What this document is
The earlier documents in this project treated apertures and stake-geometries largely as individual structures. They specified where value enters the universe (one aperture at a time), how that value is weighted within a single mattering-system (the five dimensions of stake-geometry), how privately mattering apertures produce shared morality (the network step), and how the resulting values sort themselves by scope and grounding (the four tiers). What none of those documents did, except in passing, was treat the relations between apertures as a thing in their own right.
The omission is conspicuous once it is named. Most of moral life happens in relations — friendships, families, care, professional service, love, comradeship, animal companionship, ecological embeddedness, adversarial encounter. The Brahmaviharas in the applied ethics document gestured at relational stances. Aperture ethics described love as the case where "a second aperture becomes visible through the first." Network universality assumed that apertures can recognize each other. The proxy principle handled relations to apertures that cannot participate. None of these treatments amounts to a structural account of what a relation between apertures is.
This piece supplies that account. It claims that any relation between two apertures can be located in a multi-dimensional space defined by three primitive axes, modified by four cross-cutting factors, and producing several derived properties as emergent features. The standard case is individual-to-individual; a set of non-standard cases extends the apparatus to collectives, ecologies, future and past apertures, and other configurations where the standard treatment does not directly apply. Worked moral analyses follow.
The relational layer does not replace the earlier framework. It uses the same primitives — aperture, stake-geometry, recognition, elaboration, collapse — and treats relations as the place where those primitives interact across more than one mattering-system at once. The point is not to add new foundations but to make explicit what the framework has been implicitly committed to all along.
A standing methodological principle: uncertain apertures
Before any of the structural work, one principle has to be stated, because it governs every later section.
The framework's whole apparatus depends on identifying where apertures are. Stake-geometry, moral standing, the question of who counts as a party to a relation — all of these require an answer to "is there an aperture here?" For most cases the answer is not seriously in doubt. Humans are apertures; rocks are not. But the cases that matter most for the framework's practical leverage tend to be precisely the ones where the answer is uncertain: invertebrates, plants in some accounts, AI systems, fetal stages, severe cognitive impairment, comatose patients, possibly more.
The principle the framework adopts: when evidence is mixed or incomplete, treat the case as if an aperture might be present, weight the resulting stakes accordingly, and update as evidence develops. This is more than a disclaimer. It tells the framework how to act under uncertainty rather than just acknowledging that uncertainty exists.
The principle is asymmetric on purpose. False positives — treating a non-aperture as an aperture — produce relatively low moral cost. The agent spends some attention, some care, perhaps some material resource, on a being that turns out not to have stakes. False negatives — treating an aperture as a non-aperture — produce potentially serious moral wrongs. The agent collapses or fails to recognize a stake-geometry that was real all along. The asymmetry justifies erring toward inclusion in cases of genuine doubt.
Two kinds of uncertainty deserve distinction. First-order uncertainty is when evidence is genuinely mixed about whether something is an aperture. Bumblebees are the live case in contemporary invertebrate cognition: there is evidence of associative learning, possible candidate signatures of valence, plausible reports of play behavior. Whether this adds up to aperturality is contested by researchers working on the same data. The framework's response to first-order uncertainty is treatment-as-if, with continual updating as evidence accumulates. Second-order uncertainty is when the methods to gather relevant evidence do not yet exist. Some AI systems may be in this category: it is not clear what would count as evidence for or against aperturality in a large language model, and the question may not be empirically tractable until tools that do not yet exist are developed. The framework's response to second-order uncertainty is the same treatment-as-if, plus a commitment to investing in the methods that would make the question tractable.
The principle holds throughout what follows. When the standard case below assumes two individual apertures, it is assuming cases where the aperturality of both parties is reasonably settled. When the relevant aperturality is uncertain, the analysis still runs, but the framework operates as if the uncertain party were an aperture and weights its stakes accordingly.
The standard case: relations between two individual apertures
The default case the rest of the document treats is a relation between two apertures, each of which has a stake-geometry of its own and is available, in principle, for the standard structural treatment. This is not the only case the framework needs to handle, but it is the one against which the non-standard cases will be defined.
