When patients probe the analyst: Manifestations of patient testing and its complexity – An in-depth exploration of case examples of extant research

Abstract

Patients probe the analyst with the goal of challenging pathogenic adaptations to early experiences. As the core concept of control mastery theory (CMT), testing is contextualized within psychoanalytic theory. The current work examines 29 articles illustrating therapies performed or analyzed using the CMT approach for the occurrence of testing, which takes place through interaction, self-presentation, narratives, or the use of the setting. The various manifestations of testing and their potential meanings are described. An in-depth analysis of selected testing examples is performed to compare tests within patients and across studies. The results show that patients differ in their testing strategies, shift testing strategy during the process of treatment, combine tests, and test multiple conflictual themes within a single test. Therefore, the importance of applying a case-specific approach, based on a thorough understanding of a patient, becomes evident. Recommendations concerning psychoanalytic technique, including the role of interpretation, as illustrated in case vignettes, are introduced.

Key words:

• control mastery theory
• object relations
• transference
• case studies
• psychoanalytic research

Patients bring their relational experiences, maladaptive interpersonal patterns, and conflicts into the therapeutic relationship, where they can be worked through. Control mastery theory (CMT; Weiss & Sampson & the Mount Zion Psychotherapy Research Group [hereafter referred to as Weiss & Sampson], 1986) conceptualizes transference and projective identification by its notion of testing, and theorizes that patients probe the analyst¹ to reevaluate their self and object representations. The following introduction presents CMT and its core concept, testing, which is interrelated with other components of the theory, namely interpersonal guilt, pathogenic beliefs, and the patient’s unconscious plan to get better. The postulates of CMT will then be contextualized within psychoanalytic theory and different concepts as well as clinical models of treatment. The focus lies on Freud and British object relations theorists because CMT was built up from Freud’s ego psychology and because Weiss (1994) labeled it as an object relations theory.

In the following exploration of case examples in the existing literature, various forms of testing are systematized to illustrate the complexity of the testing concept, and to provide analysts as well as researchers with a conceptualization of clinical phenomena. The notion of testing may be especially useful for treatments that seem to stand still, for ongoing clinicians, and for analysts wanting to clarify countertransference reactions.

Control mastery theory

Coming from a psychoanalytic background, Weiss and Sampson first introduced their conceptualization of the psychotherapy process in 1986, and this later became known by the name control mastery theory. The theory is based on clinical observations of patients lifting their defenses when they feel safe enough to do so (Weiss, 1971). It was proposed that previously warded-off impulses emerge not because they have broken through the patient’s defense but because the patient’s ego has judged it is safe enough to bring forth previously warded-off material (Weiss, 1971). The concept of testing the environment for safety was further studied by Weiss, Sampson, and the San Francisco Psychotherapy Research Group (SFPRG; previously the Mount Zion Research Group), who conducted several single-case studies (e.g., Silberschatz, 1986; Silberschatz, Sampson, & Weiss, 1986; Weiss & Sampson, 1986).

As a theory of psychopathology, rather than a technique or school, CMT postulates that early aversive relational experiences are transformed into pathogenic beliefs (Weiss, 1990b). They are fueled by an unconscious interpersonal guilt that has arisen out of the child’s need to protect important ties to early objects. In that sense, children cope with aversive childhood experiences by developing theories about their circumstances and origins, which are called pathogenic beliefs due to their irrational and self-blaming nature. Also referred to as pathogenic adaptations, pathogenic beliefs encompass feelings, attitudes, self-concepts, and moral judgments as well as repetitive actions and relational patterns (Rappoport, 1996b, 1997).

Weiss (1990b) specifically stresses the significance of the child’s need for safety, and asserts that the ego unconsciously evaluates the environment for signs of danger, which influences how a person interacts. Developed from an adaptation to traumatic relational experiences and internalized guilt, pathogenic beliefs persist because they convey a sense of safety (Weiss, 1992). As they were once necessary to secure bonds (Rappoport, 1996b, 1997), these hypotheses about oneself, others, or relationships continue to guide the individual’s perception, experiencing, and behavior. From the patients’ perspective, acting contradictory to these beliefs will likely lead to retraumatization or to losing important love objects (Weiss, 1993). Consequently, safety is perceived as a precondition for the emergence of repressed material.

However, the child, and later the adult, is exposed to an internal conflict between perceived dangers that pathogenic beliefs foretell and unconscious goals arising from a healthy development (Weiss, 1990a). Crediting the patient with agency and an unconscious wish to get well, Weiss (1994) posited that patients unconsciously plan their analytic work. They make efforts to move towards their goals by probing the analyst and by making conflict prominent in the therapeutic relationship based on an unconscious judgment of safety. In other words, patients are understood as having a certain control over their unconscious mental life and as being motivated to give up, and indeed capable of giving up, constricting pathological ideas, if they feel it is safe enough to do so (Weiss, 1971), hence the name control mastery.

Origins of the theory

Freud and the classical theory of psychoanalysis: repression, resistance, and repetition compulsion

Constructing their theory, Weiss and Sampson (1986) built on selected aspects of Freud’s later writings on the unconscious ego and superego (1923, 1926, 1940), on his view of defense as avoiding danger (1926), and on patients having some control over their repressions (1940). Nevertheless, in contrast to classical psychoanalytic theory, which conceptualizes repression as a defense against an impulse, and enactments as a way to maintain repressions, Weiss (1993) argues that repression is driven by guilt over potentially endangering oneself or others, and by a desire to protect the bonds to attachment figures. Repression as well as defense is thus posited as an adaptive accomplishment that provides the individual with security at the cost of inhibition and constriction. Likewise, within CMT, resistance is viewed as a person’s attempt to secure safety (Rappoport, 1996a).

In general, unconscious processes regulated by the pleasure principle are neglected in CMT in favor of a higher mental functioning hypothesis of the psychic apparatus (Weiss, 1993). The single drive that is acknowledged in CMT is the “instinct of mastery” that Freud (1920) ascribed to the boy in “Beyond the pleasure principle” as one possible motive for playing with the appearance and disappearance of an object that stands for his mother. This link to Freud seems to be the basic foundation on which Weiss and Sampson built their concept of testing. Furthermore, they take up the notion of Freud endowing patients with an unconscious motivation to master their traumas (1920) and a wish to solve problems (1926). Likewise, CMT has recourse to Freud’s (1940) understanding of the analyst as forming a pact with the patient’s ego and assisting in it, gaining back mastery over its mental processes. The aim of treatment according to CMT is to reach a larger amount of control over unconscious defenses and make them serve the ego.

For their concept of testing, Weiss and Sampson (1986) furthermore use Freud’s understanding of “experimental actions” (Freud, 1940, p. 199) and of “acting out” with the analyst (Freud, 1914, p. 151). In accord with Freud, the patient is ascribed with reproducing “what he has forgotten and repressed  …  not as a memory but as an action; he repeats it, without, of course, knowing that he is repeating it (p. 150, emphasis in the original). Nevertheless, Weiss and Sampson reject Freud’s (pp. 150–151) conceptualization of the compulsion to repeat as a resistance against remembering the past in favor of their mastery hypothesis of repetitions (e.g., Sampson, 1992).

Repetition compulsion for Freud was in no way therapeutic; for Ferenczi (1932), however, it had the function of resolving trauma by managing stimuli that rush in from the outside. In describing the devastating psychic shock of traumatic experiences, he posited that an overwhelming experience cannot be processed and is thus encapsulated in the mind. Via repetition, the patient may acquire this alienated experience that was never remembered and never thinkable. Because a memory of it was never generated, the patient must repeat a traumatic experience to make it perceivable for the first time.

