Arnoud Arntz, PhD
Department of Clinical Psychology, University of Amsterdam; Department of Clinical Psychological Science, Maastricht University
One of the most important cognitive structures conceptualized by cognitive theories of psychopathology is the schema. Beck (1967) introduced the schema concept in the context of cognitive therapy, stating that “a schema is a structure for screening, coding, and evaluating the stimuli that impinge on the organism” (p. 283). From an information processes point of view, it can be thought of as a generalized knowledge structure in memory that represents the world, the future, and the self. It is thought to govern such information-processing elements as attention (what to focus on), interpretation (what meaning is given to stimuli), and memory (what implicit or explicit memories are triggered by specific cues). Schemas can consist of verbal and nonverbal knowledge.
Core beliefs are the verbal representations of the central elements of schemas, sometimes also called central assumptions. Once a schema is activated, selective attentional processes allow much of the available information to remain unprocessed; however, a lot of meaning is added to the raw data when a schema is activated.
Because a schema steers information processing so that information that is incompatible with the schema is overlooked, distorted, or seen as irrelevant, schemas are highly resistant to change once formed. In cognitive theories, schemas bias underlying information processing.
Piaget (1923) first introduced the schema concept to psychology. He distinguished between two major ways people deal with information that is incompatible with an existing schema: accommodation versus assimilation. The default is assimilation: a new experience is transformed to match the existing schema. If the discrepancy is too large, however, accommodation might occur: an existing schema is changed to better represent reality, or a new schema is formed. A basic assumption of cognitive theories of psychopathology is that the very same phenomenon underlies the maintenance of psychopathology: people who suffer from psychopathological problems maintain their schemas by relying on assimilation instead of changing their schemas by accommodation, and it is the task of psychological treatment to help clients change their dysfunctional schemas.
Much of the research into cognitive models of psychopathology, and many therapeutic techniques of cognitive therapy, focus on biased information processes and their modification. This is somewhat surprising because the cognitive model suggests that it is better to focus on schemas rather than cognitive biases. After all, it is the schema that arguably underlies cognitive biases, and if changing cognitive biases does not result in schema alteration, change will be fragile and risk of relapse might be large. While it is true that correcting cognitive biases might lead to schema change (accommodation) if disconfirming information cannot be ignored, schema change or the formation of new schemas is difficult and thus may need to be facilitated by guided work.
Before addressing schema change, it is helpful to distinguish three layers of beliefs. At the core are unconditional beliefs, which represent basic assumptions about the self, others, and the world. Examples are “I am bad,” “I am superior,” “Others are irresponsible,” “Other people are good,” and “The world is a jungle.” The first layer around the core consists of conditional assumptions, which are beliefs about conditional relationships that can be formulated in “if…, then…” terms; for example, “If I let other people discover who I really am, they will reject me”; “If I get attached to other people, they will abandon me”; “If I show weakness, others will humiliate me.” So-called instrumental beliefs, which represent beliefs about how to act to avoid bad things and acquire good things, constitute the outer layer. Examples include “Check the hidden motives of others,” “Avoid showing emotions,” and “Be the boss.” This ordering of beliefs not only reflects different types of beliefs, but it also distinguishes what is apparent at the surface (observable behaviors reflecting instrumental beliefs) and what is behind the surface.
Cognitive theory posits that it is necessary to change the behavioral and cognitive strategies that are governed by the outer layer of instrumental beliefs before change at the level of core beliefs is likely to occur. In large part, the strategies that follow from instrumental beliefs determine what situations clients will enter; how they will manipulate the situation, and thus how other people will behave; or what information they will pick up. Thus, without changes in strategies, information that disconfirms the existing conditional and core beliefs will not be available or processed and therefore cannot lead to schema change.
Schemas and core beliefs start to develop very early in life, even at preverbal levels. A well-known example is attachment. Based on an inborn need for proximity to and soothing behavior from caregivers, especially at moments of stress, babies start to develop attachment representations that can have a lasting influence on later development, including that of self-esteem, emotion regulation, and intimate relationships. For instance, children who experience a secure attachment to caregivers tend to develop healthy self-worth and positive views of others, implying that they tend to trust others and equally respect their own and other people’s needs. Children who experience insecure attachment tend to develop negative views about self and others. But later-formed schemas, and therefore core beliefs, can also contain nonverbal meanings. Although we can describe core beliefs in words, this does not necessarily mean they are represented in a verbal way in memory. One implication is that pure verbal ways of trying to change beliefs might fail (clients might say, “I see what you mean, but I don’t feel it”), and other methods are needed.
