 |
 |
|
|
||
|
 |
Department of Experimental Psychology |
|
|
||
|
||
| University of Cambridge > School of the Biological Sciences > Department of Experimental Psychology > Academic Staff |
A. Dickinson, FRSProfessor of Comparative Psychology, Tel: 01223 333577 email:ad15@cam.ac.uk
My current research interests lie in three areas:
Associative learning is the most fundamental form of learning in that it enables us and other animals to learn about the causal and predictive structure of our environment through the formation of associations between representations of stimuli, actions and outcomes. Except for certain forms of learning, such as evaluative conditioning (Dickinson & Brown, 2007), it is generally accepted that associative learning occurs when the outcome is surprising or unexpected, thereby generating a prediction error (Schultz & Dickinson, 2000). My recent research has been concerned two aspects of the role of prediction errors.
Neural encoding of prediction errors. I have collaborated with Wolfram Schultz (University of Cambridge) on his research project studying the role of the dopamine system in encoding prediction errors. Although a variety of brain systems show response characteristics that suggest the neural encoding of prediction errors, the most compelling evidence comes from studies of mid-brain dopamine neurons. Although these cells are activated by unpredicted rewards, which should generate a positive prediction error, they do not respond to that same reward when it is predicted by a conditioned stimulus. Moreover, associative learning fails to occur in the absence of dopamine activation in a blocking paradigm (Waelti, Dickinson, & Schultz, 2001), whereas the depression in the baseline level of activity of the dopamine neurons caused by the omission of an expected reward is associated with inhibitory learning (Tobler, Dickinson, & Schultz, 2003). The concordance between prediction errors and dopamine activity is a striking example of the synthesis of learning theory and neuroscience.
Retrospective revaluation. Although associative theory provides a powerful account of causal and predictive learning, it imposes a major restriction of the ability to integrate information across different learning experiences. Consider the case in which a compound of two stimuli is paired with an outcome (AX+) before one of these stimuli is presented without the outcome (A-). The rational inference from this experience is that X was the reliable cause or predictor of the outcome during the compound training.
Although the omission of the outcome on A- episodes should generate a negative prediction error, standard associative theory does not allow for retrospective revaluation of the status of X given further information about A because learning can only occur when a stimulus is present. I have proposed a revision and extension of associative learning theory (Aitken & Dickinson, 2005; Dickinson, 2001), which assumes that learning about associatively retrieved stimulus representations, an idea that we have tested within the context of human causal learning (Aitken, Larkin, & Dickinson, 2000; Larkin, Aitken, & Dickinson, 1998). Moreover, I have collaborated with Paul Fletcher (University of Cambridge) and Phil Corlett (University of Cambridge) on their project demonstrating that dorsal PFC is a site of the neural signatures of prediction-error processing during retrospective revaluation (Corlett et al., 2004) and that abnormalities in this processing may be relate psychotic-like delusions (Corlett et al., 2007)
Aitken, M. R. F., & Dickinson, A. (2005). Simulations of a modified SOP model applied to retrospective revaluation of human causal learning. Learning and Behavior, 33, 147-159.
Aitken, M. R. F., Larkin, M. J. W., & Dickinson, A. (2000). Re-examination of the role of within-compound associations in the retrospective revaluation of causal judgements. Quarterly Journal of Experimental Psychology, 53B, 59-81.
Corlett, P. R., Aitken, M. R. F., Dickinson, A., Shanks, D. R., Honey, G. D., Honey, R. A. E., et al. (2004). Prediction error during retrospective revaluation of causal associations in humans: fMRI evidence in favor of an associative model of learning. Neuron, 44, 877-888.
Corlett, P. R., Murray, G. K., Honey, G. D., Aitken, M. R. F., Shanks, D. R., Robbins, T. W., et al. (2007). Disrupted prediction-error signal in psychosis: evidence for an associative account of delusions. Brain, 130, 2387-2400.
Dickinson, A. (2001). Causal learning: An associative analysis. Quarterly Journal of Experimental Psychology, 54B, 3-25.
Dickinson, A., & Brown, K. J. (2007). Flavor-evaluative conditioning is unaffected by contingency knowledge during training with color-flavor compounds. Learning and Behavior, 35, 36-42.
Larkin, M. J., Aitken, M. R. F., & Dickinson, A. (1998). Retrospective revaluation of causal judgments under positive and negative contingencies. Journal of Experimental Psychology: Learning, Memory and Cognition, 24, 1331-1352.
Schultz, W., & Dickinson, A. (2000). Neural coding of prediction errors. Annual Review of Neuroscience, 23, 473-500.
Tobler, P. N., Dickinson, A., & W. Schultz, W. (2003). Coding of predicted reward omission by dopamine neurons in a conditioned inhibition paradigm. Journal of Neuroscience, 23, 10402-10410.
Waelti, P., Dickinson, A., & Schultz, W. (2001). Dopamine responses comply with basic assumptions of formal learning learning. Nature, 412, 43-48.
|
|
|
|