Testing of the therapist to reduce maladaptive interactional patterns: Categorizing patients’ tests

ABSTRACT

Objective

This empirical study undertakes a categorization of the core concept of Control Mastery Theory: mostly unconscious testing of pathogenic beliefs that patients exhibit in relating to their therapist to work on their problems. The focus lays on latent meanings of manifest tests.

Method

We qualitatively analyze transcripts of 172 psychotherapy sessions with 23 patients for sequences in which significant patient-therapist interactions occur, and systematize identified tests into thematic categories based on what tests intent to achieve (ICC = .68). Guided by theory, the analysis is attending to complexity, individuality, and the unconscious.

Results

Tests circle around striving for independence, deserving/self-worth, acceptance, and entitlement. Individual tests have various underlying meanings, are interrelated, and may be multidimensional.

Conclusion

Meanings of tests must be confirmed within the psychotherapeutic process. Incorporating the treating clinician thus seems important.

Keywords:

• Control Mastery Theory
• psychotherapy process
• qualitative research
• psychoanalytic research
• patient-therapist interactions
• transference

Clinical or methodological significance of this article: Understanding the potential significance of testing situations in patient expressions, behaviours, and attitudes, clinicians can respond appropriately and thus facilitate the patient in letting go of pathogenic beliefs that fuel maladaptive interactional patterns. Focusing on clinical descriptions, this study provides a detailed overview of critical interactions in the therapeutic dyad, and enhances the understanding of own reactions to patients.

Categorizing Patients’ Tests

Introduction to Control Mastery Theory (CMT)

Rooted in psychoanalytic theory, Control Mastery Theory (CMT) developed and empirically tested by Weiss, Sampson, and the San Francisco Psychotherapy Research Group re-conceptualizes certain concepts. It explains a phenomenon observed in clinical practice in which patients determine whether it is safe to bring previously warded-off content into consciousness and to pursue healthy developmental goals (Weiss, 1990b, 1993). It is assumed that patients unconsciously know what they want to achieve in treatment (i.e., the patient's plan), and that they provide the therapist with information to that effect (Weiss, 1994).

Weiss (1990a) views psychopathology as derived from and sustained by so-called pathogenic beliefs that stem from unconscious irrational guilt, internalized conflicts, or attempts to adapt to early shock or strain trauma (Weiss, 1990a, 1993, p. 9). By blaming oneself and absolving important others from guilt, pathogenic beliefs provide an explanation of why certain experiences with others occur while securing emotional bonds to early love objects. Pathogenic beliefs are thus tied to infantile object-relations (Weiss, 1994), meaning inner (partly distorted) mental images of the real self and real others. Both repressed content as well as content that was never conscious but has been internalized as implicit, nonverbalized, embodied memories of past interactions play an important role in the formation of pathogenic beliefs. Encompassing feelings, attitudes, judgments, and self-concepts they are more than conscious thoughts (Rappoport, 1997).

Moreover, Weiss (1996) acknowledges that what occurs in the therapeutic relationship is often unstated and unconscious. Freud’s (1915) main discovery that unconscious psychic material is yet potent, influential, and operant is thus continued in CMT. Although the theory recognizes internalization of early object-relations and unconscious processes, such as repression, transference, and countertransference, its adapted terminology and reduction in theoretical complexity lead to a disregard in the psychoanalytic community. On the other hand, basic assumptions of CMT became accessible to clinicians of all orientations. The psychoanalytic term transference, for example, is re-conceptualized as an unconscious yet intentional strategy (Silberschatz, 2008) to activate early childhood conflicts and to make new experiences by means of testing.

Testing and its Purpose for the Patient

Starting to believe their distorted ideas about themselves, others, and the relationships they encounter, the child and later the adult acts as if pathogenic beliefs were true. During treatment, the patient applies pathogenic beliefs to the therapist to determine mostly unconsciously if they still have merit in the current relationship, which was termed testing (Weiss, 1990a, 1993). Testing is defined as direct or indirect expressions of pathogenic beliefs via the patient’s attitudes, communications, and behaviours with the aim to overcome pathogenic beliefs (Gazzillo et al., 2019a). In addition, tests may entail the patient re-enacting traumatic experiences, provoking certain relational constellations, or assigning a role to the therapist (see also Sandler, 1976).

In testing, patients mostly unconsciously assess whether the pursuit of individual goals poses a danger to them or others. Tests are employed in the hope to replace pathogenic beliefs with healthier alternatives (Weiss, 1994) that would follow a regular human development, such as establishing sufficient self-worth, striving for autonomy, or undertaking fulfilling relationships (Gazzillo et al., 2021). By putting focus on the adaptive function of repetition (Weiss, 1990b) rather than on its pathological character in the sense of a repetition compulsion (Freud, 1914), CMT diverges from classical analytic thinking. Its main pillars are that patients are motivated to master their conflicts and that they have some control over their psychic processes, which become manifest in testing.

Types of Tests

An initial differentiation was made between transference tests and passive-into-active tests (e.g., Weiss, 1990a). In the first case, patients identify with their traumatized self (Gazzillo et al., 2022), whereas in the latter, they identify with a potentially traumatizing other by assuming the role of early caregivers. By placing the therapist in situations similar to those patients themselves experienced as children (Rappoport, 1996), they invite the therapist to feel as the patients experienced themselves at the time, which is strikingly similar to some communicative conceptualizations of projective identification (Bion, 1959) and to the patient's contribution in enactments.

