Introduction

The language of agency is often associated with intention. In ordinary human contexts, an agent is someone who acts for reasons, forms goals, makes decisions, or deliberately brings about an outcome. Applied to biology, the same language can therefore suggest that genuine agency requires beliefs, desires, representations, intentions, or some other form of cognition.

APS does not use biological Agency in this sense.

Biological agency is viability-oriented organisational activity (Spencer 2026).

This definition identifies what living systems do without requiring that they represent what they are doing, understand the conditions under which they act, or consciously intend the outcomes towards which their activity contributes.

The distinction matters. Living systems regulate internal conditions, maintain boundaries, repair damage, alter activity under changing circumstances, and interact selectively with their environments. Such activity can be organised relative to conditions of viability without the system forming an intention to survive or representing viability as a goal.

Biological Agency therefore precedes the question of whether a living system is cognitive. Cognition, representation, intention, and conscious deliberation require further explanatory commitments. They should not be built into the definition of Agency itself.

This article explains why.

Where This Article Fits

This article is a Clarification article within the APS Agency corpus. It does not establish the canonical definition of biological Agency or determine where Agency should be attributed. Those functions belong to Biological Agency as the Activity of Self-Maintenance, the Agency Anchor.

Nor does this article ask why Agency can be described as the defining activity of Life. That question is addressed in Agency as the Defining Activity of Life.

The present article addresses a narrower boundary: why does biological Agency not require intention, representation, cognition, deliberation, conscious goals, or other mental states?

APS answers this question by distinguishing viability-oriented organisational activity from the increasingly demanding capacities through which living systems can integrate significance across time, represent aspects of their circumstances, form intentions, or reflect upon possible courses of action.

The distinction does not deny cognition, representation, or intention where evidence warrants them. It establishes only that none is constitutive of biological Agency.

1. Agency Does Not Begin with Intention

An organism can act in ways relevant to its continued viability without intending to remain viable.

Consider the difference between two claims:

A living system performs activities that contribute to maintaining conditions of its viability.

and:

A living system intends to maintain conditions of its viability.

The second claim is much stronger.

It attributes not merely organised activity but a particular relation between the system and an outcome towards which that activity is directed. Depending upon what is meant by intention, such a claim may require cognition, representation of an outcome, selection among alternatives, deliberation, conscious awareness, or other capacities that cannot simply be inferred from biological regulation.

The first claim requires none of these.

Organisational approaches to biological agency have likewise sought criteria grounded in features such as individuality, normativity, asymmetry, and the organisation of action rather than beginning from human-like intention (Barandiaran, Di Paolo, and Rohde 2009).

A living system can alter its activity as conditions change. Regulatory processes can counter perturbations. Damage can initiate repair. Boundaries can be maintained selectively. Physiological conditions can be restored following disturbance. These activities can contribute materially to maintaining or re-establishing conditions under which living organisation continues.

Calling such activity agential does not imply that the system mentally represents the outcome or acts because it consciously desires continued existence.

Agency therefore should not be understood as a weakened version of human intention. The APS claim is not that simple living systems possess tiny versions of beliefs, desires, plans, or conscious purposes.

The claim is biological:

Living organisation can be active in relation to the conditions of its own viability without representing or intending those conditions.

This is sufficient to separate biological Agency from an intention-based conception of agency.

2. Viability Without Mental Goals

If biological Agency does not depend upon intention, what gives its activity its directed character?

APS locates the relevant distinction in viability.

Living organisation can continue only under some conditions. Changes in temperature, chemical composition, resource availability, structural integrity, internal regulation, or environmental relations can affect whether that organisation continues functioning as that organisation.

These conditions are not equivalent for the living system.

Some support continued functioning. Some impair it. Some changes can be accommodated or compensated for; others exceed the organisation’s capacity to maintain or re-establish viable conditions.

APS describes this asymmetry as viability-relative normativity.

Normativity in this biological sense does not mean that the organism consciously judges one state to be better than another. It does not imply moral value, preference in the psychological sense, or an internal statement that survival is desirable.

It identifies a materially grounded asymmetry in the relation between conditions and the continuation of living organisation.

That distinction allows biological activity to be oriented without being intentional.

Regulatory activity can counter a perturbation because of the organisation through which the relevant processes are coupled, not because the system represents the perturbation as undesirable. Repair can restore damaged organisation without the system conceptualising an undamaged state. A change in activity can improve the conditions of viability without the organism consciously selecting viability as a goal.

Goal language can sometimes provide useful shorthand for such organisation, but the explanatory claim should remain clear. Viability orientation does not by itself establish a represented goal.

Biological Agency is therefore normatively structured without having to be mentally directed.

3. Agency and Biological Evaluation

The distinction between biological Agency and mental evaluation requires a further clarification.

APS uses Biological Evaluation in a specifically biological sense:

Biological Evaluation is the process through which agency generates significance.

This does not make Biological Evaluation synonymous with Agency.

Agency identifies viability-oriented organisational activity. Biological Evaluation concerns the process through which conditions, changes, and relations acquire significance relative to that organisation.

