Predictive processing models and affective neuroscience.

Area of Science:

• Neuroscience
• Affective Science
• Cognitive Neuroscience

Background:

• The neural underpinnings of affective experience, particularly valence and arousal, remain largely undetermined.
• Previous research focusing on specific brain regions or multivariate analyses has yielded limited progress in understanding affect.
• Existing functional neuroimaging practices in affective neuroscience often rely on assumptions incompatible with emerging theoretical frameworks.

Purpose of the Study:

• To critique current functional neuroimaging research practices in affective neuroscience through the lens of predictive processing models.
• To highlight inconsistencies between prevailing research methods and predictive processing principles.
• To propose implications of predictive processing for future research in affective and cognitive neuroscience.

Main Methods:

• A critical review of existing functional neuroimaging research in affective neuroscience.
• Analysis of how predictive processing models can reframe understanding of neural computations underlying affect.
• Examination of the implications for experimental design and interpretation in cognitive and affective neuroscience.

Main Results:

• Prevailing functional neuroimaging research in affective neuroscience often employs rigid assumptions that conflict with predictive processing.
• Predictive processing models offer a novel framework for understanding the neural basis of affect.
• The review identifies limitations in current research practices and suggests avenues for improvement.

Conclusions:

• Predictive processing models provide a valuable theoretical lens for advancing affective neuroscience.
• A reformulation of "reverse inference" is necessary within cognitive neuroscience.
• Increased emphasis on external validity in experimental design is crucial for future research.