

Session 23

We selected this session because it was at her average ODF level, including many minor image-
distorting and disavowal defenses, which are characteristic of antisocial and narcissistic personality
[5]. The patient showed 34 defenses. Figure 25.6 displays the evolution of the associated defense
level scores across the session. The trend reflects a decrease in her level 3 defenses by the middle
and level 4 defenses by the end of the session, with a concomitant increase in levels 6 and 7 at the
440 J.C. Perry et al.

end. The therapist was very active with a total of 85 interventions, 16 of which were interpretations,
two of which were transference interpretations. The topics of the session concerned working with
her husband and a coworker who liked to pit her against her husband, who paradoxically often sided
with the coworker. This brought up issues of self-esteem, aggressive feelings, and competition ver-
sus how she actually wanted to show a softer side at times.

Selection G-1

This selection occurred toward the end of the interview after the therapist had made a number of
interpretations concentrating on revealing hidden affects and defenses against awareness, with sev-
eral interpretations speaking to underlying motives as well. In this selection, the patient had just
revealed how her husband had a very confrontational style, based on seeing others as the ones with
problems, not himself. The therapist then used the recent material as a foil for making essentially a
summary interpretation of the most important issues of the session.
Cl I: That this is how he sees things.
S: Yeah, that he sees me as a truly sick person and that he’s not at all, not sick, would be that
he doesn’t have any of these common problems and neither does his family and so these
are super people, super humans and [DEVALUATION-SELF] it makes me feel very shitty.
It makes me feel lower [devaluation ends].
SS I: I wonder though if there’s another way of framing it which is that you have become per-
haps more knowledgeable than he has about the varieties of human experience, if you
like.
S: Yes, I’m aware of that and, you know, that’s good. I’m very happy that I’m aware of that,
but it makes me feel sad that the person that I’m with is not aware of that.
Rf I: And defines you as someone who’s sick.
S: Yes.
Assoc I: But it takes a fair amount of self esteem to stand up to that and [unclear].
S: Like everything with him, you know, I’d love to talk like this to him, but everything with
him is a confrontation. And whenever, you know, when he wanted to get married, a few
years ago when we broke up and he wanted to get back together and he wanted to get
married and I didn’t, he told me that he went to see a family friend that was a psychiatrist
and this family friend told him, you know, how to run his – you know, what to do. You
know, “You should tell her that she has to get married,” and this, you know,
[DEVALUATION-OTHER] and I thought that was crazy, you know. I mean, he went to
see a psychiatrist in order to back him up to confront me [devaluation ends]. You know,
and this is something that I don’t want to confront him. I never, you know, with what I
learn here I never go out there wanting to confront him, you know, and I just feel that
everything is a confrontation and, you know, I mean, I’d love to share on a much more
open, but somehow it’s a confrontation of I know and I don’t know if I’m making any
sense or…
Ack I: Uh-huh, uh-huh.
T-4 [T-4 begins] No, I think I understand what you’re saying, and, uh, it sounds like – well, I
guess that maybe your theme of what you’re talking about today is really that you can
threaten people, even without meaning to. If you show your rough side, your rough side,
that can threaten people; it threatens your husband’s partner and you’re worried that it
might threaten or upset me or [unclear]. That if you show some of your knowledge or
some of your own self esteem, I am not a sick person, [part of sentence unclear] that you
can be a type of [unclear] I guess and that, that concerns you sometimes.
S: Uh-huh.
25 Accuracy of Defense Interpretation in Three Character Types 441