One way schemas form is through direct (sensory) experience. Classical and operant conditioning play a role; for example, when a child is repeatedly punished when expressing negative emotions, it may result in core beliefs like “Emotions are bad” and “I am a bad person (because I experience these emotions).” A second way is through modeling: seeing how other people act offers a schematic model the child internalizes. A third way is through verbal information, such as stories, warnings, or instructions. Lastly, because people try to make sense of experiences and information, the way the individual reasons plays a role in the formation of schemas. This means that intellectual capacities and therefore all the influences on these capacities, such as developmental phase, culture, education, and so on, play a role. But this final way also implies a certain coincidence; there is a chance factor in what people make of new information that is condensed in a schematic representation.
Understanding the factors that contribute to this “making sense of experiences” is helpful in bringing about change in core beliefs. For example, when mistreated by parents, it is common for children to conclude that they themselves must be bad. Childhood and adolescence are developmental phases in which basic schemas form, but even though schema change is more difficult during adulthood, it is not impossible. Psychological therapy is a method designed to do just that.
In clinical practice, the therapist needs to discover and adequately formulate the core beliefs that underlie the clients’ problems to adequately address them. How is this accomplished?
One way, suggested by Padesky (1994), is that the therapist can directly ask about core ideas the client might have about the self (“What does this say about you?”), others (“What does this say about others?”), and the world (“What does this say about your life/the world/how things generally go?”). To get to the real core beliefs, and to prevent avoidance, it might be important that enough affect is activated while discussing the specific problem.
Another way is to use a structured cognitive technique called the downward arrow technique. The starting point is an automatic thought or an emotion that is triggered in a concrete situation. The therapist then asks what this thought or emotion means for the client (the therapist might add “if that were true”) and continues asking until detecting an unconditional basic idea that apparently lies at the root of the emotional response in the starting situation. Here’s an example:
Client: I was rejected for a job promotion.
Therapist: What does that mean to you?
Client: I don’t meet the expectations.
Therapist: [If that were true…] What does that mean to you?
Client: I make a mess of everything.
Therapist: [If that were true…] What does that mean to you?
Client: I am a loser.
Therapist: [If that were true…] What does that mean to you?
Client: I am nothing.
Note that the therapist doesn’t challenge the intermediate ideas expressed by the client but accepts them for the moment until the core belief is identified. A very similar process can be used to elicit core beliefs about other people (“What does that mean about other people?”) and the world in general.
An additional approach is to ask clients to imagine the situation at the root of the present problem, and ask them what they are feeling and thinking. For example, the therapist might ask the client who was rejected for a job promotion to close his eyes and imagine again the situation in which he got the negative feelings related to learning that he was rejected for the promotion. The therapist instructs the client to imagine the situation as vividly as possible, and next focus on emotions. Then, the therapist instructs the client to let the image go but to stay with the emotion and see if any early (childhood) memory pops up spontaneously. If so, the therapist instructs the client to relive the experience by focusing on perceptual details, emotions, and thoughts. These thoughts might reveal core beliefs; if not, the therapist can ask the client what the experience means for him. Returning to the example of the client who didn’t get the job promotion, he reported that he got a memory of his father ridiculing him as a child about his “stupid” interest in a specific kind of sport, giving him the feeling that he was worthless—“a nothing.” A similar imagery technique can be used to focus on traumatic experiences and discover the “encapsulated beliefs” associated with these experiences.
In identifying core beliefs, it can be helpful to ask clients how they would like to view themselves, and how they would like other people and the world to be. These wishes usually form the opposite of the negative core beliefs of clients. For instance, the client who was rejected for the job promotion might say he would like to see himself as somebody with clear capacities that other people welcome and acknowledge, and that the world should be just.
Belief and schema questionnaires can also be helpful as a starting point to discuss what core beliefs played a role in elevated scores. Exploring particular items that were highly rated can give important clues as well.
It is important that core beliefs be worded in ways that make sense to the client: the therapist should work with the client to find the best formulation, asking the client to rate the believability of it (e.g., How would you rate the belief “I am nothing”?) on a scale from 0 to 100, where 100 is the highest believability. If the rating is not very high, the formulation should usually be adapted—it doesn’t yet reflect a core belief. Sometimes people have dual belief systems, however, believing the core belief in certain conditions but not in others. In that case, it is important to get both believability ratings. For instance, a panic client might state that she fully believes she has a healthy heart, but when experiencing specific physical sensations she believes that she has a dangerous heart condition like angina pectoris.
Three common ways to change core beliefs are with reasoning, empirical testing, and experiential interventions.
Using Socratic dialogues and other rational ways to stimulate clients to reflect on their core beliefs, therapists can cast doubt on these beliefs and bring about a change process. For instance, the arguments in support of and against the belief can be reviewed (pro and con technique), a reinterpretation of the original situation or situations that underlie the belief can be made, and so forth (see chapter 21 for more examples of techniques). The following three specific techniques might be especially useful in changing core beliefs.
