The core ideas of the embodiment perspective propose that affective and bodily components underlie adaptive behaviors and contribute basically to the understanding of the ongoing relationship with the environment. When the body moves, the components involved account for the fundamental interaction in progress with the world by taking the online relation into consideration, hence contributing—in part—to building our subjective feelings and thoughts (Semin & Smith, 2008). In this way, the embodiment theories establish some interdependent links among motor actions and affective and motivational states, as well as cognitive processes, and support the fundamental idea that cognition co-performs with the body (conceptual processing, Barsalou, 1999, 2008; judgment, Cretenet & Dru, 2004; language, Glenberg & Kaschak, 2002; Zwaan & Taylor, 2006).
Following this orientation, the studies below aim to analyze precisely how multiple behavioral and motor variables influence cognitive functioning. The main question is indeed to determine the global and predominant mechanisms by which different motivational bodily components match and impact cognitive functioning with flexibility measures. Within this scope, a recent research article (Cretenet & Dru, 2009) has shown that a single combination of laterality (the side where the motor behavior is performed) and arm flexion versus extension variables, connected to the fundamental motivational and affective systems of approach and avoidance, determines flexibility functioning. To reveal a potential overall mechanism of embodiment in line with these findings, and to call upon more bodily components, the two studies below aim to examine the impact on flexibility functioning of a more aggregate combination of peripheral and motivational cues, that is, some bilateral behaviors of flexion versus extension of the arms.
1.1. Flexibility: Conceptual view and dimensions
To do this, we will refer to flexibility as an indicator of the core mechanisms that underlie cognitive functioning. The concept of flexibility can be defined as an ability to change one’s beliefs, attitudes, and personal habits (Rokeach, 1960; Schultz & Searleman, 2002). This multidimensional concept (and its opposite, rigidity) does not cover just a single process; it belongs to more integrative cognitive processes. For example, cognitive flexibility commonly refers to the ability to switch and shift the thoughts in progress to arouse new responses and/or approaches (e.g., Schooler & Melcher, 1995). Flexibility correlated with fluency: Fluency is one indicator of the continuous and effortless flow within which cognitive processes occur and are performed (e.g., Alter & Oppenheimer, 2009). In flexibility tasks, fluency is thus assumed to reflect the ability to switch more or less rapidly, and it accounts for the speed of the process at work (e.g., Liefooghe & Verbruggen, 2009; Oppenheimer, 2008; Rubinstein, Meyer, & Evans, 2001). Facing the demands of the flexibility task and the different strategies they require, two forms of cognitive flexibility have been distinguished (e.g., Piguet et al., 2005; Tomer, Aharon-Peretz, & Tsitrinbaum, 2007): one reactive and the other spontaneous (Eslinger & Grattan, 1993; Grattan & Eslinger, 1989). Reactive flexibility requires shifting thought and behavior in relation to external cues; that is, to particular demands and contexts. In a given framework of stimulus and activations, one must change the set of responses in accordance with some particular cues relevant to the task. Spontaneous flexibility refers to the ability to generate a diversity of thoughts, ideas, and/or solutions in response to a single question. According to Eslinger and Grattan (1993), this form of flexibility requires one to “generate diverse and sometimes creative solutions by mounting strategies to move among classes and categories of knowledge” (p. 18). It reflects internally driven strategies (Piguet et al., 2005) and is used as a measure of divergent thinking (Guilford, Christensen, Merrifield, & Wilson, 1978).
1.2. Positive affect and aversive states connected to flexibility
In studying the different relationships between affective processes and cognition, some research has examined the influence of various emotional and motivational states on flexible functioning. One of the most prominent works in this theoretical domain was conducted by Isen and her colleagues (for two reviews, Ashby, Isen, & Turken, 1999; Isen, 2000). The main orientation of this research was to focus on the influence of positive affect on cognitive functioning. For example, it has been demonstrated that positive feelings could lead to an increase of inclusive categorization in analogous exemplar rating tasks (Isen & Daubman, 1984) and also to some changes of strategies in decision-making tasks (Carnevale & Isen, 1986). More interesting for our purpose, Isen, Daubman, and Nowicki (1987, see also Estrada, Isen, & Young, 1997) found that positive affect increases cognitive flexibility, while they did not find any effect of negative affect on a test which required association of words in a flexible way. However, a major part of this research focused on the influence of positive affect, but rarely on the influence of negative feelings on cognitive functions (Freitas, Katz, Azizian, & Squires, 2008). Recent studies (Friedman & Förster, 2005a; Förster, Friedman, Özelsel, & Denzler, 2006; see also Friedman & Förster, 2001) showed that aversive motivational states, compared with approach motivational states, can impact some cognitive processing like perceptual and conceptual attention. This research did not focus precisely on cognitive flexibility; it was just suggested that aversive motivational states can influence cognitive functioning in an opposite direction to the effect of positive affect.