I: It’s almost as if you feel that you can’t let yourself kind of go and be the most impressive that
you can You sometimes have to hide your abilities and your knowledge and your capacities or
just certain aspects of yourself.
S: Uh-huh.
I: And yet there are aspects that could be threatening to other people.
S: Uh-huh.
I: Because, I suppose, because they contain a component of aggression and a well-worked com-
ponent of aggression and a refined component, but that, that’s, there’s an aggression there and
that [unclear] you feel that people will become defensive and fight you back because they feel
the aggression.
S: Uh-huh.
I: Rather than enjoying it or learning from it.
S: Right, uh-huh.
I: I suppose that’s, you know, part of – maybe one of the more difficult parts of becoming a more
confident person, a more successful person is that other people may be intimidated by you
[unclear].
S: Uh-huh.
I: If you do feel good about yourself [unclear] how are people going to handle [unclear] I don’t
know what they’ll do if they’re you [laughs], but you know, how are you doing to handle that
aspect, the fact that you may intimidate people [rest of sentence unclear]
S: Uh-huh. No.
I: Sometimes I wonder if your retreat into depression or tears or whining is really an attempt not
to use yourself as much as you can [T-4 ends].
S: Yes. For me it has been that. I don’t, you know, I don’t like it and I think that was a very interest-
ing conversation last week or two weeks ago, whatever, because, you know, we sort of talked
about that, you know, the two sides and which one do I like the most and, [UNDOING] you
know, you were even under the impression, or maybe I misunderstood you, but it seemed you
were under the impression that I liked the softer side and… [undoing ends]
[SELF-OBSERVATION] But you just put it into words very well a few minutes ago when
you said it’s – you know, there is a sort of aggression there, but a refined aggression. And this
is what I want. I don’t want to be rough and crazy and – but I certainly don’t want to have to
retreat to that and I feel that I’ve had to and that I have and that it’s not comfortable at all, you
know, and it’s – I don’t respect myself when I’m there and I don’t feel that you should respect
me or that anybody else should if I’m there too long. You know, I mean… [self-observation
ends].

Comment

The selection began with a projection (not shown) and two devaluations, one each of self and other.
The therapist then made the deepest interpretation of the session (T-4). Although it is largely an
interpretation of her defenses in relationships, because the therapist included the phrase, “it threatens
your husband’s partner and you’re worried that it might threaten or upset me,” thereby in passing
referring to the transference, it is scored as a T-4 rather than a D-4. The therapist interpreted her
inhibition of her talents (repression) for fear she would be seen as aggressive, which would then
cause others to retaliate. This makes it hard for her to enjoy herself and feel confident. She instead
turns herself into a weaker, less successful person (reaction formation) to prevent this retaliation (her
motive). This deeper interpretation has a high defense adjustment level of 1.43 because it interprets
the neurotic defenses leading to lower self-esteem, rather than the minor-image-distorting defenses
442 J.C. Perry et al.

Fig. 25.7 Ms. G Session 23: Parallel evolution of defense adjustment and dif-ODF scores across the session

used to temporarily shore up the low self-esteem. The patient responded with an opening up about
her ambivalence about herself and what the therapist pointed out about her (undoing). She then used
self-observation in an attempt to deepen her understanding of how to deal with the rough, self-
protective, and tender parts of herself.
Overall, the session had a high mean defense adjustment (1.43) and the highest mean dif-ODF
(+0.99) score of all sessions of all three patients. Figure 25.7 shows the evolution of the defense
adjustment and dif-ODF scores. In fact, for the eight paired observations with complete ratings, the
correlation of defense adjustment and dif-ODF was high (rs = 0.54, p = .17), albeit shy of power to
demonstrate statistical significance. This session shows the power of deeper defense and transference
interpretations aimed at defenses like repression that hide awareness of broader conflicts, even though
these defenses are at a higher level than those defenses used by the patient nearby. Reflected in this
high defense adjustment score is an accurate interpretation of the higher level defenses. This stimu-
lated the patient to explore the material more using obsessional and high adaptive level defenses.

Discussion Case G

Despite the combination of personality types (antisocial and narcissistic), this patient was very
actively engaged with the therapy and made great progress. Her few remaining antisocial features
ceased early on and the narcissistic and other personality issues predominated. Her initial ODF of
4.86 indicated a slight predominance of immature defenses over the neurotic and high adaptive lev-
els, although she had sufficient proportions of the latter to build upon. The therapeutic alliance
measured at 6 months of treatment was slightly above the median for the entire sample. The thera-
pist, as this session demonstrated, was very active and interpretive in ways that supported both self-
esteem and emotional growth. The mean defense-adjustment level of 1.05 indicated that, on average,
the therapist accurately interpreted the patient’s defense levels. This was associated with the highest
mean dif-ODF of the three cases (+0.26), indicating that, on average, an interpretation was followed
by an increase of one-quarter of a defense level, also the largest and only positive mean dif-ODF of
25 Accuracy of Defense Interpretation in Three Character Types 443

the three cases. Consistent with this, the patient’s mean slopes of ODF both within the sessions
(+0.012) and across all six sessions (+0.0076) were positive. She also had the largest raw improve-
ment in ODF across the 2.5 years of treatment of all three cases (+0.51), leading to a final ODF of
5.37, at the high end of the neurotic range by that time. The patient had 178 sessions over 4.26 years.
We considered that Ms. G had a very successful treatment and outcome.

