Why Compare APS and the Free Energy Principle?

Among contemporary attempts to understand living systems, few frameworks have attracted as much attention as the Free Energy Principle (FEP). Developed initially within theoretical neuroscience and subsequently extended to biology, cognition, and complex systems, FEP proposes a mathematically unified account of how self-organising systems maintain themselves in uncertain environments. Through the language of variational free energy, active inference, and organism–environment coupling, it seeks to explain how living systems preserve their organisation while continually interacting with changing conditions.

The comparison follows the protocol established in Comparative Explanatory Methodology in Theoretical Biology. It begins by distinguishing explanatory targets, organising principles, and explanatory priorities before assessing convergence, difference, or complementarity.

The Agency–Process–Scale (APS) framework approaches living systems through a distinct explanatory architecture. It proposes that life is viability-oriented, constraint-closed organisation and treats organised persistence as the problem of how that organisation maintains and re-establishes the conditions of its continuity despite change. APS therefore gives explicit explanatory priority to living organisation, biological agency, viability, and the relations disclosed through Agency, Process, and Scale.

Because APS and FEP both address adaptive and self-maintaining living systems, their explanatory targets may overlap substantially. The relevant question is not therefore whether one is simply a mathematical description and the other a biological explanation, nor whether they must be competing or complementary. It is how their targets, organising principles, conceptual relations, and explanatory priorities compare in particular cases.

This article therefore does not assume a fixed division of explanatory labour between APS and FEP. It identifies areas of conceptual overlap and difference and asks what explanatory significance, if any, follows from those differences. Formalism, biological interpretation, constitutive organisation, adaptive dynamics, and organism–environment coupling may be related differently depending upon what is being explained.

The purpose of the comparison is accordingly diagnostic rather than adjudicative. It clarifies the explanatory architecture of each framework and identifies where target-matched comparison would be required before complementarity, redundancy, independence, tension, or comparative gain could be established.

The Free Energy Principle: A Formal Framework for Self-Maintaining Systems

The Free Energy Principle (FEP) is a general theoretical framework that seeks to explain how self-organising systems maintain their integrity while continually interacting with changing and uncertain environments. Originally developed by Karl Friston within theoretical neuroscience, it has subsequently been extended to biology, cognition, and complex adaptive systems as a proposed unifying principle for understanding adaptive organisation.

At its core, FEP proposes that living systems maintain their organisation by remaining within a restricted set of viable states despite continual environmental perturbation. Formally, this persistence is described through the minimisation of variational free energy, a mathematical quantity that provides an upper bound on the surprise associated with a system’s sensory states. Although living systems cannot directly minimise surprise itself, they can minimise variational free energy through continual adjustment of their internal states and interactions with the environment.

Within this framework, organisms are understood as maintaining themselves through continuous cycles of perception and action. Perception updates the system’s internal states in response to sensory input, while action changes the environment in ways that reduce discrepancies between predicted and encountered states. These coupled processes have become formalised within the broader framework of Active Inference, which describes adaptive behaviour as the continual reduction of uncertainty through coordinated perception and action.

A central feature of the Free Energy Principle is its treatment of organism–environment boundaries. These are commonly described using the concept of Markov blankets, statistical boundaries that separate internal states from external states while allowing structured exchange through sensory and active states. Within FEP, Markov blankets provide the formal architecture through which organisms maintain distinguishable identities while remaining dynamically coupled to their environments.

The explanatory ambition of FEP extends beyond particular biological mechanisms. Rather than proposing a theory of metabolism, genetics, development, or evolution individually, it seeks a common mathematical framework capable of describing adaptive self-organisation across multiple domains. This broad scope has contributed to its influence within theoretical biology, cognitive science, neuroscience, and philosophy of science, where it has been presented as a candidate unifying principle for understanding living systems.

