Guiding spatial attention by multimodal reward cues.

Our attention is constantly captured and guided by visual and/or auditory inputs. One key contributor to selecting relevant information from the environment is reward prospect. Intriguingly, while both multimodal signal processing and reward effects on attention have been widely studied, research on multimodal reward signals is lacking. Here, we investigated this using a Posner task featuring peripheral cues of different modalities (audiovisual/visual/auditory), reward prospect (reward/no-reward), and cue-target stimulus-onset asynchronies (SOAs 100-1,300 ms). We found that audiovisual and visual reward cues (but not auditory ones) enhanced cue-validity effects, albeit with different time courses (Experiment 1). While the reward-modulated validity effect of visual cues was pronounced at short SOAs, the effect of audiovisual reward cues emerged at longer SOAs. Follow-up experiments exploring the effects of visual (Experiment 2) and auditory (Experiment 3) reward cues in isolation showed that reward modulated performance only in the visual condition. This suggests that the differential effect of visual and auditory reward cues in Experiment 1 is not merely a result of the mixed cue context, but confirms that visual reward cues have a stronger impact on attentional guidance in this paradigm. Taken together, it seems that adding an auditory reward cue to the inherently dominant visual one led to a shift/extension of the validity effect in time - instead of increasing its amplitude. While generally being in line with a multimodal cuing benefit, this specific pattern highlights that different reward signals are not simply combined in a linear fashion but lead to a qualitatively different process.

Effect of non-instructed instrumental contingency of monetary reward and positive affect in a cognitive control task.

In recent years, we observed a strong interest in the influence of motivation and emotion on cognitive control. Prior studies suggest that the instrumental contingency between a response and a rewarding or affective stimulus is particularly important in that context-which is resonating with observations in the associative learning literature. However, despite this overlap, and the relevance of non-instructed learning in real life, the vast majority of studies investigating motivation-cognition interactions use direct instructions to inform participants about the contingencies between responses and stimuli. Thus, there is little experimental insight regarding how humans detect non-instructed contingencies between their actions and motivational or affective outcomes, and how these learned contingencies come to influence cognitive control processes. In an attempt to close this gap, the goal of the present study was to test the effect of non-instructed contingent and non-contingent outcomes (i.e. monetary reward and positive affective stimuli) on cognitive control using the AX-continuous performance task (AX-CPT) paradigm. We found that entirely non-instructed contingencies between responses and positive outcomes (both monetary and affective ones) led to significant performance improvement. The present results open new perspectives for studying the influence of motivation and emotion on cognitive control at the insertion with associative learning.

Neural underpinnings of valence-action interactions triggered by cues and targets in a rewarded approach/avoidance task.

Incentive-valence signals have a large impact on our actions in everyday life. While it is intuitive (and most often beneficial) to approach positive and avoid negative stimuli, these prepotent response tendencies can also be maladaptive, as exemplified by clinical conditions such as overeating or pathological gambling. We have recently shown that targets associated with monetary incentives can trigger such valence-action biases (target condition), and that these are absent when valence and action information are provided by advance cues (cue condition). Here, we explored the neural correlates underlying the absence of the behavioral bias in this condition using fMRI. Specifically, we tested in how far valence and action information are integrated at all in the cue condition (where no behavioral biases are observed), assessing activity at the moment of the cue (mainly preparation) and the target (mainly implementation). The cue-locked data was dominated by main effects of valence with increased activity for incentive versus no-incentive cues in a network including anterior insula, premotor cortex, (mostly ventral) striatum (voxel-wise analysis), and across five predefined regions of interest (ROI analysis). Only one region, the anterior cingulate cortex, featured a valence-action interaction, with increased activity for win-approach compared to no-incentive-approach cues. The target-locked data revealed a different interaction pattern with increased activity in loss-approach as compared to win-approach targets in the cerebellum (voxel-wise) and across all ROIs. For comparison, the uncued target condition (target-locked data only) featured valence and action main effects (incentive > no-incentive targets; approach > avoid targets), but no interactions. The results resonate with the common observations that performance benefits after incentive-valence cues are promoted by increased preparatory control. Moreover, the data provide support for the idea that valence and action information are integrated according to an evolutionary benefit (cue-locked), requiring additional neural resources to implement non-intuitive valence-action mappings (target-locked).

Shared Neural Representations of Cognitive Conflict and Negative Affect in the Medial Frontal Cortex.

Influential theories of Medial Frontal Cortex (MFC) function suggest that the MFC registers cognitive conflict as an aversive signal, but no study directly tested this idea. Instead, recent studies suggested that nonoverlapping regions in the MFC process conflict and affect. In this preregistered human fMRI study (male and female), we used MVPAs to identify which regions respond similarly to conflict and aversive signals. The results reveal that, of all conflict- and value-related regions, only the ventral pre-supplementary motor area (or dorsal anterior cingulate cortex) showed a shared neural pattern response to different conflict and affect tasks. These findings challenge recent conclusions that conflict and affect are processed independently, and provide support for integrative views of MFC function.SIGNIFICANCE STATEMENT Multiple theories propose that the MFC, and the dorsal ACC in particular, integrates information related to suboptimal outcomes from different psychological domains (e.g., cognitive control and negative affect) with the aim of adaptively steering behavior. In contrast to recent studies in the field, we provide evidence for the idea that cognitive control and negative affect are integrated in the MFC by showing that a classification algorithm trained on discerning cognitive control (conflict vs no conflict) can predict affect (negative vs positive) in the voxel pattern response of the dorsal ACC/pre-SMA.