Similarly, CMT asserts that patients repeat certain maladaptive patterns in an unconscious attempt to master situations. They aim to relinquish their underlying pathogenic beliefs, which cause them suffering and constriction (Sampson, 1992). Since these hypotheses about the interpersonal world are maladaptive and use up psychic energy (Rappoport, 1996b), which makes them burdensome, they are likely to become manifest in the therapeutic relationship, where their validity is examined (Weiss, 1993). In essence, testing can be conceptualized as expressions of unconscious ego-functioning regulated by a safety principle. The ego must find a compromise between a person’s own drives and the needs and wishes of others (Foreman, 2018). CMT thus supplements classical theory that focuses on conflicts between id impulses and superego prohibitions by viewing psychopathology as a conflict between individuation impulses and attachment needs (Varga, 2006).

Object relations theory and important psychoanalytic concepts

Transference

Drawing on the concept of transference, CMT presumes that patients bring past patterns of object relations into treatment and assign a certain role to the analyst. Transference for Freud (1905) was mainly a transferred image of an object and a displaced affect (1912) from this object of the past to a current object or, to put it succinctly, a displacement of one person for another. Klein (1952) broadened the concept by postulating that what are transferred are total situations including internal and external object relations, and also unconscious elements, such as anxiety and defenses. These concepts of transference are adopted in CMT in that patients are presumed to project internal representations from prior relationships to current ones. However, CMT conceptualizes transference as actualizing the past in the therapeutic relationship with the unconscious purpose of disproving pathogenic beliefs, which has been termed testing (Weiss, 1990a). From the CMT perspective, transference is understood as an unconscious yet active strategy to reflect on past relational experiences.

Projective identification

Foreman (1996) drew an explicit connection between Ogden’s (1979) conceptualization of projective identification and one type of testing in which patients elicit certain feelings, thoughts, and reactions in the analyst that stem from the patients’ experiencing. Klein (1946) introduced projective identification as a defense mechanism that entails the patient’s phantasy of splitting off aspects of the self, projecting them into an object, and then identifying with it. Taking his definition a step further, Bion (1959) conceptualized projective identification interpersonally. Rather than patients merely seeing certain aspects of themselves in an object, Bion posited that, due to the projection, the object would feel manipulated and pressed into acting a certain way – a postulation continued in CMT.

Ogden (1979) defined projective identification as a psychological process with which patients defend themselves from unwelcome feelings or impulses. Projective identification may furthermore constitute a primitive form of relating to others and of communicating troublesome affective states by inducing these feelings in the other. Finally, projected emotions are a “pathway for psychological change” (p. 362) in that patients reinternalize what the analyst has previously processed in a more mature way than the patient would have been able to.

CMT seems to elaborate on this definition by postulating that patients project to gain mastery over pathogenic adaptations when they employ tests. More specifically, patients unconsciously observe how analysts react to their devaluation, guilt-induction, or threats, with the aim of internalizing the analyst’s way of handling the experience. Furthermore, patients unconsciously try to detect how the therapist views them (Rappoport, 1996a), and they attempt to learn something about the analyst that may enable the patients to relinquish pathogenic beliefs (Sampson, 1992).

Countertransference

The term countertransference is used inconsistently in the analytic literature. Freud (1910) subsumed under this term mostly analysts’ unconscious reaction to patients, which should be overcome. Ferenczi and British object relations theorists drew on Freud by placing significance on the analyst’s unconscious as a receiver for the patient’s unconscious. The Kleinians came to include under this concept the emotional reactions induced in the analyst by the analysand’s projective identifications. Heimann (1950) initiated a new totalitarian perspective of countertransference encompassing all feelings towards patients, including subjective emotional reactions, which could be called the transference of the analyst. She also conceptualized countertransference as created by the patient, and consequently postulated it an instrument for the analytic work.

The complementary countertransferences (Racker, 1957) are the emotional reactions in analysts that patients evoke in them by means of projective identification (Bion, 1959). The analyst’s impulses and feelings are in accord with how the patient relates to the analyst as a projected internal object (Racker, 1957). Sandler (1976) suggested that some patients induce the analyst to play certain roles. Put differently, the patient creates countertransferences and the analyst will feel pressured or manipulated to play the role the patient has stipulated. Likewise, in testing, patients unconsciously elicit a powerful emotion (Weiss, 1993) in the analyst, who in turn will feel a certain pull to respond (Gazzillo et al., 2019), which leads to the conclusion that the concepts of projective identification, role induction, countertransference, and testing are interrelated. Furthermore, within CMT, complementary countertransference is adopted as a means of detecting tests and ascribing meaning to them (see the Discussion section). A relational view of countertransference as co-constructed in the therapeutic dyad is, however, not incorporated in the original theory of CMT. What the analyst contributes to the patient’s transference is thus not explicitly taken into consideration.

Introjected objects

Moreover, CMT is connected with object relations theory by asserting that the internalization of early relational experiences dictates how the individual will act in future relationships. It could be argued that pathogenic beliefs constitute self and object representations. While CMT clearly emphasizes real experiences (e.g., Sampson, 1992), Klein’s (1946, 1959) unconscious phantasies² about an object are neither specifically included in CMT nor denied.

The work of Fairbairn, who views psychopathology as rooted in an identification with bad objects, is, however, continued in CMT (Sampson, 1992). Building on Klein’s model of psychic reality with the internalization of external objects that are split off into good and bad part objects, Fairbairn (1952) emphasized the motives for internalization in a desire to understand the world. Observing that children tend to ascribe guilt to themselves³ rather than to their environment, Fairbairn concluded that “taking upon himself the burden of badness,” the child assures that the object remains good and the environment continues to convey a sense of security, which the child is naturally in need of (p. 67). It is “better to be a sinner in a world ruled by God than to live in a world ruled by the Devil,” Fairbairn put it pointedly (p. 67). Whereas Klein (1946) postulated the internalization of both good and bad objects, Fairbairn claimed that primarily and initially bad objects, which are unsatisfying in a libidinal way, are internalized (1952), which he termed “moral defence against bad objects” (p. 18). In therapy, the patient may identify with any split-off aspect of objects or of the self, which have resulted from the splitting of the object. Finally, patients may project any of these aspects on to the analyst (Fairbairn, 1952).

Another overlap with Weiss’s (1990a) concept of pathogenic beliefs can be found with Winnicott’s (1960) conceptualization of “a compliant False Self” that is built from the internalization of the environment and “reacts to environmental demands” (p. 146). Put succinctly, if a child has to react to environmental demands, they cannot act according to their own spontaneous impulses and thus cannot develop a True Self. Furthermore, CMT adopted Winnicott’s assumption of an adaptive purpose of pathological behavior and experiencing. It remains to conclude that the False Self will become manifest in the patient employing tests.

More explicitly, Sampson (2007) drew on Winnicott in presuming that patients are motivated to come to therapy due to an internal caretaker, which in CMT is conceptualized as the patient’s unconscious plan. Winnicott (1960, p. 142) used the term “caretaker self” for his patient to describe an aspect of a person which led her to psychoanalysis and to hand “over its function to the analyst.” She desperately longed to feel real and existing, which are aspects of the True Self having been replaced by a False Self that reacts and complies. While Winnicott (1960, p. 142) discerned that patients “test the analyst’s reliability” as an object, the concept of testing pathogenic adaptations to the environment is fully developed foremost by Weiss and Sampson (1986).

The reception of CMT

While praising the direction Weiss took in adapting psychoanalytic technique, Wachtel and DeMichele (1998) criticize Weiss’s idiosyncratic articulations of testing and planning, which they suggest lead to a guarded interest within the psychoanalytic community. Their main critical point was Weiss’s assumption of patients’ plans for treatment; his proposition that patients have a certain prescience on how to get better remains a point of contention. Instead of referring to therapists deducing the patients’ plan, Wachtel and DeMichele suggest that therapists infer unconscious fears or expectations.