Despite this breadth, the Free Energy Principle does not prescribe a single biological mechanism through which adaptive organisation is achieved. Instead, it provides a formal language for describing how self-organising systems can remain within viable regions of their state space while continually responding to environmental uncertainty. Whether this formal description should also be regarded as a constitutive biological explanation remains an active topic of scholarly discussion. Whatever verdict is reached concerning its constitutive explanatory status, the Free Energy Principle remains one of the most ambitious contemporary frameworks for formally characterising adaptive self-organisation and relating those dynamics to living systems.

APS: Biological Organisation as the Target of Explanation

The Agency–Process–Scale (APS) framework approaches the study of living systems through its own explanatory architecture. APS investigates the biological organisation that constitutes living systems and makes adaptive activity possible. Its central explanatory target is the organisation characteristic of living systems and the conditions through which that organisation persists despite change. Agency, Process, and Scale are complementary analytic projections of one viability-oriented, constraint-closed organisation. Agency concerns what living systems do; Process concerns how continuity is maintained despite change; Scale concerns where persistence is organised across spatial and temporal extents. The distinctions are analytic; the organisation and activities they disclose are biologically real.

APS proposes that life is best understood as viability-oriented, constraint-closed organisation. Living systems are not defined by particular molecules, energetic processes, or informational architectures, but by the continual organisation through which they maintain and re-establish the conditions of their own persistence. This organisation is dynamic rather than static, existing as an ongoing activity rather than as a fixed structural arrangement.

Within APS, biological agency is the present-tense activity through which this organisation is enacted and sustained. Agency does not denote conscious choice, deliberation, or representation. Nor is it an additional causal force acting alongside biological mechanisms. Rather, it identifies the viability-oriented organisational activity through which living systems regulate themselves, repair themselves, and respond to changing circumstances while preserving their own continuity.

APS interprets many of biology’s central concepts—including function, development, evolution, ecology, cognition, and communication—in relation to living organisation and organised persistence. This does not require every such explanation to collapse into a single APS account or to be complementary in every explanatory context.

This emphasis on organisation also shapes APS’s account of biological normativity. The significance of biological states and processes is not derived from optimisation criteria, predictive success, or informational quantities. Instead, significance arises because biological organisation differentiates between conditions that contribute to continued viability and conditions that undermine it. Function, evaluation, and purpose are therefore grounded in the organisation of living systems themselves rather than imposed by external observers or abstract formal principles.

APS consequently presents itself not as a replacement for existing biological theories but as a comparative explanatory methodology. It investigates how the diverse explanatory achievements of contemporary biology bear upon questions of living organisation and organised persistence. In this sense, APS investigates the organisation of living systems as an explicit explanatory target while comparing how other biological explanations bear upon that target.

Comparing Explanatory Priorities: Organisation, Formalism, and Adaptive Dynamics

APS and the Free Energy Principle share important concerns. Both address adaptive, self-maintaining systems, organism–environment relations, and persistence under changing conditions. Their explanatory architectures nevertheless differ in significant respects.

FEP employs a highly general formal apparatus involving variational free energy, generative models, active inference, and Markov blankets. These resources have been used not merely to model behaviour but to formulate broad claims about adaptive self-organisation, cognition, biological identity, and living systems.

APS begins instead from its substantive account of life as viability-oriented, constraint-closed organisation and asks how organised persistence is enacted, maintained, and transformed. Its explanatory grammar therefore foregrounds biological agency, process, scale, viability, and organisation.

These differences should not be converted prematurely into a fixed distinction between “formal description” and “biological explanation.” A formal framework may support explanatory claims about biological organisation, and an organisational framework may itself employ abstraction and formalisation. The relevant issue is what explanatory work each framework performs for a specified target.

APS therefore distinguishes its own constitutive claims about living organisation from mathematical or informational descriptions of biological activity, but that distinction alone does not establish that FEP is merely descriptive or non-constitutive. Nor does FEP’s mathematical generality establish that its formal quantities exhaust the biological organisation they are used to explain.

The comparative question is consequently target-relative. For some explananda, APS and FEP may address substantially different questions. For others they may overlap directly. Where the targets coincide, explanatory adequacy must be assessed rather than inferred from the frameworks’ different formalisms or conceptual vocabularies.

