Metacognitive theory and therapy views the persistence of negative beliefs and thoughts as a result of metacognitions controllingcognition. This paper describes, with reference to the treatment of generalized anxiety disorder (GAD) and social phobia, howmetacognition contributes to cognitive stability and to change. Metacognitive therapy offers a level of formulation and intervention thatdoes not focus predominantly on challenging the content of negative thoughts and beliefs that are emphasized in traditional cognitivetherapy. The focus of treatment in GAD is on erroneous beliefs about worry and unhelpful mental regulation strategies. In treating socialphobia, a greater emphasis is placed on modifying attention and worry processes and on configuring processing during and afterbehavioral experiments.
THIS PAPER is based on the premise that resistance tocognitive change is a normal feature of information
processing, and in cognitive therapy is a result ofincomplete formulation of the internal factors ormetacognitions involved in controlling and modifyingcognition.
Theory and research in metacognition evolved in theareas of developmental and cognitive psychology (e.g.,Flavell, 1979; Nelson, 1984; Nelson & Narens, 1990), andhas recently been developed as a basis for understandingand treating psychological disorders (Wells, 2000; Wells &Matthews, 1994, 1996). Metacognition refers to cognitionapplied to cognition and may be defined as any knowl-edge or cognitive process that is involved in the appraisal,control, andmonitoring of thinking (e.g., Flavell, 1979). Itis multifaceted and a basic distinction has been madebetween metacognitive knowledge, which is informationthat individuals have about their own thinking and aboutstrategies that affect it, and metacognitive regulation,which are the strategies used to change the status ofthinking.
Our metacognitive theory (Wells, 2000; Wells &Matthews, 1994) describes the generic cognitive andmetacognitive factors underlying emotional vulnerabilityand psychological disorder maintenance. In this theorywe argued that disorder is associated with a nonspecificstyle of thinking that we termed the cognitive-attentionalsyndrome. This consists of repetitive and difficult-to-
-7229/07/18–25$1.00/0006 Association for Behavioral and Cognitive Therapies.ished by Elsevier Ltd. All rights reserved.
control thinking in the form of worry/rumination, anattentional style of threat monitoring, cognitive resourcelimitations, and use of coping behaviors that fail to modifynegative beliefs. Many of the coping behaviors aremetacognitive in nature because they involve mainte-nance of patterns of thinking or attempts to controlthoughts that are unhelpful. The syndrome is evident inthe dwelling and the brooding quality of cognition seen inpsychological disorders. A further marker for thissyndrome is the presence of heightened and difficult-to-control self-focused attention.
According to the theory, much of the knowledge onwhich processing depends is metacognitive in nature. Sothe activation and persistence of the cognitive-attentionalsyndrome in response to stress is dependent on maladap-tive metacognitive knowledge (beliefs). We have sug-gested that metacognitive knowledge (or beliefs) shouldbe formulated separately from the usual beliefs (i.e.,schemas) typically emphasized in cognitive therapy (e.g.Beck, 1976; Beck, Emery, & Greenberg, 1985).
The distinctness and importance of metacognitiveknowledge can be illustrated with reference to general-ized anxiety disorder (GAD). Cognitive models that basethe disorder on general beliefs about the self and world(e.g., “ITm vulnerable; the world is a dangerous place”)fail to account for difficult-to-control, pervasive worry, thecentral cognitive feature of the disorder. Indeed, ageneral schema such as “The world is a dangerousplace” might be associated with behavioral avoidance ofaspects of the environment rather than worry, and yet thelatter is more characteristic of GAD. Traditional schemasmight explain the negative content of thoughts but they
Figure 1. Cognitive model of GAD. From Cognitive Therapy ofAnxiety Disorders: A Practice Manual and Conceptual Guide (p.204), by A. Wells, 1997, Chichester, UK: Wiley. Copyright 1997by John Wiley and Sons, Ltd. Reprinted with permission.
19Cognition About Cognition
do not explain the presence of worry. Worry must arisefrom metacognitions that lead to that particular style ofthinking. As discussed later, the person with GAD haspositive metacognitive beliefs about the use of worry as ameans of coping, but also has negative metacognitionsconcerning its uncontrollability and potential harmfuleffects. This leads to unhelpful patterns of metacognitiveregulation causing pathological worry and anxiety.