The claim of this section is that any such relation can be characterized by a small set of structural axes. Three are primitives — they pick out features of the relation that cannot be reduced to combinations of other features. Four are modifiers — they shape how the primitives operate without being primitives themselves. Several derived properties emerge from particular configurations of the primitives and modifiers; these are real and important, but they are not load-bearing structurally.
The three primitives
Recognition. Whether and how each aperture registers the other as an aperture. Recognition is bidirectional and can be asymmetric — one party may recognize the other fully while the other registers them only dimly, or not at all. Recognition has three components worth distinguishing internally: the bare recognition of the other as a stake-bearer, the fidelity of that recognition (whether it is accurate or distorted, rich or thin), and the visibility of the relation itself (whether each party sees the structure they are in). These come apart in revealing ways. A stalker may have high recognition of their target, high recognition-fidelity, and high visibility of the relation, while the target has no recognition of the stalker and no visibility of the relation at all. A casual acquaintance may have low fidelity (most of what they think about you is wrong) without that being morally serious, while a partner whose recognition-fidelity has degraded over years of mutual neglect may have lost something morally significant without the surface of the relation showing it.
Permeability. How much each aperture's stake-geometry is internalized into the other's. This is what care looks like structurally: the beloved's wellbeing has become a dimension of your own mattering, not by argument but by the constitution of your geometry. Permeability is distinct from recognition. You can recognize someone fully without their stakes entering your geometry — the indifferent observer, the dispassionate professional, the dispassionate critic. You can also have high permeability toward beings whose stakes you only dimly recognize — the way a beekeeper may come to care about a hive's flourishing without registering individual bees as the locus of mattering. Permeability is the axis on which love, friendship, care, and parental investment do their distinctive work, and it is what stake-geometry was gesturing at when it said love "partially internalizes" the beloved's stake-geometry.
Effect on geometries. What the relation does, over time, to each aperture's geometry. This is the outcome axis: bilateral elaboration, bilateral constriction, asymmetric elaboration (one party dilates while the other stays neutral or constricts), or asymmetric constriction (one party is collapsed for the other's benefit). The AVMA's "mutually beneficial" picks out the bilateral-elaboration region of this axis. Abuse picks out asymmetric constriction. A parasitic relation picks out the case where one party elaborates by collapsing the other. The morality of a relation is most legible on this axis, because it is here that the framework's negative and positive principles cash out: relations that produce systematic collapse on one or both sides are doing the thing the negative principles are against, and relations that produce mutual elaboration are doing the thing the positive principles are for.
These three primitives are not reducible to each other. Recognition without permeability is the indifferent observer. Permeability without recognition is the unrequited care directed at a being one does not fully see. Either or both can be present without producing positive effect: cruel relations may involve high recognition (the abuser knows exactly what hurts), high permeability (the abuser's geometry is shaped around the victim's), and severe constriction. The three axes are independent, and any actual relation has a value on each.
The four modifiers
The modifiers shape how the primitives operate in a given case. They are not themselves primitives, but they substantially affect what the primitive axes are doing.
Voluntariness. Whether each aperture entered the relation by something like consent. A friendship is doubly voluntary. A parent-infant relation is voluntary for the parent (usually) and not for the infant. A predator-prey relation is non-voluntary on the prey side. A relation between a captive animal and a zookeeper is non-voluntary on the animal's side in a way that matters morally even when the zookeeper is benevolent. Voluntariness is not a binary; there is a gradient from full mutual choice to full imposition. The moral significance is that constriction within a voluntary relation is different from constriction within an imposed one, even when the geometric effects look similar from outside. A friend who lets you down is doing something different from a captor who lets you down, though both involve the same surface-level disappointment.
