Motor and cognitive integration: Effect of bilateral behaviors on judgment

Highlights

• Approach–avoidance behavior influences the way individuals process one informer for judging.

• Information Integration Theory examines the judgment rules when two informers were available.

• We study how bilateral approach–avoidance behaviors determine integration rules for judging.

• It was the symmetry of motor behaviors which influence the implementation of particular rules.

• Motivational direction, within complex movement, was not sufficient for explaining the results.

Abstract

Performing approach vs. avoidance behaviors (arm flexion vs. arm extension) on the one hand, and lateralized peripheral activations (left side vs. right side) of the motivational systems of approach vs. avoidance, on the other hand, have been shown to impact on cognitive functioning (Cretenet, & Dru, 2009), mainly in judgment tasks. When a unilateral motor congruent behavior; that is, a behavior that activates the same motivational system (e.g., flexion of the right arm) was performed during a judgment task, participants' use of complex, interactive information integration rules was facilitated. No effect was, however, found when simpler, additive rules were involved (Mullet, Cretenet, & Dru, 2014). Three experiments are reported here that examined the effect of bilateral motor behaviors (e.g., flexion of the right arm and extension of the left arm) on the implementation of information integration rules. In Studies 1 and 2, two judgment tasks similar to the ones used by Mullet et al. (2014) were used: (a) a complex task in which participants judged a person's attractiveness from personality information, and (b) a simpler task in which they attributed blame according to bad deeds. It was found that similar motor behaviors performed by the two arms (e.g., flexion of both arms), in contrast to dissimilar ones, facilitated the use of complex, interactive information integration rules. No effect was found in the case of simpler integration rules. In Study 3, these results were replicated in a judgment task in which the complexity of the integration rule varied depending on the instructions given. Overall, when bilateral motor behaviors were performed during judgment, facilitation in the use of complex integration rules no longer depended on motivational congruence as in the case of unilateral motor behavior. It depended on symmetry/similarity of behaviors.

Introduction

The present study examined the way information about a stimulus (e.g., a person's attributes) and motor activation at the time of judging (e.g., left arm flexion) are integrated during a judgment process (e.g., judging the person's attractiveness). The present study differed from previous studies into motor and cognitive integration in human judgment in that: (a) information about the stimulus was complex; that is, it was provided by several independent pieces of information, and (b) motor activation was also complex; that is, it corresponded to the activation of both arms.

The simplest type of judgment process is when only one piece of information is available, and no specific motor activation is produced. In this case, the judgment process simply consists in associating a value to the piece of information, in terms of the kind of judgment that is expected (Anderson, 2013). For example, if one is told that a person called Claudia is concerned by others' well being, and if the judgment task is to judge Claudia's attractiveness, then the response is straightforward: Claudia will be judged as rather attractive.

Researchers have examined judgment processed in the case where two (or more) pieces of information were available, and no specific motor activation was produced. In the case of two pieces of information, and according to Anderson, 2008, Anderson, 2013, the judgment process consists (a) in associating a value to the first piece of information, (b) in associating a value to the second piece of information, and (c) in integrating both values into a unitary response, in terms of the kind of judgment that is expected (e.g., a person's attractiveness). Previous studies have shown that, in this kind of situation, information integration processes obeyed simple cognitive rules such as addition, multiplication and averaging (Anderson, 2013). In an early study, Birnbaum (1974) showed that when participants were instructed to judge a person's attractiveness according to two pieces of information about this person, they integrated the information in a complex, interactive way, which is graphically depicted in the left part of Fig. 1. In this panel, the curves ascend: The higher the value of the first piece of information, the higher the attractiveness judgment. The curves are clearly separate: The higher the value of the second piece of information, the higher the attractiveness judgment. Curves diverge on the right: The effect of one piece of information on judgment depended on the level of the other piece of information. It was this divergence of curves that attested the interactivity of the judgment process in this study. Other studies have found simpler information integration processes, like the one that is shown in the right panel of Fig. 1 (e.g., Leon, 1982). In this case, the curves were parallel; that is, the effect of both pieces of information was additive.

Researchers have also studied the effect of motor activation on judgment. In some studies, only one piece of information was available and only one aspect of motor activation was examined: arm flexion vs. arm extension. In this case, as a specific motor activation was produced at the time of judging, then the judgment process consists (a) in associating a value to the piece of information, (b) in “interpreting” the information that comes from the arm considered, and (c) in integrating both elements into a unitary response, in terms of the kind of judgment that is expected. Previous studies have shown that some bodily cues (e.g., arm extension) can activate respectively the systems of approach or avoidance linked to affective valence, triggering inclinations that operate independently from affect (e.g. Cacioppo, Priester, & Berntson, 1993) and self-perception mechanisms. For example, Centerbar and Clore (2006) found that motor approach (vs. avoidance) determined more positive evaluations of stimuli when these stimuli were respectively compatible (positive vs. negative) with the behavior performed.