Possible outcomes include complementarity, partial overlap, independence, tension, redundancy, APS-favouring explanatory gain, FEP-favouring explanatory gain, or a qualified result in which neither framework displaces the other.

Normativity, Agency, and Biological Significance

The differing explanatory priorities of APS and the Free Energy Principle become most apparent in their treatment of normativity: the question of what makes particular biological states, processes, or outcomes matter to a living system.

Within the Free Energy Principle, adaptive behaviour is formally characterised through the minimisation of variational free energy. Organisms maintain themselves by remaining within a restricted range of viable states, and the mathematical framework describes how perception and action contribute to this continual regulation under conditions of uncertainty. In this sense, normativity is expressed through the formal dynamics of adaptive self-maintenance.

APS formulates normativity differently. It locates biological normativity in viability-relative asymmetry: conditions and activities differ according to how they bear upon the maintenance and re-establishment of organised persistence. This relation is part of APS’s substantive account of living organisation rather than something derived from an externally imposed optimisation criterion.

Within APS, biological agency is viability-oriented organisational activity. Agency is therefore not defined by prediction, inference, representation, or optimisation, although such concepts may describe or explain particular forms of biological activity. APS asks how living organisation itself generates the viability relations within which states, activities, and environmental conditions acquire biological significance.

APS likewise characterises biological significance through biological evaluation: living organisation differentiates conditions according to their consequences for viability. Function is understood as present-tense viability-relative organisational contribution, while biological purpose refers to viability-oriented organisation without requiring conscious intention.

These claims establish the APS position. They do not, however, demonstrate that FEP lacks resources for explaining normativity, agency, or biological significance, nor that FEP merely describes patterns which APS alone explains. FEP and active-inference approaches have substantive ambitions concerning adaptive organisation, and their relation to APS must therefore be assessed at the level of the particular explanatory claim.

The relevant comparison is not simply formal dynamics versus biological meaning. It is whether, for a matched explanandum, the different explanatory resources of APS and FEP identify the same dependencies, different dependencies, or differently useful representations of those dependencies.

APS may therefore maintain that viability-oriented organisation provides its account of the biological basis of significance without inferring that this account is explanatorily preferable merely because it is expressed in explicitly organisational rather than variational or inferential terms.

Explanatory Architecture and Theoretical Biology

The comparison between APS and FEP illustrates a broader methodological issue in theoretical biology: frameworks can overlap in biological subject matter while organising explanation differently.

Biological explanation is plural. Evolutionary, physiological, developmental, ecological, molecular, mathematical, informational, mechanistic, organisational, and other approaches may each provide powerful explanations of particular biological questions. Their adequacy cannot be ranked independently of the explanandum.

The same principle applies to theoretical frameworks. Sharing a biological subject matter does not establish that two frameworks compete, while differing in formalism or conceptual vocabulary does not establish that they are complementary. Their relationship depends upon the explanatory target and the work each framework performs with respect to it.

APS and FEP provide an especially demanding comparison because both make general claims concerning adaptive, self-maintaining living systems. FEP provides a mathematically articulated framework for adaptive self-organisation and active inference. APS develops an organisational account centred on viability-oriented, constraint-closed organisation and organised persistence. The difference is real, but its explanatory significance remains a comparative question.

One useful distinction concerns the relation between a formal representation and the biological organisation represented. Formal success does not establish ontological identity between mathematical quantities and living organisation. But neither does that distinction establish that formal frameworks cannot provide constitutive, causal, or otherwise biologically substantive explanations.

APS therefore treats explanatory architecture itself as an object of comparison. The task is to distinguish explanatory targets, organising principles, conceptual relations, methodological commitments, and evidential limits before asking whether a framework supplies explanatory gain.

On this basis, theoretical biology need not seek a single universal explanatory language. Nor should explanatory plurality be converted into automatic complementarity. Different frameworks may cooperate, overlap, remain independent, conflict, or prove differentially adequate depending upon the question.

The APS–FEP comparison should therefore be understood as a case of controlled explanatory comparison rather than as a demonstration that formal and organisational approaches occupy predetermined complementary roles.