Yet, regardless of whether patients are attributed with unconsciously planning and striving for developmental goals, clinical evidence of a certain interactional phenomenon can be found – be it rather purposeful (as Weiss assumes) or be it in the simple expression of conflict (as Wachtel and DeMichele suggest). For example, patients may test a pathogenic belief dictating them to avoid independence; however, patients may simply both wish to become more independent and fear the consequences of their independence (i.e., losing important relationships, hurting, or even annihilating others). The analysis of vignettes illustrating a clinical phenomenon nevertheless remains worthwhile.

Transference tests and passive-into-active tests

In the original theory, two types of test, namely transference tests and passive-into-active tests, were conceptualized. Rappoport (1996a, 1997) further distinguished between transference tests by compliance and by noncompliance in regard to what the patient believes the therapist is expecting from them.⁴ In this manuscript, we use the word mode to classify each test according to this categorization. It is important to differentiate between test modes because they contribute to understanding a patient and give hints on how to respond to them.

In treatment, patients may experience themselves in a similar way to the way they did in their past, and may show a comparable reaction to the therapist that they did to early attachment figures. When using transference tests by compliance, patients assess whether therapists seem gratified when patients attend to the therapists’ presumed needs. If they learn that their adaptations are neither required nor desired, they can start to let them go (Rappoport, 1996a).

Nevertheless, patients may also break their patterns in the therapeutic transference and experiment with new attitudes or new ways of relating. Thus, a noncompliant transference test aims to assess whether the therapist is threatened, upset, or critical when patients defy the therapist’s presumed expectations (Silberschatz & Curtis, 1986).

Besides transference tests, patients may employ passive-into-active tests when they are potentially hurtful or threatening to analysts by treating them in the same or in a similar way to how the patient was treated as a child (Sampson, 1990), thereby testing whether therapists will feel safe enough to protect themselves from the tests’ distressing effects (Rappoport, 1997). Patients hope that will therapists maintain their stance and even challenge the irrationality of the patients’ injurious behavior (Rappoport, 1997). They long for a response contrary to the one they have shown as a child and aim to identify with the analyst’s strength, assertiveness, or resilience (Sampson, 1991).

Conceptualizing the treatment process

Generally, testing is postulated to drive the psychotherapeutic process. Nevertheless, after a test has been passed, the patient does not automatically relinquish pathogenic beliefs, but a repetition of this pattern is necessary; therefore, patients test throughout the treatment process (Silberschatz, 2008) up to the final session (Kealy, Gazzillo, Silberschatz, & Curtis, 2020). Due to themes of rejection, dependency, loss, and guilt being evoked, the termination phase of treatment may be particularly susceptible to testing (Kealy et al., 2020). When analysts continuously fail tests, patients tend to stagnate or deteriorate (Silberschatz & Curtis, 1993) – a setback that could lead to them quitting treatment (Weiss, 1993).

In general, clinicians working with CMT principles hypothesize which pathogenic beliefs patients have, how they reactivate them in the transference, and how they probe the analyst to change them. By means of a preliminary plan-formulation, analysts generate hypotheses about possible tests and their potential meaning, which may also enhance the therapist’s awareness of potential countertransferences (Rappoport, 1996a). Based on their tentative hypothesis about a patient, clinicians work on increasing the patient’s safety and provide responses and interpretations that help in the transformation of pathogenic beliefs (Rappoport, 1997).

More specifically, therapeutic responses should allow patients to conclude that it is safe to give up pathogenic adaptations and that maintaining them is beneficial neither for the patients themselves nor for their environment. However, insights will not occur until the patient feels safe enough to have them (Rappoport, 1997). As the perception of safety is highly individual, CMT stresses the importance of case-specific responses and disregards the possibility of generically correct responses (Rappoport, 1997). So far, no systematic exploration of either testing or responding to tests has been performed; the current analysis, therefore, fills this gap.

The objective of the study

The current work examines empirical articles, illustrating treatments performed or analyzed using the approach of CMT, to provide a comprehensive overview of unconscious testing activity that psychoanalysts might encounter in clinical practice. The focus of the current manuscript lies in what patients bring into treatment, which will be the basis for future research on how to respond to these tests. We first examine when and how testing occurs, and then group case vignettes around different concepts to illustrate the complexity of testing. The in-depth analysis of various manifestations of testing aims to determine why a particular patient employs a specific test during a certain time in treatment.

By systematically collecting case examples of previously published studies, the current work extends the review of testing by Gazzillo and colleagues (2019a). In the present study, test examples are systematized according to test modes to obtain the potential underlying meanings they have for an individual patient. In comparison to Gazzillo et al. (2019), who linked the successful passing of a patient-initiated test to process parameters and outcome variables of the treatment, the current work focuses on the content of testing. Furthermore, by examining treatment progression within single patients, their individual testing strategies and changes therein can be determined. Systematically examining the whole process of therapy of several individual patients allows us to draw new conclusions that may have implications for clinical practice. Finally, we explore similarities and differences between patients who seem to be testing the same underlying pathogenic belief. By drawing connections between cases, this work furthermore extends the existing research on CMT.

Method

Searching the databases Ovid MEDLINE, PsycArticles, PSYNDEX, and PsycINFO for the term “control mastery theory” in titles, abstracts, and full texts, a total of 179 articles were obtained. Because the term “control mastery theory” was not explicitly mentioned in some of the earlier works, 11 publications were manually added from the homepage of the SFPRG. Included were empirical peer-reviewed publications in English or German examining individual psychoanalytic therapy or psychoanalysis on patients who were at least 18 years of age and had any disorder or difficulty. Treatment must have been performed by a single psychotherapist or psychoanalyst. If not conducted under the CMT approach, cases had to be rated using CMT principles. Patients must have tested, as conceptualized by CMT, as eliciting a response in the therapist and affecting the therapy process in some way (e.g., Silberschatz, 2008), and the content of testing must have been described.

We excluded publications if they examined solely pathogenic beliefs or guilt without mentioning testing or focusing exclusively on the therapist. Furthermore, expert opinions based on hypothetical cases or on expected but not actually documented tests were excluded. This process yielded 29 articles that were used to extract and analyze test sequences. Single case reports that portrayed the entire therapy process (Fretter, 1995; Gootnick, 1982; Pickles, 2007; Pole & Bloomberg-Fretter, 2006; Sammet, Brockmann, & Schauenburg, 2007; Sampson, 2007; Shilkret, 2002, 2008; Shilkret & Shilkret, 1993; Silberschatz & Curtis, 1993; Varga, 2006) gave a clear description of the patient’s demographics, history, and current clinical condition, as well as of treatment interventions and the post-intervention condition. Single case reports that provided only excerpts of treatments (Bugas & Silberschatz, 2000; Curtis & Silberschatz, 1986; Foreman, 1996; Gazzillo et al., 2019; Gazzillo, Silberschatz, Fimiani, De Luca, & Bush, 2020; Gazzillo, Dimaggio, & Curtis, 2121a; Gazzillo, Dazzi, Kealy, & Cuomo, 2021b; Gazzillo, Leonardi, & Bush, 2021c; Kadur, Flaig, Volkert, Sammet, & Andreas, 2018; Kealy et al., 2020; Rappoport, 1996a, 1997; Sampson, 1991; Silberschatz & Curtis, 1986; Suffridge, 1991; Weiss, 1992, 1994, 1995) offered only limited information on patients, interventions, and post-intervention conditions, while illustrating the content of testing. However, as the focus of the current work was to empirically collect instances where patients probe the analyst, all test examples, as they naturally occur in clinical practice, warranted inclusion.

The literature search was conducted independently by the first two authors. Test examples were gathered of all included studies, while case vignettes for in-depth analysis were selected by the first author to illustrate the complexity of the testing concept. Tests were identified by the original authors, who used case notes or verbatim transcripts. When not provided, but clearly implied and thus identifiable via the context, the first author performed test mode inference and elaborated on further interpretations and conclusions.

