The processing mode theory and psychopathology.
In parallel of the processual aspect (i.e., an abstract level of construal), the processing-mode theory (Watkins, 2008) also hypothesizes that structural factors (e.g., the level of psychopathology, mood state) can account for adaptive versus maladaptive consequences of RNT. The level of psychopathology would moderate the consequences of RNT, so that RNT in clinical or subclinical populations would have unconstructive consequences, whereas RNT in non-clinical populations would not always have such unconstructive consequences. Similarly, the mood state would also moderate the consequences of RNT, with RNT associated with unconstructive consequences when in a negative compared to a positive or a neutral mood (Nolen-Hoeksema, 2004; Watkins, 2008). For example, Lyubomirsky and Nolen-Hoeksema (1993) showed that rumination is associated with reduced willingness to engage in pleasant activities compared to distraction but only in dysphoric people. Similarly, Kashdan and Roberts (2007) showed that post-event rumination about personal disclosure interactions was associated with greater negative affect among individuals suffering from high levels of social anxiety but not among individuals with low levels of social anxiety. The processing-mode theory (Watkins, 2008) therefore highlights the importance of psychopathology or negative mood in negative consequences of RNT.
More recently, Watkins (2011) expanded his processing-mode theory by combining these processual and structural aspects. He suggested that clinical or subclinical populations, such as people suffering from mood or anxiety symptoms, would have a dysregulation in the level of construal adopted. These populations would adopt an abstract level of construal by default, without flexibly shifting to a more concrete level of construal when needed according to the circumstances (e.g., when negative affect is engendered in a context in which it seems difficult or complex to reduce the discrepancy between individual’s current and desired states). Consistent with this theory, when compared to depressed participants, only never-depressed participants experiencing an increase in a sad mood shifted towards the adoption of a concrete level of construal (Watkins, Moberly, & Moulds, 2011). This lack of psychological flexibility in clinical or subclinical populations in the adoption of one level of construal would be responsible for maladaptive consequences. As the dysregulation in the level of construal is not specifically limited to depression but has been shown in other emotional difficulties, such as social anxiety, GAD, or PTSD (Ehring & Watkins, 2008), this bias in adopting an abstract level of construal is considered as a transdiagnostic process causally implicated in the development and maintenance of several emotional disorders.
In summary, the processing-mode theory (Watkins, 2008, 2011) hypothesizes negative consequences associated with the adoption of an abstract level of construal in vulnerable populations (i.e., clinical and subclinical populations), compared to a concrete level. The consequences associated with an abstract or a concrete level of construal would not differ in non-clinical populations. Empirical evidence of this theory comes from two lines of research. The first line of research has examined the differential consequences of each level of construal by comparing non-clinical samples with vulnerable samples selected a priori on the basis of a psychopathology measure (e.g., the level of social anxiety or depression). The second line of research has explored the consequences associated with psychopathology, measured as a continuous variable, by differentiating abstract and concrete levels of construal.
The first series of studies has shown that the psychopathology interacts with the level of construal adopted during RNT to predict mood, valence of thoughts, problem solving, or self-judgments. An abstract level of construal during rumination was associated with more negative self-judgments in depressed participants (Rimes & Watkins, 2005) and less positive thoughts in high socially anxious individuals (Vassilopoulos, 2008), whereas a concrete level of construal during RNT was associated with improved problem solving in depressed participants (Watkins & Baracaia, 2002; Watkins & Moulds, 2005a), more positive thoughts, less anxious mood (Vassilopoulos, 2008) as well as fewer negative self-judgments in high socially anxious individuals (Vassilopoulos & Watkins, 2009). No differences between an abstract and a concrete level appeared in non-depressed (Rimes & Watkins, 2005; Watkins & Baracaia, 2002; Watkins & Moulds, 2005a) or low socially anxious individuals (Vassilopoulos, 2008; Vassilopoulos & Watkins, 2009). However, it is noteworthy that all these studies focused on the interaction effect of psychopathology and the level of construal adopted during RNT on emotion, cognition, or behavior.
The second series of studies has examined the effect of the level of construal adopted using a continuous measure of psychopathology, and more specifically, a trait rumination measure. Indeed, rumination is highly correlated with psychopathology and is considered as a transdiagnostic process implicated in psychopathology. Trait rumination can be therefore considered as an indicator of psychopathology. These studies evidenced that trait rumination interacts with the level of construal to predict mood or intrusive thoughts. Higher levels of trait rumination were associated with less positive affect (Moberly & Watkins, 2006), more negative affect (Watkins, 2004), more thought intrusions, and an increase in heart rate (Schaich, Watkins, & Ehring, 2013) when adopting an abstract level of construal. Higher levels of trait rumination were associated with less negative mood (Watkins, 2004) and a decrease in heart rate (Schaich et al., 2013) when adopting a concrete level of construal. Therefore, these studies add evidence to the hypothesis that psychopathology –or trait rumination– interacts with the level of construal to determine emotion, cognition, or behavior.
In summary, these two lines of research supported the processing-mode theory (Watkins, 2008, 2011) according to which (i) the adoption of an abstract level of construal would be associated with negative consequences in vulnerable populations, compared to a concrete construal and (ii) the level of construal would have no effect in non-clinical populations.
Repetitive negative thinking
However, one limitation of this theory concerns the maladaptive emotional consequences of the level of construal in vulnerable populations: Some studies found a difference between a concrete and an abstract level in these populations (e.g., Moberly & Watkins, 2006; Vassilopoulos, 2008; Watkins, 2004), whereas other did not (e.g., Vassilopoulos & Watkins, 2009; Watkins & Moulds, 2005a; Watkins & Teasdale, 2001, 2004). However, it is possible that cognitive consequences of the abstract level in vulnerable populations would lead to emotional consequences. For example, Rimes and Watkins (2005) showed that depressed individuals experienced an increase in their depressed mood and negative self-judgments when they adopted an abstract level, compared with a concrete level, but that the increase in the negative mood depended upon the increase in negative self-judgments. Future research is therefore needed to determine emotional consequences of the level of construal in vulnerable populations.
