Neural mechanisms of reward value, emotion, and decision-making

Edmund T. Rolls
Oxford Centre for Computational Neuroscience, Oxford (www.oxcns.org) 
University of Warwick, Department of Computer Science, Coventry, UK

In Rolls’ theory of emotion (2014) it is argued that emotions are states elicited by instrumental 
reinforcers which are the goals for action, the rewards and punishers. It is argued that emotions solve a 
fundamental problem in Darwinian evolution, for it is much more efficient for genes to specify goals for 
actions, rewards and punishers, rather than actions or responses. It is shown that the orbitofrontal 
cortex is important in emotion for it represents primary, unlearned, gene-specified, reinforcers including 
the taste and texture of food and face expression; performs rapid learning, and reversal, of stimulus-
reward associations; and with the pregenual cingulate cortex has activations that are directly correlated 
with pleasure, the conscious reports of the subjective state associated with rewards. These reward 
valuation systems in the orbitofrontal cortex provide independent representations at the neuronal level 
on a common scale but are not converted into a common currency, and provide inputs to our value 
based decision-making mechanisms in the ventromedial prefrontal cortex.
A combination of approaches including functional neuroimaging, neurophysiology, and theoretical 
physics, provides evidence that decisions in the cerebral cortex are taken by attractor networks that are 
biased by the evidence for the decision (1-2). Integrate-and-fire neuronal networks show that decision-
making is inherently probabilistic because of noise caused by the random firing times of neurons in the 
brain (for a given mean rate). It is shown that confidence in a decision is an emergent property of the 
decision-making process, and human fMRI investigations that test this have been described (1-6). 
Decisions can be in part predicted from the noise in the neuronal firing before the decision cues are 
applied. The attractor network architecture for decision-making is similar to that involved in short-term 
memory and attention (7), and instabilities in such networks in brain regions such as the prefrontal and 
temporal cortex are hypothesized to be involved in schizophrenia (8).

1. Rolls, E.T. and Deco, G. (2010) The Noisy Brain: Stochastic Dynamics as a Principle of Brain Function. 
Oxford University Press: Oxford.
2. Rolls, E.T. (2014) Emotion and Decision-Making Explained. Oxford University Press: Oxford.
3. Rolls, E.T., Grabenhorst, F. and Deco, G. (2010) Decision-making, errors, and confidence in the brain. 
Journal of Neurophysiology 104: 2359-2374.
4. Rolls, E.T., Grabenhorst, F. and Deco, G. (2010) Choice, difficulty, and confidence in the brain. 
Neuroimage 53: 694-706.
5. Grabenhorst, F. and Rolls, E .T. (2011) Value, pleasure, and choice in the ventral prefrontal cortex. 
Trends in Cognitive Sciences 15: 57-67.
6. Deco, G., Rolls, E.T., Albantakis, L. and Romo, R. (2013). Brain mechanisms for perceptual and reward-
related decision-making. Progress in Neurobiology 103: 194-213. 
7. Rolls, E.T. (2008) Memory, Attention, and Decision-Making: A Unifying Computational Neuroscience 
Approach. Oxford University Press: Oxford.
8. Rolls, E.T., Loh, M., Deco, G. and Winterer, G. (2008) Computational models of schizophrenia and 
dopamine modulation in the prefrontal cortex. Nature Reviews Neuroscience 9: 696-709.