The metacognitive analysis can be applied to alldisorders. A basic premise is that it is not particularlyuseful to consider the general self (e.g., “I'm a failure”) orworld schema as the main influence controlling cogni-tion. For example, in the metacognitive model andtreatment of depression (Wells & Papageorgiou, 2004),there are specific metacognitive schemas driving rumina-tive thoughts (e.g., “Ruminating on my feelings will helpme find answers”) that are separate from the moregeneral schemas representing the negative cognitivetriad. Furthermore, metacognitive beliefs and strategiesappear to lead to patterns of thinking that impedeemotional processing and contribute to posttraumaticstress symptoms (Holeva, Tarrier, & Wells, 2001; Roussis &Wells, 2006; Wells & Papageorgiou, 1995; Wells & Sembi,2004a).
Implications of the Metacognitive Analysis forCognitive Therapy
There are many implications of the metacognitiveapproach for developing cognitive therapy for psycholo-gical disorders. In this paper I will confine the focus toconsidering the effect of metacognitive beliefs andstrategies on opportunities for change in GAD and socialphobia. GAD is particularly interesting as a starting pointif we consider worry as a component of most types ofdisorder, because the dysfunctional metacognitionsunderlying it are likely to be nonspecific, basic patholo-gical mechanisms and processes.
Generalized Anxiety Disorder
The metacognitive model of GAD (Wells, 1995, 1997)is based on the principle that metacognitive beliefs,metacognitive appraisals, and thought control strategiesare central factors in the development and persistence ofthe disorder. The model (Figure 1) differs from othercognitive conceptualizations of GAD by emphasizing therole of metacognition rather than maladaptive beliefsabout the world or social self. This approach suggests thatresistance of the worry process to modification intreatment results from failure to modify the metacogni-tions underlying different types of worry.
In the model a distinction is made between two types ofworry, labeled Type 1 and Type 2. Type 1 worry isconcerned with external events and noncognitive internalevents (e.g., physical symptoms), while Type 2 worry
concerns negative appraisal of the individual's ownthought processes. This is essentially worry about worryor meta-worry.
Worrying is typically triggered by an initial intrusivethought that may occur as an image or in the form of a“what if” question (e.g., “What if my partner is involvedin an accident?”). External factors, such as news items orinformation, can act as triggers for these initial intru-sions. Once a trigger is encountered, positive metacog-nitive beliefs about the usefulness of worrying as astrategy for anticipating and dealing with threat areactivated. Examples of positive beliefs include, “Worryinghelps me cope”; “Worrying keeps me safe”; “If I worryITll be prepared,” Such beliefs are thought to be normaland are not specific to GAD; however, the individualwith GAD overuses worry sequences as a predominantmode of coping. Initial worry sequences consist of chainsof catastrophizing thoughts involving “what if” danger-related questions and the generation of potential copingstrategies. This process of Type 1 worrying is associatedwith the activation of anxiety and its cognitive andsomatic symptoms. When the goal of Type 1 worrying isreached (i.e., achieving a sense that one can cope),anxiety and other emotional symptoms decrease. Typi-cally the person with GAD continues to worry until he orshe assesses that he or she will be able to effectively copewith anticipated threat. This assessment is often basedon internal cues such as a “felt sense” that one will beable to cope, or the belief that all-important outcomes
20 Wells
have been considered in sufficient detail. Worrying mayalso stop when competing goals, which have processingpriority, are activated.
GAD develops when the person activates negativebeliefs about worrying. Negative beliefs include the ideathat worrying is uncontrollable and potentially harmfulfor physical, mental, or psychosocial functioning. Exam-ples of negative beliefs include, “Worrying could make mego crazy”; “Worrying is uncontrollable”; “Worrying candamage my body.”During worry episodes, negative beliefsbecome activated and this leads to negative appraisals ofthe worry process. Such appraisals known as Type 2 worryor meta-worry intensify the sense of threat and exacerbateanxiety responses. Anxious responses can themselves bemisinterpreted as a sign of imminent catastrophe such asloss of one's mind, which in turn leads to rapid escalationsof anxiety in the form of panic attacks. In this way therelationship between Type 2 worry and emotion constitu-tes a vicious cycle in which cognitive and somaticsymptoms associated with anxiety can be interpreted asevidence of loss of control and evidence of the harmfuleffects of worrying. Because negative appraisals ofworrying emerging from negative beliefs increase anxiety,it becomes more difficult for the person with GAD toobtain an internal signal indicating that it is safe to stopworrying.