There are problems in trying to change core beliefs by reasoning: clients might have limitations in their reasoning capacities, and reasoned insight might not affect the schema. For example, clients might respond with “I see what you mean, but I don’t feel it.” In such situations, empirical testing and experiential methods can help bring about change on a “feeling level.”
Experiments can be used to test the tenability of beliefs. It is important to formulate clear predictions so they can be compared with the observable outcomes of the experiment. Suppose a client believes that he has a weak side that would lead to rejection if discovered by others. The client could test this by sharing with others personal feelings that he considers to reveal his weakness, and then observing how others respond. It is helpful to have clients write out old and alternative beliefs and predictions and how they can be observed before the experiment is done, and then have them write down what they observed as a result of the test. The prediction from this client’s dysfunctional belief may be that others will reject him, resulting in criticism, the ending of a conversation, or the other person not wanting to see him anymore. The alternative prediction could be that others appreciate his openness and show acceptance by saying sympathetic things, sharing intimate feelings, or continuing the relationship. Special care should be taken to prevent clients from using safety behaviors that interfere with the test. If for instance the client only casually mentions a “weakness” while the focus of the conversation is on another topic, chances are high that others will ignore the statement. The client may later say that this proves that they reject him based on his weakness. A proper test would involve sharing his “weaknesses” when others are fully attuned to what he’s saying.
In more severe cases, clients might not yet be able to formulate alternative and more functional beliefs. In this case, a client’s core beliefs seem to be the only representation thinkable. It is best to not yet formulate alternative beliefs until existing beliefs are refuted (see Bennett-Levy et al., 2004, for an extended guide to setting up experiments for a variety of clinical problems).
Empirical tests offer powerful evidence for and against beliefs and are therefore important for belief change. Most clients will be more convinced by evidence they experience themselves than by abstract reasoning.
Experiential methods rely on the capacity of humans to imagine, bringing in new information while sensory, emotional, behavioral, and cognitive channels are activated. Experiential methods got a bad reputation in the 1960s and 1970s when they were wildly applied, but today they are fully integrated into CBT and evidence-based therapy generally. I discuss three major techniques.
If the client retrieves the memory, which is often of a (psychologically) traumatic nature, and emotional arousal is high enough, the therapist can—in fantasy—enter the image and intervene by stopping abuse and neglect, correcting the perpetrator(s), and taking care of the further needs of the child. In other words, the meaning of the original experience is corrected through the experiences of a different end in fantasy. Although the technique does not overwrite the original memory (there is no loss of memory or factual knowledge of what happened), there is often a dramatic change in the meaning of the original event (Arntz, 2012). In less severe cases, or later in treatment, the client can imagine entering the scene as an adult, confronting the perpetrator, and taking care of the child.
In historical role-plays, client and therapist play situations from the client’s past (usually childhood) that contributed to the formation of core beliefs (Padesky, 1994; Arntz & van Genderen, 2009). The client describes the situation and the behavior of the other person, usually (but not necessarily) a parent. (For convenience, I describe role-plays with a child-parent interaction.) Then, the therapist plays the parent and the client the child. This usually leads to a quick activation of the beliefs and accompanying emotions. There are two options for addressing these beliefs: drama reinterpretation and drama rescripting.
With drama reinterpretation, which is used when the child might have misinterpreted the parent, roles are switched. The therapist instructs the client to play the parent and be aware of any thought, emotion, and intention from the parent’s perspective. The therapist plays the client. Afterward, they discuss the client’s experience in the parent role and compare it to the original interpretation. The therapist highlights discrepancies, and the client is stimulated to reinterpret the original situation. With the new interpretation, a third act follows in which the client plays the child, now realizing the new interpretation and thus behaving differently toward the parent (e.g., more assertively asking for attention, because the client realizes that his dad was unresponsive because he was embroiled in his own troubles, not because he viewed his child as worthless).
With the drama rescripting option, the drama equivalent of imagery rescripting is played out. The role-play is restarted at a good moment for intervention, and the therapist intervenes, correcting the parent (stopping abuse, bringing in safety). Note that the parent is, at that moment, not played by anyone (e.g., he or she can be seated on an empty chair). Next, the therapist takes care of the child, saying soothing things, correcting misinterpretations, and offering a healthy explanation (“It is not your fault; your father has a drinking problem and loses control over his frustrations, and that is why he beats you and says these terrible things—not because you are a bad child.”). Later in therapy, or when working with healthier clients, clients can enter the play as an adult, address the parent, and take care of the child (now not played by anybody). The therapist can act as a coach for the client.