A change in the environment does not have the same biological significance for every organism. Its significance depends upon the organisation of the living system and the ways in which the change bears upon its viability. Conditions that are negligible for one system may be consequential for another.

This relational significance does not have to be consciously experienced or represented.

Biological Evaluation therefore does not mean that a living system mentally considers alternatives and judges their value. It identifies a biological process through which differences matter because of their relation to the continuing organisation of the system.

This distinction is important for understanding Agency without intentions.

If evaluation were defined as conscious judgment, then viability-oriented Agency would appear to require mentality. But biological significance can be generated through the organisation of living activity before there is evidence for conscious judgment, deliberation, or representation.

Agency and Biological Evaluation can therefore participate in a non-mentalistic account of living organisation without being collapsed into one another.

4. Agency Does Not Require Representation

A similar problem arises with representation.

It is tempting to infer that a system responding selectively to environmental conditions must possess an internal representation of those conditions. A bacterium moving differently under different chemical conditions, for example, can easily be described as detecting what is beneficial and acting accordingly.

Such descriptions may be convenient, but they do not by themselves establish representation.

A living system can be materially sensitive to environmental differences and alter its activity through organised relations coupling environmental conditions to internal processes. The resulting activity can affect viability without requiring the further claim that the system represents the environment internally.

APS therefore does not make representation a condition of biological Agency.

This is a boundary claim, not a general denial of biological representation.

Some organisms may possess representational capacities, and representation may become important for explaining particular forms of cognition or behaviour. Whether a particular biological system represents something is an additional question requiring its own criteria and evidence.

The important point for biological Agency is narrower:

Agency attribution does not itself entail representation.

The same restraint applies to artificial systems. An artificial system may behave responsively, optimise performance, pursue externally specified objectives, or display behaviour readily described in agential language. None of these features alone establishes biological Agency. The detailed comparison depends upon questions concerning living organisation and endogenous viability and is addressed separately in Why AI Is Not Biological Agency.

Representation and sophisticated behaviour therefore cannot substitute for the organisational conditions governing biological Agency.

5. Agency Is Not Cognition

Removing intention and representation from the definition of Agency raises an important boundary question. If viability-oriented activity is biological Agency, should all such activity also count as cognition?

APS says no.

Life does not require cognition; cognition presupposes life.

Agency and cognition consequently cannot be treated as synonyms.

Biological Agency concerns viability-oriented organisational activity. A living system can regulate conditions, respond to perturbation, repair damage, or alter its interaction with the environment without this being sufficient to establish cognition.

Cognition requires additional organisation.

The detailed criterion belongs to the APS account of cognition rather than to this article. For present purposes, the important point is that cognition cannot be inferred merely from increasing complexity, behavioural flexibility, environmental sensitivity, or temporal extension.

Nor should Agency and cognition be placed on a single uninterrupted scale on which sufficiently complex Agency automatically becomes cognition.

Some forms of living organisation may satisfy the additional conditions required for cognition. Cognitive organisation may subsequently support learning, representation, anticipation, planning, intention, and other increasingly demanding capacities. But each transition requires appropriate evidence and conceptual criteria.

There is therefore continuity in the biological organisation from which these capacities develop without explanatory collapse between the capacities themselves.

Agency remains possible where cognition has not been established.

That is precisely why biological Agency can be both genuine and non-mentalistic.

6. Agency Is Not Deliberate Choice

The same distinction applies still more strongly to deliberation and conscious intention.

Deliberate choice requires considerably more than activity that contributes to viability. A system capable of considering alternatives, maintaining possible outcomes in relation to one another, selecting among them for reasons, or consciously intending a future result possesses capacities that should not be projected onto biological Agency generally.

Human beings clearly exhibit forms of agency involving such capacities. Other organisms may instantiate some of them to differing degrees and in different ways.

Their existence does not determine the definition of biological Agency.

APS therefore distinguishes the broad biological category from particular elaborations of agency. Intention, conscious goal pursuit, and reflective choice are possible forms of organisation associated with some agents, not criteria that all biological agents must satisfy.

This prevents two opposite errors.

The first is anthropomorphism: interpreting basic biological regulation as though it involved human-like thoughts and intentions.

The second is anthropocentric restriction: refusing to recognise biological Agency unless human-like mentality is present.

APS avoids both by locating biological Agency in viability-oriented living organisation while requiring independent justification for stronger cognitive or mental attributions.

7. Continuity Without Collapse

The distinction between Agency and intention does not require an absolute discontinuity between basic living activity and more complex biological capacities.

Living systems have evolved and developed forms of organisation capable of integrating activity across increasingly extensive temporal, behavioural, and environmental relations. Some organisms learn. Some anticipate. Some appear to represent features of their circumstances. Some form intentions. Humans can deliberate explicitly about possible futures and reflect upon reasons for acting.

These capacities arise within living systems already characterised by biological Agency.

But historical or organisational continuity does not make the concepts interchangeable.

