What Sentience is For
A note to the reader. This essay is blunt about what is done to animals, and it does not soften the descriptions. The argument turns on the reality of the pathology, and sanitizing it would reproduce the very looking-away I am trying to explain. Some readers — vegan readers especially — will find parts of it hard to sit with; that discomfort is not incidental to the argument, it is close to the point.
It also draws, in places, on evidence produced by experiments that harmed the animals studied: isolation-reared primates, stereotypy blocked to measure the stress its prevention raises, bees frightened to establish that they can be frightened. I want to be plain about the tension — some of the clearest causal evidence that animals have needs was extracted by violating those needs. I cite these findings not to vindicate those experiments but against them: read differently, what was recorded as data about mechanisms is testimony about what these beings are and what was done to them. The harm is already done, and refusing to use what it exposed would return nothing to the animals; turning it toward their defense is the one use of it that might. Where naturalistic or observational evidence carries the same weight, I have preferred it.
A sow in a gestation crate bar-bites for hours. The behavior is rhythmic, invariant, sustained; she appears to have difficulty stopping. The pattern develops over weeks: what begins as short bursts of aggressive activity directed at the bars becomes prolonged bouts of less aggressive, more repetitive behavior, eventually performed at all times of day. New components appear as old ones rigidify. The behavior detaches from whatever originally triggered it.[1]
The shape of the behavior tells you what went wrong. Bar-biting comes from thwarted foraging: the sow's oral behavior, redirected toward the cage, substitutes for the consummatory activity her confinement prevents. Pacing in zoo animals traces thwarted ranging. The oral stereotypies of the milk-fed veal calf trace its thwarted drive to suck. Self-mutilation in isolation-reared primates traces thwarted social contact. In each case, the form is derived from the behavior that would have addressed the need, so at onset it indexes which motivational system was frustrated — a strong signal, though once a stereotypy is established and detaches from its eliciting cause the form no longer maps one-to-one; what it registers unmistakably is that a need is being overridden.
An animal in a barren cage with adequate food, water, and temperature control can still develop severe stereotypies, because the arrangement forecloses the strategies — foraging, ranging, social interaction — through which its needs for autonomy, connection, and subsistence would be met. The stereotypy is not a sign of insufficient provision. It is a sign that the provision addresses the wrong thing.
We already know this about humans. Solitary confinement provides food, water, shelter, sanitation, temperature control — every physiological requirement met. It is recognized as torture. The reason is precisely the one my needs framework identifies: the arrangement forecloses the strategies through which the prisoner's needs for connection, participation, autonomy, and understanding would be met. The pathology that results — hallucinations, self-mutilation, stereotypic pacing, cognitive deterioration — is not a failure of provision but a consequence of what the arrangement prevents. The isolation cell and the barren cage produce analogous damage through the same mechanism: adequate maintenance of the body while systematically denying the conditions under which the organism develops and functions species-typically. In both cases, what the arrangement preserves is the capacity to experience; what it destroys is what that capacity is for.
Welfare science can measure the pathology. It can document the cortisol levels, the immune suppression, the behavioral abnormality. It can register that a life pushed below a threshold is worse for the animal — that much its own measures do. What its framework is not equipped to do is ground that evaluation or explain the arrangement behind it: it can tell you the sow deviates from a metric, but not why the threshold sits where it does, whose interests the arrangement serves, or why it persists. For that, you need a framework that evaluates arrangements by whether they address the needs of those they structure, and a structural account of why certain arrangements systematically frustrate those needs.
The arrangements at issue are not marginal. Billions of animals are held in them at any given moment — the body's maintenance provided, nearly everything the animal is built to do prevented — and they are held there by industries organized for the purpose, not by oversight.[2] The evaluative premise the case turns on, meanwhile, is one most readers already hold: anyone who condemns solitary confinement on the grounds just given — that an arrangement can be condemned for what it forecloses and not only for what it withholds — has accepted the premise that decides the animal case. What is left to settle is whether that premise is followed where it leads, or halted at a species boundary that would then have to be argued for rather than assumed.
Two things turn on the answer. The first is what counts as a remedy. If the pathology comes from foreclosure rather than from insufficient provision, the reform program built on welfare metrics is not merely slow; it is aimed at the wrong object. Enrichment and larger enclosures do move animals further from the threshold, and that is not nothing — but they leave standing an arrangement whose defining feature is that the strategies through which needs would be met are foreclosed by design, and it is the arrangement that produces the damage. The second is the coherence of a politics. A critique that grounds itself in the systematic frustration of needs and then restricts its scope to one species, without saying what warrants the stop, is not applying a principle but halting at an inherited border.
This essay develops what it means to say that non-human animals have needs: what cross-species needs are, where they apply, how class power reorganizes animal activity through the same mechanisms it uses on human labor, and what my materialist needs framework adds to existing approaches. The title names the claim that organizes the rest: the free exercise of an animal's species-typical capacities is not a luxury added on top of survival but what sentience is for, and the arrangements analyzed here keep animals alive while destroying exactly that.
The argument runs in seven steps. Section I presses the consistency test: whoever grounds needs biologically cannot stop at their own species without an argument for stopping. Section II sets out what cross-species needs are and the four independent lines of evidence that they exist. Section III fixes where the evaluative operation applies — at sentience, the capacity for experience that feels good or bad — and why a gradient of experience licenses no hierarchy of standing. Section IV takes the hardest case for an account grounded in species-typical needs: the animal whose species-type has itself been engineered. Section V follows class power into the managed hive, the dairy, and the work animals do, and reads their resistance as evidence about the arrangement rather than as pathological outburst. Section VI measures the existing cross-species political theories — capabilities, political membership, sentientist cosmopolitanism, care — against the gap they share. Section VII states what follows.
"Where the Logic Is Most Complete" and "Bred Into Existence" stand on the argument developed here rather than the other way round; nothing below depends on having read them. The first identifies speciesism as the fullest expression of meritocratic reasoning; the second applies social reproduction theory to animal exploitation at the extreme end of the incorporation gradient. Both presuppose that nonhuman animal needs are real and evaluatively significant. This essay makes both claims explicit and defends them.
I. The consistency test
The metaethics essay grounds moral evaluation in species-typical needs, identified biologically.[3] The evaluative operation runs on arrangements: does this arrangement address the needs of those it structures? The grounding is biological, not contractual: needs are features of the kind of animal, not products of agreement or consensus. They are identified through the biological architecture the species possesses: its neurobiological systems for affect, motivation, social bonding, threat-response, reward, and play. These systems produce the conditions under which its members develop species-typically. Needs are those conditions. Their frustration produces identifiable pathology, physiological, behavioral, and developmental, regardless of whether the arrangement provides a vocabulary for describing what went wrong.
One distinction is central to what follows. Needs are few, broadly shared within a species, and grounded in its biological architecture. Strategies are the concrete, historically and environmentally variable ways in which needs are met.[4] The need for connection is species-typical for wolves (and many other animals); the pack is a strategy for meeting it. The need for autonomy is species-typical for pigs (and many other animals); rooting and foraging are strategies for meeting it. The same need can be met by many different strategies, and whether a particular strategy is available depends on the conditions the organism finds itself in. What matters evaluatively is whether the arrangement permits the need to be met, not whether the specific strategy the species would use under unmanaged conditions is available. A stereotypy is itself a strategy: the animal's body attempting to address a need that the arrangement has made unaddressable. The form of the stereotypy preserves which strategy was frustrated at onset, and thus which need; its intensity and persistence register how strongly the animal is motivated to deploy it. This is why the sow bar-bites and the ranging animal paces.
If the grounding is biological, it cannot be species-specific to Homo sapiens. The neurobiological architecture underlying affect, motivation, and social bonding is homologous across mammals, structurally analogous in birds, and functionally convergent in lineages as distant as cephalopods and social insects. The specific systems differ: mammals share limbic, oxytocin-mediated, and dopaminergic circuits; octopuses built complex brains through independent protocadherin diversification; social insects coordinate through neurochemical systems with no mammalian homologue. But wherever the architecture produces valenced experience, motivational orientation, and social responsiveness, the organisms that possess it have conditions under which they develop species-typically and conditions under which they don't. The wolf's needs for autonomy, order, and connection (expressed through ranging and pack coordination), the octopus's for play, understanding, and engagement (expressed through exploration and object manipulation), the bee's for social order (expressed through the hive's architecture): each is grounded in the species' own biological architecture, identifiable through the pathology that follows when it is frustrated. The species-specific content differs; the principle is the same.