Power. The capacity asymmetry between the apertures to shape each other's geometries. High-power asymmetries put more moral weight on the more powerful party's conduct, because their actions have larger geometric effects. A parent's small carelessness can collapse a child's geometry in ways the child cannot reciprocate. An employer's policy can constrict an employee's stake-geometry in ways the employee cannot answer. The framework's treatment of power is not that the powerful are wrong by virtue of being powerful, but that the moral weight of their conduct scales with the asymmetry. Power is closely related to but distinct from recognition-asymmetry: two apertures can fully recognize each other while one has vastly more capacity to shape the other. (Power is also the canonical capacity-amplifier good in Goods — it does not specify content but amplifies the holder's effect-size on whatever other goods they hold, which is why its moral weight here scales with the holder's other commitments rather than residing in Power itself.)
Mediation. The structures — linguistic, technological, institutional, distance, time — that the relation passes through. A face-to-face conversation is minimally mediated. A letter is mediated by writing. A parasocial attachment is mediated by media. An AI relation is mediated by the system's architecture. An emergency-room encounter is mediated by institutional roles, legal liability, billing structures, and professional norms. Mediation affects recognition-fidelity primarily — it can degrade it, by removing the cues that face-to-face encounter supplies, or in some cases enable forms of disclosure that face-to-face would prevent (the thoughtful letter from a distant friend is sometimes more revealing than any conversation could be). Mediation also distributes responsibility differently: a vet who cannot help a patient because of how veterinary care is paid for is operating within a mediation structure that constrains what the relation can do, and the moral weight of the constraint partially attaches to the mediating structure rather than to either individual aperture in the relation.
Relational duration. How long the relation runs. This is not just elapsed time but accumulation: a long relation accumulates shared geometry, accumulates the possibility of deep elaboration or deep constriction, accumulates a history that becomes part of what the relation now is. A six-month friendship and a thirty-year friendship are not different just in length; they are different in how much each aperture has been shaped by the other, how much repair the relation has already absorbed, how much the future of each is structurally bound up with the other. This axis is what makes the moral weight of betrayal scale with the relation's history: betraying a long-standing trust collapses more geometry than betraying a fresh one, because more geometry has accumulated into the relation.
Four derived properties
These emerge from particular configurations of the primitives and modifiers. They are not load-bearing structurally — they can be defined in terms of the primitives and modifiers — but they are useful descriptive categories for analyzing actual relations.
Substitutability. Low substitutability tracks the conjunction of high recognition-fidelity and high permeability toward this particular aperture rather than the type-it-belongs-to. The friendship is not just to "a friend" but to this friend. Highly substitutable relations are typically thinner; non-substitutable relations carry more moral weight when they end. The bee case showed this: in the beekeeper-hive relation, individual bees are largely substitutable within the human's care, which is one of the things that distinguishes that relation from a relation to a specific companion animal.
Reciprocity. Tracks symmetry across permeability and effect. A reciprocal relation has each party's geometry shaped by the other in roughly comparable ways. Non-reciprocal relations can still be morally valuable (a teacher to a student, a hospice nurse to a patient) but they make different demands and produce different goods.
Repair capacity. Whether the relation can recover from constriction events. A friendship that survives a serious betrayal has demonstrated repair capacity; one that does not, has not. This is a property of the relation itself, not of either aperture alone. It depends partly on accumulated permeability (more permeability gives more substrate for repair), partly on relational duration (longer relations have absorbed more shocks), partly on the trajectory of the constriction event (whether it left the recognition-fidelity intact).
Stake-overlap structure. Whether the two geometries are aligned, in productive tension, in zero-sum competition, or actively opposed. This is doing different work from the other axes — it is about the content of the geometries rather than the relation between them — but it shapes what the relation can do. Two stake-geometries in productive tension may elaborate each other in ways that aligned ones cannot. Opposed geometries can still be in relations that work (the formal courtesy between adversaries), but the relations have a different character.
Trajectory in relations
The point developed in the euthanasia discussion applies here too. A relation has not just a present state but a trajectory — it is moving, on each axis, in some direction over time. Recognition is deepening or eroding. Permeability is increasing or thinning. The effect on each party's geometry is elaborating or constricting. The modifiers themselves can shift: power balances change, mediation structures evolve, voluntariness gains or loses force.