To explore and clarify this last proposition, recent models and research have been oriented to study the effect of motivation direction on flexibility and fluency performance (Gable & Harmon-Jones, 2010; Papousek, Schulter, & Lang, 2009) rather than the effect of valence on cognitive functioning. These authors proposed that affective states should be studied with an approach versus avoidance dimension. Following this proposition, research has been developed, studying not only the influences of conscious affective states or motivational orientation on flexibility but also how peripheral activations of motivational cues induce cognitive functioning (Gable & Harmon-Jones, 2010).
1.3. An embodiment view of motivational effects on flexibility
One question about these innovative studies is whether it is possible to influence cognitive functioning in a nonconscious manner, with the use of some peripheral activations. It has been shown that some bodily and motivational cues can activate the systems of approach or avoidance, triggering inclinations that operate independently of affect (e.g., Cacioppo, Priester, & Berntson, 1993) and self-perception mechanisms (Friedman & Förster, 2000; Strack, Martin, & Stepper, 1988). In a seminal paper inspired by the original work of Solarz (1960; see also Chen & Bargh, 1999), Cacioppo et al. (1993) found that performing an arm flexion—as an approach behavior toward the self—influenced positively the judgment of neutral stimuli, while an arm extension—as a behavior moving away from the self—influenced negatively the judgment of these stimuli. This flexion-extension paradigm paved the way for numerous studies. Some of them (Freina, Baroni, Borghi, & Nicoletti, 2009; Van Dantzig, Pecher, & Zwaan, 2008) looked differently at motor behaviors associated with approach and avoidance; arm flexion and extension were considered as approach or avoidance behaviors depending on the distance between the participant and the object he would like to reach (see also Seibt, Neumann, Nussinson, & Strack, 2008) or depending on its form (a closed or open hand; Freina et al., 2009); these results focused on a cognitive interpretation of the motor behavior performed, although an embodied perspective could also be applied to these behaviors, when multiple bodily variables are involved. Beyond the review of the affective and cognitive consequences of the peripheral and behavioral inputs, the impact of the bodily components themselves on cognitive functioning has increased the understanding of the mechanisms of embodiment at work. So, and more appropriately for the purpose of this article, the impact of the effective approach/avoidance motor behaviors on various cognitive performances, such as those associated with the different facets of creativity (e.g., Baas, De Dreu, & Nijstad, 2008), has been studied. For example, by the use of the Gestalt Completion Task (Ekstrom, French, Harman, & Dermen, 1976), Friedman and Förster (2000, see also Friedman & Förster, 2002 and Friedman, Förster, & Denzler, 2007) showed that arm flexion, compared with extension, enhances cognitive restructuring associated with the break of a context-induced mental set. In sum, it appeared that the motivational direction of the motor behavior performed led to an increase (flexion-approach) or a decrease (extension-move away) in flexible functioning. For example, a recent study (Friedman & Förster, 2005b), directly in line with our purpose, examined peripheral activations of the approach and avoidance systems and their effects on flexibility (while focusing mainly on attentional processes). The authors have shown (Study 1) that manipulating exteroceptive motivational cues associated with the approach system, compared with those associated with the avoidance system, increased the ability to shift the focus of attention in a Stroop task. Friedman and Förster (2005b) considered this task as assessing attentional flexibility, whereas others (Schultz & Searleman, 2002) have described this test as a measure of cognitive flexibility or rigidity, because the Stroop task requires shifting and maintaining a perceptual set while suppressing a usual response in favor of a novel response.
All these studies have usually found that different conscious affective states did not mediate the effect of motor behaviors on flexibility (see also Friedman & Förster, 2000; Strack et al., 1988). They also showed that no self-perception mechanisms (inferring what one feels from what one has done; Bem, 1967) could explain the results, as participants did not rate the motor behaviors performed as difficult or agreeable, depending on the experimental conditions. These results suggested that motor behaviors, connected to motivational orientations of approach and avoidance, directly influenced flexible functioning.