My needs framework either applies wherever the relevant biological architecture exists, or its grounding is not biological but covertly anthropocentric, a disguised appeal to human specialness that contradicts its own premises. (Where that architecture — specifically, valenced experience — actually extends is the empirical threshold question Section III takes up; the disjunction demands only that the grounding be applied consistently across its real extent, not to every organism whatever.) The consistency argument is not an optional extension. It is a test of whether my framework means what it says.
The evaluative operation, too, is the same. An arrangement that systematically frustrates needs is evaluable as exploitative regardless of whether the beings whose needs are frustrated can articulate what is happening to them. The question is not whether the sow can formulate a grievance but whether the arrangement addresses or overrides her needs. The stereotypy answers that question. The capacity for self-articulation is a feature of the human strategy for participation; it is not a precondition for having needs.
As Francione notes, we do not make this demand elsewhere. Because he grounds moral status in sentience alone, the demand never arises for him; Regan, by contrast, working within a rights framework, had to route the same beings through the category of "moral patients." Infants cannot articulate their needs, and neither can people with severe cognitive disabilities. In neither case do we conclude that the capacity for self-articulation is what makes their needs evaluatively significant, or that arrangements structuring their lives escape evaluation because the people affected cannot file complaints. We conclude the opposite: that their inability to contest the terms of an arrangement makes the obligation to evaluate that arrangement more urgent, not less.
But a distinction developed in the companion essay on the family sharpens what "we conclude" actually means here.[5] The care extended to infants and moral patients typically operates not through the pre-evaluative register — responsiveness to need without asking whether the being deserves it — but through the innocence perception: evaluation is suspended because the being is classified as not responsible for their condition. The innocence perception produces real care, but its structural properties are different. It is fragile and revocable: it breaks when the being does something that looks willful, when the classification is contested, or when the cost of care exceeds what the arrangement deems warranted. The institutional treatment of people with severe cognitive disabilities confirms the prediction — care is conditional on the classification, administered within an arrangement the care-receiver cannot contest, and withdrawn when institutional resources are redirected. Under class rule, "their inability to contest makes the obligation more urgent" degenerates into "their inability to contest qualifies them for the innocence classification" — which is a different thing, because the classification is revocable and the obligation it generates is conditional.
For animals as property, even this conditional logic is inverted. The same incapacity for self-articulation that qualifies human moral patients for the innocence perception is treated as grounds for denying that obligations exist at all — or for leaving their determination entirely to the owner's economic calculus.
Two things are moving here that are worth separating, because they can and do move independently: whether a being is classified as not responsible, and whether anything it does obliges anyone. The innocence classification buys care at the price of agency, and what it buys is revocable. The farmed animal is where the two come apart completely — the classification is not revoked but never extended, and no refusal it gives obliges anyone, so nothing needs voiding that property has not already done. Which is also why the same species boundary is drawn differently for the dog and the pig, in the same household, by the same person.[6]
The most common objection runs the other way: that human and animal needs genuinely conflict — that we are predators, and predation is natural. But the predator/prey relation is not a needs-level conflict; it is the limit case of strategy scarcity. The obligate carnivore's biology supplies no alternative subsistence strategy, which is the only thing that makes its subsistence run through another animal's death. That constraint does not hold for the humans who invoke it: where alternative strategies exist, the arrangement is asymmetric by choice, not by necessity. And the constraint is not even absolute for predators — a just-fed leopard has been filmed sheltering the day-old infant of the baboon it had killed rather than eating it, and a lioness adopted a succession of calves of a prey species even as she stopped hunting and began to starve.[7] The caregiving and subsistence systems coexist in the same animal; the predatory response is not a hardwired reflex but a system that can be suspended (satiety, in the first case) or even overridden (misdirected care against subsistence, in the second). "It's natural" imports the one condition — no available alternative — that the human case lacks. This defeats the appeal to naturalness without yet settling whether predation as such falls within my framework's reach: a wolf taking a deer is not an arrangement structured by class power, and Section VI takes up why that places wild predation outside the evaluative operation entirely.
II. Needs across species
The generative principle
My needs framework does not require a universal list of needs that applies identically to all species. It requires a generative principle. Where the biological architecture is homologous across species, the need-dimensions are shared; where species differ in ecology and social organization, the strategies for meeting those needs differ. The principle generates a species-specific needs profile without requiring that we start from human needs and ask which other species "also" have them.
The thirteen dimensions the human-focused analysis identifies (subsistence, protection, connection/belonging, recognition, understanding, participation, play, creation, identity, autonomy/freedom, order/harmony, efficacy/power-to, mourning/celebration) were derived in "Why Competing Solidarities Tend to Win" from Max-Neef's nine, with additions grounded in the cross-species evidence this essay develops.[8] But the principle applies cross-species. For any species with the relevant biological architecture, the question is the same: what conditions does this architecture require for members to develop species-typically? The answer is species-specific in content but universal in form.
The concepts themselves must be calibrated. The method is three steps: describe the behavior using our existing concept; investigate the behavior's mechanisms and compare cross-species; revise both the concept and the attribution together. "We will decide both what we mean by planning and whether the squirrels' behavior counts."[9] A needs framework applied to animals cannot take human-derived categories and ask which other species satisfy the human version. The need for connection is met through pack structure in wolves, matrilineal kin networks in elephants, pair bonding in corvids, shoaling in fish, the hive's social architecture in bees. The need for protection is met through a shell in tortoises, herd vigilance in zebras, burrowing in rabbits, ink-jetting in cephalopods, claw display in crabs. The need-dimension is shared; the strategy is species-specific.
A subtler error runs in the opposite direction. Since needs and strategies are distinct, the cognitive sophistication of a particular strategy is irrelevant to whether the need exists. Defining a need by its most cognitively elaborate human strategy and then concluding that animals — and, by the same move, "simpler" humans — "lack" the need is a confusion of levels — what Buckner, as discussed by Andrews, calls "anthropofabulation": the overintellectualization of human cognition that makes animal cognition look deficient by comparison.[10]
The generative principle, stated so far in functional terms, also has an experiential dimension. What it looks like from the inside is play. The free exercise of an entity's most complex powers or capacities tends to become an end in itself — what Graeber, drawing on Kropotkin, calls the "principle of ludic freedom."[11] The exercise of species-typical capacities is not a luxury added on top of survival but is what survival is for. This is consonant with welfare science's neutrality ceiling: survival-critical provisions produce at best a neutral affective state.[12] Play is not only one need alongside twelve others. It is the experiential register in which needs-fulfillment is experienced as an end in itself rather than as an administered provision.[13] An arrangement that keeps animals alive while systematically preventing that exercise is not meeting their needs at a reduced level. It is destroying what sentience is for.
Four convergent lines of evidence
Four bodies of evidence converge on the claim that needs are cross-species. The evidence is richest for mammals and birds, because ethology has studied them most intensively, but the principle extends to any organism with the relevant architecture. Section III addresses where the threshold falls.
Emotional continuity. The neurobiological architecture underlying emotional experience, including the limbic system, the oxytocin system, dopaminergic reward circuits, and shared neurotransmitters, is homologous across mammals and structurally analogous in birds. Grief, play, empathy, fairness-sensitivity, and social bonding are not human achievements projected onto animals but shared evolutionary inheritances expressed in species-specific forms.[14] Hawaiian Crows and other corvid species display grief, consolation behavior, and attendance at the deaths of conspecifics, mediated by homologous neurochemical systems.[15] These systems share a common evolutionary origin and work through the same neurochemical mechanisms. Emotions are functional states oriented toward need-satisfaction: fear orients the organism toward protection, attachment toward connection, play toward positive affect and social learning. Their cross-species homology is evidence that the needs they serve are also cross-species.
