Instrumental Performance following Reinforcer Devaluation Depends upon Incentive Learning