Trainability and Modification of Subpersonal Priors: A Hierarchical Perspective

Written by

in

Subpersonal priors—the unconscious expectations operating below conscious awareness that shape perception and cognition—can indeed be trained and modified, though with important qualifications and constraints. These implicit probabilistic beliefs form a crucial component of how our brains process information, guiding everything from basic sensory processing to complex social cognition. Understanding their malleability has significant implications for learning, habit formation, and clinical interventions.

Hierarchical Organization and Differential Plasticity

Subpersonal priors exist within a hierarchical structure, and their susceptibility to modification varies significantly depending on their position within this hierarchy. According to predictive processing frameworks, priors operate at multiple levels of cognitive processing, with differential capacities for updating and change.

At intermediate levels of processing, subpersonal priors function as “empirical priors” that are regularly updated through incoming sensory evidence. These priors demonstrate significant plasticity, as “they are updated by evidence from lower levels” and depend upon experience1. This aligns with the fundamental predictive processing principle that “today’s posteriors become tomorrow’s priors,” indicating an ongoing process of adjustment and refinement1.

However, not all subpersonal priors share this flexibility. Particularly at lower levels of processing, certain priors are held with greater precision and demonstrate remarkable resistance to updating. For example, “innate subpersonal priors that underwrite homeostasis” are “clearly less amenable to updating”1. These deeply embedded priors serve fundamental biological functions and thus resist modification even in the face of contradictory evidence.

Precision Weighting and Updating Mechanisms

A key mechanism governing the plasticity of subpersonal priors involves precision weighting—the brain’s assignment of confidence levels to both predictions and sensory evidence. The philosophical literature supports that “on the PP approach, subpersonal priors are also rationally adjustable in light of contrary (sensory) evidence”6. This suggests that even without conscious intervention, subpersonal systems can adjust their expectations based on prediction errors.

The precision assigned to these priors directly influences their resistance to change. When precision is high, the prior exerts greater influence and resists updating; when precision is low, the prior becomes more amenable to modification through incoming evidence. This precision weighting mechanism explains why some subpersonal priors remain stubbornly resistant to change while others demonstrate remarkable plasticity.

Pathways for Training and Modification

Several specific mechanisms facilitate the training and modification of subpersonal priors:

Experience-Dependent Neural Plasticity

Experience-dependent neural plasticity represents a fundamental mechanism through which subpersonal priors can be modified. Research on brain reorganization following damage illustrates how experiences can shape neural architecture. Training produces measurable “neuroanatomical plasticity” including increased synaptic densities and the proliferation of specific synapse subtypes7. These structural changes likely provide the physiological substrate for modified priors.

This experience-driven plasticity interacts with reactive neural plasticity to create growth-permissive environments in the brain that are more sensitive to behavioral experiences7. This enhanced sensitivity facilitates the learning of new behavioral patterns, which subsequently reinforces and further shapes the underlying neural architecture and associated priors.

Cultural Transmission and Supra-Personal Control

Higher-level subpersonal priors show particular susceptibility to modification through social and cultural processes. As noted in the research, “our prior expectations at this level of control are malleable and largely determined by our culture”5. This suggests that cultural learning represents a powerful mechanism for shaping and modifying many subpersonal priors.

The concept of “supra-personal control” illuminates how messages from others can alter private cognitive processes5. This cultural transmission involves two critical processes: converting private cognitive representations into public forms that can be communicated, and the reverse process through which public information modifies private processes. These interactions “at the top of the hierarchy of control create and maintain cultural priors”5.

Interestingly, while high-level priors may resist modification through bottom-up evidence, they “can be very quickly changed by top-down messages from other people”5. This asymmetric response to modification attempts reflects an adaptive strategy: “We can get more precise priors from other people who have had more experience. We can get even better estimates from our cultural milieu because this encompasses the experience of many people over a long time”5.

Automatization of Conscious Processes

Another pathway for modifying subpersonal priors involves the automatization of initially conscious processes. With practice, “cognitive processes cease to be controlled and become automatic”5. This transition from controlled to automatic processing is accompanied by “a reduction of activity in frontal cortex presumably because monitoring and control is no longer needed”5.

This mechanism explains how intentional practice can eventually reshape subpersonal priors, as deliberate behaviors become habitual and the corresponding neural patterns become encoded at a subpersonal level. For example, in the context of ethical behavior, “selfish behavior has ceased to be the default behavior and altruistic behavior has become habitual”5 through repetition and practice.

Resistance and Constraints to Modification

Despite their potential for change, subpersonal priors demonstrate notable resistance to modification under specific conditions:

Motivated Resistance to Updating

Priors that align with an agent’s motivations or desires demonstrate particular resistance to updating. As noted in research on stereotype formation, “higher order priors can be less amendable to update if the existing higher order predictions are positive for the agent and the incoming evidence is negative for the agent”1. This suggests that subpersonal priors that serve beneficial functions for the individual may resist modification even in the face of contradictory evidence.

The research notes that “only when encountering information highly contradictory to group-based priors do perceptually implemented stereotypes/prejudices become amendable to update”1. This indicates that the threshold for evidence required to modify deeply held priors can be extraordinarily high, particularly when those priors serve psychological or social functions.

Level-Specific Constraints

Different levels of subpersonal priors demonstrate different constraints on modification. While “higher level priors can be quickly changed by top-down messages,” lower-level perceptual priors often demonstrate remarkable stubbornness. Some visual priors, for example, are “remarkably stubborn and difficult to override, suggesting they’re deeply encoded in the architecture of the visual system”1. This architectural constraint reveals how certain subpersonal priors can operate independently of personal-level cognition and resist modification.

Conclusion

The evidence clearly indicates that subpersonal priors can indeed be trained and modified, though with important qualifications regarding their hierarchical level, precision, and functional role. Higher-level priors demonstrate greater susceptibility to modification through cultural learning and social interaction, while lower-level priors often show remarkable resistance to change, particularly those involved in fundamental biological functions.

This understanding of subpersonal prior modification has significant implications across domains from education to clinical interventions. The multiple pathways for modifying priors—through experience-dependent plasticity, cultural transmission, and the automatization of conscious processes—offer potential avenues for intentional intervention and modification of maladaptive priors. However, the stubborn resistance of certain priors, particularly those aligned with an agent’s motivations or serving fundamental functions, presents a significant challenge that requires targeted, persistent approaches to overcome.

These insights not only illuminate the complex nature of cognitive change but also suggest practical approaches for facilitating learning, habit formation, and behavioral change through the strategic modification of underlying subpersonal priors.