The evidence extends beyond the mammal-bird divide. Bumble bees subjected to simulated predator attacks develop anxiety-like states — learned avoidance, scanning behavior at each flower, and "false alarms" (rejecting safe flowers while searching for absent predators) — that persist twenty-four hours after the attack with no predator present.[16] Honey bees shaken to simulate a predator attack display pessimistic cognitive bias, interpreting ambiguous stimuli as threatening at higher rates than undisturbed controls; bumble bees given a small surprise sugar reward display the opposite bias. The cognitive-bias paradigm is the standard diagnostic for assessing emotional states in vertebrates; applied to insects, it produces the same results. These are not homologous systems — insects lack limbic architecture entirely — but functionally convergent ones: the insect central complex integrates external stimuli, internal states, and past experience in ways that produce valenced responses to the world. If emotions are functional states oriented toward need-satisfaction, their independent emergence in insects is evidence that the needs they serve are not confined to the lineages where the mammalian version of the architecture evolved.
Social norms and moral behavior. Social animals regulate their behavior through other-regarding norms — norms about harm, benefit, fairness, and cooperation. Wolves enforce play rules and ostracize cheaters. Capuchins reject unequal rewards. Elephants assist disabled herdmates. Rats refuse to press a food lever when doing so shocks a neighboring rat, and will work to lower a distressed, suspended companion. Rhesus monkeys have starved themselves for up to twelve days to avoid shocking a companion.[17] The content of these norms is welfare: "harm and benefit are the basic units of moral currency." If welfare is already operative as a regulatory standard within animal communities, then the claim that animals have needs is not an external projection but a recognition of something animals themselves track and enforce. Coalitionary enforcement producing normative expectations operates cross-species, not only in Homo sapiens.
Ravens extend the evidence across the mammal-bird divide. They maintain a social prohibition against pecking the eyes of conspecifics, and enforce it lethally. When a stranger introduced to an established group repeatedly raided food caches, all four residents killed her: puncture wounds covered her body, and her eyes were deliberately pierced — the specific taboo overridden as punishment. They did not eat her flesh. The enforcement was coalitionary: when one bird attacked, its mate and then all others joined in, "obliterating the birds' original individual tolerances."[18] The dead-conspecific response is further evidence of a species-typical norm: ravens given a dead raven or crow reacted with alarm calls and refused to eat the body, which rotted in place; given a dead black chicken, they ate it immediately. One bird approached the dead crow, touched it gently with his bill, and called directly at it with contact calls normally meaning "I'm here."
The elephant case sharpens the developmental point. When poaching and culling disrupt elephant society, young males deprived of appropriate social conditions develop symptoms resembling PTSD: pathological aggression, abnormal startle response, violence against other species.[19] Social conditions are strategies for meeting needs like connection and protection, not just the context in which other strategies operate. Evolutionary history produced the social architecture through which elephants meet those needs; disrupt it and development goes wrong.
Positive affect and funktionslust. The morally relevant capacity is not only the ability to suffer but the ability to experience positive states: play, comfort, social pleasure, the satisfaction of exercising species-characteristic capacities. The German word Funktionslust, the pleasure of doing what you do well, names the last of these.[20] A needs framework that captures only negative states (freedom from pain, fear, frustration) is systematically incomplete. It grounds only negative obligations (don't cause suffering) but not the positive obligation to maintain conditions under which animals can experience the states their biology makes possible. The distinction between biological existence (meeting physiological requirements) and biographical existence (a life with experiences, relationships, history) marks the difference between a minimal welfare standard and a genuine needs standard.
Funktionslust extends into the social register. Much animal cooperation has nothing to do with survival or reproduction but is a form of pleasure in itself: vultures wheeling for entertainment, hares boxing with other species, flocks performing military-style maneuvers, squirrels wrestling.[21] "To exercise one's capacities to their fullest extent is to take pleasure in one's own existence, and with sociable creatures, such pleasures are proportionally magnified when performed in company." Cooperation-as-pleasure, not cooperation-as-strategy. Play and sociality connect here: collectively governed practices addressing multiple needs simultaneously are not instrumental means to an end but are what species-typical functioning consists of.
Ravens demonstrate both dimensions. Heinrich's long-term behavioral study documents snow-sliding repeated six consecutive times, invention of a ball-and-tube game with no food reward, aerial acrobatics with a twenty-foot plastic streamer passed between birds in flight while non-participants perched to watch, and bathing as social pleasure — experimental tests showed that dirty birds do not bathe more, and temperature had little effect.[22] The cooperation-as-pleasure point receives its strongest cross-species confirmation from the raven-wolf symbiosis. Ravens evolved with wolves in a mutualism millions of years old, and the relationship is characterized by play: tail-pulling, mock-stalking, tolerance at wolf dens where ravens walk behind pups and gently yank their tails. In experimental trials at Yellowstone, ravens refused to feed at unattended carcasses (zero of twenty-five within the observation period) but fed without hesitation when wolves were present; numbers at wolf kills averaged thirty-two. "Ravens don't feed next to wolves because they have to, they do it because they want to."
The principle reaches into lineages where play was long considered impossible. Bumble bees given access to small wooden balls will roll them repeatedly, with no food reward attached; younger bees roll more than older ones; the behavior satisfies the operational criteria for play (spontaneous, self-rewarding, developmentally patterned).[23] In a lineage that diverged from vertebrates over five hundred million years ago, under no selective pressure to charm human observers, the same principle holds: the free exercise of species-typical capacities becomes its own end.
Behavioral needs evidenced by their frustration. The stereotypy evidence introduced in the opening generalizes beyond the individual cases. The experimental confirmation is precise: selectively blocking jumping stereotypy in bank voles by lowering cage ceilings produced elevated corticosteroids. The stress response rose when the stereotypy was prevented. The three voles that developed a new stereotypy showed faster return to baseline. The need found a new expression. This is the reading the opening note promised: what the experiment logged as a corticosteroid curve is, read the other way, testimony to the need it had to override to produce the measure. You can suppress a strategy; you cannot suppress the need.
But frustration is not the only tell. The same need shows itself, unpathologically, in the behavior of an animal whose conditions let its strategies run — and in the grief and searching when a bonded conspecific is taken — so the diagnostic runs through normal valenced behavior as much as through the disordered kind.
A distinction from welfare science sharpens the point.[24] Survival-critical affects (thirst, hunger, pain, breathlessness) are genetically embedded motivational systems essential for survival; they can at best be neutralized, never made positive. Situation-related affects (anxiety, frustration, boredom, depression) arise from the animal's perception of its external circumstances, and these can be replaced by positive affects when conditions improve. Only providing opportunities for rewarding behavior (exploration, foraging, play, social bonding) crosses the threshold from survival into a life worth living. Full compliance with welfare codes focused only on survival-critical provisions produces, at best, what the welfare science literature calls "a life worth avoiding" approaching neutrality from below.
The framework that cannot see play
The ethological evidence for cross-species needs is abundant but has been systematically underreported, and the reason is built into the discipline's explanatory framework. An analysis of animal behavior does not count as scientific unless the animal is assumed, at least tacitly, to be operating according to the same means/end calculations that one would apply to economic transactions.[25] Behavior must be explained as serving survival, territory, dominance, or reproduction, or it does not count as explained. Play is an "intellectual scandal" under this assumption: an expenditure of energy toward no identifiable instrumental goal. The result: the field understudies play, undertheorizes positive affect, and routinely cannibalizes the evidence that does exist into attempts to demonstrate an ultimate survival or reproductive function. The obvious rival explanation — that play is simply harder to operationalize and fund than survival-linked behavior — does not compete with this account so much as fall out of it: what counts as tractable to measure and worth funding is itself shaped by what the framework counts as explanation.
What kind of operation this is becomes clearer from its counterpart at the scale of a single person. A companion essay on the capacity to refuse finds the same adjudication running on an individual's own states: a preference will not be heard as a complete reason and has to be converted into something auditable — a conflicting obligation, an illness, a prior claim — before it counts, and an activity whose whole account is that it is its own reason gets reclassified into a hobby, into self-care, into a side hustle.[26] What the discipline demands of play is what that proceeding demands of a want. Because I want to and for its own sake fail the same test, and fail it for the same reason: neither converts. The demand for an ultimate function is not a scientific scruple that happens to fall hardest on play; it is the conversion requirement operating where a discipline sets its standard of explanation rather than where a person weighs their own preference.