Two further mechanisms contribute to problemmaintenance. These are labeled “behavior” and“thought control” in Figure 1. Subtle behaviors involvingavoidance of situations that might trigger worryingcontribute to the problem since the individual fails todiscover that worrying is subject to personal control.Reassurance seeking may also be used as a means oftrying to decide that there is really nothing to worryabout. This transfers the control of worry to otherindividuals, removing opportunities for the individual todevelop beliefs about self-control. Information search isanother behavioral strategy and consists of reading booksand/or surfing the Internet to find information that mayreduce worrying. Unfortunately, this process can backfireand increase the range of threat-related information theindividual is exposed to, widening the range of worrytriggers.
Thought-control strategies used by the individualconsist of trying to suppress thoughts about worry triggers.Suppression strategies are rarely totally successful andsome empirical evidence suggests that they may becounterproductive (Purdon, 1999; Wegner, Schneider,Carter, &White, 1987). Thus, individuals are engaged in acontrol activity that generates information concerning ageneral inability to control thoughts in a desired way. Thiseffect reinforces negative appraisals concerning mentalcontrol and contributes to negative metacognitive beliefsin this domain. Another strategy important in maintain-
ing the problem is a failure to interrupt the worry processonce it is activated. Because the individual believes thatworrying is uncontrollable, desirable, or part of his or herpersonality, few concerted efforts are made to interruptthe catastrophizing process once it is initiated. Thisdeprives the individual of control experiences thatmight otherwise modify negative beliefs about uncontroll-ability. Even on the occasions when worry is avoided orsuppression is successful, the termination of the worryprocess means that beliefs concerning the dangers ofworrying are not challenged.
Metacognitive Therapy for GAD
The pervasive and difficult-to-control worry processderives from the individual's metacognitive knowledgebase and strategies. Patients often have limited insightinto the metacognitive beliefs that feed such mentalprocesses. Elsewhere we have suggested (Wells & Mat-thews, 1994) that much of this knowledge should beconsidered procedural in form as plans or programs fordirecting cognitive processing.
This model emphasizes the importance of examiningthe patientTs metacognitions that drive the implementa-tion of maladaptive coping. This focus is different fromexisting cognitive therapy, which might focus on challen-ging the content of Type 1 worrying. Such attempts areoften met with resistance and/or the phenomenon ofworry substitution in which one worry concern is replacedwith another concern. The problem is a manifestation ofthe control that metacognitions have over processing styleand regulation strategies.
Metacognitive therapy for GAD focuses first onchallenging negative metacognitive beliefs concerninguncontrollability of worrying. Initially, the therapistconstructs and shares the metacognitive formulation ofthe problem. The therapist emphasizes that worrying is anormal process but has become problematic because ofthe negative and positive beliefs that the person holdsabout worrying and unhelpful strategies that are used toregulate it. Socialization to the metacognitive perspectiveis achieved by contrasting positive and negative beliefsabout worrying and asking the patient what theconsequences of such conflict might be for the regula-tion of worry processes. The therapist asks if worryingwould be a problem if the patient no longer believedthat it was uncontrollable and harmful. Finally, athought suppression experiment is used to illustratehow some thought control strategies are not particularlyeffective.
The first target of treatment is modification ofuncontrollability metacognitions. Evidence and counter-evidence supporting uncontrollability is reviewed. Thepatient is asked how it is that worry ever ceases if it istruly uncontrollable. The worry postponement
21Cognition About Cognition
experiment is then introduced. This consists of askingthe patient to do the following for homework: (a) noticea trigger for worrying and then, without suppressing thetrigger, (b) postpone the Type 1 worry process. Thisshould be postponed until a period of time later in theday, which can be designated as a worry time lastingapproximately 15 minutes. The therapist makes a cleardistinction between suppression and postponement ofthe worry process. The patient is not being asked toremove the content of an initial worry-intrusion fromconsciousness merely to interrupt the Type 1 worryprocess normally engaged as a means of problem solvingor coping. The allotted worry time is not compulsoryand the patient is informed that he or she does not haveto use it.