1.4. The match hypothesis and its embodied view
As the previous research into the emotion-flexibility links has studied the influence of single affective or motivational activations (valenced affective states and motivational approach vs. avoidance cues), some recent experiments have examined the combination of various motivational cues on flexibility. One set of experiments looked at how a fit between motivational states and other elements in the environment can enhance flexibility (Friedman et al., 2007; Grimm, Markman, Maddox, & Baldwin, 2008; Maddox, Baldwin, & Markman, 2006; Markman, Baldwin, & Maddox, 2005). They found that a match between an induced regulatory focus (prevention vs. promotion) and a reward structure (respectively loss vs. gain) led to greater cognitive flexibility, compared with a mismatch of these conditions. This effect has been extended to a task assessing set shifting, the Wisconsin Card Sort Task (WCST; Maddox, Filoteo, Glass, & Markman, 2010). This matching effect has also been found in an interference task requiring attention allocation (De Lange & Van Knippenberg, 2007).
Recent work on embodiment cognition (Cretenet & Dru, 2009) corresponds to these results and stresses that flexibility could be influenced through the involvement of multiple motivational and motor activations. Added to a psychological perspective which examines the influence of approach versus avoidance behaviors (arm flexion vs. extension; Cacioppo et al., 1993) on flexibility, it has been shown in the neuropsychological domain, that two basic motivational response systems associated with cerebral asymmetries are activated by some lateralized motor behaviors (e.g., Harmon-Jones, 2006; Sutton & Davidson, 1997). Davidson (1984, 2004) showed that for right-handers, motivational approach states are linked with a higher activation of the prefrontal cortex of the left hemisphere, whereas motivational avoidance ones are linked with a higher activation of the opposite prefrontal cortex. Because of the spread of cerebral activations linked to the unilateral muscle contractions (Kinsbourne, 1982, 2006) and the closeness of the cerebral regions involved in motor and emotional processing (Kinsbourne & Hicks, 1978), it appeared that contracting the right hand (for right-handers) led, for example, to the production of a more positive story on the base of an ambiguous picture, and that contracting the left hand led to the production of a more negative one (Schiff & Lamon, 1989, 1994). This rationale was proposed because unilateral muscle contractions activate the contra-lateral motor cortex (the left and right hands are controlled by the right and left hemispheres, respectively). These motor cortices are also located in neighboring anterior regions involved in motivational processing (Harmon-Jones, 2006; Schutter, de Weijer, Meuwese, Morgan, & van Honk, 2008). Casasanto (2009) also recently found that right-handers were different from left-handers in their spatial location of emotional valence, showing that laterality helped to embody abstract concepts associated with valence. Laterality could then be understood as handedness, or as the side on which a motor behavior is performed; that is, performing motor behaviors either on the left or right side.
Cognitive performance was also associated with hemispheric asymmetry. For example, Badre and Wagner (2006) found that the left prefrontal cortex helped to resolve conflict in a switching task. Elsewhere, Gray, Braver, and Raichle (2002) found integration in the lateral prefrontal cortex of motivational states and cognitive functioning in a working memory task, without especially involving flexibility measures; similarly, Watanabe and Sakagami (2007; Sakagami & Watanabe, 2007) have also shown that motivational and cognitive information is integrated in the lateral prefontral cortex (LPFC) with the use of a go/no go task. However, these studies did not focus on flexibility functioning. One possible hypothesis would be that the contra-lateral activation of the prefrontal cortex resulted in revealing the cerebral asymmetry in motivational processing and in cognitive set shifting also. Konishi and colleagues did this (Jimura, Konishi, & Miyashita, 2004; Konishi et al., 2002) by demonstrating that flexible adaptation to changing contexts is one of the central functions of the prefrontal cortex. They found that when negative feedback was administrated in the Wisconsin Card Sorting Test (Milner, 1963)—a test that required flexible adaptation to changing contexts—the right lateral frontal cortex was activated, and that the left lateral frontal cortex was activated during the updating of the response, hence linking the flexible adaptation engaged with the positive approach system predominantly located in the left hemisphere. Whenever updating responses in a switching task cannot be considered only as a flexibility measure, it appears that the asymmetry of the prefrontal cortices is involved in cognitive set shifting and that the left prefrontal cortex is associated with flexible behavior. Set shifting is linked to the same frontal region associated with the processing of motivational approach states. However, the possible connectivity of lateral hemispheric areas responsible for flexibility with motor cortices has not been fully studied yet. It seems interesting to examine this question, bearing in mind that the motivational systems of approach and avoidance would be involved in this connectivity.