Overall Analysis of the Hypotheses

These three cases provide an opportunity to examine the potential value of our hypotheses about
change in defense mechanisms in psychotherapy and to determine the feasibility of the design and
methods used. Our approach takes the justification perspective of discovery rather than validation,
as a case series of three is limited in statistical power and generalizability. Despite this, we have
extended the usage of validated methods (i.e., DMRS, PIRS) to examine the process in which patient
and therapist respond to one another centered on the interpretation of defenses. The examination of
the three cases together provides a preliminary evaluation of the stated hypotheses and this approach
to examining them.
Table 25.2 displays the individual scores and a summary row for each case with mean, SD, and
the percentage of each rating that were positive. The two rightmost columns summarize the defense
scores as calculated by individual linear regression models for all the sessions rated for defenses for
each case. While all three cases evidenced improved scores by about 2.5 years, the rates and amounts
of change varied by factors of 22 and 10, respectively. Thus, the three cases provide a good range of
the outcome of interest: change in ODF. To examine our hypotheses, we correlated each of the mea-
sures of interest summarized for the group of sessions for each case with the above mentioned rate
of change in ODF across the 2.5 years. We then relate each statistic to the relevant hypothesis as a
basis for discussion. We calculated both Spearman rank order rs and Pearson’s r, viewing the former
as more conservative, but the latter as potentially more informative, given its sensitivity to the mag-
nitude of differences. Significance testing is omitted, as the sample size (N = 3) is inadequate. The
following results are summarized in Fig. 25.8.
We first hypothesized that the rate of change within sessions would relate to the overall rate of
change across sessions. We provide two tests of this. Both the mean rate of change in ODF within
sessions (rs = 1.00 and r = 0.91) and, to a lesser extent, the percentage of sessions in which the rate
of change (rs = 1.00 and r = 0.65) were positively correlated with the overall rate of change in ODF
over 2.5 years. This is consistent with our hypothesis. The rate and direction of change within the
individual session are apparently fractal phenomena which, when aggregated over time, correlate
with the overall rate of change in ODF. The variability of rates of change across individual ses-
sions suggests that this overall relationship includes sessions that are better or worse. Some ses-
sions ending with regressed defensive functioning were found in all three cases, suggesting that
some occasional regression is compatible with overall positive change. Two cases, E and F, actu-
ally had a high proportion of sessions with some regression, which suggests that additional factors
may be necessary to produce overall change. This leads to consideration of our second hypothesis,
discussed below.
The corollary to the first hypothesis was that large changes in defensive functioning within a ses-
sion would identify “hot spots” in which something was affecting defensive functioning. Our textual
selections from the cases provided some examples. However, we cannot provide a broad test of the
factors associated with these dramatic shifts from selections alone, other than those systematically
studied in the second hypothesis. Conceptually, these factors may include one or more of the follow-
ing: a) internal stress, anxiety, and conflict which lead to a shift in one’s defensive state of mind
(exploratory, inhibited, counter-attacking, disorganized); b) recounting a recent or past vignette that
444 J.C. Perry et al.

Hypotheses Results

1. Defensive changes within sessions Yes

predicts overall defensive change across • The mean rate of change of ODF within
sessions sessions correlated highly with overall rate of
change of ODF across sessions, r = 0.91
• The proportion of sessions with a positive rate
of change of ODF correlated moderately with
the overall rate of change in ODF across
sessions, r = 0.65

2. Accuracy of defense interpretation

predicts change in defensive functioning
within sessions

a. On average within a session, the Yes

accuracy of interpretation will be reflected • Within sessions: Correlations between
in the direction and amount of change in accuracy of interpretation (defense
patient defensive functioning adjustment) and immediate change in ODF
(dif-ODF) varied from r = 0.20 to 0.80
(median r = 0.57) for 4 sessions studied
• Using the means from all 3 cases, based on all
14 sessions rated:
♦ Mean defense adjustment and mean
dif-ODF correlated r = 0.62
♦ Mean defense adjustment and mean
overall slope of ODF within sessions
correlated r = 0.96

b. The trend within a session in shifts Yes

in defensive functioning from before to after • Using the means from all 3 cases, based on all
defense interpretations will mirror the 14 sessions rated:
overall change in defensive functioning ♦ Mean defense accuracy scores
within the session correlated with rate of change in ODF
across 2.5 years r = 0.76

