The Architecture of Agency › Volume 5 › Consent and Property

Consent and Property

Two load-bearing definitions

This chapter is a review — it is readable but still changing.

“Sign this or we’re letting you go.” The employee signs. Ask the company and you will hear that she agreed — there is her signature. Ask her and you will hear something else. A user clicks “accept” on a sixty-page terms-of-service document; has he consented to anything? A squatter has held a house for a decade; is it his? A wolf pack patrols a territory and drives off every intruder; do the wolves own it?

These questions cannot be settled by intuition, because the intuitions conflict — that is what makes them disputes. They can only be settled by definitions, and consent and property are two of the most load-bearing words in the moral vocabulary that almost nobody bothers to define. Both do foundational work in ethics, law, medicine, sex, and governance, and both are routinely treated loosely or simplistically. The looseness is not harmless. When the definitions are vague, the words become weapons: “she consented” launders submission into agreement, and “it’s my property” launders possession into right.

The remedy is the same one I applied to coercion: state a definition whose every element is necessary, then test each element with cases that pass and cases that fail. The counterexamples are not decoration. They are how you find out whether a criterion is doing real work — a criterion that nothing fails is not a criterion.

Consent is decision-specific authorization, given intentionally and voluntarily by an agent with sufficient capacity and material understanding, within a stated scope.

Every element is necessary. Revocation applies prospectively where continuation remains possible, while reliance, completed transfers, and emergencies can change what withdrawal can still undo. Take the elements in turn.

Capacity and continuity. Capacity is decision-specific rather than a global badge: the agent must be able to understand, appreciate, choose, and communicate sufficiently for the matter at hand. Valid authorization can persist through later loss of consciousness. A competent patient’s advance directive or preoperative consent is not erased by anesthesia because the continuing agent authorized the later action while capable. Children and impaired adults require supported decision-making and tightly bounded surrogate or fiduciary authority rather than the fiction that present incapacity erases preferences or standing.

Intentional. Consent must reflect a deliberate decision. Someone who is asked to share their location and explicitly says “yes, I agree” has performed a clear volitional act. Someone who mumbles “okay” while distracted, not realizing they are agreeing to surveillance, has not — there was no deliberate act, only a noise that got counted as one. The same goes for consent inferred from silence in a complex legal setting: silence is not necessarily intentional, and treating it as agreement manufactures consent out of its absence.

Materially informed. The agent must understand the nature, scope, material risks, and realistic alternatives relevant to the decision. This does not require omniscience or disclosure of every remote possibility. It requires information a reasonable person in that position would treat as important, plus attention to what this particular agent asks or plainly needs. A passenger who agrees to a ride-share but is not told about hidden audio recording has not authorized the recording. A sixty-page terms-of-service dump can document notice without producing understanding; a ritual of disclosure is not disclosure.

Voluntary. Authorization must not be produced by a threat or dependency that defeats meaningful refusal. Pressure, need, unequal bargaining power, and unattractive alternatives matter, but they do not make every hard choice non-consensual. The baseline and authority matter. An employer may condition an optional role or benefit on relevant terms; threatening an independent setback, breaching an existing entitlement, exploiting control of an essential dependency, or fabricating penalties is different. A detainee’s apparent agreement to a search under threat of increased charges is presumptively involuntary because the authority threatening the penalty also controls the supposed alternative. Voluntariness is evidence-sensitive, not a magic property of a signature.

Scoped and prospectively revocable where possible. Consent authorizes the described actor, act, purpose, duration, and material risks—nothing more. Ongoing participation generally requires a meaningful way to withdraw. Withdrawal ends authorization for continuation; it does not retroactively convert completed authorized acts into violations or necessarily unwind irreversible reliance and transfer. Agreement to begin is not automatically agreement to continue.

Put the pieces together and the composite is unmistakable:

A competent adult, after reading full documentation and receiving verbal clarification, signs a form agreeing to a research study, knowing they can opt out at any time and facing no consequences for refusal.

Capacity, intention, material understanding, voluntariness, scope, and prospective exit: valid consent in this case. Remove an element and the authorization may fail wholly or only outside its proper scope. The definition keeps authorization distinct from mere compliance without pretending every defect has identical consequences.