Tracking of belief in uncontrollability indicates theeffectiveness of the worry postponement experiment. Theexperiment is refined and repeated in order to maximizebelief change. In the next step the therapist asks thepatient to use the allotted worry time and during thatperiod to deliberately try to lose control of the worryprocess. This experiment may then be further refined andthe patient is subsequently asked to deliberately losecontrol of worry at the time it is triggered rather thanpostponing the activity.
Following the effective modification of erroneousbeliefs about uncontrollability and the introduction ofalternative metacognitive strategies, treatment focuseson challenging beliefs concerning the danger of worry-ing. Verbal reattribution focused on these metacogni-tions can be used to weaken such beliefs. Patients oftenequate worry with stress and have an oversimplistic viewthat stress and therefore worry is harmful. One strategyis to de-couple the concepts of worry and stress.Moreover, the oversimplistic view of anxiety as a harmfulagent should be challenged. For instance, stress andanxiety responses are part of the individual's survivalprogram and they would not have functioned effectivelyin evolutionary terms if they had led to death, seriousillness, or psychological breakdown. Behavioral experi-ments are then used in which the patient attempts toproduce feared negative outcomes as a result ofperiodically intensifying the worry process. For instance,one patient was fearful that she could induce a mentalbreakdown through worrying. This was operationalizedas losing touch with reality and seeing things that otherpeople could not see (i.e., hallucinations). In thetreatment session, the therapist worked with her to tryand induce hallucinations by having intense periods ofin-session worrying.
In some cases patients believe that worrying isabnormal and must be a sign of psychological instability.In these instances, the process of worrying can benormalized by asking clients to conduct a mini-survey in
which they ask a range of other people if they engage inworry, how frequently, and if they are distressed byworrying. Often patients are surprised to discover thatother people who do not have GAD experience frequentworry and find worrying difficult to control.
Later in treatment, positive metacognitions about theneed to use worrying as a means of preparation andcoping are the target of therapeutic modification. Somepatients are reluctant to completely abandon worry as acoping strategy. However, the therapist has a range ofcognitive treatment techniques to choose from. First, areview is undertaken of the counterevidence suggestingthat worrying does not assist coping. Second, themechanism by which worrying improves outcomes canbe explored and challenged. Next, the therapistintroduces the worry mismatch strategy in which thepatient is invited to write, in detail, the contents of arecent worry sequence. This is followed by writing outthe “reality” script, which is a description of what trulyhappened in the “worried about” situation. The worryscript is compared against the reality script, with the aimof emphasizing the discrepancy. Identification of such adiscrepancy allows the therapist to question how usefulworrying can be when it does not accurately depictreality. Worry modulation experiments are also used. Here,the client is asked to increase worry on some days anddecrease the frequency of worry on other days whileobserving the effects on coping outcomes. For example,a person can be asked to determine if they cope betteror perform more effectively on the days when they worrymore.
Treatment normally concludes with reviewing alter-native strategies for dealing with intrusions and stressesthat trigger worrying. This consists of building up analternative strategy base (i.e., new knowledge of strate-gies) so that individuals can develop a greater range andflexibility of responses to intrusions and stress that do notnecessitate engagement in catastrophizing. This can beviewed as helping the patient develop an alternative setof plans or subroutines that can be called to directprocessing.
SummaryIn this section, I have described how the metacognitive
analysis of knowledge and strategies informs a particularformulation and treatment of difficult-to-modify worry inGAD. The emphasis is placed on the control of thinkingby beliefs that are metacognitive in nature. Using thisapproach it is not necessary to formulate and treat other(nonmetacognitive) belief domains.
The principles of metacognitive therapy can beapplied to the cognitive-attentional syndrome in otherdisorders. In the next section I will briefly describe howcognitive therapy for social phobia has been abbreviated
Figure 2. Cognitive model of social phobia. From CognitiveTherapy of Anxiety Disorders: A Practice Manual and Con-ceptual Guide (p. 169), by A. Wells, 1997, Chichester, UK:Wiley. Copyright 1997 by John Wiley and Sons, Ltd. Reprintedwith permission.