Overall summary • The rate and direction of change of defensive

functioning within sessions are apparently
fractal phenomena, which, when aggregated
over time, correlate with the overall rate of
change in ODF
• Some sessions show regression in ODF, which
is still compatible with overall positive change
over time
• Within and across sessions, accuracy of
interpretation (defense adjustment) correlates
with immediate change in defenses (dif-ODF)
• Therapists appear to select which of a patient’s
defenses to interpret
• Accurate interpretation of defenses correlates
with rate of change in defenses over time
• Further research on this approach is
warranted

Fig. 25.8 Summary of hypotheses and results

25 Accuracy of Defense Interpretation in Three Character Types 445

includes substantially different levels of functioning from the present; or c) response to an external
stimulus such as time of session, an interruption (e.g., checking who called on a cell phone), or a
therapist’s response. The latter includes defense interpretations.
Our second hypothesis examined the interpretation of defenses, specifically that the accuracy of
defense interpretation, as measured by defense adjustment, would be associated first with the direc-
tion and amount of change within individual sessions and across sessions. To explore part A of this
within the sessions, we examined the correlations between defense adjustment and dif-ODF scores.
For the four individual sessions reported in the above case examples, we found Spearman correla-
tions of 0.80 and 0.20 (Ms. E), 0.60 (Mr. F), and 0.54 (Ms. G). The median correlation was 0.57.
This is consistent with our hypothesis. To explore this for the averages across the sessions for all
three cases, we correlated the summary statistics in Table 25.2 (all correlations n = 3). First, we cor-
related mean defense adjustment with mean dif-ODF, which was positive: rs = 1.00 and r = 0.62. We
then correlated the mean defense adjustment score with the mean slope of ODF within sessions:
rs = 1.00 and r = 0.96. Both are consistent with hypothesis 2a: change in ODF within sessions is
related to defense adjustment.
For hypothesis 2b, we correlated the mean defense adjustment score for each case with the rate
of change in ODF across the 2.5 years, obtaining rs = 1.00 and r = 0.76. This is consistent with the
hypothesis that, on average, defense adjustment within sessions relates to the overall rate of change
in ODF across sessions. Of course, as observational not experimental data, these findings are consis-
tent with suggesting but not validating that therapist accuracy of interpretation might be a mecha-
nism of change for patient defensive functioning.
The three cases differed in mean ODF for the rated sessions; notably, Case E with borderline
personality had markedly lower mean ODF in the rated sessions. This led us to conduct an explor-
atory analysis, correlating the mean session ODF with mean defense adjustment scores: rs = 0.50 and
r = 0.98. This correlation is open to several possible interpretations. First, the ODF of the patient
moderates the therapist’s selection of the level of defense adjustment; for instance, low patient ODF
pulls for the therapist to interpret toward a lower defense adjustment. Second, the finding could
simply reflect differences in the therapists’ responses to their patients, rather than a systematic mod-
erating effect (e.g., therapists randomly differed in the likelihood of interpreting lower level defenses
when they occurred). A larger sample will be required to indicate further the likelihood that any or
all of the above affect defense adjustment and the patient’s responses to interpretation.

Overall Discussion

Overall, we balance the coherence of the above findings against the limitations of examining our
hypotheses on only three cases. Our findings are consistent with both hypotheses but from an explor-
atory not validating perspective. Nonetheless, we consider them good evidence of the value of pur-
suing this further on a larger sample, which we have available. Furthermore, we conclude that our
methods are feasible and practicable for conducting this type of psychotherapy process and outcome
research.
Kazdin [36] has encouraged those examining the process of psychotherapy to move from examin-
ing correlates and predictors of change in psychotherapy toward mechanisms of change. The present
report suggests that we have candidate measures that may fit this aim. We can assess defenses on a
moment-to-moment basis and obtain evidence of change within sessions. By comparing defenses
before and after an intervention, we can judge the patient’s apparent response to the intervention. We
can then summarize these measurements as the mean level of defensive functioning for the session,
and change in the mean level of defensive functioning across the session, and across the interventions
of a given type within the session. We can then relate these summary scores, like those in Table 25.2,
446 J.C. Perry et al.