Further distinctions are made between transference tests by compliance (TC) and transference tests by non-compliance (TNC) in regards to the perceived expectation of others (Rappoport, 1996), or in regards to the pathogenic belief itself (Gazzillo et al., 2019a). Patients unconsciously repeat their habitual ways of experiencing themselves and interacting with others. Acting differently would in fact result in a perceived danger to the patient or to relationships with others. Nevertheless, when feeling secure and stable enough, patients may also break their patterns in the current therapeutic relationship and experiment with new ways of interacting and experiencing that were previously considered dangerous and thus run counter to their pathogenic beliefs. At this point an essential aspect of the testing concept becomes apparent. Patients express directly what they want to achieve (TNC) or the very opposite thereof (TC), hoping that the therapist will point out their irrationality. Similarly, some patients do not directly say what they want to accomplish but subtly show it.

By passive-into-active-tests by compliance (PIAC), patients behave in a manner consistent with interactions with early caregivers, and test whether therapists will react as the patient’s pathogenic beliefs dictate and thus whether it is appropriate to have them. Gazzillo et al. (2019a) propose that some patients treat the therapist in the opposite manner, that is, as the patient would have liked to be treated, which defines passive-into-active testing by non-compliance (PIANC).

A classification of tests is necessary in order to subsequently determine adequate interventions since tests of different types call for different therapeutic responses (Rappoport, 1996; Weiss, 1990a, 1993) that must furthermore be specifically tailored to the individual patient (Weiss, 1994).

Clinical Use of the Concept

Firstly, the treating clinician will feel a certain pull or an inclination to respond to the patient’s tests (Silberschatz, 2008). In response to a transference test, the clinician usually feels helpful; in contrast, passive-into-active tests (by compliance) are more likely to elicit feelings of confusion, guilt, or being controlled (Weiss, 1990a, 1993). They may provoke feelings similar to those experienced by the patient in the past (Gazzillo et al., 2022). Yet, these rules of thumb can not be rigidly applied. Therapists may experience negative feelings of being scared, sad, angry, worried, or disoriented in transference tests as well, and may experience a pull to respond in either re-traumatizing or transformative ways (ibid).

Secondly, the therapist’s countertransference is used as epistemological tool to identify and categorize tests, according to the previously described typology (Rappoport, 1996), while the clinician's contribution in the form of individual tendencies and own transference to the patient is acknowledged (Gazzillo et al., 2022).

A test is considered passed when a pathogenic belief is disconfirmed and the patient’s unconscious plan is supported (Weiss, 1994). The treating clinician may accomplish this in various ways: by plan-compatible interpretations, the provision of corrective emotional experiences, and the overall therapeutic attitude (see Weiss, 1993, 1994 for examples). An active technique in which therapists convey interest, unburden, and support is proposed (Gazzillo et al., 2019a; Rappoport, 1996) but should not generally be applied to every patient. Shilkret (2006) shows that while in some cases a nonjudgmental therapeutic attitude may suffice, other patients need therapists to display a certain attitude, for example that they would not worry. A supportive attitude for this particular patient would likely confirm his pathogenic belief that he was the cause for extensive worry. Similarly, some patients benefit from interpretations, while others are endangered by them, which Shilkret (2006) demonstrates especially for narcissistic vulnerable and severely disturbed patients.

Finally, it is proposed that when the therapist continuously passes tests, patients can move towards overcoming their maladaptive patterns of interacting and experiencing (Rappoport, 1996; Silberschatz, 2008).

Significance and Operationalization of the Theory

The significance of CMT and its testing concept becomes apparent through various single case studies (reviewed by Gazzillo et al., 2019a and Novak et al., 2022) and one naturalistic outcome study (Silberschatz, 2017) showing that passing a test is significantly correlated with therapy progress and outcome variables. Moreover, several efforts have been undertaken to operationalize concepts of CMT, some of which are introduced below.

Plan-formulation method

In conceptualizing a case, the therapist or independent researcher can rely on plan- formulations. Based on what patients reveal and what can be inferred from their narration within the initial sessions, the patients’ goals, obstructions (i.e., pathogenic beliefs and early traumata that contribute to their development), potential tests, and useful insights to be gained are listed (Curtis et al., 1994). Gazzillo et al. (2019b) recommend adding corrective emotional experiences while Rappoport (1996, 1997) includes therapeutic attitudes to patient’s tests. Gazzillo et al. (2019b) demand plan-formulations to be comprehensive and explicative in providing a coherent picture of what obstructions hinder patients in their pursuit of goals and how they will be tested in therapy. They should be specifically tailored to the individual patient in interconnecting the previously mentioned aspects.

An important consideration is that readily expressed goals, such as becoming more autonomous or avoiding anger, may represent an attempt to provide information about the plan but may likewise already constitute a test of whether the patient is expected to have certain goals (Weiss, 1998). Similarly, some patients express or hint at goals to test whether they are allowed to have certain goals without endangering the relationship to their therapist (Weiss, 1994, 1998).

A plan-formulation aims to raise awareness of the therapist's countertransference, to anticipate potential testing, and to facilitate appropriate response (Rappoport, 1996). As a kind of roadmap or compass, it serves to guide the treatment process (Gazzillo et al., 2019b). Its accuracy is indicated when patients show improvement following interventions that are consistent with their plan. Nonetheless, Weiss (1994) considered it a tentative hypothesis about a patient and pointed out that it may need to be adjusted as treatment progresses. Therefore, a plan-formulation is always supplemented with the observation of the patient’s reactions to interventions to assess whether a test has been passed (Gazzillo et al., 2019a) and whether the intention of testing has been correctly hypothesized.