22 Wells
by focusing on dysfunctional cognitive processes andmetacognitive beliefs driving them.
Social Phobia
The cognitive model of social phobia that we proposed(Clark & Wells, 1995) contains elements that werediscovered as a result of using the metacognitive theoryto guide interviews with patients. Searching for thecognitive-attentional syndrome led to the discovery thatpatients focused attention inward onto an “observer”image of themselves in social situations, and they engagedin worry/rumination in the form of anticipatory proces-sing before social encounters and as a postmortem. Theseprocesses can impede and slow down cognitive change inthe treatment of social phobia.
On entering feared social situations the person withsocial anxiety activates dysfunctional beliefs about thesocial self and performance, leading to concern aboutfailure to create a favorable impression. This is associatedwith a shift in the direction of attention to self-focusedprocessing. Such self-processing is typically characterizedby processing an image of the self from an “observer”perspective, that is to say, from another person's vantagepoint. In this observer image anxiety symptoms and signsof embarrassment or failed performance are seen as highlyconspicuous. This image usually represents an exaggera-tion, but it is assumed by the individual to be accurate.
As a result of negative thoughts and the negative self-image, the person engages in coping behaviors that areintended to avert feared social catastrophe and improveself-presentation. These (safety) behaviors are, however,problematic in several respects and contribute to apersistence of negative beliefs and anxiety. The mechan-isms linking safety behaviors to problem maintenance areas follows:
1. The behaviors support an attributional bias in whichthe nonoccurrence of social catastrophe, such as beingcriticized, is attributed to the use of the behavior andnot the fact that this is unlikely to happen.
2. Some behaviors increase self-focused attention as theyrequire self-monitoring for their execution. Thisinterferes with task-focused processing, renderingsocial performance more difficult, and interferingwith the processing of social feedback that coulddisconfirm negative beliefs.
3. Some safety behaviors exacerbate unwanted symp-toms; for example, wearing extra layers of clo-thing to conceal sweating increases body heat andsweating.
4. Some safety behaviors contaminate the social situationand make the person with social phobia appear aloof,unfriendly, and disinterested in others, which canreduce the opportunity for social disconfirmation of
negative beliefs. The model is presented diagramma-tically in Figure 2.
Apart from the in-situation processing depicted inFigure 2, the person with social phobia tends to engage inworry before and after social encounters. These processeshave been termed anticipatory processing and the postmor-tem. Anticipatory processing typically involves thinkingabout what can go wrong in the situation and planning orrehearsing coping strategies. Such preparation canprevent the person from discovering that social cata-strophe is unlikely. Another problem with anticipatoryprocessing is that it can cause the individual to enter thesocial situation already in a state of heightened anxietyand self-consciousness.
The postmortem consists of mentally reviewing andruminating about what happened in the situation.Unfortunately, because the person was largely self-focusedthere is little encoded in memory that can modify thepersonTs negative self-appraisal. Instead, the postmortemfocuses on negative feelings and negative self-perceptionsand can strengthen beliefs about poor social performanceand the negative self-image.
23Cognition About Cognition
A Metacognitive Approach to Treatment
The focus of initial treatment based on this model hasbeen predominantly cognitive rather than metacognitive,and this has met with demonstrated success (Clark et al.,2003). However, we (Wells & Papageorgiou, 2001)considered it possible to abbreviate treatment withoutloss of overall effectiveness by taking a stronger metacog-nitive approach.
The earlier treatment had a limited emphasis onmetacognitive strategy and no consideration of metacog-nitive beliefs that may impair change. We modified theemphasis to focus more on strategy and on metacognitivebeliefs. This enabled us to enhance cognitive change(e.g., Wells & Papageorgiou, 1998) and develop anefficient, brief treatment (Wells & Papageorgiou, 2001).
Components of brief metacognitive-focused treatment. Thebrief treatment uses the maintenance component of themodel as depicted in Figure 2 as a basic case formulation.According to the metacognitive approach negativebeliefs/thoughts about the social self are products ofthe style of processing driven by the personTs metacogni-tions. Thus, treatment focuses on modifying the style ofprocessing before, during, and after social situations. Anindex of change in this style is the patientTs level of self-reported self-consciousness in the past week in socialsituations. We decided to use this rating as a treatmenttermination criterion rather than general (social) beliefratings. Self-consciousness is a marker for the cognitive-attentional syndrome in the metacognitive theory, andthe hypothesis is that removal of the syndrome shouldfacilitate change, so this is an important measure.