Results

Occurrence of testing

In general, tests can be categorized into narratives, self-presentations, interactions, and use of the setting. Identified as tests were dramatic provocative expressions, such as voicing criticism (e.g., Silberschatz & Curtis, 1986) or threatening to quit treatment (Shilkret, 2002), as well as less spectacular behaviors, such as repeated statements that transpose an attitude about oneself (e.g., Sammet et al., 2007) or invite a certain view of oneself (e.g., Shilkret, 2008). Another form of subtle testing can be the patients’ narration about themselves (e.g., Shilkret, 2008) or others (e.g., Pole & Bloomberg-Fretter, 2006). Sometimes, patients assign a value to their narration. A patient described by Pickles (2007), for example, talked about difficult childhood experiences while denying their abusive nature and then started to question her memory.

Tests via narration and self-presentations

Patients unconsciously observe the analyst’s reaction not only, to their narration, but also to their self-presentations mostly revolving around guilt (e.g., Pickles, 2007) and autonomy (e.g., Fretter, 1995). Some patients may test by denying previous goals and expressing doubts about them (Weiss, 1994). Other patients exhibit persistent negative circular thinking to determine whether therapists agree with the patient’s perceived pathological guilt that justifies their ruminating (Gazzillo et al., 2021c). Moreover, patients who perceive themselves as burdensome may test whether they are worthy of attention by presenting long and complex dreams to test whether the therapist remains interested (Bugas & Silberschatz, 2000; Gazzillo et al., 2020). Patients may likewise test whether the analyst will force them to work hard and not enjoy leisure time or whether the analyst will protect them from self-destructiveness (Weiss, 1994).

From the patients’ perspective, testing can be stressful since they do not know whether therapists will pass their tests (Bugas & Silberschatz, 2000). Therefore, some patients set up testing situations by giving a relevant history or by attuning therapists and providing instructions to facilitate therapists passing their tests, which is termed coaching (Bugas & Silberschatz, 2000). Patients may coach directly by warning therapists about their habitual behavior or by giving instructions, but may also coach implicitly, through contradictive behaviors, exaggerations, or a narrative about an incident outside therapy that illustrates what the patient would find repellant or frightening (Bugas & Silberschatz, 2000). Similarly, patients who are hesitant to bring up directly what they want to discuss may instead present a dream⁵ hinting at a certain topic (Gazzillo et al., 2020).

Tests via interaction

Some patients test by directly enacting an early relationship and casting the analyst in a role, such as the position of the inadequate father or servile husband (Varga, 2006). Patients may also be dependent or needy (Pickles, 2007; Silberschatz & Curtis, 1993) or may be overly accommodating (Rappoport, 1996a) in trying not to upset or hurt the analyst (Weiss, 1992). They may furthermore entice the therapist to take the lead (Rappoport, 1996a), ask for advice (Weiss, 1995), and demand information (Pole & Bloomberg-Fretter, 2006). One patient, for example, asked for the clinician’s opinion about the patient’s ability to return to work, thereby testing whether his doubts about being ready had merit and whether he deserved the therapist’s care (Kadur et al., 2018). Similarly, one patient tested her pathogenic belief of not deserving the analyst’s time by complaining about the late beginning of sessions (Weiss, 1992).

Other types of interaction tests can be found when patients reject analysts and invite their rejection of the patient. To illustrate in detail how an interaction test may be presented, consider the following test sequence occurring at the beginning of a treatment. A patient was inviting rejection by mentioning that she had seen another clinician and that she was not sure with whom to start therapy. The therapist’s hypothesis was that the patient might have been driven by the underlying pathogenic belief that she did not deserve to receive care. When the therapist explicitly encouraged her to show up to the next session with her, she was challenging the patient’s pathogenic belief (Shilkret, 2002). In the subsequent session, the patient went on to praise the other clinician and said she liked that one better. By making it very easy and reasonable to encourage the patient to go to the other clinician, the patient invited her therapist to reject her (Shilkret, 2002).

Tests via the use of the setting

Finally, patients may use the setting for testing by missing (Gootnick, 1982; Suffridge, 1991), canceling (Shilkret, 2008), or rescheduling (Pole & Bloomberg-Fretter, 2006; Varga, 2006) sessions; by coming late (Gootnick, 1982) or trying to extend the session (Bugas & Silberschatz, 2000); or by delaying or refusing payment (Foreman, 1996; Suffridge, 1991; Varga, 2006). They may also threaten to quit therapy (Gootnick, 1982), actually quit therapy (Shilkret, 2002), inquire about treatment termination (Kealy et al., 2020; Rappoport, 1997; Silberschatz & Curtis, 1986; Weiss, 1994), or ask for an extended treatment (Sammet et al., 2007).

Test examples according to test mode

Each of these ways to probe the analyst may be used in either of the previously described modes of transference tests or passive-into-active tests. Some patients tested their pathogenic beliefs in compliant mode by presenting themselves as undeserving (Shilkret, 2002), fragile (Silberschatz & Curtis, 1993), overly attentive (Sammet et al., 2007), or unexpressive (Rappoport, 1996a, 1997) – unconsciously hoping that analysts would not require these self-presentations and expressions.

In contrast, patients mentioning their achievements (Fretter, 1995) or demonstrating their intelligence (Bugas & Silberschatz, 2000) were testing in a noncompliant transference mode whether the therapist would be threatened by the patients’ accomplishments or punish them for their independence. Similarly, posing questions and taking the initiative (Silberschatz & Curtis, 1986), attending to their own needs (Rappoport, 1996a), and disagreeing (Pickles, 2007; Weiss, 1994, 1995) were examples of transference tests by noncompliance via interaction for the particular patients described in these studies.

Patients who acted as demanding (Gootnick, 1982; Varga, 2006), critical (Silberschatz & Curtis, 1986), or guilt-inducing (Pickles, 2007; Shilkret & Shilkret, 1993), thereby eliciting stronger countertransference reactions in the therapist (Rappoport, 1996a), were mostly testing using passive-into-active tests. Other illustrations of this mode are complaining or doubting the analyst (Foreman, 1996). A patient presented by Bugas and Silberschatz (2000) continued his narration after the session ended and showed signs of relief when the therapist passed his test by setting limits and challenging unreasonable demands.

Gazzillo and colleagues (2019a) further expanded the theory on testing by adding noncompliant passive-into-active testing, which is illustrated through a case vignette. By telling her therapist he should be proud of what he accomplished with her in therapy, a patient identified with the perspective of her mother (i.e., passive-into-active), yet she acted contrary (i.e., noncompliant) to her pathogenic belief of not being entitled to be proud of herself. When the therapist put the focus on her achievement, rather than being proud himself, he actually confirmed her pathogenic belief that one should not be proud of an accomplishment.

In-depth analyses of tests – getting to the meaning of tests

Differences between patients

Testing is complex and may have different meanings across patients. For example, a patient voicing criticism about their analyst may be testing, in transference by noncompliance mode, by experimenting with being self-directed and independent (Sammet et al., 2007) or, in passive-into-active mode, by trying to make the analyst weak and devalue them, as the patient was disparaged in childhood (Pickles, 2007). Similarly, some patients presenting themselves as fragile tested compliantly whether therapists would require them to be weak (Rappoport, 1996a; Silberschatz & Curtis, 1993), whereas one patient tested noncompliantly whether he was entitled to be weak (Shilkret, 2008). Furthermore, this patient was canceling and missing sessions during times in treatment when he felt particularly burdensome (i.e., testing compliantly), while other patients test in passive-into-active mode by missing sessions to put therapists in a situation over which they do not have control (Suffridge, 1991). They may also test this way to determine whether the analyst would worry about the patient, as the patient described by Gootnick (1982) felt he had to worry about his father. The analyst understood his use of the setting as an invitation to worry and to feel guilty and thus stopped reminding him of sessions or discussing his tardiness (Gootnick, 1982).