Systematization of pathogenic beliefs and testing

A systematic way for generating themes that are likely to be tested was developed by Sammet and colleagues (2007a), who compiled a self-reported list of patients’ pathogenic beliefs. Items circled around self-doubt, doubts about others, expressions of anger, fear of close relationships, guilt about success, and responsibility guilt. Another systematization of pathogenic beliefs resulted in slightly different items: avoiding relying on others, being undeserving, and having a sense of interpersonal guilt (Aafjes-van Doorn et al., 2021).

The systematization of testing began with its categorization into test types as introduced in the introduction. Moreover, a previously performed literature review showed that testing occurs on a continuum of subtleties, verbally or nonverbally. Interactions with the therapist, the use of the setting, as well as the patients’ narratives or self-presentations may constitute tests (Novak et al., 2022). Since tests are shaped by the patients’ interpersonal history, psychic structure, and personality style, testing is highly individual.

Objective

The test concept is of theoretical and clinical importance. Since previous research on the content of tests is idiographic, the comparison between patients is limited. Other concepts of CMT, pathogenic beliefs and guilt, were examined nomothetically, which allowed to relate them to other measures. A major limitation of previous research is however its focus on directly expressed self-reported content and thus inevitably not on the unconscious. The current study fills this gap by integrating the treating therapists and their interpretations of the tests. This allowed us to go beyond a mere descriptive observation and to include underlying psychodynamics. By focusing on unconscious testing activity, we have not abandoned its psychoanalytic roots. We distinguish between how a test manifests itself and what patients may actually intent to achieve with the test (i.e., its latent meaning for the individual). Without neglecting the complexity of the concept, we present a systematization of tests that allows for more systematic and comparative research.

Attempting to soften the outdated separation of deterministic quantitative and interpretative qualitative research and their mutual animosity, the current study is twofold and entails first a thorough theoretically guided analysis of test sequences in transcribed psychotherapy sessions according to the plan-formulation method. Second, after differentiating the underlying meanings of tests, categories of test themes are empirically generated.

The focus of the current work is to provide therapists with an overview of potentially critical situations. Clinicians can gain an increased understanding of the inner world of their patients and may obtain valuable insights into their own associations, emotions, and impulses. Elicited in the therapist during testing, these countertransference reactions, especially its subtler forms, are sometimes overlooked or rationalized. We aim to sensitize clinicians for the possibility of mostly unconscious testing activity, which is assumed to drive therapy progress.

Method

Patients, Therapists, and Treatment

We analyzed 172 transcribed therapy sessions of a randomly selected subsample of 23 out of 100 patients obtained from the Munich Psychotherapy Study (Huber et al., 2012). Analyses were based on two to three consecutive sessions each from the beginning, the middle phase, and the final phase of treatment. The 18 female and 5 male patients were between 24 and 43 years old, primarily diagnosed with a depressive disorder (mild or severe episode), and had agreed to audio-recordings during the entire course of treatment. Patients received either psychoanalytic (n = 8), psychodynamic (n = 9), or cognitive–behavioural (n = 6), therapy, which was funded by the German healthcare system and provided by 16 well-experienced psychotherapists (mean experience: 20 years), who were not familiar with CMT. This is compatible with the assumption that patients test regardless of whether therapists are aware of it and independent of their clinical orientation.

Identifying Tests in the Transcripts

Prior to identifying tests, we established a set of rules. Tests must be observable in the patient-therapist interaction. Self-presentations when expressed within the therapeutic relationship were included in the rating. However, a mere narration was not sufficient to be included. Likewise, we refrained from rating tests by the use of the setting if these incidents were not followed up in the session and thus the potential underlying motivation was not expressed neither by the patient nor the therapist.

In a first step, plan-formulations were gathered for 12 patients using the first two available sessions (session nr. 3 in 16 cases, nr. 4–8 in 7 cases). Three independent raters listed traumata, pathogenic beliefs, possible tests, and insights to be gained. All listed items were added up across categories and subsequently examined on agreement. Since raters use their idiosyncratic language in plan-formulations, the clinically experienced last author decided whether descriptions referred to the same construct. The overall ICC amounted to .85. Due to the excellent reliability according to Cicchetti (1994), only the first author gained plan-formulations for the remaining 9 cases.

Next, transcripts were independently read and analyzed by (A.N.N., training psychoanalyst with 2 years experience in rating CMT cases, rated 23 cases. J.K., training cognitive behavioural therapist with 7 years experience rated 12 cases). This process was subsequently discussed and supervised by S.A., psychodynamic therapist with 22 years clinical experience.

A directed mixed content analysis approach was used (e.g., Schreier, 2017). We deductively identified tests based on the expected tests listed in each plan-formulation. Between two raters over 12 patients, we computed an ICC of all passages that were either identified as tests or not as tests, which amounted to .72, considered a good level of agreement according to Cicchetti (1994). The remaining 9 cases were analyzed by the first author, who was psychoanalytically supervised.

Qualitative Data Analysis

After transcripts of an entire case were read, three independent raters gathered the meaning of tests by the patient’s intention for testing. The content of test examples was concisely summarized with a sentence. The meaning of a test was primarily derived from (a) the context of sessions, (b) the personal history reflected in the plan-formulation, and (c) the response of the treating clinician. Below is a condensed extract of a transcript to exemplify the analytic process (session 216):

P:

at the moment, I have the feeling as if I was in late puberty, I am in defying silence, and all these negative things, I fell apart and no one puts me back together.

T:

and I should put you back together?

P:

it just went to fast. I have the feeling that the big bang is missing, waking up and being a new woman!