In other studies, another aspect of motor activation was examined: whether activation was produced by the left or by the right hand. For example, Schiff and Lamon, 1989, Schiff and Lamon, 1994 examined how unilateral muscle contraction (left side vs. right side) leads to different evaluations of stimuli. Previous research has suggested that more positive evaluations were associated with activation of the left cerebral hemisphere, and more negative evaluations were associated with activation of the right hemisphere (Davidson, 1984, Davidson, 1992, Davidson et al., 1990). This cerebral asymmetry, located in the anterior prefrontal region (Sobotka et al., 1992, Sutton and Davidson, 1997), reflects two basic response systems in the brain, which are associated with either approach or withdrawal. Schiff and Lamon, 1989, Schiff and Lamon, 1994 found that right-handed participants contracting muscles in their left hand; that is, stimulating the right hemisphere — linked to the avoidance system, reported more personal negative feelings, whereas participants contracting muscles in their right hand; that is, stimulating the left hemisphere — linked to the approach system, experienced more positive feelings.

In summary, previous studies have shown that, depending on the type of movement — flexion vs. extension or right arm vs. left arm -, the final response was higher or lower than the one that would be expected without the presence of specific motor activation. In other words, motor activation influences the final response. These findings supported the idea that motor activation conveyed information in terms of approach–avoidance that was integrated with the given piece of information at the time of judging.

In yet other studies, researchers have assessed the combined effect of different aspects of motor activation, namely the effect of flexion–extension in combination with the effect of laterality on judgment when only one piece of information was available. In this case, where one piece of information (e.g. Chinese stimuli, facial expression or picture) was available and complex motor activation was systematically produced, the judgment process consisted (a) in associating a value to the piece of information, (b) in interpreting and combining the information that comes from the arm, and (c) in integrating both elements into a unitary response. Cretenet and Dru (2004) demonstrated that when right arm flexion was performed; that is, when there was congruence of an approach–approach type in the motivational system, judgments were more positive. When left arm extension was performed, that is, when there was congruence of an avoidance–avoidance type, judgments were also more positive. In contrast, when right arm extension was performed; that is, when there was incongruence of an approach–avoidance type or when left arm flexion was performed; that is, when there was incongruence of an avoidance–approach type, judgments were more negative.

What these studies illustrated was that motor integration did not obey an additive rule of the kind: Laterality + Flexion/Extension. Motor integration was interactive: (a) when motor activation was congruent in terms of what each component meant at a motivational level, the judgment was more positive and (b) when motor activation was incongruent, the final judgment was more negative.

Finally, researchers have assessed the combined effect of different aspects of motor activation on judgment when two pieces of information (not just one) were available (Mullet, Cretenet, & Dru, 2014). In this case, the judgment process consisted (a) in associating a value to each piece of information, (b) in interpreting and combining the information that comes from the arm, and (c) in integrating all the information — cognitive and motor — into a unitary response. In this higher-level case, several judgment processes can be envisioned but the simplest one would be: (a) integration of cognitive information followed by (b) integration of motor information, followed by (c) integration of both products of previous integrations. Mullet et al. (2014) showed that in this complex case, the cognitive integration process itself was affected by motor activation. Motor congruence, as defined above, facilitated the use of interactive — conjunctive and disjunctive — information integration rules (e.g., in an attractiveness judgment task). Motor incongruence hindered the use of interactive information integration rules. In addition, motor congruence or incongruence had no effect when the judgment process was expected to be non-interactive (e.g., a blame attribution task). In other words, motor congruence, as an interaction between two motor and motivational tendencies, impacted on the judgment process when interactive rules were used, while no effect was observed when additive rules were used. These findings were consistent with the findings from previous studies (Cretenet & Dru, 2004) that have shown that at each level of complexity, the combination rule that is operative cannot be simply deduced from the combination rules that have been observed at lower levels of complexity. Mullet et al. (2014) explained their findings by suggesting that non congruent motor behavior hindered the use of cognitive resources, owing to motor mismatching in terms of the motivational system, whereas congruent motor behavior facilitated the use of cognitive resources, owing to motor matching in terms of the motivational system.