This is not incidental. The neo-Darwinist framework that treats life as an instrument for the propagation of genetic sequences (DNA as what Dawkins called "successful Chicago gangsters," ruthlessly expanding their territory) is itself a product of the social order I analyze elsewhere. Where Victorian Social Darwinism treated nature as a marketplace, neo-Darwinism treats it as a site of capitalist accumulation: the drive has no limits, life is a mere instrument, and purposeless expenditure is a scandal. The companion essay "Matter Out of Place" provides the theoretical account of why: Bataille's distinction between the intimate order (sacred, excessive, anti-utilitarian — the animal "in the world like water in water") and the order of things (profane, useful, durational) identifies play as the intimate order's surviving expression. The disciplinary framework's difficulty registering play as anything but scandal is the "complete reduction to the order of things" operating at the level of scientific methodology — the commodity form's taboo-structure letting anti-utilitarian expenditure appear only as an anomaly to be explained away, not as a phenomenon in its own right, within its own categories.
The continuity runs the length of the gradient. What "Matter Out of Place" traces at the collective-human scale — the festival's non-productive expenditure expelled, the taboo against "wasting time" installed in its place — is the same foreclosure this essay traces at the animal scale: the managed hive stripped of swarming, the broiler bred past the capacity to move, the barren cage that forecloses foraging. Capital forecloses the free exercise of capacities as an end in itself wherever it reorganizes life, from the bee rolling a ball to the suppressed saint's-day. Play is not a human luxury the analysis extends charitably to animals; it is the one thing the order of things cannot represent, foreclosed by the same operation across every organism it subsumes.
The economistic assumptions that colonize biology are an instance of what I analyze elsewhere as ignorance production through disciplinary fragmentation: the institutional organization of knowledge production makes it structurally difficult to produce the knowledge that would make cross-species needs visible. The suppression is not merely structural but active: when Heinrich submitted experimental evidence that naive ravens solve multi-step string-pulling problems through mental simulation — acting on first exposure as though they had already run the trials in their heads — the paper was rejected five times; one reviewer invoked Freud, another called the findings "incredible leap of faith" and "matters of the heart."[27] The companion NVC essay identifies the interpersonal analogue: evaluative language replaces needs-perception with desert-classification through the same structural logic — institutional conditions that train people to read behavior as evidence of what someone deserves rather than as an expression of what they need. The disciplinary framework and the evaluative vocabulary are two registers of the same suppression. The evidence is abundant; the disciplinary framework is disposed to underweight it and to convert what it cannot ignore back into function.
This screening is not confined to the disciplinary framework; the same perceptual mechanism operates wherever a subordinate being's higher-level needs must be read from a position of power. Experimental psychology has documented it on marginalized human populations: people systematically underestimate subordinate groups' higher-level psychological needs while acknowledging their physical ones (Schroeder and Epley 2020). Donors overwhelmingly choose gifts that cover physiological needs over gifts that would allow recipients to exercise autonomy, despite recipients far preferring the latter — because the donors underestimate how much autonomy, self-esteem, and respect matter to people they perceive as worse off. The mechanism is the same one that makes the barren cage appear adequate: the animal's physiological requirements are met, and the needs that confinement forecloses — for autonomy, social complexity, exploration, play — are not perceived as needs because the being occupies a subordinate position in the perceiver's evaluative framework. That welfare science focuses overwhelmingly on physical welfare indicators and harm-avoidance measures while struggling to operationalize positive affective states, social complexity, and autonomy reflects three things at once: these needs are genuinely harder to measure, since gauging how well they are met means inferring the animal's inner state rather than counting feed or floor space; the same positionally generated distortion that makes donors choose blankets over cash makes higher-level needs harder to see from a position of power; and fully recognizing them would collide with the extractive use to which the animals are put. The perceptual difficulty and the interest in not seeing reinforce each other.[28]
How the taxonomy works
My needs framework does not specify a single cross-species needs list. It assembles a species' needs profile from three components already in hand: the generative principle (shared biological architecture produces the conditions of species-typical development), the per-species evidence base (ethology, welfare science, and behavioral ecology supply the empirical content of what a given species needs), and the needs/strategies distinction (the need identified by the architecture, the strategy by the species' ecology and social organization) — which lets it evaluate arrangements without prescribing strategies: what matters is whether the arrangement permits needs to be met, not whether it supplies the specific strategy the species would use in unmanaged conditions.
For a given species, the needs profile is assembled empirically but organized by the generative principle. The Five Domains model is the closest existing framework to this operation: it assesses physical and functional states, identifies the affects those states generate, and evaluates whether the animal's life falls above or below the threshold of "a life worth living." What it cannot supply is the structural analysis of why the arrangements that push animals below the threshold exist and persist.
III. Where the line falls
The generative principle grounds needs in biological architecture. But which organisms have needs in the evaluatively relevant sense, where frustration registers as something bad for the organism? A plant can be damaged but not harmed in the morally relevant sense, because there is no subject for whom the damage registers as bad.[29] A sow can be harmed, because the frustration of her foraging need registers as distress that she experiences. The threshold falls at sentience: the capacity for valenced experience, experience that feels good or bad.[30] The consistency argument that refused to stop at the human (Section I) might seem to press further still: if the grounding is biological, why stop at sentience rather than extend to everything that lives? It stops here because valenced experience was the ground from the start, not bare functioning. Damage relative to an organism's telos is real — the plant can be set back — but it becomes harm, a setback that matters to someone, only where there is a subject for whom it registers; that is what sentience marks and mere life does not. The consistency argument therefore reaches exactly as far as valence does, and no further.[31]
This definition is narrower than "phenomenal consciousness" (which includes all subjective experience, including perceptual experience that may not have a valence) and broader than "pain" (which is one form of negatively valenced experience). It does not presuppose a specific theory of what grounds consciousness. It asks a functional question: does this organism have the capacity for experiences that register as positive or negative? If yes, it has needs that can be satisfied or frustrated, and my framework's evaluative operation applies.[32]
Where sentience falls
The question of which organisms are sentient is empirical, not philosophical, and the scientific community has not reached consensus. Five positions on what neural structures suffice for sentience range from the most restrictive (distinctively primate mechanisms required) to the most inclusive (midbrain or functional analogues, including the insect central complex, sufficient).[33] My framework does not need to commit to a specific position. It adopts a precautionary structure: any organism where there is a realistic possibility of sentience that it would be irresponsible to ignore falls within its scope.
On this basis, all adult vertebrates are sentience candidates (the least contested case). Coleoid cephalopods (octopuses, squid, cuttlefish) and decapod crustaceans are sentience candidates.[34] Adult insects are sentience candidates on the most inclusive positions, and recent experimental evidence strengthens the case considerably. Honey bees possess segregated neural pathways for nociception (the neural detection of tissue-damaging stimuli, independent of whether it is felt as pain), and their alarm pheromone floods the system with endogenous painkillers that make guard bees "oblivious to their battle injuries" — context-dependent pain modulation of the kind that distinguishes pain perception from simple nociception.[35] Bumble bees display, besides the pessimistic and optimistic cognitive biases already noted (Section II) — the standard vertebrate diagnostic for emotional states — cross-modal object recognition (recognizing by sight an object they had previously touched only in the dark, and vice versa — implying a mental representation not tied to a single sense), metacognition (opting out of difficult discrimination tasks at rates that track task difficulty, "as if they were aware of their own uncertainty"), and self-awareness of their own body dimensions. Each of these capacities, when demonstrated in mammals, is treated as evidence of sentience without controversy. If the same behavioral and cognitive criteria are applied consistently, bees qualify.