T:

there is an aspect of taking stock of therapy, with some dissatisfaction, such as that something is missing!

P:

yes, I think a lot is missing!

T:

and like this, I just let you go!

First, this paragraph was highlighted as a potential test since the clinician seemed to experience a certain pull (Gazzillo et al., 2019a), expressed by: “I should put you back together”. In a subsequent step, the context of the current session was incorporated to determine the potential meaning of this interaction. Later in the transcript, it becomes clear that the stipulated termination approaches. Employing the metaphor of food, the therapist hypothesized that the patient was not yet full to which she responded, she was not sure if anything else would even fit, suggesting that the time to end treatment may have come. Later in the session, a contradiction in the patient’s expression about terminating becomes apparent.

Furthermore, the context of previous sessions and the plan-formulation are incorporated. Taking into consideration that this particular patient previously expressed difficulties with separating from her parents and struggled with being dependent on others, it becomes plausible that in this sequence, she expresses her ambivalence of wanting to care for herself and not knowing whether she was able to without endangering the self or others. This conclusion is further supported by the concept of separation/disloyalty guilt for becoming autonomous (O'Connor et al., 1997). It can be argued that she was careful not to hurt the analyst by leaving and tested the pathogenic belief of not being allowed to be independent.

This hypothesis becomes even more valid when considering the clinicians response to her test. He verbalizes potential unfulfilled hopes the patient might have had about treatment as well as potential disappointment and anger. He also broaches the issue of cutting a connection as a necessity in order to become free. Offering the hypothesis that the patient may be in a loyalty conflict, he proposes she might also feel destructive and annihilating, should she agree to leave treatment. Lastly, he highlighted the importance on reflecting on their parting.

After a discussion of three independent researchers, this sequence was ultimately coded as a transference test by compliance circling around striving for independence. By being reluctant to end treatment, the patient tested whether the therapist would let her go and whether he would trust her to manage on her own without being personally affected.

Incorporating the Unconscious

Unconscious aspects of tests can only be gathered via their observable derivatives within the therapeutic interaction. For example, the manifest content of the patient stating that something was missing is complemented by the latent meaning that she must not be independent, which was verbalized by her therapist. This step of including latent meaning is necessary since the mere manifestation of what the patient said or did is, even if it leads to highly reliable ratings, rather meaningless. An examination that simply collects what patients directly and consciously express would not do justice to the concept of testing and is therefore not valid. However, to adhere to an empirical research methodology that is traceable and does not heavily rely on subjective interpretations, we solely focus on latent meanings expressed in the patient's and therapist's communications that were identifiable in the transcript. Deeper unconscious meanings that therapists may have hypothesized but did not express would have to be inferred from the text and were thus not included. This study bases on interpretations of well-experienced therapists, who may have gained latent meanings via the application of theory, their countertransference, or reconstruction. Therefore, the therapist’s work may very well contain unconscious aspects of tests.

Empirical Systematization of Test Categories

The content analysis was inductive in that clinical manifestations of testing were systematized according to thematic categories derived from the material itself (Schreier, 2017). The formation of a typology, was carried out via a systematic comparative analysis (Juettemann, 1990) of 23 single cases. After detailed individual analyses examining the peculiarities of each case, overarching resemblances and regularities of tests were gathered in comparing and contrasting cases. Based on similarities in selected characteristics, tests were grouped into types that differ from each other as much as possible, according to pathogenic beliefs, underlying guilt, dominant interactions, and comments of the therapist.

The ICC for this step of the analysis across three raters before discussing 12 cases amounted to .68, considered a good level of agreement according to Cicchetti (1994). The ICC bases on binary nominal data meaning agreement or non-agreement of assigned category that indicates underlying meanings of tests. If no consensus was reached, we listed both interpretations in the discussion.

Results

Almost all of the analyzed patients exhibited testing behaviours during their entire course of treatment. While one third of patients tested primarily in the middle and end phase, only a few patients tested exclusively in the initial, middle or final phase. The intensity of testing increased for several patients when the topic of ending treatment arose. Most patients employed about 4–6 tests in 6–9 sessions. While the highest test rate amounted to 12 tests, only one patient did not test at all.

As depicted in Table I, we categorize tests by using the patient's intention for testing to gain (1) independence, (2) self-worth and a sense of deserving, (3) acceptance, and (4) a sense of entitlement. Concerning the overall frequency, the subcategories securing bonds without sacrificing the self (acceptance TC tests: 12 cases) as well as control and rejection PIAC tests, circling around entitlement: (11 cases) and self-worth tests (9 cases) were found most often in the sample, whereas actively striving for autonomy (TNC) were the least frequent tests (7 cases). All patients tested in more than one category. Especially tests of being quiet and withdrawing were found in many different thematic categories thus pointing to various underlying latent meanings of these tests.

Table I. Test situations categorized into test themes and test type.