Similarly, patients may have different pathogenic beliefs underlying the same testing behavior of inquiring about treatment termination. One patient tested his pathogenic belief about being unworthy of treatment and not having anything interesting to offer, whereas a different patient brought up the topic of ending therapy as a way to devalue the treatment and the therapist (Kealy et al., 2020). Patients may invite the therapist to reject them and to accede to termination in the hope that the therapist would continue to treat them (Kealy et al., 2020). Conversely, a patient who had been working on becoming more autonomous and reducing feelings of guilt for pursuing her goals tested by being reluctant to end therapy. Instead of taking her initial reluctance and regret to end treatment at face value, the clinician understood it as a test of whether she was entitled to follow her own direction (Kealy et al., 2020).

Differences within a patient during the course of treatment

The underlying meaning of tests may differ across patients but also within a patient. In-depth analyses of case reports lead to another clinically relevant result: a patient may simultaneously test two different themes within a test using different test modes, which call for opposing therapist responses. A patient’s wish to extend stationary therapy and a subsequent suicide threat (Sammet et al., 2007) could be viewed from two different perspectives. It was assumed that she tested to see whether the clinic personnel would take care of her (i.e., transference test by noncompliance) rather than whether they would set boundaries, as she had failed to do when she was a child (i.e., passive-into-active test). Consequently, when the therapist was not giving in to the patient’s wish under her threat, the raters evaluated this intervention as confirming her pathogenic belief of not deserving treatment. However, there was some evidence pointing to the passive-into-active aspect of her test. Even during the crisis, her self-efficacy levels increased, and she seemed less dependent, which was interpreted as a reaction to the therapist setting boundaries and refusing to take on irrational guilt (Sammet et al., 2007).

Furthermore, Gazzillo and colleagues (2020a) found that some patients rapidly switch test modes within one testing sequence in a given session. In other words, what appears to be a single test may actually be a composition of tests, all of which call for different interventions. For example, a severely traumatized patient with a personality disorder asked to skip some sessions to focus on an exam and then accused the therapist of wanting to get rid of her when he allowed her to miss sessions. Subsequently, she became very aggressive and blamed him for tormenting her. The therapist conceptualized this sequence as a composition of three tests, starting with a transference by noncompliance test (revolving around being allowed to accomplish something), followed by a transference test by compliance (circling around being worthless and a burden), finally leading to a passive-into-active test (trying to figure out how to handle being accused and humiliated).

Another patient presented by Gazzillo and colleagues (2019a) tested two diverse pathogenic beliefs in two different test modes by presenting the same testing behavior at different times during treatment. By repeatedly calling her therapist during vacation times, she initially tested in compliant transference mode, challenging her pathogenic belief that her needs were burdensome. As therapy progressed, she continued to call but shifted to passive-into-active mode when she projected the pathogenic belief of having to care for others onto the therapist.

Discussion

This in-depth exploration of case vignettes was executed to provide an overview of the various clinical manifestations of patient testing, which was defined as an unconscious transference enactment that is performed to disconfirm pathogenic beliefs (Weiss, 1990b). Testing occurs directly when patients interact with analysts or use the setting to create this interaction. Other tests, such as patients presenting themselves a certain way or disclosing a narrative, are more subtle.

Concerning test modes, the in-depth analysis of test examples confirms previous assumptions by Rappoport (1997) that compliance tests are employed most frequently in the early stages of treatment because they are the safest. Noncompliant testing constitutes the more courageous mode since the patient acts in a way presumed undesirable for the therapist, and thereby risks retraumatization or the loss of emotional bonds (Rappoport, 1997).

Shilkret (2008) further postulates that severely impaired patients may need longer to develop the necessary sense of safety to test more boldly. They may also use a variety of test modes (Pickles, 2007) or switch modes very rapidly (Gazzillo et al., 2021b). Similarly, patients who suffer from severe traumatic childhood experiences will generally feel less safe in therapy because of their lack of reassurance that they deserve good treatment (Shilkret, 2008). These patients may prefer passive-into-active testing. By being injurious themselves and avoiding assuming a vulnerable position, they reduce the risk of being hurt and retraumatized.

Nevertheless, Rappoport (1997) also argues that patients must have a compelling reason to use passive-into-active testing and that they will outweigh presumed risks with informational gains, which might not be attained otherwise. Consider a patient who needs to refute beliefs about criticism and its consequences who cannot easily test these in the transference relationship because competent therapists will not likely be seduced into criticizing;⁶ therefore, to work on this issue, the patient has to use passive-into-active testing by being critical themself (Rappoport, 1997).

Differences between patients

Other pertinent results of the case analysis were that tests are flexibly performed and work in a variety of ways. One single test may have several different meanings across patients and may be motivated by different underlying pathogenic beliefs. For example, by experimenting with his emotional experiencing, a patient presented by Shilkret (2008) was trying to oppose his belief of not being entitled to be weak rather than acting in accordance with it, as a restrained patient would (Rappoport, 1996a). For this particular patient, challenging his fragility would have been detrimental (Shilkret, 2008), whereas for a restrained patient, this intervention would help to refute a pathogenic belief (Rappoport, 1996a).

Likewise, one patient who threatens to quit can be employing a passive-into-active test, assuming the role of the traumatizing parent, threatening to remove themself from the therapeutic relationship. This patient could hold the belief that others will abandon them, as they are abandoning the therapist. A different patient, however, may show the same behavior, but test whether the therapist will bring the patient back due to the pathogenic belief that they do not deserve treatment. Similarly, declining a request for prolonged treatment was determined to confirm one patient’s pathogenic belief of not deserving care (Sammet et al., 2007). However, terminating on time for two different patients who asked for an extension disconfirmed their pathogenic beliefs of hurting others by being independent, and actually supported their striving for autonomy (Kealy et al., 2020; Silberschatz & Curtis, 1986).

Differences within a patient

The complexity of testing becomes apparent not only between patients, but also within a patient. A patient’s single pathogenic belief of not deserving to fulfill their own wishes can be tested in various ways during the course of treatment: (1) compliantly by being overly accommodating; (2) noncompliantly by being assertive and insisting; or (3) passive-into-actively by denying the therapist any gratification of their wishes. In the newer perspective, the patient may furthermore use (4) compliant passive-into-active testing by encouraging therapists to fulfill their wishes – something that patients would have desired for themselves in early childhood (Gazzillo et al., 2019).

In contrast, one single testing behavior of a patient may have different underlying themes (see the case by Sammet et al., 2007). Moreover, what seems to be one test may actually be a composition of several smaller tests. Furthermore, the test may change its theme in the course of therapy and may thus call for opposing therapeutic responses. Because the patient described by Gazzillo and colleagues (2019) employed similar tests of calling her therapist with different underlying pathogenic beliefs, the therapist might have missed her switching modes. While in the initial phase of treatment giving in to her requests was determined to oppose her pathogenic beliefs of being burdensome, later in therapy, however, the same therapist reaction was confirming her pathogenic belief of irrational responsibility for others. Therefore, analysts should be cautious of being confronted with a different test mode that may call for a different therapeutic intervention than the previous one. In fact, testing appears to change dynamically when patients make progress, calling for therapists to be attuned and keeping the dynamic in mind.

Implications for clinical practice

The current work focuses mainly on the clinical phenomenon of testing; how the analyst actually responds to these tests is largely determined by the clinician’s theoretical orientation. The theory does not prescribe a certain technique but illustrates therapeutic responses in various case vignettes. In general, it seems that therapists have a continuum of possible responses. Fretter (1995) claims that patients can be treated with a sound case-specific plan-formulation that yet allows individuality in regard to therapeutic style.

In testing, patients unconsciously evaluate the reality basis for the dangers their pathogenic beliefs foretell, which could be conceptualized as a form of reality testing (Freud, 1940). In line with Freud’s structural model, CMT also aims for patients to gain ego strength and to enhance reality testing, thereby achieving insight into their illusions and repetitions. Similarly, Winnicott (1969) describes the achievement of distinguishing self from other as moving from relating to subjective objects to using objective objects, which could be perceived as another way of reality testing.⁷ His proposed emancipation from the False Self as curative element is similar to the relinquishment of pathogenic beliefs in CMT. Both are accomplished by enabling the patient no longer to have to comply or, as Winnicott (1960) put it, not having to self-manage by reacting to demands.