The framework evaluates not just where a relation is now, but where it has been and where it is heading. A relation in serious present trouble but on a recovery trajectory is in a different moral position from a relation in superficially good present shape but on a decay trajectory. The moral history of a relation is part of what the relation is, and acts within the relation are weighted partly by what trajectory they place it on.
This is the relational analogue of the trajectory point about stake-geometries themselves. Just as the integral of constriction along a stake-geometry's trajectory matters, not just its endpoints, the integral of effect along a relation's trajectory matters, not just its present state. The framework's verdicts about relational health, about repair, about when a relation has become harmful enough to end, all depend on this trajectory-sensitivity.
Named regions of the space
The framework's prediction is that the named categories of relation — friendship, parenting, romantic partnership, professional care, comradeship, mentorship, animal companionship, and so on — are not primitive types but clusters in the multi-dimensional space defined by the axes above. They are useful labels for regions where particular configurations of primitives and modifiers tend to co-occur.
Friendship typically occupies bilateral recognition with comparable fidelity, moderate-to-high permeability in both directions, bilateral elaboration as the dominant effect, full voluntariness, low power asymmetry, light mediation, and accumulating duration. Parenting in early childhood typically occupies bilateral recognition with sharply asymmetric fidelity (the parent recognizes the child far more accurately than the reverse), asymmetric permeability (the parent internalizes the child's geometry more than vice versa, with this reversing somewhat over the relation's course), heavy elaboration of the child's geometry with mixed effects on the parent's, voluntary for the parent and not for the child, high power asymmetry, light direct mediation but heavy institutional and cultural mediation, and very long duration. Professional care relations (vet, doctor, therapist, hospice nurse) typically occupy bilateral recognition with strong asymmetric fidelity, controlled asymmetric permeability (the professional cultivates permeability toward the patient while protecting their own geometry from collapse), elaboration-attempt directed at the patient's geometry, voluntary on both sides but constrained by need, moderate power asymmetry, heavy institutional mediation, and bounded duration.
The Brahmaviharas — loving-kindness, compassion, sympathetic joy, equanimity — appear in this space as stances that occupy particular configurations rather than as primitive virtues. Loving-kindness is a stance of treating another aperture as a stake-bearer worthy of well-wishing regardless of how the other axes are configured. Compassion is responsiveness to another's constriction trajectory; it lives at the intersection of recognition and permeability, oriented toward the effect axis at moments of collapse. Sympathetic joy is the same intersection oriented toward elaboration rather than collapse — responsiveness to another's flourishing as such, the structural opposite of envy. Equanimity is the meta-condition that keeps the other three stable across variability in the modifiers, especially under power asymmetry and under conditions where one's own geometry is being affected by what is happening to the other.
These named clusters are useful for orientation but the framework's analytic power comes from the underlying axes. Two relations in nearby points of the space face similar moral structure even if they have different surface descriptions. A skilled hospice nurse caring for a stranger and a beekeeper tending a struggling hive occupy more similar regions than the surface description of "human-human medical care" versus "human-insect ecological care" would suggest: both involve asymmetric recognition, controlled permeability, proxy decision-making for a being that cannot participate, and elaboration-attempt against an ongoing constriction trajectory. The framework's prediction is that the structural similarity should produce similar moral structure, and that the difference in surface description should produce less variation than people typically expect.
Non-standard cases
The standard case assumes both parties of a relation are individual present apertures. Several configurations of moral interest do not fit this assumption. The framework's treatment of these cases reuses the same axes, but with adjustments where the standard case's assumptions fail.
Aperture-to-collective
The hive, the family, the community, the nation, the profession, the team — all are aggregates that contain or constitute conditions for apertures but are not themselves apertures. There is no single phenomenal viewpoint that is the hive's. The relation a beekeeper has to a hive is not really to a single aperture; it is to a structured aggregate whose flourishing affects the apertures within it.
