Understanding the Brain’s Predictive Abilities

Predictive coding, also known as predictive processing, is a theoretical framework that has gained significant attention in the field of psychology. It proposes that the brain is constantly engaged in making predictions about the world based on existing knowledge and sensory input. This paradigm has provided valuable insights into perception, cognition, and even mental disorders. A key problem for organisms is processing the dynamic data consistently flowing from the environment. The great waves of stimuli contain more information than our available finite resources can process.

Lisa Feldman Barrett, PhD., a University Distinguished Professor at Northeastern University with appointments at the Massachusetts General Hospital and Harvard Medical School, wrote that prediction is the brain’s most important job. She explains that “it’s not rationality. Not emotion. Not imagination, or creativity, or empathy. Your brain’s most important job is to control your body—to manage allostasis—by predicting energy needs before they arise so you can efficiently make worthwhile movements and survive” (Barrett, 2020).

The Basics of Predictive Coding

At its core, predictive coding suggests that the brain is an active prediction machine, constantly generating hypotheses about the environment. According to this framework, perception is not a passive process where we simply receive and analyze sensory input; rather, it involves the brain actively predicting what our sensory organs will detect. We then compare predictions to incoming sensory information, the brain updates its understanding of the world and make necessary adjustments.

Daniel Williams described brain processing in predictive coding theory as fundamentally “a probabilistic prediction machine, continually striving to minimize the mismatch between self-generated predictions of its sensory inputs and the sensory inputs themselves” (Williams, 2020).

How Predictive Coding Works

One of the earliest mentions of predictive coding can be traced back to a published article in 1982. The authors, Mandyam Veerambudi Srinivasan, Barry Simon Laughlin and A. Dubs explained that the brains predictive coding works through an antagonistic surround technique that helps to enhance the detection of fine details by reducing noise and increasing the ability to distinguish different images. It does this by using statistical predictions based on neighboring receptors and subtracting the predicted value from the center signal. This way, the interneuron can focus its dynamic range on encoding a smaller range of intensities, improving sensitivity to subtle variations in the visual world” (1982).

Predictive Errors

Instead of predicting everything afresh, the brain start with a predictive value based on past memories, than corrects for errors. The errors are a fundamental point in the theory of predictive coding. . In a recent peer reviewed research article, the authors explain “the brain’s (bottom-up) evoked response to a stimulus should reflect prediction errors. That is, the size of this bottom-up evoked response should reflect the size of the prediction error” (Bowman, Collins, Nayak, & Cruse, 2023). Prediction errors in many ways are similar to cognitive dissonance, where conflicting elements create arousal and demand rectifyiing.

Perhaps, this is why when life significantly derails from expectation we experience disrupting heightened arousal. Our brain has significant corrections to make, requiring sizable cognitive resources that we budgeted for other processes. The shift is discomforting, especially for those that struggle with variability, and possess limited cognitive flexibility.

When a prediction matches the incoming sensory input, little prediction error is generated. However, when there is a mismatch between the predictions and the actual sensory data, a significant prediction error occurs. These prediction errors are then transmitted upward through the hierarchy, leading to an update in the predictions at each level. This iterative process allows the brain to make more accurate predictions over time, optimizing its understanding of the world.

Predictive coding operates on the idea of a hierarchical predictive system within the brain. This system consists of multiple levels, with each level responsible for generating predictions and receiving prediction errors from the level above it. Higher levels in the hierarchy generate more abstract predictions, while lower levels process more specific sensory information.

Applications of Predictive Coding in Psychology

Predictive coding has proven to be a valuable framework for understanding various psychological phenomena. Here are a few areas where predictive coding has found applications:

Perception and Attention

Predictive coding provides insights into how we perceive and attend to sensory information. By emphasizing the role of top-down predictions, this framework helps explain phenomena such as perceptual illusions, selective attention, and the integration of sensory modalities. Since we already call up predictive images from the past, new stimuli doesn’t require attention to the entire framework, leaving more cognitive resources to focus on minute details.

Cognitive Processes

Predictive coding also sheds light on higher-level cognitive processes, such as decision-making and memory formation. By highlighting the brain’s predictive abilities, researchers are able to better understand how we form expectations, make choices, and remember information. Barrett proposes it is a survival mechanism, reducing resources for processing incoming stimuli. She explains we do this by prioritizing one movement over another, she explains that “movement should be worth the effort, economically speaking. That is a prediction, based on past experience, to prepare a body for action.” She clarifies that this isn’t conscious, thoughtful decisions but “that something must occur inside a creature to predict and launch one set of movements rather than another” (Barrett, 2020).

And this cognitive process is some form of predictive coding. Barret emphasizes the importance explaining, “creatures that predicted correctly most of the time, or made nonfatal mistakes and learned from them, did well. Those that frequently predicted poorly, missed threats, or false-alarmed about threats that never materialized didn’t do so well. They explored their environment less, foraged less, and were less likely to reproduce (Barrett, 2020).

Psychopathology

Predictive coding has implications for psychopathology as well. It has been suggested that certain mental disorders, such as schizophrenia, might involve disrupted predictive coding mechanisms. Understanding these disruptions can lead to new interventions and treatment approaches.

Conclusion

Predictive coding offers a compelling perspective on how the brain processes information and perceives the world. By considering the brain as a prediction machine, this framework provides valuable insights into various psychological processes. From perception to cognition and psychopathology, the study of predictive coding continues to expand our understanding of the complex workings of the mind.

References:

Barrett, Lisa Feldman (2020) Seven and a Half Lessons About the Brain. Houghton Mifflin Harcourt.

Bowman, H., Collins, D., Nayak, A., Cruse, D., & , (2023). Is Predictive Coding Falsifiable?. Neuroscience & Biobehavioral Reviews, 105404. DOI: 10.1016/j.neubiorev.2023.105404

Srinivasan, Mandyam Veerambudi; Laughlin, Simon Barry; Dubs A. (1982). Predictive coding: a fresh view of inhibition in the retina Proceedings of the Royal Society, London ppg 216,427–459. DOI: 10.1098/rspb.1982.0085

Williams, Daniel (2020). Predictive coding and thought. Synthese, 197(4), 1749-1775. DOI: 10.1007/s11229-018-1768-x