Even though not acknowledged, the relinquishing of pathogenic beliefs, which are emotionally laden assumptions rather than mere cognitive thoughts, is also in agreement with Klein’s (1963) view of acquiring insight into psychic reality (p. 279) and inner processes (p. 294). The goal of treatment is to become conscious of phantasies and ultimately to modify these thoughts pervaded with emotions (Klein, 1946, p. 107). By letting them go, the patient can become more in touch with reality, which is essentially the aim of treatment in CMT.

While CMT and transference-focused therapy (TFP; e.g., Kernberg, Yeomans, Clarkin, & Levy, 2008) both emphasize interpersonal interactions and countertransference development, these two approaches have more differences than similarities as they unfold in a session. First, they are based on different models of psychopathology. Whereas CMT entails the assumption that the internalization of traumatic early object-relations may lead to a distorted view of reality, and consequently to psychopathology, TFP is based on Kernberg’s structural theory of the interdependence of affective disposition and early interactions with others. In order to avoid psychopathology, the individual must clarify the borders between self and other, overcome splitting, and balance libidinal and aggressive impulses. Therefore, TFP therapists focus on primal phantasies more than is the case for therapists operating within the framework of CMT. Overall, clinical practice in the former is characterized more by technical neutrality, transference analysis, and reconciling good and bad object representations. The aim of treatment is to integrate split-off aspects of the personality that Kernberg presumes patients express consecutively and alternately (Kernberg et al., 2008). In showing that the patient identifies with both roles, and in developing tolerance for ambiguity, integration can be achieved. CMT has similar visions for a successful treatment, but is open to supportive interventions.

The description of the clinical phenomenon of patients establishing a certain interaction in the therapeutic dyad of both approaches however, seems congruent. For example, patients placing themselves in the role of the powerless child and the therapist in the role of the indifferent caretaker (Kernberg et al., 2008), could be described as compliant transference testing in CMT and as projection in TFP. The reversal of roles could be characterized as passive-into-active testing and as projective identification, respectively. Nevertheless, CMT assumes that patients’ tests have a purpose.

Moreover, CMT is connected with the relational school of psychoanalysis in that the cure is provided through the therapeutic relationship. Yet the proposition in CMT is not the sole provision of corrective emotional experiences (Weiss, 1992), as in self-psychology after Kohut or in intersubjective approaches. While agreeing with Alexander and French (1946) in placing importance on the notion of corrective emotional experiences, Weiss (1992) distances CMT from the idea of the imprudent use of adopting a role that is in opposition to that of the early pathogenic caretaker as this approach poses the danger of inauthentic role playing. In contrast, Weiss ascribes great importance to the analyst being guided by the unconscious testing of the patient and points out the necessity to reflect on countertransferences in determining how to react to a patient’s test.

It could be argued that CMT is in accord with the view of Sandler (1976), who alerts analysts to a type of countertransference being evoked by the use of projective identification (Bion, 1959). Concerning technique, Sandler suggests implementing a “free-floating responsiveness” (p. 45) to the patient’s induced roles that analysts should embody.⁸ Subsequently, the analyst should carefully observe the reaction of the patient to an intervention.

CMT aims to break patterns of maladaptive interpersonal relationships by providing new experiences that patients unconsciously seek and that enable them to absorb an interpretation (Weiss, 1992, 1994). Corrective emotional experiences are thus provided by passing the patient’s tests and by offering plan-compatible interpretations.

Perhaps CMT is also in line with Ferenczi (1928), whose emphasis shifts from the correctness of the interpretation to letting the patient have a healing experience without providing fast remedies or avoiding feeling any pain. A pioneer of developing psychoanalytic technique, Ferenczi (1932) ascribes great importance to the patient’s experience in the sessions rather than to the reconstruction of the past. While he assumes that the past influences the patient’s current experiencing, Ferenczi is convinced that the analysand must affectively live through an experience analogous to one from the past in order to absorb an interpretation. While Ferenczi is well known for his active technique, he stresses the importance of the patient being active (1928). Similarly, in CMT, the postulate is to encourage patients to test and thereby to repeat past constellations and object relations without the conditions of trauma in a safe and trustworthy environment. Ferenczi (1933) specifically encouraged the expression of negative transference feelings and began to challenge the traditional concept of resistance, by suggesting keeping an open mind and trusting patients in their feelings and criticisms.

Building on Ferenczi, Curtis & Silberschatz (1986) suggest that what commonly appears as resistance may very well be conceptualized as a test. Patients who hold themselves back, give short replies, bring no or few spontaneous associations, but are frequently silent may be testing by showing their inhibition to realize their own wishes (Curtis & Silberschatz, 1986). Similarly, Gazzillo and colleagues (2019b) interpret a patient’s refusal to talk about a certain topic as a test circling around the entitlement to be autonomous instead of being avoidant. On the other hand, Curtis and Silberschatz (1986) acknowledge that patients’ statements cannot necessarily be taken at face value but might contain an underlying meaning. Patients who present a concern they want to work on could in fact be testing to gain assurance that the therapist will not be fooled or overwhelmed (Curtis & Silberschatz, 1986; Gazzillo et al., 2019). Again, CMT is in congruence with Ferenczi, who advises the analyst should “keep one eye constantly open for unconscious expressions  …  and  …  bring them remorselessly into the open (1928, p. 93).

Pro-plan responses and the role of interpretation

CMT credits patients with an unconscious wish to get well and with unconscious long-term goals, which are subsumed under what Weiss (1994) conceptualized as the patient’s plan. Plan compatibility is thus a term used to describe the suitability and responsiveness of an intervention in regard to the particular patient’s problems and goals. Weiss (1992) claims that pro-plan responses help patients to carry out their plans and to pursue healthy developmental goals forbidden by pathogenic beliefs and guilt. The analyst’s task is guiding patients to an insight concerning the nature and consequences of their unconscious pathogenic beliefs. This is accomplished by allowing the patient to test, by providing pro-plan interpretations (Curtis & Silberschatz, 1986), and by the analyst’s pro-plan attitude (Weiss, 1994). Failing a test was furthermore defined as neglecting a key problem of the patient that the patient hinted at while focusing on something else they had said (Silberschatz & Curtis, 1993).

Weiss (1992, 1994) recognizes interpretation as an essential technique and considers the sole application of noninterpretative interventions to be partial and incomplete. Furthermore, he posits that, in some instances, the latter may be needed to form a relational foundation that makes patients more receptive to interpretations. For example, a patient who challengingly threatens to quit may improve most from an analyst who initially encourages them to continue the current treatment rather than making a referral to another clinician. After demonstrating that they will not reject the patient – which is what the patient is provoking – the analyst will then provide interpretations circling around rejection (Weiss, 1992).

Consequently, Weiss (1992) advises the analyst to take on an active role, thereby acknowledging taking sides with some of the patient’s impulses while discouraging others.

For example, to answer the patient’s test of whether the analyst would protect the patient from his own destructiveness, Weiss (1994) clearly set boundaries and clarified that the patient could not remain in treatment should he continue with his destructive behavior. While the patient initially reproached the analyst for failing to maintain an analytic attitude, he later allowed into consciousness childhood memories that entailed not being protected by his parents. Therefore, Weiss does not renounce the use of authority, and does not aim for protection of the patient’s autonomy at all cost. Since this patient was suffering from loyalty guilt and had to self-sabotage his success, he benefited from an analyst who was confrontational rather than neutral, which does not imply fulfilling patients’ wishes (Weiss, 1994).

