Parallel action preparation has previously been

shown in PMd (Cisek and Kalaska, 2005) and PRR (Scherberger and Andersen, 2007), but in those studies the actions were specified by distinct stimulus cues. Here, Klaes et al. show that a single stimulus can specify two actions, revealing the simultaneous application of two different transformation rules in parallel. Interestingly, the direct goal engaged neural activity earlier than the inferred, consistent with prior studies showing that responses oriented directly toward stimuli are processed more quickly than responses requiring remapping (Crammond and Kalaska, 1994). This suggests that the information for specifying the direct goal may be processed along a simple parietal-to-frontal route, while information for Talazoparib cell line selleck inhibitor specifying the inferred goal may need to pass through prefrontal cortex and then be sent back to premotor and parietal regions. Indeed, an earlier study from the same lab showed that unlike direct goals, inferred goals

were represented in PMd before appearing in PRR (Westendorff et al., 2010). Of course, in many situations, we make decisions that are unrelated to any particular action. When choosing between university courses, one presumably is not planning routes for walking to class. Obviously the brain is capable of making abstract decisions that do not involve action, and many studies have examined the neural mechanisms which may be involved. For example, in a paradigm similar to that used in Klaes et al., 2011 and Bennur and Gold, 2011 compared how monkeys judged the direction of visual motion when they either did or did not know what saccadic response would be used to report their decision. It was found that even before a saccade plan could be made, some cells in parietal cortex were selective for the motion direction of the visual stimulus. In the reach-planning system, Nakayama et al. (2008) showed that premotor activity is selective even when monkeys are only given a “virtual” action plan, specifying whether the rightmost

or leftmost of two stimuli will be the target for movement but the locations of the stimuli themselves are still not known. In fact, the very same monkeys studied by Klaes et al. were very familiar with this kind of situation, having previously been trained on heptaminol tasks in which the rule was indicated before the spatial target (Westendorff et al., 2010). In those cases, one might imagine the competition took place between the rules, and then later, also between the actions (Figure 1C). Since animals are clearly capable of making decisions between abstract rules, then why should they, in situations such as the experiment of Klaes et al., bother to simultaneously apply two rules to prepare two actions, only one of which can physically be performed? One answer, as Klaes et al. suggest, may be that doing so allows animals to make more informed choices.