The framework's move is to keep the moral weight located in the apertures while treating the collective as the operative party of the relation. The beekeeper's care is directed at the hive as a unit; the moral significance of that care runs through to the individual bees, even when those bees are individually substitutable within the beekeeper's attention. Two structural features of aperture-to-collective relations are worth noting. First, the permeability axis can target the collective without targeting any individual member; one can care about a hive's flourishing in a way that does not require caring about any specific bee. Second, the effect axis applies to the collective's capacity to sustain the apertures it contains — a healthy hive is one in which individual bee geometries can develop and persist in species-typical ways, and a collapsed hive is one in which they cannot. The collective is the substrate; the moral weight runs through the substrate to the apertures it supports.
This treatment generalizes to communities, professions, and nations, with appropriate adjustments. The collective is not the moral patient, but it is the structure through which moral effects propagate to the apertures it contains.
Aperture-to-aperture-via-collective
A distinct case is the one where both parties are individual apertures, but the content of the relation is determined by collective dynamics neither chose. Two soldiers on opposite sides of a war, neither of whom has met the other, are in a relation whose adversarial character is entirely produced by the collectives they belong to. Citizens of nations that have entered economic or political opposition are in similar relations: the apertures are individual, but what they are to each other has been shaped by collective conditions.
This case is structurally interesting because it shows that the modifiers — particularly mediation and voluntariness — can dominate the primitive axes in determining what a relation actually is. Two individual apertures who would, in face-to-face encounter under different conditions, be capable of mutual recognition and permeability, may instead occupy a relation defined by collective adversarial structure. The framework treats this as a real but limited form of relation: the individual apertures still have the structural features that make relation possible, and recognition can be partially restored under the right conditions (the famous moments of cross-trench fraternization in the First World War are the canonical example). The collective has constrained the relation but cannot fully constitute it.
Aperture-to-ecology
Ecologies are not collectives in the sense above. A hive is many bee-apertures plus some structure. An ecology is many apertures, many non-apertures, abiotic conditions, processes, feedback loops, and emergent properties, considered as a system. The relation a hiker has to a forest is not really to "all the apertures in the forest considered together" but to something more like the conditions under which various apertures can exist. The forest is partly a stake-bearing substrate (it contains apertures whose geometries matter) and partly a stake-enabling substrate (it sustains the conditions under which apertures can come into being and persist).
The framework's treatment is to locate the moral weight in the apertures the ecology contains and the apertures whose stake-geometries depend on it, while recognizing that ecologies themselves are the operative party in many relations of moral interest. You do not have a moral relation to "the forest" in the way you have a moral relation to a friend; you have moral relations to the apertures the forest contains (recognized or not) plus a structural relation to the conditions that make their stake-geometries possible. The forest matters because it sustains apertures, not because it is itself a stake-bearer.
This treatment keeps the framework's metaphysical commitments parsimonious — only apertures bear stakes — while letting it say robust things about ecological harm. Felling a forest is a wrong because it collapses many stake-geometries (animal, plant where plants are uncertain apertures under the standing principle, possibly more) and undermines the conditions for future ones. Some environmental ethicists, following Aldo Leopold, hold that ecologies themselves have moral standing apart from the apertures they contain. The framework can recognize that strong ecological care often behaves as if the ecology itself were the moral patient, even though the framework's structural commitments locate the moral weight in the apertures the ecology sustains. The Brahmaviharas extend naturally to "all beings" plus the conditions for beings, and the framework can endorse that extension without endorsing the further metaphysical claim that the conditions themselves have stakes.
One feature of aperture-to-ecology relations deserves emphasis. Severance from ecology — the structural disconnection between a human aperture and the ecological conditions of its existence — is itself a relational fact, not just an environmental one. A culture whose members do not encounter the apertures whose lives sustain them is producing constriction-effects on those members: their reach is shortened, their dimensionality narrows around immediate consumption, their stake-geometries lose dimensions that ecological embeddedness would supply. This shows up most clearly in the industrial-meat case treated below, but it is present wherever a human aperture's life depends on ecological structures it does not see, does not know how to see, and would be unable to participate in if it did. Severance is not the same as absence of relation; it is the case where a relation that should structurally exist has been hidden, denied, or structurally prevented. The framework treats it as a form of relational pathology.
