Trends in Cognitive Sciences
ReviewHow Do Expectations Shape Perception?
Section snippets
Expectation in Perception
Humans, like many other species, are ‘anticipatory systems’ [1]. They construct predictive models of themselves and their environment, allowing them to quickly and robustly make sense of incoming data. In line with this notion, the brain has been described as a ‘prediction machine’ [2] that attempts to match incoming sensory inputs with top-down expectations. Although the concept of the predictive brain is not new, dating back at least to Helmholtz [3], the neural implementation of such a
Perceptual Consequences of Expectation
We live in a highly predictable world, in which most objects remain stable and things change only slowly over time. This allows us to build internal models that can predict upcoming input on the basis of past and present input. Such expectations may prepare sensory cortex for processing, thereby increasing perceptual sensitivity for expected stimulus features. However, what are the consequences of expectation for perception?
There are several ways in which expectation can influence perception,
Where Do Expectations Come From?
The brain can predict future input by learning about and exploiting statistical regularities in its inputs [20] – but how does it achieve this? Because such regularities come in different shapes and forms (Box 2), the neural mechanisms likely depend on the type of regularity.
Arguably the simplest regularity in our sensory input is that particular features appear more often, and are thus generally more likely than others. For instance, cardinally oriented (i.e., horizontal and vertical) lines
How Do Expectations Modulate Sensory Processing?
Expectations, alongside other cognitive factors such as attention and reward, strongly modulate the responsiveness of even the earliest sensory regions [64]. Overall, stimuli that are expected evoke a reduced neural response. This pattern has been observed when stimuli are validly predicted by a preceding, arbitrarily paired, stimulus 65, 66, 67, 68, 69, when stimuli comply with a higher-order pattern such as a shape or scene 33, 34, or when stimuli are the predictable result of an animal’s own
Expectation in Computational Models of Perception
While it is clear that neural responses are heavily influenced by prior expectations, the computational role of these modulations is not yet fully understood. In general, expectations figure prominently in computational theories that cast perception as a process of probabilistic inference. Because the information conveyed by our senses is both noisy and ambiguous, perception has often been conceptualized as a process of probabilistic inference in which the system decides on the most probable
Alterations of Expectation in Psychopathology
Expectations may not only be important to understand how the human brain implements perceptual inference but also for understanding various psychopathological conditions. In particular, neurodevelopmental disorders such as autism spectrum disorder 132, 133 and schizophrenia 134, 135 have been linked to an atypical integration of prior and incoming information, with autism even being cast as a ‘disorder of prediction’ [136]. While both conditions are linked to aberrant expectations, they are
Concluding Remarks and Future Perspectives
In this article we have discussed how the brain capitalizes on prior knowledge to facilitate the neural computations underlying sensory processing. While there are many different forms of prior knowledge, which can be neurally implemented in distinct ways, there appears to be a common currency in terms of their modulatory effect on target regions involved in the processing of sensory data. An interesting question for future research could be how all the prediction signals from different sources
Acknowledgments
This work was supported by The Netherlands Organisation for Scientific Research (NWO Vidi grant to F.P.d.L., NWO Research Talent grant to M.H., NWO Rubicon grant to P.K.), the James S. McDonell Foundation (JSMF scholar award to F.P.d.L.), and the EU Horizon 2020 Program (European Research Council starting grant 678286 awarded to F.P.d.L.). We thank Matthias Fritsche, Eelke Spaak, Chaz Firestone, and two anonymous reviewers for helpful comments on an earlier version of the manuscript.
References (160)
Computational psychiatry
Trends Cogn. Sci.
(2012)- et al.
Computational psychiatry
Neuron
(2014) Believing is seeing: expectations alter visual awareness
Curr. Biol.
(2008)Changing expectations about speed alters perceived motion direction
Curr. Biol.
(2011)Cross-modal prediction changes the timing of conscious access during the motion-induced blindness
Conscious. Cogn.
(2015)- et al.
View from the top: hierarchies and reverse hierarchies in the visual system
Neuron
(2002) - et al.
The role of context in object recognition
Trends Cogn. Sci.
(2007) - et al.
Shape perception simultaneously up- and downregulates neural activity in the primary visual cortex
Curr. Biol.
(2014) A Bayesian perspective on magnitude estimation
Trends Cogn. Sci.
(2015)Shaping of object representations in the human medial temporal lobe based on temporal regularities
Curr. Biol.
(2012)