The independent-evolution evidence bears on the threshold question directly. Cephalopod sentience, if real, evolved independently of vertebrate sentience: the common ancestor, roughly 600 million years ago, was almost certainly not sentient, probably a small flattened worm with at most a simple nerve net.[36] Protocadherin molecules, which build neural complexity, diversified at least three times independently: in octopuses, in squid and cuttlefish, and in vertebrates. The octopus-cuttlefish split (~270 million years ago) has roughly the same historical depth as the mammal-bird split (~320 million years ago), meaning parallel brain-building occurred within the cephalopod lineage as well as between cephalopods and vertebrates. If sentience can arise through radically different neural architectures, phylogenetic proximity cannot serve as a proxy for moral consideration. The morally relevant criterion is functional (does this organism have the capacity for valenced experience?), not phylogenetic (how closely is it related to us?).
The independent-evolution evidence also bears on the depth of sentience. The "primordial emotions" (thirst, breathlessness, pain) are, in Derek Denton's account, the subjective side of the homeostatic instincts — interoceptive states that register a deviation of the internal milieu and carry a compelling intention toward the act that would correct it; the sensation, when severe, is "imperious."[37] What makes them phylogenetically old is not their felt urgency but their function and their seat: they are the regulatory imperatives of the internal milieu, in Denton's account subserved by the brain's most ancient regions rather than by late-arising cortex, and life-critical in the earliest vertebrates as much as in us. This is why he takes them to mark where consciousness dawned: reflexive homeostatic regulation became felt before it became thought. If valenced experience is thus the old and widespread thing, sentience preceded the cognitive architectures that later transformed it — simple feeling came first, and complex cognition is what happened to it afterward.
The gradient and non-hierarchy
Sentience is probably gradient rather than binary: a continuum from simple detection to rich experience. That gradient has most often been put to work in one direction: to argue that only beings capable of complexity matter morally, and that "simpler" beings — animals, but historically also women, children, the enslaved and colonized, the poor, the "uneducated" — matter "less," and, where interests conflict, in practice not at all. It also creates a tension with my framework's claim that needs are non-hierarchical. The resolution: having needs is threshold-dependent (the threshold is valenced experience) — the "simple detection" end of the continuum is the sub-valence region the precautionary line screens off, so the threshold names the onset of valence rather than a sharp cut in an otherwise smooth gradient. The range of needs varies by the species' cognitive, affective, and social architecture. A chimpanzee's needs profile is richer than a crab's, including dimensions (creation, identity) that the crab probably lacks. But this does not make the chimpanzee's needs more important. It makes them more numerous. And with that, the crab's need for protection is not less evaluatively significant than the chimpanzee's need for connection, because both are conditions of species-typical development. What varies is the complexity of the profile, not the weight of its elements.
The reason lies one level below the needs, in the need-bearer. The sentience threshold does not measure how much of a subject an organism is; it marks whether there is a subject at all — someone for whom things go well or badly. Experiential richness is scalar and shades in near the threshold, but being a subject with a stake at all is not a further magnitude layered on top of it: it is what the threshold marks, and the margin where even that is genuinely uncertain is what the precautionary line screens. Standing, once secured, does not come in degrees. A crab is not a partial subject and a chimpanzee a fuller one: each is wholly a site for whom its own needs matter, and the evaluative operation ranges over every such site without weighting it by the richness of what it finds there. Richness varies the content the operation addresses, never the standing of the one it is addressed to. This is what blocks the recurring temptation — reinvented whenever it suits the one doing the ranking — to convert a longer list of capacities into a weightier claim on concern. My framework can refuse that conversion where a utilitarian equal-consideration principle[38] cannot, because its operation is not aggregative: it does not sum needs into a quantity a richer profile would swell, but asks, of each subject an arrangement structures, whether that subject's needs are addressed or overridden.
This resolves the gradient problem without reintroducing the cognitive hierarchy that the capacities tradition uses to rank species against each other.
IV. When the species itself has been reshaped
A further problem arises for organisms whose biology has itself been reshaped by class power.
Phillippa Foot and Michael Thompson's natural-historical propositions are contaminated for humans because all observation has occurred under class-structured conditions. For domesticated animals, the contamination is deeper: their biology itself is an artifact of class power. The Holstein cow's body, shaped by centuries of selective breeding for maximum milk production, is not a natural-historical form but a product of real subsumption. Her "species-typical" biology includes udders prone to infection, bones prone to fracture, and a metabolism prone to collapse. What counts as a species-typical need for a being whose species-type has been engineered?
Posed flatly, this reads as a neutral puzzle about baselines. It is not neutral. It is the capacity-ranking of the previous section returning in its most active form. There the temptation was to convert a thinner profile of capacities into a weaker claim on concern — a conversion my needs framework refused and a utilitarian equal-consideration principle could not.[39] Here the same logic reappears, but the thinner profile is no longer merely found and downranked: it is deliberately produced, and then read back as the animal's own nature, the ground on which its treatment is justified. This is where ableism shows itself as the conceptual infrastructure of speciesism.[40] The same capacity-based logic that devalues cognitively disabled humans (who supposedly lack "morally relevant" capacities) licenses the devaluation of animals — the same meritocratic reasoning I trace elsewhere, in which a creature's condition is read as evidence of what it is and deserves rather than of what has been done to it. Challenging speciesism without challenging ableism leaves the underlying evaluative structure intact, and vice versa.
But the domesticated animal marks a step beyond ordinary ableist devaluation. Ordinary ableism takes a disability as given and holds it against the one who bears it. Here the disability is first produced — bred into the body — and only then held against the animal: the same modification that makes the body profitable is the one that makes it broken, and the brokenness is then read as the creature's own deficient nature. The deficiency is manufactured at both ends — engineered into the animal to extract value, and imputed to the animal to license the extraction.
My constructive answer begins with the stereotypy evidence. The chicken bred to ovulate 250 times per year is still driven to dust-bathe and forage. The sow in a gestation crate is still driven to root and nest-build. The veal calf in a stall is still driven to suckle. The motivational systems persist regardless of how thoroughly the body has been reshaped for production. Selective breeding can redesign the body; it cannot eliminate motivational architecture that is phylogenetically far older than domestication. The bank vole experiment makes the same point at the individual level: block one stereotypy and the stress response rises; allow a new one and it falls. The need persists through the deformation; only the strategy changes.
The disability produced by selective breeding is not a side effect of industrial agriculture but its mechanism.[41] These animals are "simultaneously disabled and hyperabled," made disabled by the very enhancements that make them profitable. The battery hen's 250-egg-per-year ovulation rate (versus the roughly 60 her body can sustain without pathology) is simultaneously an "enhancement" from the industry's perspective and a disability from the hen's: osteoporosis, broken bones, chronic pain. The "enhancement" and the "disability" are the same genetic modification. There is no version of this animal that is both profitable and healthy. The dairy cow's lameness, affecting roughly 60 percent of dairy cows, is not an unfortunate consequence of milk production but a structural product of breeding for maximum yield. The disability is the mechanism of profit. And the pathology doubles as its own concealment. The conditional care of the innocence perception is graduated: because it answers to a display of helplessness rather than to need as such, it tracks how pitiable the creature appears, thinning as the animal seems to cope and, past a threshold of visible degradation, curdling into disgust rather than sympathy. The reshaped animals sit at exactly that threshold — bodies made permanently leaking, dependent, and visibly broken trigger the avoidance body's disgust response rather than its sympathetic one, the same circuit the companion essay "Matter Out of Place" traces for human gradient-bottom populations, where suffering produces degradation that blocks the empathic response to it. The deformed species-being is invisible not despite but because of the deformation.
The deformation doesn't eliminate needs. It frustrates protection directly — chronic pain, lameness, reproductive breakdown — while foreclosing the strategies through which needs like autonomy and connection would be met. The needs persist; the strategy-space has been deliberately narrowed to the point where need-satisfaction is structurally impossible. Evaluate against the needs that persist, evidenced by their frustration. Not against an idealized pre-domestication baseline — which in any case is not there to be recovered: species are not sharp natural kinds but fuzzy, shifting populations, and this one's very "type" has been reshaped by breeding. That the species-type is unavailable as a yardstick is no loss to my framework, because it never measured against a type. It measures against what the frustration registers in the living individual — and that evidence is not confined to pathology. The stereotypy, where a blocked strategy turns back on itself, is one form of it; the cow's grief when her calf is taken — the searching, the days-long bellowing, the refusal to settle — is another, and no weaker for being an intact, unpathological response rather than a disorder. Not every sign of a thwarted need is a symptom; some are simply what a creature does when the thing its need reaches for has been removed. The aurochs tells you what cattle were; the stereotypy and the grief together tell you what the cow still needs, and that she suffers under the use, confinement, and disposal of herself and her young whenever the owner finds it profitable.