On the other hand, for a patient who tested whether the analyst would induce him to work hard and to refrain from enjoying his leisure time, Weiss simply showed interest in whatever the patient brought up. Yet a different response is needed for patients who deny their goals or express doubts about them, hoping the analyst will allow them to have these goals (e.g., leaving home, getting a job). In these cases, the analyst should challenge the patient’s objection to previously voiced goals or plans (Weiss, 1994).

Besides interpreting, gaining insight, and challenging, CMT thus sometimes favors a more active technique in which therapists unburden and support. For some patients, Rappoport (1997) specifically points out the value of noninterpretative responses, such as encouragement or self-disclosure. Therapists may also give information, make comments, and display affect (Rappoport, 1997). Similar to Weiss, he emphasizes the importance of conveying an active interest by asking questions, making suggestions, as well as initiating interaction and recognizing the patient’s accomplishments (Rappoport, 1996a).

Finally, Weiss (1994) is in accordance with Winnicott (1960) and Ferenczi (1928) in advising waiting with an interpretation until the patient is no longer endangered by it. Up to this point, the analyst should pass the patient’s test by showing a pro-plan attitude. The analyst may also voice the pathogenic beliefs the patient holds and the struggle they have in relinquishing them (Weiss, 1994). However, it must be acknowledged that patients may benefit even more from a clinician who allows for ambivalence and who opens a reflective space by asking why patients think it is important to act a certain way, rather than from a clinician who resolves an underlying conflict for the patient by rapidly delivering a corrective experience. While Weiss (1992) asserts that insight is necessary for change, he also proposes that insight can be achieved by the analyst’s behavior and attitude.

Getting to the definition of pro-plan responses

Concerning interpretation, Weiss (1992) ascribes a greater importance to them being pro-plan than being tactful or complete. The remaining question is what a plan-compatible interpretation entails, which is illustrated in several case vignettes. For example, a patient suffering from the pathogenic belief of having to be accommodating and testing this by avoiding any criticism of the analyst would benefit more from interpretations that connect their behavior with the analyst’s perception of them attempting not to hurt the analyst. In contrast, Weiss (1992) claims that the patient could regard an interpretation that does not stress their unconscious attempt as the analyst’s request to be noncritical.

Likewise, when a patient repeatedly complained about the analyst starting sessions several minutes late, Weiss (1992) focused on her underlying guilt of her negative feelings towards him and her pathogenic belief that she deserved to be denied her entitled time. He advised that merely pointing out she was transferring the negative feelings of her mother onto him would not help her to refute pathogenic beliefs and would not support her plan to relinquish them. Weiss concludes that she was later able to bring into memory the fact that she sometimes provoked her mother to cut her short.

To provide plan-compatible responses for transference tests, therapists generally show that they do not expect the pathogenic adaptions that parents used to require (Rappoport, 1997). CMT posits that analysts should not be gratified by patients’ solicitous and accommodating behavior. Furthermore, it is beneficial when analysts recognize testing situations and are attentive to avoiding unconsciously undermining the patient unconsciously experimenting with new ways of interacting or experiencing themselves that were previously disregarded as causing danger. Rappoport (1997), for example, favors attributing a value to certain patient’s attitudes or behaviors during noncompliant transference tests. While agreeing with classical theory to frustrate patients’ demands during passive-into-active tests, CMT predicts that a frustration of impulses for patients using certain transference tests would support rather than disprove a pathogenic adaptation (e.g., Silberschatz et al., 1986). Therefore, CMT sometimes seems to strain away from abstinence and technical neutrality, thereby contradicting traditional theory.

The response to a passive-into-active test has been vividly discussed. Generally, this type of test should not be discouraged and is usually not interpreted. Strong countertransference reactions to passive-into-active testing can, however, confound initial plan-formulations as well as therapeutic reactions to tests, especially when therapists become defensive (Rappoport, 1996a). Therefore, anticipating passive-into-activing testing allows the clinician to avoid colluding with the patient in their pathogenic adaptation. Remaining stable and empathically attached should usually suffice (Foreman, 1996). In their suggestions to remain calm and confident while avoiding being troubled, defensive, or rejecting (Rappoport, 1997), practitioners working with CMT are in agreement with Ferenczi, who suggests the analyst should “not show any trace of irritation or offence” (1928, p. 93).

In the original studies on Mrs. C. (e.g., Silberschatz, 1986) plan compatibility was defined as counteracting or remaining neutral to patients’ demands, which was criticized in later research (Gazzillo et al., 2019). While maintaining a neutral stance might be sufficient for a specific patient, it cannot be presumed that it is in accord with another patient’s plan (Gazzillo et al., 2019).

Determining the meaning of tests

Returning to the different pathogenic beliefs underlying testing, the importance of recognizing the mode of testing, in order to respond adequately to a test, must be highlighted. Primarily, the potential underlying meaning of a test is determined by the therapist’s countertransference reaction to a test (Rappoport, 1996a). During a transference test, analysts usually feel helpful, whereas during passive-into-active tests, they may feel responsible or guilty, but also confused and treated like a child (Weiss, 1990a, 1993). Patient attitudes that occur during passive-into-active testing include acting in a demeaning, disparaging, and guilt-inducing manner, as well as being rejecting, critical, and distant. Therefore, these tests often produce intense countertransference reactions, such as feeling disempowered or inadequate (Rappoport, 1996a).

Secondly, clinicians take a thorough patient history and observe the patient’s reactions. Weiss (1994) claims that many patients directly state what they want to work on within the first few sessions, while others may refrain from doing so out of fear the analyst will oppose their plans. These patients will, however, reveal their unconscious plans, by hinting at them or by exhibiting a contradiction (Weiss, 1994).

According to CMT, patients will become calmer, gain new insight, and bring new material when an intervention has been plan-compatible (see Gazzillo et al., 2019 for a review on process variables; see also Silberschatz & Curtis, 1993). Subsequent interventions are then oriented to the patients’ reaction. More precisely, a positive reaction should be characterized by autonomy (Rappoport, 1997). Behaving in a friendly or accommodating way, and even seeming to be working on their problems, could merely be a test of whether being solicitous is required. Similarly, what appears to be a negative alliance may actually be a passive-into-active test, and when passed, patients will improve (Curtis & Silberschatz, 1986). Thus, the quality of the alliance does not automatically predict successful treatment but may reflect testing.

Within several weeks of responding to tests, there should be progress, otherwise the analyst may need to revise or adapt the plan-formulation (Weiss, 1994). To determine whether a test mode has been correctly hypothesized, Shilkret (2002) proposed considering the uncoiling of several sessions. To illustrate, a patient’s decision to move out of town could be understood as a test demonstrating assertiveness, over which she generally feels guilty and being what she has been testing in the past (Shilkret, 2002). Nevertheless, Shilkret determines that her testing of whether she could force her decision on her therapist has more merit. The successful passing of the test (i.e., encouragement to remain in treatment) was followed by the patient bringing up previously warded-off material. She remembered earlier violent childhood experiences and her mother’s inability to protect her; consequently, she refrained from moving until her treatment would be finished even when another possibility of moving arose. Her telling her therapist of her decision to move was actually a test reflecting an underlying worry of not being protected from her own decisions (Shilkret, 2002). Therefore, in some instances, the meaning of a test does not become apparent immediately but only during the following interaction. The development of a session or subsequent sessions is thus essential in determining whether a test has been correctly hypothesized (Shilkret, 2002).

When uncertain about the mode of testing, therapists should carefully try out a response and monitor the patient’s reaction to determine whether an intervention was plan-compatible (Weiss & Sampson, 1986). It may be safer to categorize a test as a transference test because failing to recognize a noncompliant transference test is more detrimental than failing a passive-into-active test. In the former the patient is more vulnerable than in the latter (Foreman, 1996), which was illustrated in the case of Sammet and colleagues (2007). Furthermore, patients are likely testing the passive-into-active element again in the course of therapy once the analyst has passed the transference element (Foreman, 1996).