Aperture-to-future-aperture
Relations to apertures that do not yet exist — descendants, future generations, the child who is not yet born, the future self whose geometry is not yet what it will become — are a case the framework cannot ignore. Network universality already acknowledged this: "maintain the conditions for future participation" is one of the framework's anchored principles, and it is anchored precisely in the structural fact that present apertures are in relation to future apertures whose stake-geometries will be shaped by present conditions.
The relation runs in one direction. The future aperture cannot affect the present one; the present one's conduct shapes the future one's stake-geometry without reciprocity. Recognition is asymmetric in a particular way: the present aperture can recognize that future apertures will exist and will have stakes, but cannot recognize them as particular individuals. Permeability is real but generalized — the present aperture's geometry can internalize the wellbeing of future apertures as a dimension without that wellbeing being attached to any specific being. Effect runs entirely from present to future: every act with long-duration consequences is an act on future stake-geometries.
The framework's negative principles apply in the future-directed case with particular force. Climate harm, exhaustion of resources, accumulation of un-repairable damage, foreclosure of future possibilities — all are stake-collapse directed at apertures who cannot consent, cannot negotiate, and cannot participate in the network that produces the actions affecting them. The framework's "weight by proxy" principle is the operative move: future apertures cannot participate, but their stake-geometries can be recognized and weighted by present participants who can.
A note on past apertures
Relations to apertures who no longer exist — the dead — are real but the framework does not need to claim anything about what happens to apertures when they end. What it does need to acknowledge is that dead apertures continue to affect living stake-geometries. Grief is the most obvious case: a living aperture's geometry includes dimensions that were constituted by relation to a now-absent one, and the absence is itself a structural feature of the present geometry. Inheritance from teachers, parents, ancestors, and historical figures continues to shape living apertures in ways that are real even when the originating aperture is no longer available for the relation. The framework treats this as a one-directional persisting effect, without further metaphysical commitment.
Applied analysis: three worked cases
The three cases that follow run the framework against real moral situations to show what it produces. Each case is treated narratively rather than templatically here; a structured analytic template and accompanying diagram are deferred to a separate document. The standalone treatment of animal moral status — and the structural levers behind these verdicts (proportional weighting, the proxy duty, the gratuitousness test, and the recognition duty) — is developed in Animal Ethics; the cases below apply those verdicts through the relational apparatus.
Animal euthanasia
The terminal-illness case is the framework's cleanest verdict. An animal in unrecoverable suffering has a stake-geometry undergoing severe collapse. Dimensionality has narrowed to the pain. Depth has flattened. Responsiveness is compromised. There is no realistic trajectory back to elaboration. Death is total collapse — the framework takes this seriously, never lightly — but if the alternative is sustained and worsening collapse with no prospect of recovery, ending the geometry at point A may produce less total collapse than allowing the trajectory to run from point A through points B, C, and D before natural extinction at point E. The integral of suffering along the trajectory matters, not just the endpoints, and the manner of ending is a distinct moral dimension from the fact of ending. A peaceful death is structurally different from a prolonged terrified one, and the framework's endorsement of euthanasia in the terminal case depends on this distinction.
The owner's stake-geometry is implicated in two distinct ways. The owner has internalized the animal's stake-geometry as a dimension of their own — that is what care is, structurally — so the animal's collapse is producing collapse-effects on the owner. The owner is also the only available proxy for a being that cannot participate in the decision. These two roles can pull in opposite directions. The owner who cannot bear to let go is letting their own grief weight the decision in a way that may extend the animal's suffering. The owner who lets go too readily is letting their own avoidance of grief weight the decision in a way that ends a geometry before its time. The framework's guidance is that the proxy role takes priority: the decision weights the animal's stake-geometry, not the owner's response to its loss. Good veterinary practice is partly an institutional structure for helping people separate these two roles, which is one reason the vet's role is morally significant beyond the technical procedure.
