Search

Blunted motivation is a core symptom of Major Depressive Disorder (MDD). Although the empirical picture is mixed, cognitive processes that can be collectively referred to as reward processing have been found to be consistently muted in MDD; most notably, reward sensitivity and reinforcement learning. Works on the modulation of response vigor in individuals with MDD have examined various types of reward, but recent research has shown that in the general population, response vigor is not modulated by type of reward on tasks that are highly similar to those used in these experiments. The present study implemented a form of non-reward related reinforcement which has repeatedly been shown to modulate response vigor in the general population. It investigated whether modulation of response vigor by this type of reinforcement would be effective in individuals with MDD. Clinically depressed individuals (N = 121; 76 post-exclusion) engaged in a task in which their responses led to predictable and immediate sensorimotor effects, or no such effects. Response vigor increased when responses led to sensorimotor effects, which was comparable to the increase found in the general population. These findings support the utility of isolating the computations leading to different reinforcement types and suggest that motivational deficits in MDD may be specific to the type of reward (i.e., hedonically or otherwise explicitly desired stimuli). These results contribute to the literature by suggesting that the reinforcement from sensorimotor predictability stems from processes devolved to motor control, whereas reinforcement from rewards may depend on more general-purpose processes. [ABSTRACT FROM AUTHOR]

An important model for explaining humans' feeling of agency—the Comparator model—draws on ideas used to explain effective motor control. The model describes how our brain estimates the degree of control over the environment offered by a specific motor program (in short, an action's effectiveness). However, given its current level of specification, the model is at best vague on how (or even whether) the prediction of effectiveness of an action is dynamically updated. To test the issue empirically, our participants performed multiple experimental blocks of a task (reliably shown to measure reinforcement from effectiveness) in which blocks with and without action-effects (or with spatially unpredictable feedback) were interlaced. This design creates a sinusoidal-like objective increase or decrease in effectiveness (quantified as the n-trials back probability of receiving feedback), which participants were unable to report. As previously found, response speed indexed reinforcement from effectiveness. The results suggest that reinforcement from effectiveness is sensitive to both the degree and trend of effectiveness; that is, reinforcement is sensitive to whether it is increasing, decreasing, or is unchanged. Given the previous links made between reinforcement from effectiveness and the computation of effectiveness by the motor-system, the results are the first to show an online, dynamic and complex sensitivity to a motor-programs' effectiveness that is directly translated to its production. The importance of testing the so-called sense of agency in a dynamic environment and the implications of the current findings for a dominant model of the sense of agency are discussed. [ABSTRACT FROM AUTHOR]

Automatic imitation refers to the act of unintentionally mimicking observed actions. Inspired by a theoretical framework that allows for controlled yet unintentional processes, we tested whether automatic imitation depends on the task relevance of the to-be-imitated movements. Replicating previous results, we find that movements that are part of the participant's task set unintentionally influence response. Our key finding is that participants generally do not imitate similar and highly familiar movements that are not part of the task set and hence are task-irrelevant. Furthermore, the results of computational data modeling are consistent with the notion that task-relevance modulates the mental activation of information, as posited by the above theoretical framework. Our findings are not predicted and cannot be explained using current accounts of automatic imitation, such as Associative Sequence Learning or Theory of Event Coding. At a broader level, the key contribution of this study is in challenging the empirical basis for automatic imitation by showing that the effects interpreted as imitation occur only for task relevant responses. This pattern lends itself to a different interpretation which is not related to imitation, automatic or otherwise, but rather to the general phenomenon of response compatibility effects. [ABSTRACT FROM AUTHOR]