However, it is likely that in the intact animal, there is a dynamic interdependency between goal-directed and habitual systems and that control is likely Akt inhibitor to emerge simultaneously and competitively (Wassum et al., 2009). If habit and goal-directed processes indeed act concurrently, then this invites questions regarding what precisely are the factors that influence the integration and competition between the two systems. We return to these issues below. It is also worth noting here that although

goal-directed or response-outcome learning has a strong declarative flavor, it is conceptually distinct from a hippocampal-dependent stimulus-stimulus form of learning. There are some alluring parallels with this account of the evolution from goal-directed to habitual responding. One is the transfer of control of a simple spatial behavior (turning in a “plus” maze) from a hippocampal-dependent, allocentric, reference frame to a striatum-dependent, egocentric one (Packard and McGaugh, 1996). Similar double dissociations arise from reversible lesions in these two regions at different time points, for example early or late, during learning.

The other parallel is with the transfer over the course of experience from allocentric to egocentric reference frames of a sequence of manual button presses (Hikosaka et al., 1999), although this was proposed to depend on two separate cortical regions that both interact with the basal ganglia. Recent lesion studies have examined more sophisticated representational issues, Gefitinib for Carnitine palmitoyltransferase II instance, comparing the sort of stimulus-response associations that underpin habits to a hierarchical association scheme in which the presence of a certain stimulus implies that a response leads to a particular outcome (Bradfield and Balleine, 2013). Control apparently based on the latter representation

is compromised by lesions to posterior dorsomedial striatum, whereas in complex circumstances, lesions to dorsolateral striatal actually enhanced learning, suggesting that a form of competition might be at work. The rich backdrop of animal experiments has inspired a collection of studies that address the architecture of human instrumental control, often employing straightforward adaptations of successful animal paradigms as well as seeking and exploiting homologies (Balleine and O’Doherty, 2010 and Haber and Knutson, 2010). Many of these have involved the use of fMRI in order to investigate the neural representation of the value of stimuli and actions to see whether or not they are affected by devaluation. We consider two studies of particular interest in this context that respectively target goal-directed and habitual choice (Valentin et al., 2007 and Tricomi et al., 2009). Valentin and colleagues trained human subjects on a task in which two different instrumental actions resulted in two distinct food reward outcomes (Valentin et al., 2007).