to the overall change in defenses over time, either using mean session scores, or defenses as rated
outside of sessions, say, independent dynamic or RAP interviews [13]. The capabilities of our mea-
sures put us within conceptual and methodological reach of one aspect of Kazdin’s suggestion. Does
defense interpretation with certain characteristics (e.g., accuracy as measured by defense adjustment)
cause improvement in defensive functioning within and across sessions?
Elsewhere, we have categorized 74 hypotheses about change in defenses in psychotherapy, of
which 19 we considered highly warranted for immediate further study [21]. The accuracy of inter-
pretation, as measured by defense adjustment, is one of these. We posited several different designs,
including the experimental manipulation of defense interpretation, that may bring us closer to under-
standing how defense interpretation may in fact be a causal mechanism in producing change in
defenses, which, in turn, are mechanisms underlying the level of symptoms and functioning. Thus,
the study of defenses in and outside of psychotherapy may offer a very robust opportunity to tie
process and outcome in a theoretically coherent way, an important aspect of the validation of causal
mechanisms [36]. Our next step is to examine the present hypotheses in our larger sample, and then
as many of the other 19 as resources allow. Our most difficult challenge will then follow: attracting
financial support for the further study of the causal role that changing defenses in therapy plays for
overall improvement. For this, we will need to rely on our best defenses.

References

1. Freud S. The neuro-psychoses of defence, The complete works of sigmund freud, vol. 2. Standardth ed. London:
Hogarth Press; 1962.
2. Davidson K, MacGregor MW. A critical appraisal of self-report defense mechanism measures. J Pers. 1998;
66:965–992.
3. Perry JC, Ianni F. Observer-rated measures of defense mechanisms. J Pers. 1998;66:993–1024.
4. Hoglend P, Perry JC. Defensive functioning predicts improvement in major depressive episodes. J Nerv Ment Dis.
1998;186(4):1–7.
5. Perry JC, Presniak M, Olson T. Defense mechanisms in borderline schizotypal, antisocial and narcissistic person-
ality disorders; 2011.
6. Vaillant GE, Drake RE. Maturity of ego defenses in relation to DSM-III Axis II personality disorders. Arch Gen
Psychiatry. 1985;42:597–601.
7. Vaillant GE. The wisdom of the ego. Cambridge: Harvard University Press; 1993.
8. Hersoug AG, Sexton HC, Hoglend PA. Contribution of defensive functioning to the quality of working alliance
and psychotherapy outcome. Am J Psychother. 2002;56:539–554.
9. Perry JC. A pilot study of defenses in adults with personality disorders entering psychotherapy. J Nerv Ment Dis.
2001;189:651–660.
10. Roy C, Perry JC, Banon E, Luborsky LL. Changes in defensive functioning in completed psychoanalyses: the
Penn Psychoanalytic Treatment Collection. J Am Psychoanal Assoc. 2009;57:399–415.
11. Kramer U, de Roten Y, Stiglar M, Despland, J-N. Change in defense mechanisms and coping over the course of
short-term dynamic psychotherapy for adjustment disorder. J Clin Psychol. 2010;66:1232–1241.
12. Perry JC, Bond M. Change in defense mechanisms during long-term dynamic psychotherapy and outcome. In:
Paper presented at annual meeting of the Society for Psychotherapy Research, June 22–25, 2010, Asilomar
(Monterey); 2010.
13. Perry JC, Beck SM, Constantinides P, Foley JE. Studying change in defensive functioning in psychotherapy, using
the defense mechanism rating scales: four hypotheses, four cases. In: Levy RA, Ablon JS, editors. Handbook of
evidenced based psychodynamic psychotherapy: bridging the gap between science and practice. New York:
Humana Press; 2009. p. 121–153.
14. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC:
American Psychiatric Press; 1994. p. 751–757.
15. Cramer P. Coping and defense mechanisms: what’s the difference? J Pers. 1998;66:919–946.
16. Banon E, Semeniuk T, Perry JC, Bond M, Hersoug AG, Despland J-N, et al. Therapist interventions using the
Psychodynamic Interventions Rating Scale (PIRS) in dynamic interviews and therapy, psychoanalysis and CBT,
under review.
25 Accuracy of Defense Interpretation in Three Character Types 447

