What it means to call the brain a prediction machine
Predictive processing describes the brain as an engine of probabilistic hierarchical inference: it builds models of what is likely to be causing its sensory input and updates those models when predictions fail [2]. The gap between what was predicted and what arrived is a prediction error, and the brain works to minimize it, either by changing its model or by acting on the world. This single idea is proposed as a general theory of brain function, and it has been extended to questions of conscious experience, including qualia and the sense of self [2].
The appeal for consciousness science is that it offers a way to map biological mechanisms onto the felt properties of experience rather than treating consciousness as one single target [5]. That framing matters for emotion and self-awareness, because it suggests these are not separate mysteries but different layers of the same predictive architecture.
Emotion as the brain's interpretation of body and self
One strand of work places emotion at the very center of conscious experience, arguing that affective states act as a primordial prediction layer that turns abstract pattern recognition into embodied meaning, arising before conceptual thought [1]. In this view emotion is not a reaction to reality but part of how the experiential world is generated [1]. That is a strong theoretical claim, and the paper presents it as a framework with testable predictions rather than an established result [1].
A more mechanistic account focuses on how emotions are generated and regulated. Standard predictive processing explains emotion as prediction-error minimization, but that alone does not explain why different prediction errors feel different. A self-predictive model addresses this by treating emotional valence as an interpretive process of the self-model, which assigns significance based on goals, values, and autobiographical context [4]. In plain terms, the same bodily stir can feel like excitement or anxiety depending on what the self-model predicts it means.
The body supplies the raw material. Bodily organs such as the heart, lungs, and stomach interact continuously with the brain, and their signals modulate electrophysiological dynamics that shape how emotional and self-related information is integrated [3]. This is why interoception, the sensing of internal bodily states, sits at the junction of emotion and selfhood [3].
Where the sense of self and of being present comes from
If the brain is constantly predicting, the sense of self can be modeled as one of its own predictions. Work connecting predictive processing to literature argues that literary texts can serve as designed environments for exploring conscious experience, including the sense of self [2]. Reading is used here as a way to probe how predictions about self and world are built and revised, not as direct evidence about brain mechanisms [2].
Bodily self-consciousness is tied more directly to the body's internal signals. Because interoceptive signals feed into high-level brain functions, they contribute to emotion, self, and bodily consciousness together [3]. The same paper notes that altered interoceptive responses might underlie deficits in psychopathology, which is a hypothesis about mechanism rather than a demonstrated cause [3]. Evidence on how interoceptive electrophysiology develops during infancy and adolescence remains limited, and attempts to establish causality are still early [3].
How firm is the predictive picture of emotion and self?
Much of this extension is theoretical. The claim that emotion is ontologically primary and generates experience is presented as a reconceptualization with testable predictions about emotional precision, priming, and narrative intervention, not as confirmed findings [1]. The self-predictive model of emotion regulation is likewise offered as a framework that opens avenues for computational modeling and empirical research [4].
The interoception work is the most grounded in measurement, drawing on electrophysiological studies of cardiac, respiratory, and gastric signals [3]. Even there, the authors describe causality work as attempts and highlight important areas for further research [3]. Taken together, the extension of predictive processing to emotion, presence, self-recognition, and bodily self-consciousness is best read as a coherent and generative research program whose central claims are still being tested.
Sources used in this answer
Emotion as Constitutive of Consciousness: A Predictive Processing Framework for Autonomous Experience
Proposes that emotion acts as a primordial prediction layer that generates embodied meaning and the experiential world through predictive processing, offered as a framework with testable predictions.
Designs on consciousness: literature and predictive processing
Argues predictive processing can model conscious experience including qualia and sense of self, and that literature can serve as a designed environment for exploring consciousness.
Neurophysiological dynamics of visceral signals in emotion, self and bodily consciousness
Reviews electrophysiological evidence that cardiac, respiratory, and gastric interoceptive signals contribute to emotion, self, and bodily consciousness, with causality work still early.
On Valence: A Self-Predictive Processing Model of Emotion Regulation
Proposes a self-predictive processing model in which emotional valence arises from the self-model interpreting prediction errors using goals, values, and autobiographical context.
Predictive processing as an empirical theory for consciousness science
Argues progress in consciousness science needs theories mapping biological mechanisms to phenomenological properties, and that predictive processing is a promising candidate.