In symbolic role-plays, the therapist and client set up a situation that has symbolic relevance for the core belief but has never happened nor will ever happen. An example is the court play, developed to challenge core beliefs about responsibility (Van Oppen & Arntz, 1994). In this role-play a specific accusation related to the core belief is played out as if it has been brought before a court (e.g., “The defendant is guilty of the pedestrian’s death because he had the intrusive thought that a pedestrian might be killed by a car the pedestrian didn’t see, but he didn’t act on the thought and prevent the accident”). The client and the therapist can play different roles (the public prosecutor, the defendant’s advocate, the judge, the jury) and exchange arguments. Experiencing different views on the (fantasy) case helps clients to reconsider their original belief.
Lastly, core beliefs can be tested in present-focused role-plays. In a sense, this is a behavioral experiment done in role-play, in which clients can change roles and take different perspectives, which helps them to discover how they come across to others.
The therapeutic methods described in this chapter are known to be clinically helpful because they change core beliefs (e.g., Wild, Hackmann, & Clark, 2008). A broader focus on the kind of process-oriented research discussed in this volume will be needed to see if methods such as imagery rescripting also alter such processes as cognitive defusion (see chapter 23), self-acceptance (see chapter 24), or mindfulness (see chapter 26), but the earliest steps in that direction support the possibility (e.g., Reimer, 2014).
Core beliefs can be addressed by many interventions, and the position taken here is that it is good to use different channels of change: reasoning, empirical testing, and experiential intervention. Clients probably differ in their sensitivity to each intervention, so it is good to have a choice of interventions and to integrate various channels. In this chapter I stressed the importance of experiencing disconfirming information, and not just trying to convince clients with verbal reasoning. The reason for this is that although the therapist and client can formulate core beliefs in words, these representations aren’t always open to verbal arguments. Clients often need to experience disconfirmation on a sensory and emotional level.
The current thinking regarding the effects of psychological treatment is that old (dysfunctional) schemas and new (functional) schemas compete for retrieval (Brewin, 2006). In other words, with each encounter with a relevant cue, there is a chance that the old schema is activated and the dysfunctional core belief dominates the person. However, basic research suggests that it might be possible to change the meaning of the original knowledge representation (Arntz, 2012). If so, this will have important implications for practice, as changing the original representation is preferable to building a new representation that has to compete with the old one. For example, relapse chances are much higher when two representations have to compete than when the original representation can be changed. Future research will shed light on this issue.
Arntz, A. (2012). Imagery rescripting as a therapeutic technique: Review of clinical trials, basic studies, and research agenda. Journal of Experimental Psychopathology, 3(2), 189–208.
Arntz, A., & van Genderen, H. (2009). Schema therapy for borderline personality disorder. Chichester, UK: Wiley-Blackwell.
Arntz, A., & Weertman, A. (1999). Treatment of childhood memories: Theory and practice. Behaviour Research and Therapy, 37(8), 715–740.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. Philadelphia: University of Pennsylvania Press.
Bennett-Levy, J., Butler, G., Fennell, M., Hackmann, A., Mueller, M., & Westbrook, D. (Eds.). (2004). Oxford guide to behavioural experiments in cognitive therapy. Oxford: Oxford University Press.
Brewin, C. R. (2006). Understanding cognitive behaviour therapy: A retrieval competition account. Behaviour Research and Therapy, 44(6), 765–784.
Hackmann, A., Bennett-Levy, J., & Holmes, E. A. (Eds.). (2011). Oxford guide to imagery in cognitive therapy. Oxford: Oxford University Press.
Holmes, E. A., & Mathews, A. (2010). Mental imagery in emotion and emotional disorders. Clinical Psychology Review, 30(3), 349–362.
Padesky, C. A. (1994). Schema change processes in cognitive therapy. Clinical Psychology and Psychotherapy, 1(5), 267–278.
Piaget, J. (1923). Langage et pensée chez l’enfant (1st ed. with preface by É. Claparède). Paris: Delachaux et Niestlé.
Reimer, S. G. (2014). Single-session imagery rescripting for social anxiety disorder: Efficacy and mechanisms. Doctoral dissertation, University of Waterloo, Ontario. Retrieved from UWSPACE, Waterloo’s Institutional Repository. (hdl.handle.net/10012/8583).
Van Oppen, P., & Arntz, A. (1994). Cognitive therapy for obsessive-compulsive disorder. Behaviour Research and Therapy, 32(1), 79–87.
Wild, J., Hackmann, A., & Clark, D. M. (2008). Rescripting early memories linked to negative images in social phobia: A pilot study. Behavior Therapy, 39(1), 47–56.