Test category	Test type	Underlying guilt & intention
(1) Independence
(1a) actively striving for independence by non-complying with pathogenic beliefs
express autonomy without being overly concerned with the T stand up for own wishes/express what they need or do not want (also by use of the setting: e.g., session times) actively decide what they want to talk about (act astonished when being told it is up to them) (tentatively) express own opinions (also about not wanting to come to PT) stop being overly considerate	TNC	separation guilt/disloyalty guilt
(1b) striving for independence by complying with pathogenic beliefs
present themselves as incapable of knowing answers or finding solutions ascribe omnipotence to their T/put excessive value on the T's opinion dependent on T to provide answers/solutions/confirmation ask overly for advice (what to do/how to approach an issue)/constantly hand over own responsibilities present themselves as helpless or insecure voice concern or doubts about ending PT (indicate that something is missing/that PT is not yet completed) present symptoms in light of ending PT (state that the T cannot leave the P in this state of mind) inquire extension of treatment while hoping the T believes the P is capable/ready to take care for themselves never miss a session (put PT above travel plans, job, family etc.)	TC	lose fear of disrupting ties, of hurting/losing others by being autonomous
(1c) closeness and distance
comment on the physical distance try to be overly close to the T (be like them, use same expressions) or not allow any closeness present an ambivalence about closeness and distance inquire about termination or miss sessions to keep distance to the T	all	balancing closeness and autonomy
(2) Deserving and gaining Self-worth		survivor guilt/self-hate
express doubts about themselves, worry about not being right or making themselves ridiculous want to get rid of feelings or personal characteristics (as expected by others) wonder whether they are capable to change, worry about not accomplishing anything in PT belittle their own importance for their T and believe to be exchangeable withdraw from the T in complying with not deserving attention/care/interest/fulfilling relationships deflect the importance of the T or PT to protect themselves from disappointments deflect importance of assaults and past adverse experiences (talk about trivialities, change topic, and use empty phrases, or question their memory, do not want T to conclude they had a bad childhood) act guilty/state that nothing will change and that they have to accept their guilt	TC	become entitled to feel assaulted (vs. need to feel guilty), test whether expression of self-worth endangers self or others
ask whether they have to change (as expected by others) non-compliance with unjust others (ascribe responsibility to others, bring documents of reported assaults)	TNC
(3) Striving for Acceptance
(3a) criticize T overtly or covertly		survivor guilt/self-hate
accuse T of not being able to understand or not being interested, disparage T's’ capabilities express doubts or concerns about T, PT (verbalize doubts of family/friends) voice /show discontent with PT progress, question why they come to PT or whether it is good for them present symptoms to show that PT is ineffective reject T's comments and interventions, instruct T	PIAC	hope not having to feel responsible for criticisms of others
(3b) invite rejection
reject or criticize T, depreciate PT, voice PT is a waste of time withdraw, present themselves unproductive, pause or threaten to quit PT	TC	hope T will not reject the P
(4) Striving for Entitlement
(4a) secure bonds without sacrificing the self		omnipotent responsibility guilt
overly accommodating, adaptive, apologetic (ask how they should behave and what the T would like to hear) quiet, hesitant, reserved out of fear to be burdensome, state not having anything interesting/important to say (anymore), state it is always difficult to start, that they came up with many things that are now gone concerned about taking too much time/inquire about time worry about doing something wrong/not performing adequately and thus losing the T's benevolence worry about boring or annoying the T, try to be overly positive to avoid burdening the T with their emotions excessively praise the T or overly thankful want to end sessions early out of fear to be burdensome, convinced not having more hours/time available propose or agree to short-term PT, inquire about termination or a reduction of sessions per week	TC	hope not having to be accommodating and losing fear of being a burden
(4b) setting boundaries
demand interest and attention (demanding attitude) expect T to provide answers to every question, to receive solutions, or to be freed from problems provoke anger, bring T in a position in which they feel inclined to reprimand the P control and dominate (instruct T, inhibit T by being passive aggressive or overly submissive) invade and encroach put T in a situation they cannot control (threaten to hurt themselves) entice T to break boundaries	PIAC	hope not endangering self or others by setting boundaries themselves

Note. T = therapist, PT = psychotherapy, P = patient, TC = compliant transference test, TNC = non-compliant transference test, PIAC = compliant passive-into-active test.

Striving for Independence

As can be seen in Table I, Independence tests are split up into tests in which patients (1a) actively strive to be autonomous by non-complying with pathogenic beliefs, and tests in which patients (1b) are reluctant to express any independence out of fear to disrupt ties with others. They also strive for autonomy but do so by complying with pathogenic beliefs. Some patients who used to be overly attentive and who tended to neglect their own wishes begin to express their autonomy once they experience that the therapist does not require any accommodating attitude (1a: TNC). Other patients present themselves as incapable of allowing answers or solutions to emerge within themselves, while they ascribe this capability to their therapist (1b: TC). They excessively elevate the therapist's importance and make themselves small. Last, tests circling around closeness and distance (1c: all test types) were placed in the category independence since patients test whether being independent while sustaining a relationship is possible and whether closeness leads to a loss of autonomy.

Deserving and Gaining Self-Worth

While the subcategory (1b) encompasses the neglect of own wishes and opinions for the sake of securing bonds, the category (2) deserving/self-worth entails tests in which patients tend to believe they do not deserve anything good for themselves. Patients test whether they have the right to express self-worth or whether they endanger others or themselves by doing so. Two patients who were convinced they needed to let go of their anger (TC) were unconsciously testing the therapist’s stance on being angry. One patient bought books on anger, another patient said the therapist would never see her angry, thereby testing whether the therapist would find this change necessary. Patients belittling past traumatic experiences (TC) may test whether they have the right to feel assaulted, whether they are overreacting and in fact guilty themselves, and whether they can trust their memories. Similarly, in acting guilty (TC) patients test whether they are right in blaming themselves or whether they can dare to ascribe responsibility to others. In the latter case they would be non-complying with unjust others (TNC).