In the first treatment session the therapist introducesthe idea that attentional strategies of self-focusingcontribute to problem maintenance by increasing aware-ness of symptoms, impairing performance, and divertingattention away from disconfirmatory external informa-tion. For example, patient's are asked how they can knowthat “everyone is looking” at them if they avoid eyecontact and focus on symptoms. The therapist thenexamines the positive metacognitive beliefs held aboutfocusing attention on the self and the negative beliefsheld about focusing on the external social environment.The disadvantages of maintaining an attentional strategyof this kind are reinforced.
The effect of attentional focus on symptoms andperformance is contrasted by asking the patient toperform an anxiety-provoking social task under twoconditions. In the first the patient is asked to engage inusual coping behaviors and self-monitoring/control. Thepatient is then asked to predict what will happen if he orshe repeats the experiment while engaging in external-focused attention, dropping all coping behaviors. Thepatient is videotaped performing these tasks in the firstsession (this video is to be used later). Typically, the
patient discovers that contrary to predictions there is nomore evidence of failed or inadequate performance in thesecond condition when compared to the first. In manycases, anxiety and performance “feel better” in thesecond condition. For homework, patients are asked topractice external-focused attention in social situationsand nonsocial situations.
Exercises are provided to help anchor attentionexternally. These include asking patients to try and“feel” the texture of different objects by looking atthem, and paying attention to other peopleTs facialexpressions to determine if they appear tired or rested.
The next step in treatment focuses on the concept ofanticipatory processing and the postmortem. Here thetherapist introduces the idea that worry and ruminationare examples of self-focused attention that do not provideevidence that can modify negative beliefs about thesituation and the self. Next, the advantages of anticipatoryprocessing are elicited as a means of identifying positivemetacognitive beliefs about the process. Such metacogni-tions are then challenged. Similarly, the advantages of thepostmortem are elicited and challenged. The patient isthen asked to modify anticipatory processing so that it istime-limited and not focused on predicting negativeoutcomes. The postmortem is dealt with by reviewing theadvantages and disadvantages, reinforcing the latter, andasking the patient to ban it.
Following from this, video feedback is used tochallenge the validity of the negative observer imageand to illustrate how using self-focus on an internalimpression as a means of judging the self can lead toerroneous conclusions. As in the original cognitivetreatment, the therapist works with the patient toconstruct a detailed and objectified list of the exagger-ated nature of symptoms that the patient expects to seeon the video. The aim is to show that a discrepancyexists between the patientTs internal image and the trueimage.
The next phase of treatment consists of interrogatingthe social environment by displaying signs of anxiety/failed performance and testing predictions concerningthe consequences of doing so. The important feature ofthese experiments that renders them metacognitive isthat the therapist aims to expose the patient to socialencounters while controlling on-line cognitive processesin a way that facilitates belief change. To do so, behavioralexperiments follow the P-E-T-S protocol (Wells, 1997) asoutlined below:
1. Prepare (P): The therapist elicits and rates a negativebelief to be modified in the social situation. Thebelief is configured as a prediction concerning theobservable effects of showing anxiety/poor perfor-mance. The therapist introduces the idea of external
24 Wells
attention focusing on social cues as a processimportant for testing the belief.
2. Expose (E): The patient is exposed to the anxiety-provoking social situation (e.g., a crowded bar).
3. Test (T): The patient is asked to execute a specific testor disconfirmatory strategy. This involves externalattentional monitoring while reversing coping beha-viors or engaging in embarrassing behavior (e.g., Payclose attention to the reaction of others while spilling adrink in a bar).
4. Summarize (S): The results of the experiment aresummarized in terms of the initial prediction andbelief level is re-rated. The therapist refines theexperiment and repeats it as necessary. The patient isasked to ban further postmortem processing followingexposure.
