Figure 1.

Nested temporal horizons for computing and signaling average rewardand opportunity. Each temporal period is associated with availability of reward opportunity, E(V), that should presumably motivate appropriate appetitive behavior in response. However, the value of that opportunity is contingent upon the average reward in the broader, containing temporal frame, denoted by $\overline{\text{r}}$. Average reward is presented as a progressive set of nested temporally defined contexts that determines the value of opportunity and resources expended in shorter periods of time. The extent to which DA transmits opportunity, E(V) or average reward, $\overline{\text{r}}$, by “phasic” or “tonic” DA cell activity or by alterations in the DA substrate itself is tentatively indicated by vertical bars on the right.