Expecting the good: Symbolic valence signals provoke action biases and undermine goal-directed behavior.

Positive and negative events are known to trigger opposing action tendencies (approach vs. avoidance). Recently, we found that advance monetary incentive cues can override such valence-action biases. In the present study we tested whether symbolic emotional valence cues can lead to similar adjustments and facilitate performance regardless of the required action. To this end, we performed three closely related experiments in which valence prospect (positive vs. neutral; indicated by stimulus color) and action requirements (approach vs. avoid; indicated by stimulus shape) were manipulated in a trial-to-trial fashion. Orthogonal to this, valence prospect was either embedded in the cue or target stimulus in discrete blocks (cue-valence vs. target-valence blocks). Actual valence was presented in the form of emotional face stimuli after response execution, which mirrors monetary incentive manipulations. In two of the experiments, we observed a positive-approach bias in form of performance benefit for positive versus neutral valence trials, which was exclusive for approach actions. Although numerically more pronounced in target-valence blocks, the bias was not significantly diminished in cue- versus target-valence blocks. This opposes our prediction that emotional valence cues can diminish such biases and instead highlights the robustness of inherent mappings between emotional valence and action tendencies - even if this goes against the task goal.

Interactions between incentive valence and action information in a cued approach-avoidance task.

Environmental stimuli can provoke specific response tendencies depending on their incentive valence. While some studies report positive-approach and negative-avoidance biases, others find no such mappings. To further illuminate the relationship between incentive valence and action requirement, we combined a cued monetary incentive paradigm with an approach/avoidance joystick task. Incentive type was manipulated between groups: The reward group won money, while the punishment group avoided losing money for correct and fast responses to targets following incentive cues. Depending on their orientations, targets had to be 'approached' or 'avoided'. Importantly, incentive valence (signaled by cue color) was orthogonal to action requirement (target orientation). Moreover, targets could carry valence-associated information or not (target color), which was, however, task-irrelevant. First, we observed that both valence cues (reward/punishment) improved performance compared to neutral cues, independent of the required action (approach/avoid), suggesting that advance valence cues do not necessarily produce specific action biases. Second, task-irrelevant valence associations with targets promoted action biases, with valence-associated targets facilitating approach and impairing avoid responses. Importantly, this approach bias for valence-associated targets was observed in both groups and hence occurred independently of absolute valence ('unsigned'). This rather unexpected finding might be related to the absence of a direct contrast between positive valence and negative valence within groups and the common goal to respond fast and accurately in all incentive trials. Together, our results seem to challenge the notion that monetary incentives trigger 'hard-wired' valence-action biases in that specific design choices seem to modulate the presence and/or direction of valence-action biases.

Task switching in autism: An EEG study on intentions and actions.

One of the main symptoms of Autism Spectrum Conditions (ASC) is experiencing cognitive inflexibility when adjustments of behaviour are required. While this so-called behavioural rigidity is broadly recognised in ASC, finding evidence for the underlying neurocognitive mechanisms remains challenging. In this electroencephalographic (EEG) study, participants with ASC and matched controls were instructed to choose between two cognitive tasks in each trial, and to respond to the subsequently presented target stimulus according to their task choice. While doing so, we tracked the frontally distributed contingent negative variation (CNV) during the task preparation interval as a measure of intentional control, and the posteriorly measured P3 during the task execution interval to monitor the translation of intentions into actions. The results support the notion of intentional control difficulties in ASC, where the CNV was attenuated in the ASC group compared to the control group. Furthermore, the CNV was differentiated between the tasks and transition types in the control group only, suggesting that the ASC group was less fine-tuning the required amount of intentional control to contextual circumstances. In contrast, the P3 showed no significant differences between the groups. Together, these findings highlight the importance of intentional control mechanisms as a crucial future route for a better understanding of cognitive flexibility and behavioural rigidity in ASC.

Understanding Behavioural Rigidity in Autism Spectrum Conditions: The Role of Intentional Control.

Although behavioural rigidity belongs to the core symptoms of autism spectrum conditions, little is known about its underlying cognitive mechanisms. The current study investigated the role of intentional control mechanisms in behavioural rigidity in autism. Autistic individuals and their matched controls were instructed to repeatedly choose between two simple cognitive tasks and to respond accordingly to the subsequently presented stimulus. Results showed that autistic participants chose to repeat tasks more often than their controls and when choosing to switch, they demonstrated larger performance costs. These findings illustrate that when required to make their own choices, autistic people demonstrate rigidity at different performance levels, suggesting that intentional control mechanisms might be important for a better understanding of behavioural rigidity in autism.