A multiple attribute model resolves a conflict between additive and multiplicative models of incentive salience

Abstract: A model of incentive salience as a function of stimulus value and interoceptive state has been previously proposed. In that model, the function differs depending on whether the stimulus is appetitive or aversive; it is multiplicative for appetitive stimuli and additive for aversive stimuli. The authors argued it was necessary to capture data on how extreme changes in salt appetite could move evaluation of an extreme salt solution from negative to positive. We demonstrate that arbitrarily varying this function is unnecessary, and that a multiplicative function is sufficient if one assumes the incentive salience function for an incentive (such as salt) is comprised of multiple stimulus features and multiple interoceptive signals. We show that it is also unnecessary considering the dual-structure approach-aversive nature of the reward system, which results in separate weighting of appetitive and aversive stimulus features.