The mechanism need not be genetic. Captive Hawaiian Crows ('Alalā) have lost parts of their vocal repertoire because there are no free-living adults to learn from.[42] No body has been redesigned. The neurobiological architecture for vocal learning is intact, but the social conditions required for cultural transmission — the conspecifics from whom the strategy is learned — have been destroyed. The domesticated cow's deformation and the captive crow's silence are produced by different mechanisms, but both frustrate needs through the same structural operation: foreclosing the conditions under which species-typical development occurs.
V. How capital reorganizes animal activity
The ethological literature documents what self-directed activity looks like across species. Wolves range over wide territories, hunt cooperatively, and maintain pack structure through negotiated social relationships. Elephants migrate seasonally, maintain multi-generational matriarchal kin networks, and mourn their dead. Corvids cache food, monitor competitors, and innovate tool designs. Octopuses build and rearrange dens, use tools such as carried shells, and solve novel problems like opening jars. In each case, the animal's activity is organized by its own motivational systems in response to its environment, not by an external agent for external purposes.
Capital reorganizes animal activity through the same mechanisms it uses on human labor — and does so because the activity is labor: living, value-producing activity appropriated through ownership rather than purchased through the wage.[43]
Deskilling through selective breeding. The modern Holstein has been bred to be large, placid, and tractable, her capacity for independent action systematically reduced — a raised threshold of provocation, not a diminished need: the docile cow needs no less, she is merely harder to trip into defensive or self-directed behavior. This is the animal analogue of deskilling proper: not the body reshaped for output (that is real subsumption, below) but the capacity for self-directed activity bred down as a condition of exploitability, reducing the organism's autonomy.
Spatial control through confinement. The gestation crate, the battery cage, the feedlot eliminate the environmental complexity required for self-directed activity. The animal cannot range, forage, explore, or maintain its characteristic social structure. Concentrating bodies in space maximizes managerial control over the production process.
Temporal control through hormonal synchronization. Lutalyse induces farrowing on schedule — ensuring sows give birth during the regular working day rather than on their own biological timetable — while PMSG synchronizes estrus across the breeding herd. Light manipulation overrides hens' seasonal reproductive cycles. Biological time is subordinated to industrial time.[44]
Real subsumption of the body. Domestication had already reshaped these bodies over millennia; capitalist real subsumption goes further, redesigning the body itself for maximum output at a depth and scale — and on genetics decoupled from any individual body — that pre-capitalist breeding never reached. The broiler bred to grow so fast its legs cannot bear its weight, and the roughly 300% increase in laying-hen ovulation rate, are not welfare problems within an otherwise acceptable arrangement. They are the arrangement. Deskilling and real subsumption both operate through breeding, but they optimize different things — the one for tractability, the other for yield — and only the second reshapes the body itself.
Confinement at industrial scale destroys social structures and forecloses the strategies through which animals would otherwise manage conflict and maintain social order. The result is pathological aggression: feather-pecking and cannibalism in hens, tail-biting in pigs, horn-goring in cattle. The industry's response is preventive mutilation — debeaking, tail-docking, dehorning — to limit the property damage. The needs that produce the behavior aren't invisible; the welfare science exists. But the property relation ensures they register only instrumentally, weighed against the owner's economic interest.[45] Debeaking is cheaper than reducing stocking density. The industry recognizes the need for protection as a variable in a cost calculation the animal has no standing to contest.
The managed hive
The same mechanisms operate on eusocial insects, where capital has not reshaped the genome but has reorganized the colony's activity and environment to maximize output.[46] Wild honey bee colonies maintain themselves at roughly one per square kilometer, nest in small cavities (~40 liters) high in trees with thick insulating walls, coat their nest interiors with antimicrobial propolis, produce drone comb at ~17% of total area, and swarm annually as their mode of reproduction. Managed colonies are housed at densities orders of magnitude higher (<1 meter apart), in hives three to five times the volume of natural cavities, with thin uninsulated walls, on manufactured comb foundation that suppresses drone production, and are prevented from swarming by management intervention. Wild colonies survive the Varroa destructor mite without treatment; managed colonies die at roughly 40% per year. The difference is not the parasite. It is the conditions of life — the hive-reorganization mechanisms catalogued next, to which industrial forage stress (migratory trucking into pesticide-laden monocrops) adds a further burden.
Each of the four mechanisms is visible. Deskilling operates through environmental design rather than genetic modification: comb foundation dictates what kind of comb the colony builds, swarm prevention blocks the colony's evolved reproductive strategy, and miticide treatment arrests the development of behavioral disease resistance — the grooming, hygienic brood removal, and cell-recapping behaviors through which wild colonies manage mite loads. Spatial control operates through apiary density: in an experiment comparing crowded colonies (<1 meter apart) with dispersed colonies (~34 meters apart), none of twelve crowded colonies survived two years; five of twelve dispersed colonies survived — the mechanisms are precise, as crowding produces drifting (nearly half of drones entering wrong colonies), queen loss during requeening, and horizontal disease transmission, including cascading "mite bombs" when collapsing colonies' parasites are redistributed by robber bees. Temporal control operates through stimulative feeding that forces brood production on the beekeeper's schedule and migratory pollination contracts that override the colony's relationship to local seasonal rhythms. Real subsumption operates far less on the genome than in birds and mammals (open mating and haplodiploidy sharply limit selection, so managed bees remain close to their wild genotype despite queen-breeding programs) and mainly on the colony's structure: its population is inflated by preventing swarming, its comb composition dictated by manufactured foundation, and its nest designed for managerial access rather than colony function.
The pathologies parallel those of confined birds and mammals: mite population explosions the colony cannot control because it cannot swarm or maintain a small population; virus hyper-virulence evolved through the horizontal transmission that crowding facilitates; immune overactivation in the absence of propolis envelopes (smooth lumber walls fail to trigger propolis deposition behavior); and brood developmental damage from temperature fluctuations in thin-walled hives, where even one degree Celsius of deviation from optimal brood-nest temperature produces measurable neural damage. The colony's self-organized behavioral repertoire constitutes the strategies through which its members' needs are met; management forecloses those strategies to drive the colony's labor into honey production — well past what the colony would store for itself — subordinating the colony's own functioning to yield.
The bee case adds a dimension the bird and mammal cases do not: need-satisfaction is both individually and collectively organized — each bee has valenced experience, and the colony's self-organization is itself a strategy for meeting those individual needs that no bee can meet alone. Disrupt the collective strategy and individual needs go unmet.
Resistance as evidence
Animals resist the conditions of their captivity. Zoo escapes, circus attacks, farm animal refusals: these are not pathological outbursts but purposive responses to specific conditions.[47] Ken Allen, an orangutan at the San Diego Zoo, escaped his enclosure repeatedly using plans that required foresight, patience, and tool use. Tilikum, an orca at SeaWorld, attacked trainers after years of overwork and coercive food-deprivation discipline. Farm animals balk, refuse to move, kick handlers, break equipment.
The resistance is best explained by the animal's having something like an orientation toward its own conditions, which is what "having needs" means; the planful, conditions-tracking structure of Ken Allen's escapes fits that reading better than the undirected pathology institutional accounts substitute for it ("aggression," "behavioral disorder," "failure to adapt"). Read against the grain, the pattern is the same as worker resistance: not a failure of adjustment but a response to arrangements that frustrate needs.