Striving for Acceptance

Acceptance tests (3) entail the latent intention to receive acceptance while the manifest interaction is (3a: PIAC) characterized by patients rejecting and criticizing therapists but also by (3b: TC) inviting their rejection of the patient. The majority of criticism was covert. One patient talked about a book that served her needs better than the therapist did. Another patient brought an article on therapeutic interventions thereby instructing the therapist how to perform. Covert criticism also became manifest in withdrawing by claiming not having anything to talk about. Similarly, when in fact they fear being rejected, certain patients (3b) test whether the therapist is enticed to end therapy early, when they claim not having much to say or that they do not want to come to treatment. Patients unconsciously invite rejection by actively rejecting the therapist or by voicing that therapy was a waste of time. Other patients directly use the setting and express they want to quit or pause treatment, but unconsciously hope that the therapist will question their proposal for leaving therapy early.

Striving for Entitlement

The category (4) entitlement entails tests employed (4a: TC) in the hope to secure bonds without sacrificing the self, and to lose the fear of being burdensome. Patients are thus overly accommodating and adaptive but also (4b: PIAC) place therapists in situations in which they need to set boundaries for not being subjugated and controlled by the patient. Exhibiting a demanding attitude, certain patients, expect the therapist to provide solutions to their problems. One patient asked what she should do with her daily frustration and insisted the therapist find a solution. She furthermore asked for advice knowing that she would not follow it. This patient was testing whether the therapist would counter her attempts to exploit and belittle him. By remaining strong and assertive, while reflecting on their interaction, he managed to sensitize the patient for her unconscious attempt to devalue him. Another patient overly demanded interest and attention from her therapist, asked whether she remembered contents of the last session, and insisted that she watch the patient’s TV appearance. She insistently instructed her therapist how to proceed. Yet, only one patient was extremely controlling and encroaching. She claimed to know what the therapist was feeling and thinking, bluntly demanded her concentration, and gave orders on what needed to be written down.

Discussion

Enhancing and refining an existing model of the therapeutic process, the goal of this study was to systematize tests into categories resulting in an overview of potential testing situations that therapists might encounter in clinical practice.

Categories and the Complexity of Testing

First, the complexity of testing may not become readily evident due to the breaking down of tests into categories. In their simplified presentation in Table I, it may seem that tests may be easily assigned to a category. However, a thorough analysis of the individual case is necessary to determine the underlying meaning of a test. For example, the manifest doubts of the patient introduced in the method section were her way of testing whether the therapist believed she was ready to distance herself from the therapist. She was striving for independence (the intention of her test) while being reluctant to end treatment by stating that something is still missing (the manifestation of the test). Moreover, similar manifest tests of other patients may have entirely different underlying meanings. The importance to distinguish between manifest content and latent meaning is furthermore illustrated by the following example.

Patients generally use doubts about themselves to determine whether they deserve to gain more self-worth (TC). When doubts turn into ruminating and hopelessness, however, it is important to consider the possibility that patients may induce guilt in their therapists (PIAC). This type of test was found in several cases in the literature (e.g., Pickles, 2007) when patients place therapists in a situation in which they have no control.

A specific case analyzed in the current study of a patient with persistent negative thinking may constitute such a test. Losing his hope and sense of self-worth towards the end of therapy was a test that spanned over at least three sessions. It is likely that the patient shifted his guilt onto the therapist who became too entangled in the interaction to reflect the situation. Nevertheless, the therapist continued to reassure the patient, which the latter constantly demanded and then rejected. The therapist’s rational for this intervention might be quite different from the one proposed here. Since the therapist did not reflect on the patient’s ruminating, the interpretation of induced guilt must remain a conjecture.

Another interpretation could be that the patient tested whether the therapist would stay hopeful, confident, and assertive while the patient lost all hope in the treatment and trust in himself. By employing this test, patients may also try to determine whether the therapist cares or whether boundaries are enforced (Sammet et al., 2007b). Yet, other patients test this way to determine whether the therapist will worry and feel guilty (Gootnick, 1982). Due to the complex nature of the human psyche and the interrelatedness of conflictual themes being tested, it is likely that this test had multiple aspects, as was already proposed by Sammet et al. (2007b). The therapist in the current study remained optimistic and confident and thus supported the patient. However, he did not discuss the patient’s extensive hopelessness; therefore, he may have failed other aspects of the test circling around control and guilt.

In general, the use of the setting is particularly suitable for diverse intentions of testing. Therefore, cancelling or rescheduling sessions, and wanting to leave early, as well as pausing treatment or reducing hours need to be viewed under the perspective of having a latent meaning, such as fearing to be burdensome and thus endangering bonds. While these tests may simply constitute manifest expressions of autonomy for patients working on separation guilt and self-worth, they may likewise be fuelled by a belief of not deserving treatment. Consequently, they constitute a latent invitation to the therapist to reject the patient. Yet, other patients may avoid closeness by these test and determine whether the therapist would collude. Arriving late or missing sessions without prior notice is also used to test whether the therapist excessively worries, which would imply that the patient himself should worry for others as well (Gootnick, 1982). Expressing threats to hurt oneself including suicide threats may be employed to control the therapist or to check for boundaries but likewise to test whether the therapist cares (Sammet et al., 2007b).

The most common manifest expression across the thematic categories was that one does not have much to say. Most patients did not want to burden their therapists, others tormented therapists and put them in a situation they could not control. Behind this recurring theme can also be an underlying criticism or an invitation for the therapist to reject the patient. Nevertheless, the patient’s hesitance to talk might not be a test at all but reflect the result of the therapist failing to discuss or to invoke the fundamental rule for patients to freely express their thoughts and feelings without censoring or withholding. This can only be an assumption since we did not analyze every session.

Moreover, we hypothesize that presenting empty phrases and narrations that sidetrack, or dropping important topics may be used to test whether the therapist participates in avoiding. For most tests involving manifest withdrawal, no sufficiently robust evidence of their underlying significance could be obtained from either the treating clinician's remarks or the therapeutic interaction.