Because these overt approach versus avoidance and laterality variables have been shown as both impacting flexible functioning and creative thinking, the effect of their combination constitutes a relevant support for studying an embodied phenomenon linked to different peripheral and motivational cues. The effect of the interaction between these behavioral and motor components on evaluative judgment (Cretenet & Dru, 2004; Dru & Cretenet, 2008) and especially flexible functioning (Cretenet & Dru, 2009), showed a congruence mechanism linked to the activation—by one lateralized variable and one flexion/extension variable—of the same motivational and affective system (Motor Congruence Effect; Cretenet & Dru, 2004). Performing a right flexion of the arm (two motor and behavioral variables that activate simultaneously the same system of approach) or a left extension (two motor and behavioral variables that activate simultaneously the same system of avoidance/withdrawal) increases perceptual, behavioral, and cognitive flexibility, whereas performing a left flexion or a right extension (two motor and behavioral variables that activate simultaneously the two different systems) decreases flexibility (Cretenet & Dru, 2009). It appears that the quality of congruence, whatever the fundamental system that is activated, is one mechanism of embodiment that accounts for the contribution of the motor system to cognitive functioning. No conscious affective states or self-perception mechanisms mediated the effect of motor congruence on flexibility.
1.5. Bilateral and motor activations of flexibility: The match with congruency
As unilateral motor congruent behaviors, compared with noncongruent ones, seem to influence flexible functioning, what results will be obtained when bilateral motor behaviors are involved; and which mechanisms might explain the results? The extension of this kind of internal phenomena of agreement and concordance with bilateral motor activations should reveal the existence of a phenomenon of overall congruence between the cognitive influences of the two unilateral behaviors. Such a mechanism about the influence of two unilateral motor behaviors performed on evaluative judgment has indeed been found by Cretenet and Dru (2008). They showed that it was the match or mismatch between the qualities of congruence of each unilateral motor behavior performed together that determined the evaluative judgment of different valenced stimuli. Performing two congruent motor behaviors together (right flexion/left extension) or two noncongruent ones (right extension/left flexion) leads to different affective judgments than performing simultaneously one unilateral congruent behavior with a noncongruent one (double extension or double flexion). Right flexion and left extension match one another, because they both determine a quality of congruence; left flexion and right extension match one another because they both determine a quality of noncongruence. Double flexion or double extension determines a mismatch condition, because each unilateral behavior involves a different quality of congruence (congruence/noncongruence and vice versa).
In line with these findings, we proposed that the match between the respective quality of congruence of each arm flexion or/and extension increases flexible functioning, whereas mismatch between these qualities impairs flexibility. In sum, the use of bilateral motor behaviors and the knowledge of their respective quality of congruence or noncongruence make possible the study of the influence of a more overall and complex arrangement of peripheral, motivational, and motor cues on cognitive functioning. Fig. 1 represents this pattern of expected results as the main hypothesis of the research of this article.
1.6. The present research: Bilateral motor behaviors and measures of behavioral and cognitive flexibility
The two studies conducted in this article aim mainly to examine the influence of bilateral and motor activations of the fundamental motivational and affective systems in two tasks involving flexibility. As the flexibility notion is usually connected to different mechanisms and methods of measurement, it has usually been equivalent to the ability to shift from one activity to another (e.g., Schaie, 1955; Schaie, Dutta, & Willis, 1991) and to the ability to generate a diversity of thoughts, ideas, and/or solutions in response to a single question (Eslinger & Grattan, 1993). On the one hand, reactive flexibility requires shifting thought and behavior in relation to external cues; that is, to particular demands and contexts. Thus, the first study focused on Motor Cognitive Flexibility by the use of the Capitals test (Schaie, 1955), which gives an account of effective adjustment to shifts in familiar patterns, as a measure of reactive flexibility. It accounted for the ability to shift from a cognitive activity to a motor one in a task in which they are linked. On the other hand, spontaneous flexibility refers to the ability to generate a diversity of thoughts, ideas, and/or solutions in response to a single question. This form of flexibility (Eslinger & Grattan, 1993) requires the generation of diverse and sometimes creative solutions by mounting strategies to move among classes and categories of knowledge. So the second study involved a cognitive flexibility task (the Alternate Use Task, AUT; Christensen, Guilford, Merrifield, & Wilson, 1960), as an operational measure of spontaneous flexibility. It must be noted that reactive flexibility was assessed through a behavioral, motor-cognitive task, whereas spontaneous flexibility was measured with a pure verbal, cognitive one.
Several mechanisms would be tested whenever previous research found a direct effect of motor behaviors on cognitive functioning. Various emotional states associated with the bilateral arm flexion and/or extension performed could mediate the link between bodily activations and flexible functioning. In addition, a conscious experience of ease, low effort, and speed could also explain the potential effects. Fluency mechanisms, conceptually close to flexibility (De Dreu, Baas, & Nijstad, 2008; Hirt, Devers, & McCrea, 2008), such as subjective or conscious fluency, could therefore help to understand the effects depending on the effort and agreeableness associated with the different behavioral conditions (Winkielman, Schwarz, Fazendeiro, & Reber, 2003).