What the resistance is not is a set of temperaments. The forms it takes are ordered by what the arrangement leaves open, which is why the repertoire is so consistent across cases that share nothing else: balking, dragging, refusing to move, the task done wrong, the escape, and — where even those are closed — the body's own routes, which is where the stereotypy sits.[48] Two things follow. Institutional accounts read the form as a property of the animal ("aggression," "behavioral disorder") when it is a reading of the openings; and the preventive mutilations described above confirm the reading from the other side, since they operate on the form rather than on what produces it.
One form is absent from the repertoire, and its absence is produced rather than natural. A joint refusal requires an organization that exists before the refusal is needed, and confinement at industrial scale is among other things the destruction of exactly that: herds and flocks assembled from strangers, resorted at intervals, and never permitted the social structure through which a coordinated response could form. Where such an organization is not optional, the point becomes visible directly, which is what the managed hive above showed: swarming is the colony's collective departure, not any individual bee's to take, and preventing it leaves the organization standing while removing what it could do with itself.
The "tie judgement" and its limits
Non-exploitative animal labor is possible in small-scale farming, or so the counter-argument runs. The "tie judgement" names the genuine recognition that develops between a farmer and specific animals in long-term working relationships.[49] The Dejours preface to her edited volume Animal Labor frames domestication through a revealing question: "Everything that animals put into their living work with humans is unnatural. Where does this animal will to live with humans come from?" The question frames domestication as a mystery of voluntary relationship rather than a historical process of capture, confinement, and selective breeding. That such a question can be asked at all is itself a product of the arrangement rather than an observation about it. Where no threat is made and none is needed, the compliance that follows is available to be read as agreement — and in a domain that admits no refusal, all compliance is of that kind, so there is nothing for the reading to be checked against.[50]
The relationships Porcher describes are not unreal. But they operate within arrangements structured by property relations the animal cannot exit or renegotiate. The farmer may recognize the cow, and the cow genuinely trusts the farmer — cooperates in handling, settles into routine, permits care. But the trust is honored conditionally: maintained while the cow's productivity warrants her upkeep, withdrawn when it does not. The cow is still property. Under class conditions — where the animal's body is owned, its reproduction controlled, and its life terminable when productivity declines — no animal labor relationship is non-exploitative. Not because the cow cannot refuse: she can, and the balking and kicking of the previous section are refusals. But behavioral refusal is not the structural capacity to renegotiate or exit the terms of ownership, and the property relation forecloses both. It is worth being exact about which does the work, because framing the loss as an inability to exit understates it and misdescribes it. A refusal that can be given only by leaving was never a refusal within the relation to begin with; what voids the cow's balking is not that she cannot leave afterwards but that nothing in the arrangement is obliged to receive it. It arrives, and is re-filed as a handling problem.[51] Yet contest-capacity is not itself what makes a relationship non-exploitative — the wage relation stays exploitative even where workers can contest it, and its absence is not what makes this one exploitative either. What makes it exploitative is that the arrangement overrides needs. The property relation overrides the need for protection directly: the body is owned, confined, and terminable at the owner's discretion — and, as the care-ethics discussion below develops, bodily integrity is among the strategies for meeting that need, so an arrangement that can set it aside at will frustrates the need however attentive the care within it.
Porcher's own cases demonstrate this. The military and police dog chapters show animals exercising initiative, intelligence, and judgment within arrangements they cannot renegotiate or exit. The guide dog exercises "initiative-spirit" in navigating traffic, but cannot refuse the work, choose a different handler, or determine when its career ends. Relationship quality and structural exploitation are compatible. A factory worker can have a genuine relationship with a supervisor; the wage relation is still exploitative.
This is not a condemnation of individual farmers. It is an evaluation of the arrangement, the same operation my framework runs on every other domain of class-structured social life.
VI. What the existing frameworks miss
Several cross-species political theories share my needs framework's orientation toward evaluating arrangements rather than calculating welfare. None of them has a class analysis.
The capabilities approach
The capabilities approach extends species-specific capabilities lists to animals, shares the orientation toward evaluating arrangements by their effects on what beings need to flourish, and rejects the requirement that only participants in the social contract count as subjects of justice.[52] Three structural differences sharpen in the cross-species context.
Grounding. The capabilities lists are grounded in Aristotelian wonder and overlapping consensus. My framework grounds its needs profiles in biological architecture: the neurobiological systems whose presence and frustration are empirically identifiable. For human capabilities the difference is easy to mistake for a merely philosophical one, since those whose needs a consensus misdescribes can organize to contest it; for animal capabilities it becomes unmistakably practical, because those subject to the list cannot. Capabilities lists assembled case by case through consensus among informed citizens will reflect what those citizens (all of whom benefit from animal exploitation) consider important. The generative principle provides a criterion that does not depend on the evaluators' interests.
The deformed species-being problem. The species norm is evaluative: it identifies which innate capabilities are good and central to a flourishing and dignified life for a creature of that kind. But for domesticated animals, the "innate capabilities" include capabilities engineered by class power. The Holstein's capacity to produce enormous quantities of milk, the broiler's capacity to grow at lethal speed: capabilities that are simultaneously the mechanism of profit and the mechanism of disability. A theory that cannot distinguish capabilities grounded in the species' shared biological architecture from capabilities produced by class-structured breeding cannot evaluate what industrial agriculture does to the animals it has reshaped. The treatment developed in Section IV (evaluate against the needs that persist, evidenced by the pathology their frustration produces) addresses this.
The structural gap. The capabilities approach treats the failure to protect animal capabilities as a failure of political will and institutional design, solvable through constitutional inclusion, international accords, and shifts in public attitude. The question it doesn't ask is why the failure persists: the accumulation process that profits from the systematic frustration of animal needs, the class interests that maintain current arrangements, and the institutional architecture that locks them in. Without this structural analysis, the prescriptions (better laws, more humane treatment) address symptoms rather than causes.
Two questions require engagement. On predation: "the gradual supplanting of the natural by the just" does not follow from the argument developed here. Predation is not an arrangement structured by class power; it is not evaluable by an operation that runs on arrangements. A wolf eating a deer is not such an arrangement. A factory farm is.[53] On paternalism: "species-sensitive paternalism" is compatible as a transitional principle, but under class conditions, "paternalism" toward animals is property management, not benevolent guardianship. The question is not how to be better paternalists but how to dismantle the arrangements that make paternalism the only available relationship.
Political membership across species
Donaldson and Kymlicka propose a differentiated typology: domesticated animals as citizens, wild animals as sovereign communities, liminal animals (urban wildlife, feral populations) as denizens.[54] Two contributions my framework can use. First, dependent agency: citizenship and political participation do not require autonomous rational choice. Animals consent, refuse, signal preferences, and resist, not through propositions but through behavior. The task of political institutions is to read and respond to these communications. Second, the critique of extinctionism. Donaldson and Kymlicka argue that the abolitionist aim of phasing domesticated species out of existence by ending all breeding treats their disappearance as a good in itself, discounting the interests of the animals who would otherwise live. (The abolitionist position is not that existing animals should be harmed, but that those who control domesticated animals should frustrate their reproduction.) Whatever the merits of that dispute, what I can take from it is narrower: domesticated animals have genuine interests shaped by their domestication history, including interests in human social contact, and those interests are real regardless of how the species came to have them. The question is not whether domesticated animals should exist but under what arrangements.
What this framework lacks is political economy. The analysis operates entirely within liberal political theory. It treats the transition away from animal exploitation as a problem of political will, institutional design, and cultural change, not as a problem of capital accumulation, corporate power, or class interest. Animal exploitation is not a failure of recognition but a class relation, generated by the accumulation process, maintained by property relations, and locked in by the same institutional ratchet that locks in other forms of class rule. The prescriptions (citizenship for domesticated animals, sovereignty for wild animals) describe what just arrangements would look like but cannot explain why current arrangements exist or what structural forces would resist the transition.