Distinguishing Between Tests

Certain interactions, such as asking for answers and demanding solutions can be placed in the categories independence (1b) or entitlement (4b: setting boundaries). For example, a patient expressing that she needs a starting point on how to work on her problem may comply with her pathogenic belief of having to depend on others by handing over responsibility, when in fact she aims to be self-directed (Silberschatz, 2017). Nevertheless, when pursued more intensely and aggressively, asking for answers and demanding solutions likely reflects passive-into-active testing (Gootnick, 1982) in which patients put therapists into situations similar to those the patient had to endure. The distinction between different types of tests can be achieved by considering the patient's attitude as either striving for autonomy or as being demanding. Nevertheless, asking for guidance may have yet other individual meanings. A patient who grew up with little parental help from early caregivers who experienced her as a burden might test whether she can rely on the therapist (Silberschatz, 2017). This test would have to be considered an acceptance test (TNC; securing bonds without sacrificing the self and losing the fear to be burdensome).

In using their countertransference, clinicians can determine whether they are put in a situation they cannot control and whether they are devalued or criticized. Clinicians will likely experience different reactions to an entitlement test by feeling pleased or patronized by patients who are excessively avoiding to be burdensome. Therapists’ reactions will yet again be different for patients who deflect the importance of treatment and avoid working through.

Furthermore, the timing of a test must be taken into account when making distinctions between meanings of tests. Asking for opinions and affirmations may constitute a test for independence, but it may also reflect a common insecurity early in treatment. Moreover, patients who ask whether they improved towards the end of therapy are likely to recapitulate their progress rather than test.

Special Cases—Lack of Testing and Severely Disturbed Patients

The current study revealed that one out of 23 patients did not test. However, his avoidance of establishing a relationship with the therapist could be interpreted as one big test. Since the therapist did not voice the patient’s overall reluctance to interact, it was impossible to determine underlying meanings within the established set of rules for analyzing transcripts. Another interpretation by Brockmann et al. (2018) who concluded that not every patient tests, is that some patients show lower tolerance to activate a conflict, and in consequence test. They rather use the therapeutic relationship for impulse control and stabilization. Patients who seem to assume a static defensive position and who refrain from any emotional relationship may protect themselves from re-traumatization or from the fear of becoming dependent. In addition, they may be in denial about their wish for contact or have lost all hope for it. The clinician, in turn, is likely to feel deprived of any meaning or under pressure to function like a machine.

More severely disturbed patients and patients with structural difficulties might also not test or test differently. Within our sample, this characterization might apply to one patient with a potential borderline personality organization (Kernberg, 1967), which could not be confirmed from the available patient data. This particular case was different from the other 22 cases in that almost no rational discourse seemed possible. The patient tested by claiming to know what the therapist thinks, by confronting her with commands, and generally by being extremely invading and controlling, which is in accordance with the extant research. Gazzillo et al. (2021) show that testing occurs in treatments of patients diagnosed with personality disorder. Generally, severely impaired patients switch test type rapidly (Gazzillo et al., 2019a), use a variety of test types (Pickles, 2007), tend to employ passive-into-active tests (Gootnick, 1982), or, on the contrary, may need longer to develop the necessary sense of safety in order to test (Shilkret, 2006).

Subtle Forms of Testing

The analysis of transcripts showed that some patients test intensively by bringing relational episodes outside of therapy as well as dream narrations. One patient, for example, who worked on lowering her separation/disloyalty guilt towards her mother tested by narrating episodes, where she did not allow herself to be autonomous as well as episodes where she experimented with being independent. By determining the therapist’s reaction to these episodes, she explored her need to adapt to the wishes of others.

Moreover, patients also worked through their issues by re-experiencing unconscious self-presentations, which were common tests in the extant literature (e.g., Shilkret, 2002). Similar to narration tests, self-presentation tests may be employed to determine which attitudes the therapist values in patients; however, it may simply be a symptomatic expression. For these two test strategies, the researcher must draw on other aspects, such as the patients’ history, their usual way of interacting, contradictions, and the development during the entire treatment. Without being able to question therapists directly on their countertransferences or having access to their process notes, we, as independent researchers, could only observe associations, feelings, and impulses that therapists verbalized or apparently acted upon.

Overlap with CMT Categories and Concepts

As shown in Table I, most thematic test categories reflect the type of test. Although we did not find passive-into-active tests in the category independence, we acknowledge that patients may very well test this theme by placing therapists in situations in which they are confronted with being autonomous and are either criticized for (PIAC) or supported for doing so (PIANC). Therefore, other test types within the here proposed categorical system are not to be ruled out. Patients may also employ passive-into-active tests by non-compliance to detect whether they can be proud, and whether therapists are capable to be proud of themselves and of the patient (Gazzillo et al., 2019a), which might have been tested by one patient in our sample. Since the therapist did not reflect on this potential test, it remains speculative, and we did not list this instance.

Concerning other relevant concepts within CMT, we suggest that independence tests are likely based on separation/disloyalty guilt, whereas gaining self-worth as well as striving for acceptance seem to relate to self-hate or survivor guilt – feeling guilty for getting more out of life than significant others. Entitlement tests and especially patients who are overly accommodating seem to base on omnipotent responsibility guilt (O'Connor et al., 1997).

Some pathogenic beliefs listed in previous studies are more or less directly represented in the test categories. Nevertheless, tests circling around entitlement and acceptance that manifest themselves in the patient's demands, rejections, and criticisms may not be as readily apparent from the lists of pathogenic beliefs. We therefore conclude that the current work of test categories contributes to empirical knowledge in the field of CMT.