Strengths and limitations

Important limitations of the current work constitute the exclusion of chapters and books and the selection criteria that excluded all studies examining couple and group therapies, as well as research on children. By neglecting possible peculiarities of treatments for specific patient groups, in which we suspected testing would take on a different form, we, however, collected test examples generally applicable to a large number of patients. Furthermore, the numerous case reports that provided selected excerpts of treatments pose another limitation. For most patients whose passive-into-active tests are described, the entire therapy process was not available. Therefore, it cannot be concluded that they tested solely in this mode.

Finally, it has to be taken into account that a majority of the included research was conducted by researchers affiliated to one research group (i.e., the SFPRG), whereas the studies by Gootnick (1982), Kadur et al. (2018), Sammet et al. (2007), and Varga (2006) appear to be independently performed. In some publications, therapists and researchers are two different professionals; in other studies, it is not clear. At least Fretter (1995), Pickles (2007), Shilkret (2002, 2008), and Weiss (1992, 1994, 1995) performed empirical research on previously executed treatments.

We see a strength in this overlap in that the analysts could consider their countertransference reactions more profoundly in comparison to researchers reading transcripts, which likely lead to a deeper understanding of the patient. Furthermore, this study fills a research gap by systematically examining studies and case reports that have previously been disregarded in reviews on testing. Overall, this work attempts to unite the profession of psychoanalysis with empirical research. Following the tradition of CMT that incorporated empirical investigations of psychoanalysts’ clinical work from its beginning, this study shows that research can be scientific and contain enough valuable content to be relevant for clinical practice.

Future research

While we provide an extensive overview of testing, we do not claim to cover all potential forms or meanings of tests. A striking point in the results was that few test examples were reported in which patients worked on anger or aggression. Furthermore, canceling sessions was interpreted as a test to avoid being burdensome in one study and to devalue the therapist in a different study. Another patient may, however, by not showing up for sessions, test for object constancy of whether the analyst remains an object on which the patient can rely. Therefore, it has to be concluded that testing may take on different forms, which calls for additional research in the field, especially research that does not specifically examine CMT.

Future work can focus on the questions that remain unanswered by CMT, such as why patients are attributed with an unconscious wish to master traumatic situations and to pursue goals that pathogenic beliefs warn them against. Since not all repetitions are necessarily attempts to master aversive experiences, examining cases in which patients relive conflictual or traumatic situations as a self-punishment, often derived out of guilt, may also proof useful. The reader may furthermore ask how patients know of healthier ways of interacting and thinking about themselves if they have never made an alternative experience.

Moreover, future research can systematize which interventions help individual patients with certain characteristics in the relinquishment of pathogenic beliefs and see at what point in treatment they let go of pathogenic adaptations. Finally, future work can also determine whether clinicians who have received training in detecting and adequately responding to tests achieve better treatment outcomes than clinicians without training.

Conclusion

Overall, the notion of testing makes it difficult to easily identify concrete tests, fueled by a clear pathogenic belief, and circling around a heterogeneous theme. In fact, tests may encompass different conflictual contents, and multiple modes may mix within a testing sequence. It is thus essential that researchers and clinicians have a good understanding of psychodynamic principles and of the patient’s individual biography when working with the concepts proposed by CMT.

By focusing intensively on higher mental functioning and endowing the patient with an unconscious motivation to change, Weiss and Sampson diverge from classical psychoanalytic thinking. The conceptualization of unconscious processes is limited to internalizations from early distressing experiences and unconscious testing activity. While the theory underlying the concept of testing seems reductionist and may only partially capture the patient’s motives and underlying structures that give rise to psychopathology, the conceptualization of transference and projective identification by the means of testing can enrich the way we think about patients. Passive-into-active testing, for example, can explain a variety of patient reactions and was thus deemed to be the most clinically valuable contribution of CMT (Foreman, 1996). Furthermore, the concept of testing provides a promising tool for performing clinically relevant research into psychoanalytic treatment.

To conclude, CMT principles contribute to the understanding of patients and their motives for exhibiting certain expressions, interactional patterns, or attitudes. The concept of testing may guide treatment but does not rigidly structure it by applying certain predetermined interventions (Gazzillo et al., 2021b). It provides therapists with an additional theoretical understanding based on actual clinical examples, which may be congruent with or complement other theoretical assumptions.

We intended to raise analysts’ awareness of this transference phenomenon, which can assist analysts in detecting and anticipating critical situations. The concept of testing may be especially useful for ongoing clinicians and particularly applicable to patients who do not seem to get better. By incorporating what patients bring into therapy and by working with their patterns and past object relations, analysts ensure that the intended tests do not go unanswered. Rather, they meet the unique needs of patients and respond in a helpful way.

Disclosure of interest

The authors report no conflict of interest.

Authors' Contributions

A.N.N. performed the research work and wrote the manuscript. J.L. served as second rater for inclusion of articles and assisted in research design. S.A. critically read and gave final approval of the version to be published.

Additional information

Notes on contributors

Alexandra Nicole Novak

Alexandra Nicole Novak, MS, is a second-year doctoral candidate at the University of Klagenfurt, Austria. This doctoral program focuses on psychotherapy and psychoanalysis, for which she has completed all required coursework. In addition, she has concluded two years of preliminary training in psychotherapy at a private institution and worked at an inpatient facility. She is currently training as a psychoanalyst at an Austrian institution affiliated to the IFPS.

Jonas Luedemann

Jonas Luedemann, MSc, was a university assistant at the University of Klagenfurt, Austria where he taught and performed research for the Department of Clinical Psychology.

Sylke Andreas

Sylke Andreas, DrPhil, is professor and head of the Department of Clinical Psychology at the University of Klagenfurt, Austria as well as director of research of the university's outpatient clinic and psychotherapeutic research & teaching center. She has been awarded her approbation as a psychological psychotherapist at the Lou Andreas-Salomé Institute (DPG) Göttingen and at the APH Hamburg, Germany. Currently she is a member of an Austrian psychoanalytic association accredited by the IFPS and training as supervisor for mentalization-based treatment.

Notes

1 In this manuscript we use analyst and therapist interchangeably to address both professions. In citations, the authors’ original term is employed.

2 Distinguishing them from daydreams, unconscious phantasies according to Klein (1959, p. 250) “are an activity of the mind that occurs on deep unconscious levels and accompanies every impulse experienced by the infant,” as for example hallucinating being fed and loved; however, it also includes hallucinations producing feelings of being deprived and persecuted.

3 Ferenczi (1949) had already postulated that the child introjects the adult’s sense of guilt when he brought back the reality of traumatic experience in his theory of psychopathology, after Freud had rejected his own trauma theory in favor of his seduction theory. Freud contributed to the concept of guilt in his essay “Mourning and melancholia” (1917) and developed the concept of borrowed guilt in “The ego and the id” (1923).

4 Gazzillo and colleagues (2019a) aim to make tests more easily identifiable by proposing to apply the concept of compliance and noncompliance to the pathogenic belief rather than to the expected reaction of the clinician.

5 Weiss (1993) claims that dreams do not necessarily have a wish-fulfillment component but have an adaptive function and can be considered as important messages that dreamers send to themselves.

6 Nonetheless, analysts may be seduced into allying with a strict superego.

7 Winnicott (1969) considers this an accomplishment that is not always achieved in normal development and it may thus become a task for patients who have not yet achieved it. Therapists facilitate the ability of patients to achieve the capacity to use an object. This entails enabling the patient to view an object as independently existing, rather than to simply relate to it as “a bundle of projections” (p. 712) under the omnipotent control of the child.

8 This should not be confused with Bacal’s (1998) “optimal responsiveness.” In fact, CMT is more consistent with the view of Kohut (1984), who stresses the empathic understanding of the patients’ experience and the attunement to their needs, while aiming for optimal frustration, in opposition to optimal responsiveness. CMT theorists and Kohut both ascribe an importance to gaining insight and understanding via the therapeutic relationship.