17. Foreman S, Marmar CR. Therapist actions that address initially poor therapeutic alliances in psychotherapy. Am
J Psychiatry. 1985;142:922–926.
18. Despars J, Kiely MC, Perry JC. Le développement de l’àlliance thérapeutique: influence des interventions du
thérapeute et des défenses du patient (The development of the therapeutic alliance: influence of therapist interven-
tions and the defenses of the patient). Psychothérapies. 2001;21:141–152.
19. Junod O, de Roten Y, Martinez E, Drapeau M, Despland J-N. How to address patient’s defences: a pilot study of
the accuracy of defense interpretations and alliance. Psychol Psychother Theor Res Prac. 2005;78:419–430.
20. Winston B, Winston A, Samstag LW, Muran JC. Patient defense/therapist interventions. Psychotherapy. 1994;31:
478–491.
21. Olson TR, Perry JC, Janzen JI, Petraglia J, Presniak MD. Addressing and interpreting defense mechanisms in
psychotherapy: general considerations. Psychiatry. 2011;74:142–165.
22. Rangell L. Defense and resistance in psychoanalysis and life. In: Blum HP, editor. Defense and resistance: histori-
cal perspectives and current concepts. New York: International Universities Press; 1985. p. 147–174.
23. Perry JC, Bond M. Defensive functioning [in personality disorders]. In: Oldham J, Skodol AE, Bender D, editors.
The American psychiatric publishing textbook of personality disorders. Washington, DC: American Psychiatric
Press Inc; 2005. p. 589–609 [Chapter 33].
24. Bond M. Empirical studies of defense style: relationship with psychopathology and change. Harv Rev Psychiatry.
2004;12:263–278.
25. Bond M, Perry JC. Long-term changes in defense styles with psychodynamic psychotherapy for depressive, anxi-
ety and personality disorders. Am J Psychiatry. 2004;161:1665–1671.
26. Bond M, Perry JC. Psychotropic medication use, personality disorder and improvement in long-term dynamic
psychotherapy. J Nerv Ment Dis. 2006;194:21–26.
27. Perry JC, Bond M. The sequence of recovery in long-term dynamic psychotherapy. J Nerv Ment Dis.
2009;197:930–937.
28. Perry JC, Bond M, Roy C. The natural history of long-term dynamic psychotherapy: predictors of treatment dura-
tion. J Psychiatr Pract. 2007;13:221–232.
29. Perry JC. Defense Mechanism Rating Scales (DMRS). 5th ed. Cambridge: Author; 1990.
30. Perry JC, Høglend P, Shear K, Vaillant GE, Horowitz MJ, Kardos ME, et al. Field trial of a diagnostic axis for
defense mechanisms for DSM-IV. J Pers Dis. 1998;12:56–68.
31. Cooper SH, Bond M. The Psychodynamic Intervention Ratings Scales (PIRS), 2002 revision. Montreal: authors
(1992 and revised in 1996, 1998, 2002).
32. Milbrath C, Bond M, Cooper S, Znoj HJ, Horowitz MJ, Perry JC. Sequential consequences of therapists’ interven-
tions. J Psychother Pract Res. 1999;8:40–54.
33. Keller M, Lavori PW, Friedman B, Neilsen E, Endicott J, McDonald-Scott P, et al. Longitudinal interval follow-up
evaluation: a comprehensive method for assessing outcome in prospective longitudinal studies. Arch Gen
Psychiatry. 1987;44:540–548.
34. Perry JC, Cooper S. Psychodynamics, symptoms, and outcome in borderline and antisocial personality disorders
and bipolar type II affective disorder. In: McGlashan TH, editor. The borderline: current empirical research.
Washington, DC: American Psychiatric Press; 1985. p. 21–41.
35. Stern A. The borderline group of neuroses. Psychoanal Quart. 1938;7:467–489.
36. Kazdin AE. Mediators and mechanisms of change in psychotherapy research. Ann Rev Clin Psychol. 2007;
3:1–27.