What Counts as Property

Now the second proposal, which the first turns out to need. What makes something property, rather than mere possession or temporary control? In the framework used here, a strong property claim has five features. This is an operational model for evaluating claims, not a discovery that every legal or cultural system must use the word in exactly this way.

Scarcity. The resource must be inherently rivalrous: one agent’s use necessarily diminishes its availability or utility to others. A plot of land passes — my farming it precludes yours. The open air fails, and so does a tune once everyone has heard it: my use subtracts nothing from yours, so there is nothing for a property claim to protect.

Identifiable boundaries. Property must have clearly defined boundaries — physical, conceptual, or symbolic — that allow precise delimitation and make exclusion possible. A surveyed parcel passes. “The nice part of the valley” fails: a claim without a boundary cannot say what would even count as trespass.

Excludability. It must be practically feasible to prevent others from accessing or using the resource; effective property rights depend critically on enforceable exclusion. A locked warehouse passes. A secret shouted in a public square fails — once it is out, no mechanism exists to keep others from it, and a right that cannot be enforced against anyone protects nothing.

Durability. The resource must persist through time long enough to justify enforcement, investment, and maintenance of exclusive rights. A house passes. A snowbank in spring fails: there is no point defending title to something that will not outlast the paperwork.

Transferability. A mature ownership system normally specifies how control can be voluntarily transferred through negotiation, exchange, gift, or inheritance. Deeded land is the clean case. A wolf pack’s territory is better described as de facto territorial control than as socially recognized title: the distinction concerns the kind of institution involved, not an assertion that animals lack every precursor of ownership. Voluntary transfer imports the chapter’s definition of consent. A transfer extracted at gunpoint may move possession without moving legitimate title.

With the conditions in place, ownership itself can be stated formally, as a tuple that enumerates each required component:

\[\text{Ownership} = (A, R, X, E, T)\]

where:

The tuple is not ornamentation; each slot is a place where an ownership claim can fail, and the failures are instructive. Strike out E and you have the central implication: property fundamentally depends on social, legal, or technological enforcement mechanisms, and without them ownership reduces to mere possession. This is the squatter’s situation, and the resolution of his case. His decade of holding the house is possession; it becomes property at exactly the moment some framework will enforce his exclusion of others — as adverse-possession law eventually may — and not a moment before. There is no fact about ownership sitting underneath the enforcement, waiting to be discovered.

Which is to say that property is conditional and interpretive all the way down. Readers of the earlier volumes will recognize the move: all truth is conditional, and property rights are an especially vivid case, because they require interpretative frameworks and societal norms for their existence. No molecule of the house records who owns it. The evolutionary record says the same thing from the other direction: property likely evolved from territorial behaviors, with proto-property forms among pre-human hominins — resource caching, tool ownership, territory marking — setting the stage for the fully symbolic, transferable ownership norms that emerged in humans. What was added along the way was not a metaphysical ingredient but a social technology: symbols, agreements, enforcement.

The hard cases test the model rather than merely confirming it. Animal territories meet scarcity and excludability but lack many of the symbolic and institutional features of human title, so the model classifies them as territorial control or proto-property. Intangible property runs the other way: intellectual property and digital assets are weak on natural boundaries and natural excludability, so explicit social or technological frameworks — copyright law, cryptography — construct boundaries and enforcement. Whether those constructions are legitimate remains a separate question; enforceability explains how a property regime exists, not why every rule it enforces is just.

Two Definitions, One Machine

Neither of these definitions is a discovery about the metaphysical furniture of the world. There is no consent written into nature and no ownership inhering in atoms; there are agents, agreements, thresholds, and enforcement mechanisms, all specified relative to frameworks that could have been otherwise. That is not a weakness. A definition you know is constructed is one you can state in full, test against counterexamples, and revise when it misclassifies — which is more than can be said for the intuitions it replaces.

And the definitions interlock. Property’s transferability condition presupposes consent; consent’s uncoerced condition presupposes coercion; and coercion is itself a threat of the harm this part began by defining. The wrongs we care most about — theft, fraud, assault, expropriation — are at bottom claims that someone was harmed through the violation of exactly these boundaries. With the classifications in hand, one question remains — when may force answer such wrongs — and that is the work of The Boundaries of Force.