Conclusion: Explanatory Difference Without a Predetermined Verdict

APS and the Free Energy Principle address overlapping problems concerning adaptive, self-maintaining living systems while organising explanation through different conceptual and methodological architectures. FEP develops a highly general formal framework involving variational free energy, active inference, Markov blankets, and adaptive self-organisation. APS develops an organisational account of life centred on viability-oriented, constraint-closed organisation and organised persistence.

These differences are significant, but they do not establish a fixed division between mathematical description and biological explanation. FEP has been used to advance substantive claims about living and cognitive systems, while APS advances substantive organisational claims of its own. Their relationship therefore cannot be decided simply by classifying one as formal and the other as constitutive.

APS retains its claim that biological agency, normativity, function, significance, and organised persistence are intelligible through viability-oriented organisation. That substantive position need not be weakened. What requires restraint is the comparative inference that APS therefore supplies the biological meaning or constitutive explanation that FEP lacks.

The appropriate conclusion is consequently conditional. Where APS and FEP address different explanatory targets, both may remain independently adequate. Where they address the same target, target-matched comparison is required. The outcome may be complementarity, overlap, independence, tension, redundancy, APS-favouring gain, FEP-favouring gain, or some more qualified relation.

This result illustrates the wider purpose of APS as a comparative explanatory methodology. APS does not acquire explanatory authority simply because its conceptual architecture is broad or integrative. Its substantive claims concerning living organisation remain open to assessment against the strongest relevant alternatives.

The comparison with FEP therefore establishes a genuine comparative problem rather than a predetermined synthesis. Clarifying that problem is itself methodologically useful; determining whether APS ultimately provides a preferred explanation of living organisation belongs to subsequent target-matched comparative research.

Explanatory Architecture

Central Question

How do APS and the Free Energy Principle converge and differ in their explanations of adaptive, self-maintaining living systems, and what—if any—comparative explanatory significance follows from those differences?

Architectural Role

This comparative article examines the relationship between APS and the Free Energy Principle without assuming competition, complementarity, or explanatory hierarchy. It distinguishes their explanatory targets, organising principles, formalisms, and conceptual architectures while identifying where genuine target-matched comparison would be required to establish explanatory gain. Its function is comparative and methodological rather than adjudicative.

Preceding Explanatory Dependencies

These concepts and articles establish the explanatory resources presupposed by this article. They identify dependencies within the APS explanatory corpus rather than chronological order, hierarchy, or levels of organisation.

  • What Is APS?
  • APS and Contemporary Theories
  • Comparative Explanatory Methodology in: Theoretical Biology Organised Persistence Biological Agency Viability Constraint Closure Biological Organisation Biological Evaluation Biological Significance Subsequent Explanatory Developments

The comparative distinctions established here are developed or applied in:

  • Agency and Organised Persistence
  • Architectural Dependency and Biological Explanation
  • Organism–World Coupling: Agency, Not Control
  • Explanatory Priority Is Not Ontological Priority
  • Why APS Reframes Biology
  • Related Explanatory Questions
  • How do APS and the Free Energy Principle differ in their explanatory targets and organising principles?
  • When do APS and FEP address genuinely matched explanatory targets?
  • Does mathematical formalisation itself determine the explanatory status of FEP? How do viability, active inference, agency, and adaptive self-maintenance relate across the two frameworks?
  • Under what conditions might APS and FEP prove complementary, overlapping, independent, competing, redundant, or differentially explanatory?
  • What evidence would be required to establish explanatory gain favouring either framework?

Position Within APS

This article forms part of the APS Comparative Biology series. It examines one of the most ambitious contemporary frameworks for adaptive self-organisation against APS’s account of viability-oriented organised persistence. Its role is not to establish APS as a deeper biological explanation or FEP as merely a formal description, but to clarify the conditions under which their explanatory claims can be meaningfully compared. It thereby contributes to APS’s broader comparative methodology while leaving any claim of comparative explanatory superiority to target-matched assessment.