Okinawa Institute of Science and Technology, Japan
A neuroscientist explores what brain imaging can reveal about deliberative and intuitive decision-making.
When you pick a dish from a menu, do you select it for its taste or its calculated nutritional benefits? The decision-making processes of intuition and deliberation can be considered as, respectively, model-free learning, which involves trial and error, and model-based learning — evaluating future outcomes using a pre-learned model of the results of choices. A big question is how these complementary processes are realized in the brain.
Using functional magnetic resonance imaging (fMRI) in humans, Jan Gläscher at the California Institute of Technology in Pasadena and his co-authors found neural signatures for these two modes of learning (J. Gläscher et al. Neuron 66, 585–595; 2010). The team scanned the brains of volunteers as they learned a two-step choice task. During the first part of the study, volunteers were presented with an abstract image and had to choose a left- or right-button press. Depending on which button they chose, they were then presented with another image and asked to make a second left-or-right choice to see a third image.
Over many trials, the volunteers learned the probability of a certain image resulting from a particular choice. During the half-time break, they were told that each final image would have a specific monetary reward (0, 10 or 25 cents). During the second half of the study, volunteers could use what they had learned in the first half to make profitable choices.
Analysis of the fMRI data revealed involvement of the brain’s intraparietal and lateral prefrontal cortices in model-based learning, and the ventral striatum in model-free learning. The study paints a new picture of the neuroscience of deliberation versus intuition. We should now be able to ask not only where in the brain but also by what algorithms we make decisions.