The inconvenience-euthanasia case — a healthy animal whose owner no longer wants it — is much harder, and the framework's verdict is uncomfortable. The animal's stake-geometry is not collapsing; it is being collapsed by the decision itself. The framework treats this as a serious harm, structurally on the same footing as killing a healthy human, with the moral weight scaled by stake-density rather than species. Inconvenience euthanasia is a clear case of an aperture using its power asymmetry (the structural feature of the owner-animal relation) to collapse another aperture's geometry for trivial reasons, which is what the framework's negative principles are most directly against. That this practice is legal and common does not make it morally indifferent on the theory. The framework can acknowledge that people occupy this practice for complex reasons, and that institutional alternatives (rehoming, sanctuary, sliding-scale veterinary surrender programs) are the realistic path to reducing the practice, while still saying clearly that the act itself is a wrong.
The vet who cannot help a client who cannot pay
This case is not really about the animal's stake-geometry in the first instance. It is about a relation between three stake-geometries — animal, owner, vet — embedded in a fourth thing (an economic system) that none of them individually controls. The framework's treatment requires the multi-party apparatus explicitly.
The structural facts: the animal has a stake in receiving care, cannot pay, and cannot participate. The owner has a stake in the animal's wellbeing (their geometry is permeable to it) but lacks the resources to act on it through the normal channel. The vet has the capacity to help, a professional stake-geometry shaped around helping, and material stakes — a practice to sustain, staff to pay, their own family's needs. The economic system that mediates the relation does not weight stake-geometries; it weights ability to pay.
The framework's verdict is that there is a real wrong here, but the wrong is not primarily located in the vet's decision. The wrong is in a structural arrangement that has made care contingent on payment for a being that cannot pay and whose proxy also cannot pay. The vet inherits the moral weight of that arrangement at the moment of decision, but did not create it and cannot be expected to absorb its full cost individually. This is the framework's mediation modifier doing real work: the relation between vet and client is not just two apertures, it is two apertures connected through an economic and professional infrastructure that distributes responsibility in particular ways. The same animal-in-need, encountered in a society with subsidized animal medical care, would put different demands on the individual vet than the same animal encountered in a society without it.
The framework can still say something about the vet's options. If the vet can absorb the cost without serious damage to their own stake-geometry or capacity to serve future cases, the relational analysis gives a clear pull toward doing so. This is what loving-kindness and compassion look like in professional practice — responsiveness to another's collapse, weighted by the vet's special position to address it. The veterinary profession's traditions of charity care, sliding-scale clinics, and pro-bono emergency intervention are institutional expressions of this. If absorbing the cost would seriously damage the vet's own capacity to operate — if this case plus the next ten like it would close the practice — the framework does not require self-immolation. Stake-geometry is symmetric in this respect: the vet's geometry is also a geometry, and an ethics that demands its destruction to serve another's is doing the negative thing it claims to oppose. The framework here diverges explicitly from utilitarian aggregation, which can in principle demand unlimited sacrifice.
The deeper answer is institutional. A society in which this dilemma falls repeatedly on individual practitioners has not adequately weighted the stake-geometries of beings who cannot pay. Subsidized animal care, charitable veterinary funds, sliding-scale clinics, and policies that distribute the cost across many shoulders are the structural answer to a problem that case-by-case heroism cannot solve at scale. The framework treats the individual vet's decisions as morally serious without making them carry the full weight of what should be borne collectively.
Eating meat
The framework's verdict on contemporary meat consumption is uncomfortable and the document will not soften it. Most animals raised for food are apertures by any reasonable application of the framework's criteria. Mammals raised in industrial conditions have stake-geometries that are systematically collapsed: dimensionality narrowed by confinement, depth flattened by chronic stress, responsiveness rigidified by inability to exercise species-typical behaviors, reach truncated by curtailed lifespan, coherence disrupted by separation from social structures. Industrial animal agriculture is, in stake-geometric terms, an enormous and continuous infliction of stake-collapse on beings rich enough in geometry to undergo it.