Agency is not simply the lowest point on a ladder whose successive steps are cognition, representation, intention, and consciousness. Nor does every increase in regulatory sophistication move a system further along such a ladder.

The relevant question at each boundary is whether the organisation required for the more demanding capacity is actually present.

This is especially important for cognition. APS does not define cognition merely as complex regulation or flexible Agency. Its attribution requires additional organisation of biological significance across time and viability-relevant possibilities.

The same methodological principle extends further. Representation should be attributed where representational criteria are satisfied. Intention should be attributed where the organisation required for intention is supported. Reflective Agency requires still more.

Biological continuity therefore need not entail conceptual collapse.

The existence of more elaborate forms of agency does not make intention constitutive of Agency in its biological sense.

8. Why the Distinction Matters

Separating biological Agency from intention changes how agential language can be used in biology.

It allows researchers to ask how living systems regulate and reorganise activity relative to viability without first deciding whether those systems possess minds.

It prevents normativity from being mistaken for conscious judgment.

It prevents environmental sensitivity from being treated automatically as representation.

It prevents behavioural flexibility from becoming an automatic criterion for cognition.

And it prevents sophisticated human agency from determining which forms of biological activity are permitted to count as agential.

The distinction also imposes discipline on APS itself.

Calling activity agential does not explain every feature of that activity. It does not replace the mechanisms through which regulation, repair, behaviour, development, or environmental interaction occur. Nor does Agency attribution establish that an agential description provides explanatory gain for a particular explanandum.

The claim made here is therefore deliberately bounded.

Biological Agency is real without intentions because the relevant activity is grounded in viability-oriented living organisation rather than in the possession of mental states.

Stronger claims require stronger evidence.

Conclusion

Biological Agency does not begin with intention.

Living systems engage in activity through which conditions relevant to their viability are maintained, altered, and, where possible, re-established. This activity can be organised relative to viability without the system representing viability as a goal, consciously evaluating its circumstances, deliberating among alternatives, or intending its own continuation.

APS calls this biological Agency:

Biological agency is viability-oriented organisational activity.

Its normative structure is grounded in viability-relative asymmetry rather than mental judgment. Biological Evaluation can generate significance relative to living organisation without becoming synonymous with Agency or requiring conscious evaluation. Representation, where it occurs, is an additional capacity rather than a prerequisite for Agency.

Cognition likewise requires more than biological Agency alone. Agency can therefore occur without cognition, even though cognitive capacities arise only within living organisation already capable of viability-oriented activity.

The resulting picture is neither anthropomorphic nor anthropocentrically restrictive. APS does not attribute miniature intentions to simple living systems, but neither does it reserve Agency for organisms possessing human-like minds.

The boundary is organisational:

A living system need not intend its persistence in order for its activity to be organised in relation to the conditions of its viability.

That is why biological Agency can be genuine Agency without intentions.

Explanatory Architecture

Central Question

How can biological Agency be genuine biological activity without requiring intention, representation, cognition, deliberation, conscious goals, or other mental states?

Architectural Role

This article is a Clarification article within the Conceptual Foundations cluster.

It establishes the non-mentalistic boundary of biological Agency. It explains why viability-oriented organisational activity does not require the system to represent viability, consciously evaluate circumstances, deliberate among alternatives, or intend its own continuation.

Relationship to the Agency Anchor

Biological Agency as the Activity of Self-Maintenance is the canonical Agency Anchor. It establishes what biological Agency is, where its attribution is warranted, and the distinction between Agency attribution, participation in Agency, explanatory Scale, and explanatory gain.

The present article does not independently establish those conditions.

It begins from the canonical definition of biological Agency and asks what that definition does—and does not—commit APS to concerning mentality.

Relationship to Biological Evaluation

Biological Evaluation is the process through which Agency generates significance.

It is not synonymous with Agency and does not imply conscious judgment. Biological significance can arise relationally through the bearing of conditions upon viability before cognition, representation, or conscious evaluation has been established.

Relationship to Cognition

Agency and cognition are not synonyms.

Biological Agency can occur without cognition. Cognition requires additional organisation and should not be inferred simply from complexity, flexibility, environmental sensitivity, or temporally extended regulation.

The positive criterion for cognition belongs to the APS cognition corpus rather than to this article.

Position within APS

Agency, Process, and Scale are analytic projections of one viability-oriented, constraint-closed organisation rather than components of biological reality.

The present article concerns a boundary within the Agency projection: viability-oriented organisational activity does not require intention or other mental states.

Corpus Significance

This article establishes the principal APS guardrail against both anthropomorphic and anthropocentrically restrictive interpretations of biological Agency.

Its distinctive claim is:

Biological Agency does not require intention because viability-oriented organisational activity does not depend upon mental representation, conscious goals, or deliberate choice.

For the canonical account of Agency, see Biological Agency as the Activity of Self-Maintenance. For the relationship between Life and its defining activity, see Agency as the Defining Activity of Life. For the positive boundary of cognition, see the APS cognition corpus. For the separate artificial-system comparison, see Why AI Is Not Biological Agency.