In summary, when only one piece of information is available, motor activation adds or subtracts some value to the response. In some way, it acts like a second piece of information. When two pieces of information are available, motor activation acts in another, completely different way. It alters the way the pieces of information are combined; that is, it alters the judgment process (the integration rule used). In the case of motor congruence, information integration is facilitated, whereas in the case of motor incongruence it is hindered. The present study addressed an even more complex issue: What happens when both arms perform systematic flexions or extensions? In other words, what happens at a higher level of motor activation? In Cacioppo et al.'s (1993) initial study, participants were asked to flex and extend their two arms before ranking neutral stimuli along a valence scale. Arm flexions or extensions were associated with the activation of the approach and avoidance systems with valence, which might be in contradiction with more recent studies presented above in this area (for example the hemispheric asymmetry account). Flexing or extending the two arms could, however, be considered as situations where similar behaviors are produced.

As a result, one possibility suggested by the findings of Mullet et al. (2014), is that (a) symmetry and similarity, in terms of types of behavior performed by the two arms, facilitate the use of complex integration information rules, and (b) asymmetry and dissimilarity hinder the use of complex integration information rules. The initial rationale for this hypothesized pattern of effects can be derived from evolutionary psychology. Wynn (2002) suggested that symmetric patterns in general are preferred by the organism because they serve biological functions. Reber (2002) provided evidence that stimuli are objectively and subjectively easier to perceive when they are symmetrical (see also, Makin et al., 2012, Reber and Schwartz, 2006, Reber et al., 2004). Makin et al. (2012) found that electromyographic activity in the Zygomaticus Major — the muscle responsible for smiling, associated with fluency processing (Cannon, Hayes, & Tipper, 2010) — increases when participants have to process symmetrical stimuli. Topolinski (2010) and Sparenberg, Topolinski, Springer, and Prinz (2012) found that it is not perceptual similarity per se that increases preference for a target, but similarity between motor behaviors in an action vs. observation task about this target. Sparenberg et al. (2012) found that it was effector matching (the same leg or arm of the actor and the observed stimuli) which was responsible for increased preference, instead of movement matching in terms of direction. However, no motivational information was associated to these behaviors. As a result, effector similarity/symmetry (double flexion or extension), rather than motivational similarity might also be at work for fluency and judgment when pieces of information have to be assessed.

Another rationale for this hypothesized pattern of effects can be derived from phenomenology. Common experience shows that it is easier to flex both arms or to extend both arms than to flex one arm and to extend the other. As a result, and given previous findings showing that motivational conflicts impact cognitive integration (Mullet et al., 2014), a facilitative effect in information integration should be observed when both motor behaviors are similar (performed in a symmetrical way) and a hindering effect should be observed when both behaviors are dissimilar (performed in a asymmetrical way).

This hypothesized pattern of effects is not necessarily in contradiction with previous findings. As indicated before, previous research has emphasized that, at each level of complexity, the way motor integration impacts on cognitive integration cannot be simply deduced from what has been observed at lower levels. In other words, passing from one level of behavioral complexity represented by unilateral motor behavior, as in Mullet et al. (2014) to a higher level of behavioral complexity represented by bilateral motor behavior, as in the present study, could be accompanied by a change in the pattern of effects, from a pattern interpretable in terms of motivational congruence to a pattern interpretable in terms of symmetry/similarity of behaviors. The findings by Cacioppo et al. (1993), who instructed their participants to perform a double flexion or a double extension, suggest that the latter pattern is more plausible than the one initially derived from the approach–avoidance perspective.

Study 1 tested the hypothesis that when similar/symmetrical arm behaviors are performed (e.g., extension of both arms) during a complex, interactive judgment task, information integration is facilitated, whereas when dissimilar/asymmetrical arm behaviors are performed, information integration is hindered. The judgment task examined in Study 1 was the person's attractiveness judgment task already used by Mullet et al. (2014).

Study 2 tested the hypothesis that when similar/symmetrical arm behaviors are performed (e.g., extension of both arms) or when dissimilar/asymmetrical arm behaviors are performed (e.g., left arm extension and right arm flexion) during a simple, additive-type judgment task, information integration would be neither facilitated nor hindered. The judgment task examined in Study 2 was the blame attribution task already used in Mullet et al. (2014).

Study 3 aimed to replicate the findings from Studies 1 and 2 using a single judgment task; that is, using the same stimuli and the same judgment scale. Indeed, since two different judgment tasks were used in Studies 1 and 2, it was important to check for the possible effect of the tasks' contents on the observed phenomena. In Study 3, a judgment of performance task was used, and the nature of the expected judgment process was varied using different instructions (easy vs. difficult task to be performed).