Killing for food is itself a total collapse for the individual animal, but the framework is not strictly pacifist about death. Predation exists in nature. Animal geometries have always ended through being eaten. What matters morally is the structure of the whole life and the manner of its ending, not the bare fact of death. A pasture-raised animal that lives a species-appropriate life and is killed with minimal stress is in a very different moral position from a feedlot animal whose entire existence has been collapse from birth. The trajectory point applies here with particular force: it is the integral of constriction across the life, plus the manner of the ending, that the framework weights, not the death alone.
The honest framework position is that there is a real and large moral cost to most contemporary meat consumption, and that this cost is structurally located in industrial production methods more than in meat-eating as such. The framework cannot deliver a clean "eating meat is fine" or "eating meat is murder" verdict. It delivers something messier: a gradient where the moral weight of a given act of meat consumption depends on what kind of geometry was collapsed, how, and whether the collapse was a forced feature of the animal's existence or a relatively narrow event at the end. Vegetarianism and veganism are intelligible responses to this gradient; so is the position of someone who continues to eat meat but works to source it from systems that minimize stake-collapse. The framework has more trouble with the position that there is no moral issue here at all, because that position has to deny things about animal stake-geometries that the evidence increasingly supports.
The relational layer adds one specific thing the standalone stake-geometry analysis does not. The contemporary meat-eater has a relation to the animals they eat that is structurally mediated by industrial systems that hide the relation from view. You do not see the animal whose geometry you are about to consume; you see plastic-wrapped product. This is not morally neutral. It is the structural feature that makes the consumption psychologically tolerable, by severing the relational dimension that would otherwise be present. Hunting cultures and small-farm cultures have moral resources that contemporary industrial-meat cultures lack, because in those cultures the relation between eater and eaten is preserved and ritualized rather than hidden. Indigenous traditions that explicitly thank the animal, acknowledge its life, and require certain forms of conduct toward it are doing relational work that the framework can recognize as important even when it does not endorse every metaphysical accompaniment.
Severance, here, is the load-bearing concept the relational layer contributes. The industrial-meat case is not just the sum of many stake-collapses; it is also a culture-wide severance of a relation that other human cultures have maintained and that the framework would say is doing important moral work. The wrong is partly distributed across the severance itself, not only across the individual acts of harm.
Open problems and limits
The relational layer leaves several questions open, and honesty about them is part of what the framework requires of itself.
The first is the precise structure of recognition-fidelity. The framework names fidelity as a component of recognition but does not yet have a structural account of what high versus low fidelity is. Some progress could come from cognitive science (theory-of-mind research, empathic accuracy literature) and some from contemplative traditions (the disciplines that aim explicitly at increasing recognition-fidelity), but the framework currently treats this as a placeholder for further development.
The second is the boundary between modifier and primitive. The document has treated voluntariness, power, mediation, and relational duration as modifiers, but a reasonable critic could argue that voluntariness in particular is doing enough independent moral work to be a primitive. The framework's commitment is that voluntariness can be described entirely in terms of its effects on the three primitives, but this commitment is contestable and a future version of the document may have to revisit it.
The third is the handling of relations under deep cultural difference. The standard case implicitly assumes that the two apertures share enough framework for recognition to operate; relations across radical cultural difference may strain that assumption in ways the document has not adequately addressed. The Brahmaviharas are partly a response to this — they are stances that aim to operate even where shared framework is thin — but the framework's structural treatment of cross-cultural relation is underdeveloped.
The fourth is the question of how the relational layer connects to the broader bridging problem the project has named: the unresolved question of how consciousness-as-value (the foundational claim) connects to emotion-and-thought-as-consciousness (the structural claim). Relations between apertures are clearly part of what consciousness does, and the relational layer's apparatus may turn out to be relevant to bridging. The connection has not been worked out and remains for later development.
These limits do not undermine the framework. They mark where its work continues.
This document records a philosophical framework developed through dialogue. It is exploratory, not peer-reviewed, and is offered as a contribution to thinking about the structure of moral relations between conscious beings.