Summary
The brief metacognitive-focused treatment differsfrom the longer cognitive treatment in several respects.There is a greater emphasis on modifying self-attentionwith practice of external attention monitoring within andbetween sessions from the beginning. This is coupled withan emphasis on removing worry/rumination in the firsttwo sessions. Video feedback is used to both correct thecontent of the distorted self-image and also to show howthe strategy of relying on an internally generated self-image can give rise to an exaggerated negative sense ofself. This finding is used to strengthen the case for shiftingto external-focused attention and developing new strate-gies of processing the external social environment ratherthan the inner sense of self. Behavioral experiments forchallenging predictions follow the P-E-T-S protocol, andat least two behavioral experiments are used at eachtreatment session. The brief treatment consists of little orno verbal challenging of negative thoughts, and generalnegative beliefs about the self as a social object are nottargeted. Using this approach and a termination criterionbased on self-consciousness ratings, we have deliveredtreatment in a mean of 5.5 hourly sessions and haveachieved reductions in fear of negative evaluation thatappear to be comparable to those of the full cognitiveintervention.
Conclusion
The metacognitive approach emphasizes the dynamicnature of processing in psychological disorders. In thismodel general beliefs and negative thoughts are theoutput of metacognitions that control the retrieval ofinformation from long-term memory, the direction ofattention, and the implementation of thinking styles(strategies). In this brief paper I have attempted to show,with reference to GAD and social phobia, how psycholo-gical disorder is associated with the maladaptive control
of cognition and the nature of metacognitive beliefs, andnot necessarily the content of the more traditionalschemas.
The contrast between cognitive and metacognitivetherapy can be perhaps illustrated with reference to thenature of the Socratic dialogue used in treatment.Cognitive therapy is characterized by questions such as,“What is your evidence for believing that?” In contrast,metacognitive therapy augments such questioning byintroducing elements characterized by the following:“How have you arrived at that belief?”; “What are youpaying attention to?”; “What are the internal factors thatlead you to that conclusion?” The metacognitiveapproach views the construction of dysfunctional beliefsand appraisals as the result of recurrent dynamic patternsof processing guided by metacognitions. Such metacogni-tions and patterns are central targets for change.
I have outlined some principles of metacognitivetherapy to illustrate how formulating the factors thatcontrol and appraise cognition can provide a newapproach to treatment. This approach has the potentialadvantage of not relying on the direct verbal challengingof the content of negative thoughts and beliefs in thecognitive domain. Instead, it argues for the modificationof beliefs and strategies in the metacognitive realm. Thisapproach has the potential to overcome the resistance tochange seen as recurrence of the worry process in GAD,and it has been used as a basis for accelerating cognitivechange in social phobia.
Metacognitive therapy is not confined to the twodisorders discussed here. Metacognitive models andtreatments have also been developed for posttraumaticstress disorder, obsessive-compulsive disorder, and depres-sion, and are currently under evaluation (Fisher & Wells,2005; Wells, 1997; Wells & Papageorgiou, 2004; Wells &Sembi, 2004a,b). The advantage of a metacognitiveapproach to treating these disorders is that treatmentdoes not require prolonged exposure to memories oftrauma or obsessional stimuli or challenging depressiveautomatic thoughts. A significant proportion of patientsand therapists find it difficult and time consuming tocomply with optimal exposure and cognitive therapypractices.
Metacognitive theory considers a range of factors notdiscussed in detail in this paper. There are, for example,other issues concerning the way in which knowledge isrepresented (as factual information and as tacticalknowledge) and the effect of different mental modes onchange (Wells, 2000). In recent work, Leahy (2002, 2003)has drawn on metacognitive constructs in his formulationof emotional schemas as beliefs individuals hold about thedurability, controllability, and pathology of emotions. Animportant implication of dysfunctional knowledge con-cerning emotions is that it may have an impact on oneTs
25Cognition About Cognition
willingness to accept emotions and be exposed to them,factors likely to affect emotional processing and change.
The formulation of metacognition as a powerfulinfluence on cognitive affective-change and on resistanceto change reminds us of the complexity of thinking. It is acomplexity that cognitive therapy might embrace with aview to discovering a landscape of change beyond thatoffered by the traditional schema principle.
References
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York:International Universities Press.
Beck, A. T., Emery, G., & Greenberg, R. L. (1985). Anxiety disorders andphobias: A cognitive perspective. New York: Basic Books.
