, 2008). This competition can be biased by many factors, such as expected gain (Platt and Glimcher, 1999), subjective strategies (Dorris and Glimcher, 2004), or indeed any factor relevant to the choice. Dorris and Glimcher (2004) proposed the term “relative subjective desirability” to imply that Panobinostat chemical structure what modulates neural activity during decision tasks is a subjective variable that depends upon the relative desirability of one option versus another. Klaes et al. show that the modulation of neural activity is indeed related to subjective desirability. A recent study in
our lab (Pastor-Bernier and Cisek, 2011) shows that this neural modulation is related to relative, rather than absolute, desirability. In our study, monkeys made decisions between two targets whose stimulus features indicated how many drops of Selleckchem SP600125 juice each was worth, and we examined whether neural activity in PMd reflected a competition between the two potential reaching actions. As expected, we found that neural activity increased as the value of the preferred target increased while the other target’s value was constant. We also found that if we kept the preferred target’s value constant
and increased the other target’s value, neural activity decreased, suggesting a competitive interaction. Most importantly, if only a single target was presented then neural activity was completely insensitive to its value—strongly suggesting that in all cases, activity specifying potential actions is modulated by the subjective desirability
of those actions relative to other options. This further strengthens the proposal made by Klaes et al. that the modulation of activity in PMd and PRR reflects subjective preferences for one action goal over another. The question of how the brain makes decisions is the ADAMTS5 topic of many recent and ongoing studies. Klaes et al. provide a critical piece of the puzzle by showing that the brain is capable of simultaneously applying two rules to the same sensory information in order to specify two parallel potential action goals in the sensorimotor regions of frontal and parietal cortex. They show that these activities do not simply reflect sensory information, nor do they simply reflect the motor options, but that they reveal the animals’ strategies and subjective preferences. Taken together with other studies cited here and in Klaes et al., these findings support an “intentional” framework for sensorimotor behavior (Shadlen et al., 2008), whereby the brain makes decisions about actions through a biased competition taking place within the same system that guides the execution of those actions (Cisek, 2006). Although the brain can also make purely perceptual decisions in situations where no response has yet been specified (e.g., Bennur and Gold, 2011), the strategy of specifying multiple potential actions appears to be adopted in all situations in which it is possible.