As suggested by Watkins (2008), one possible explanation of the maladaptive consequences of abstract RNT is that it promotes generalization. Generalization refers to the tendency to generalize from a single negative or positive event, such as a failure, to the entire self (Carver, 1998). Generalization, especially negative generalization, is associated with depression and anxiety (Carver, 1998; Carver & Ganellen, 1983; Fulford, Rosen, Johnson, & Carver, 2012). Adopting an abstract level of construal would lead to more generalization (i.e., to overgeneralize), therefore leading to cognitive and emotional consequences. Consistent with this explanation, experimental studies showed that the adoption of an abstract level of construal led to a greater negative generalization than the adoption of a concrete level of construal in a sample of dysphoric participants (Van Lier, Vervliet, Vanbrabant, Lenaert, & Raes, 2014) or in a non-clinical sample (Van Lier, Vervliet, Boddez, & Raes, 2015). In the same way, when thinking about a positive event, the adoption of an abstract level of construal also promotes generalization, especially in individuals with high self-esteem (Van Lier, Moulds, & Raes, 2015). Thus, applied to RNT, adopting an abstract level of construal during RNT would promote negative consequences through negative generalization (Watkins, 2008; Watkins, Moberly, & Moulds, 2008). Despite these three experimental studies, more research is needed to determine the role of generalization in negative consequences associated with an abstract level of construal as well as to determine the similarities and differences between abstraction and generalization.
In summary, maladaptive RNT is characterized by a predominance of verbal thoughts over images. This reduced imagery seems to be underpinned by a process of abstraction ( i.e., the adoption of an abstract level of construal or an abstract mode).1 Clinical or subclinical populations, such as people with high levels of depression or anxiety, tend to have abstract RNT, maintained by negative metacognitive beliefs about concrete RNT. Even if more research is needed, the process of abstraction in clinical or subclinical populations would lead to greater generalization and be responsible for negative emotional, cognitive, motivational, and social consequences. Finally, we will develop therapies targeting RNT.
Therapies targeting repetitive negative thinking
Based on his substantial work, Watkins (2008, 2011) has developed an intervention for depression based on the idea that training people to adopt a concrete mode of thinking would decrease the negative consequences of the abstract mode of thinking. This training–known as the concreteness training– has been shown to decrease rumination and depressive symptoms (Watkins et al., 2012; Watkins, Baeyens, & Read, 2009; Watkins & 1 In this dissertation, we will refer to the abstract versus concrete level of construal by using the terms abstract versus concrete mode, or abstract versus concrete level in an undifferentiated way.
Moberly, 2009). More recently, Watkins has developed a more complete therapy grounded in cognitive behavior therapy and focused on rumination: the rumination-focused cognitive behavioral therapy (RFCBT). The RFCBT targets the underlying process of abstraction as well as the avoidance function and lead to a decrease in depressive symptoms through a decrease in rumination (Jacobs et al., 2016; Watkins et al., 2007; Watkins, Mullan, et al., 2011). An internet-based RFCBT has also been created (Cook & Watkins, 2016; Topper, Emmelkamp, Watkins, & Ehring, 2017). A recent meta-analysis conducted by Querstret and Cropley (2013) showed that these therapies –among others such as those using mindful techniques– are effective at reducing RNT.
To conclude, RNT, including both rumination and worry, is considered as a transdiagnostic process implicated in many psychological disorders. It is characterized by a predominance of verbal thoughts over mental images, has an avoidance function, and is underpinned by an abstraction process. Several disorder-specific and transdiagnostic measures have been developed, with some of them differentiating between maladaptive and adaptive forms of RNT. Therapies targeting RNT and more specifically abstract RNT have been created and seem to offer a promising area for treating psychological disorders with RNT (e.g., depression, GAD). Alongside this large body of research focused on verbal thoughts, another field is interested in mental images of events, more especially in the visual perspective adopted in mental images, and will be developed.
Table of contents :
PART I. THEORETICAL SECTION
CHAPTER 1. Repetitive negative thinking
1. Definition of repetitive negative thinking
2. Repetitive negative thinking as a transdiagnostic process
3. Processes implicated in the development and the maintenance of repetitive negative thinking
4. Predominance of verbal thoughts over images
5. An underlying process of abstraction
6. Therapies targeting repetitive negative thinking
CHAPTER 2. Visual perspective in mental imagery
1. Definition of mental imagery
2. Visual perspective in mental imagery as a transdiagnostic process
3. An avoidance function
4. An underlying process of abstraction
5. Therapies targeting mental imagery
PART II. EMPIRICAL SECTION
Synthesis and overview of the empirical section
1. Repetitive negative thinking and visual perspective: A special relationship?
2. Previous studies on the relationship between repetitive negative thinking and visual perspective
3. Overview of the empirical section
CHAPTER 3. Mental rumination and visual perspective: Common function and process?
CHAPTER 4. Manipulation of the abstraction process during rumination
CHAPTER 5. Manipulation of the abstraction process
CHAPTER 6. Repetitive negative thinking and visual perspective: A special relationship? A meta-analysis
(1) A special relationship between repetitive negative thinking and visual perspective?
(2) An avoidance function
(3) An underlying process of abstraction
(4) Maladaptive consequences of adopting an observer perspective
Clinical implications and future perspectives