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. InR. Heimberg, M. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.),Social phobia: diagnosis, assessment and treatment. New York: TheGuilford Press.
Clark, D. M., Ehlers, A., McManus, F., et al. (2003). Cognitive therapyversus fluoxetine in generalized social phobia: a randomizedplacebo-controlled trial. Journal of Consulting and Clinical Psychol-ogy, 71, 1058–1067.
Fisher, P. L., &Wells, A. (2005). Experimental modification of beliefs inobsessive-compulsive disorder: a test of the metacognitive model.Behaviour Research and Therapy, 43, 821–829.
Flavell, J. H. (1979). Metacognition and metacognitive monitoring: anew area of cognitive-developmental inquiry. American Psychologist,34, 906–911.
Holeva, V., Tarrier, N., & Wells, A. (2001). Prevalence and predictors ofacute stress disorder and PTSD following road traffic accidents:Thought control strategies and social support. Behavior Therapy,32, 65–83.
Leahy, R. L. (2002). A model of emotional schemas. Cognitive andBehavioral Practice, 9, 177–190.
Leahy, R. L. (2003). Emotional schemas and resistance. In R. Leahy(Ed.), Roadblocks in cognitive behavioral therapy. New York: TheGuilford Press.
Nelson, T. O. (1984). A comparison of current measures of feeling-of-knowing accuracy. Psychological Bulletin, 95, 109–133.
Nelson, T. O., & Narens, L. (1990). Metamemory: a theoreticalframework and some new findings. In G. H. Bower (Ed.), Thepsychology of learning and motivation (pp. 125–173). New York:Academic Press.
Purdon, C. (1999). Thought suppression and psychopathology.Behaviour Research and Therapy, 37, 1029–1054.
Roussis, P., & Wells, A. (2006). Posttraumatic stress symptoms: tests ofrelationships with thought control strategies and beliefs aspredicted by the metacognitive model. Personality and IndividualDifferences, 40, 111–122.
Wegner, D. M., Schneider, D. J., Carter, S. R., & White, T. L. (1987).Paradoxical effects of thought suppression. Journal of Personalityand Social Psychology, 53, 5–13.
Wells, A. (1995). Metacognition and worry: a cognitive model ofgeneralized anxiety disorder. Behavioural and Cognitive Psychother-apy, 23, 301–320.
Wells, A. (1997). Cognitive therapy of anxiety disorders: A practice manualand conceptual guide. Chichester, UK: Wiley.
Wells, A. (2000). Emotional disorders and metacognition: Innovativecognitive therapy. Chichester, UK: Wiley.
Wells, A., & Matthews, G. (1994). Attention and Emotion: A ClinicalPerspective. Hove, UK: Erlbaum.
Wells, A., & Matthews, G. (1996). Modelling cognition in emotionaldisorder: the S-REF model. Behaviour Research and Therapy, 32,867–870.
Wells, A., & Papageorgiou, C. (1995). Worry and the incubation ofintrusive images following stress. Behaviour Research and Therapy,33, 579–583.
Wells, A., & Papageorgiou, C. (1998). Social phobia: effects of externalattention on anxiety, negative beliefs and perspective taking.Behavior Therapy, 29, 357–370.
Wells, A., & Papageorgiou, C. (2001). Brief cognitive therapy forsocial phobia: a case series. Behaviour Research and Therapy, 39,713–720.
Wells, A., & Papageorgiou, C. (2004). Metacognitive therapy fordepressive rumination. In C. Papageorgiou, & A. Wells (Eds.),Depressive rumination: Nature, theory and treatment (pp. 259–273).Chichester, UK: Wiley.
Wells, A., & Sembi, S. (2004). Metacognitive therapy for PTSD: apreliminary investigation of a new brief treatment. Journal ofBehavior Therapy and Experimental Psychiatry, 35, 307–318.
Wells, A., & Sembi, S. (2004). Metacognitive therapy for PTSD: acore treatment manual. Cognitive and Behavioral Practice, 11,365–377.
Address correspondence to Adrian Wells, Ph.D., University ofManchester, Academic Division of Clinical Psychology, RawnsleyBuilding, MRI, Manchester, M13 9WL, UK; e-mail: [email protected].
Received: January 18, 2006Accepted: January 30, 2006Available online 30 November 2006