Clinical Implications

Employing passive-into-active tests, patients determine whether therapists can persist through criticism or accusation or whether they will take on irrational guilt. While some criticisms and demands may be warranted, the possibility of a passive-into-active test in individual cases should not be ruled out. In the latter case, patients are relieved when the therapist remains calm or assertive, when the therapeutic relationship persists, and when the therapist points out the irrationality of the patient’s demands (Rappoport, 1996, 1997; Weiss, 1990a). This therapeutic attitude and intervention may however be detrimental for other patients (Gazzillo et al., 2019a) testing by non-compliance or by either type of transference test. For these tests, therapists should rather demonstrate that they do not expect pathogenic adaptations and that they are not gratified by the patients’ solicitous and accommodating behaviour. Clinicians should challenge the patients’ pathogenic beliefs, and value their experimenting with new interactions, attitudes, or feelings that were previously disregarded as causing danger (Rappoport, 1996, 1997). We explicitly advise clinicians to reflect on the possibility of patients testing by experimenting with breaking their pathogenic patterns. Since tests revolving around the active pursuit of autonomy are scarcely noticeable and may thus be easily overlooked, therapists should avoid unconsciously undermining the patients striving for healthy development goals. In addition, the patients’ strive for autonomy should not be confused with a deflection of the importance of the therapist or an avoidance of closeness.

Another important implication is that patients who seem to have what is commonly known as a negative alliance, may in fact perform passive-into-active tests (Curtis & Silberschatz, 1986) by compliance (Gazzillo et al., 2019a). Similarly, an accommodating attitude may actually be a test rather than a positive therapeutic alliance. The researcher and the clinician must therefore be receptive to two aspects of the patient's (non)verbal communication: what is being said and what is potentially meant, the conscious and the unconscious, the facts and their emotional tone. A continuous analysis of one's countertransference is thus essential. The psychotherapist should differentiate between own transference and emotional responses to the patient's transference. Another clinical task is to sustain these feelings and to reflect them rather than act on them. This way countertransference serves as a tool to understand the patient's material.

Last, a test is passed when it disproves pathogenic beliefs (Gazzillo et al., 2019a; Weiss, 1992). How a patient ultimately arrives at letting them go is less clear and may even remain a point of contention largely determined by the therapist’s theoretical and clinical background. Moreover, in many tests an ambivalence is comprised, which ought to be addressed to appropriately and fully respond to them.

Future Research

Determining adequate responses to tests is thus a topic for future research qualitatively analyzing clinical material. Observations of well-trained raters are ideally guided by theoretical constructs and include the treating psychotherapist. Clinical orientation of raters, which will influence outcomes, needs to be reflected. To gain a more comprehensive picture of how patients test, the researcher may want to consider findings of other authors who conceptualize similar interactions as acting out in the transference, actualization, or involving the therapist in enactments. Another recommendation to acquire a deeper understanding of testing is a full objective hermeneutic interpretation according to Oevermann (1979). To analyze entire treatments, the researcher may work with therapist's process notes instead of transcripts. Incorporating the treating clinician will lead to more comprehensive results. The researcher gains access to countertransferences that the treating therapist decided not to reveal in the therapeutic situation. Conducting studies with different methodologies seems essential.

Strengths and Limitations

The here proposed typological classification of testing is subjected to objectivity criteria of empirical positivism. Gathered by well-experienced treating clinicians and confirmed by independent researchers, the meanings of tests are more objective than observations of a single individual.

Nevertheless, the perspective that generating a comprehensive depiction of a complex phenomenon by researchers of various theoretical and clinical backgrounds is enforced. As a heterogeneous research team, we avoided a one-sided focus due to underlying implicit theories of researchers, which cannot be completely ruled out in observations. While our subjective countertransferences and personal inclinations to favour either self-worth or guilt themes, may have influenced the analysis of transcripts, we reflected well on their implications in our discussions. The main goals were to avoid a hasty categorization of tests and to make the manifestations of unconscious aspects in testing visible. A great strength of this study is that it leaves open the possibility for dialogue, which does justice to the object of research: complex case-specific therapeutic processes determined by interrelated and overdetermined aspects. The current work intends to encourage the clinician and the researcher to bear not knowing, to leave space for the enigmatic in order to avoid a constricted view of a patient. Remaining differences among raters may furthermore reflect the multidimensional nature of tests (Sammet et al., 2007b) and complement the clinical picture.

The scope of this study, however, only allowed to examine unconscious testing that directly involved the therapist. Tests to which the therapist did not respond were either overlooked or could not be rated according to the previously set rules. While, we have shown that unconscious testing takes place in non-analytic treatments as well, it must be concluded that we likely missed tests especially in this treatment modality. Another limitation is that we did not trace the entire therapy process.

Conclusion

In general, testing seems problematic to operationalize because it is a highly individual phenomenon entailing the difficulty of dissecting complex and interrelated tests. The patients’ intentions for testing are largely unconscious, not easily identifiable and ultimately revealed in the therapeutic process (Shilkret, 2002). Nevertheless, it is worthwhile to incorporate the concept of testing in clinical practice since it offers a comprehensive conceptualization of various patient interactions and expressions. Despite its complexity, it provides a framework to understand patients’ key problems and allows making sense of feelings elicited in the therapist. The current study shows how the gap between empirical research and clinical practice can be overcome.

Disclosure Statement

No potential conflict of interest was reported by the author(s).