The permissive stance on continued use of animal products (eggs, milk, and wool are acceptable if the animal lives a flourishing life as a citizen) inherits a deeper problem. The framework for distinguishing "use" from "exploitation" turns on whether the arrangement looks like cooperative membership to the humans who designed it, not on whether the animal's contestation is registered as anything more than the humans interpret it to be, or whether its bodily integrity is addressed or overridden. Who defined "contribution"? Who determined that the animal's reproductive output constitutes its share of cooperative life? Can the animal refuse the terms? No one poses these questions, because the humans who designed the arrangement evaluate it from their own perspective. The last of them has an answer, and it is not the one the framework's vocabulary would predict. What a refusal lacks here is not a capacity in the animal — the balking, kicking, and escaping of Section V are refusals given through channels that work — but a party obliged to receive one.[55] Citizenship without that is a status conferred on a being whose no still has nowhere to be lodged, which is a description of the arrangement the framework was trying to distinguish itself from. The care ethics discussion below develops the rest of this point.
Sentientist cosmopolitanism
A cosmopolitan framework rejects the relational typology as "morally arbitrary": a wild animal's moral entitlements should not diminish because it was born in the wild rather than bred in captivity.[56] Moral worth does not vary by relational position, but the incorporation gradient is analytically descriptive, not normatively justificatory. The gradient explains the form exploitation takes (total reproductive appropriation at the extreme versus wage exploitation at the integrated end), not the worth of the beings at different positions. Conflating diagnosis with prescription produces the objection.
An inconsistency runs through the position: the marginal-cases argument targets claims of higher moral worth for humans (infants and cognitively disabled humans cannot frame conceptions of the good either) but is not applied to the liberty question. Animals are denied an intrinsic interest in liberty on the grounds that they cannot "frame, revise, and pursue their own conceptions of the good." The rationalist autonomy standard that does this work has historically been used to justify forced labor for humans deemed insufficiently rational. The meritocratic reasoning is visible in the philosophical debate itself.
Where the positions converge: the insistence on positive duties to wild animals, duties of assistance rather than just non-interference. If wild animal habitats are structured by accumulation (and they are: enclosure, resource extraction, habitat destruction), then the evaluative operation applies to those arrangements too.
Care ethics and the consent gap
The feminist care tradition in animal ethics is the approach most likely to be confused with my framework's evaluative operation, because it emphasizes attentiveness and responsiveness to animal needs.[57] The tradition's critique of rights-based and utilitarian approaches was substantially correct: these frameworks abstract from concrete relationships and material conditions, presuppose a rationalist ontology that already excludes most of the beings at issue, and reduce moral life to a calculus that misses what actually happens between humans and animals.
Three structural problems differentiate my framework.
Centering. Care ethics evaluates caring relationships from the caregiver's phenomenological standpoint. Even "entangled empathy" operates through the human's empathic perception of the animal's needs: the animal cannot set the terms of the relationship, articulate its own interpretation, or contest the caregiver's assessment. The power to determine what the animal needs remains with the human. For human caring relationships this is a recognized difficulty; for cross-species relationships where the care-receiver cannot speak, it is structurally irresolvable within the tradition's own resources. What lets my framework press this critique without turning it on itself is not that it escapes human interpretation — its needs-reading is a human reading too — but that the reading answers to an objective criterion: frustration registers as measurable pathology independent of the evaluator's standpoint, where care ethics has only the caregiver's empathic perception to go on.
Consent and bodily integrity. Because the tradition defined itself against rights-based moral theory, it has no internal apparatus for theorizing consent violations or bodily-integrity claims. The care-receiver's refusal carries no independent theoretical weight distinct from the caregiver's assessment of what is required. Feminist critics identified this for human care relationships in the 1990s: the framework can legitimate coercion by powerful caregivers (Card); the caring ideal reproduces the expectation that boundaries should yield to the caregiver's judgment (Hoagland). Held and Tronto eventually conceded the point by arguing that care and justice are "complementary," importing the autonomy and rights concepts the tradition had defined itself against, at the point where they are indispensable.
For animals the gap is fatal. The entire apparatus of industrial animal agriculture describes itself in the vocabulary of care: "husbandry," "welfare," "stewardship." A dairy farmer who is genuinely attentive and responsive to her cows' needs within the constraints of a dairy operation can satisfy care-ethical criteria while the arrangement itself (forced insemination, calf removal, slaughter upon reproductive exhaustion) goes unevaluated. The tradition evaluates the quality of the caring practice rather than asking whether the arrangement addresses or overrides the animal's bodily integrity. Applied cross-species, the consent gap becomes the mechanism through which exploitation operates.
The gap is also specifiable rather than merely named, which matters because a gap one can locate is a gap one can say what would close.[58] A refusal functions as one only where several separable conditions hold — the animal has to produce a signal, have some form the refusal can take, and meet someone obliged to register it. Husbandry withholds the last of these structurally, and the care relation is what stands where it would go: attentiveness is supplied at precisely the point an obligation to be bound would be required, and attentiveness is what the caregiver decides to extend. The more attentive the practice, the more completely the substitution is concealed — which is why the best farmers are the hardest case for the tradition and the easiest for this one.
Cooperative membership as redescribed appropriation. The "cooperative membership" framework for distinguishing legitimate use from exploitation, and Porcher's "tie judgement" discussed above, operate on the same logic: the quality of the relationship within the arrangement substitutes for evaluation of the arrangement itself. Care ethics, cooperative membership, and the tie judgement are often treated as three distinct positions. They differ in normative valence — the tie judgement endorses animal labor, the care tradition tends abolitionist, cooperative membership is a liberal-citizenship construct — but they share a single evaluative structure under three vocabularies: relationship-quality within the arrangement stands in for evaluation of the arrangement.
Protection is a species-typical need; conditions under which the organism's body is not subject to systematic violation are among the strategies for meeting that need. An arrangement that systematically overrides these conditions frustrates the need for protection regardless of how attentive, responsive, or relationally rich the care practices within it are. My framework can make this evaluation from its own resources. Care ethics cannot, which is why Held, Tronto, and Gruen have each had to import structural analysis, justice theory, or critical theory from outside to address what the tradition's own apparatus leaves unnamed.
The shared gap
None of these frameworks provides a class analysis of why animal needs go systematically unmet. The capabilities approach locates the problem in insufficient political will. The political membership framework locates it in insufficient institutional design. Sentientist cosmopolitanism locates it in insufficient democratic representation. Care ethics locates it in insufficient attentiveness. In the companion metaethics essay I identify what they share at a deeper level. Each is arrangement-oriented relative to welfarism, yet each still routes its verdict through a status it assigns to the being — adequate-capabilities, member, worth-bearer, care-worthy — so that the being's status, not the arrangement, is where evaluation comes to rest: what I call, in that essay, the predicative structure.[59] My framework attributes a profile to beings too — needs are features of the kind of animal — but that profile is only the standard against which it evaluates the arrangement, and the standard is grounded in biological architecture rather than assigned by consensus; the object of evaluation remains the arrangement. My framework locates the problem in class power's appropriation of animal life — under capitalism, in the accumulation process specifically: the systematic frustration of animal needs is profitable, and the arrangements that produce it are maintained by class power. This is not a rival answer to the same question. It is a different question. The others ask what just arrangements should look like — and, as the preceding sections show, frequently take in only part of the arrangement at that. Mine also asks: why do unjust arrangements persist, and what structural forces would resist their transformation? Without this question, cross-species political theory describes a destination without analyzing the terrain between here and there.
The companion essay "Matter Out of Place" adds a further dimension: the political will, institutional design, and attentiveness these frameworks call for cannot be generated under class conditions because the property form's taboo-structure maintains the species boundary through affect rather than argument. The disgust that forecloses needs-perception across the species boundary operates through the same mechanism that forecloses it across human gradient positions: the avoidance body policing its own animality from inside (shame), policing others' visible animality from outside (disgust). Alongside it, contempt recodes the animal as beneath the attention its needs would command. Discrediting the justifications for speciesism no more dissolves this affective orientation than discrediting the justifications for racism dissolves the disgust at gradient-bottom human populations. The terrain between here and there is not principally a deficit of better arguments but a taboo-structure that operates below the level argument can reach. But that structure is installed rather than fixed — laid down late and unevenly, shallow in some people and toward some animals (a dog's needs registered, a pig's refused), deeper in others — and "Matter Out of Place" develops how the relational rather than instrumental proximity that de-installs it reaches a latent openness argument alone cannot.
VII. What follows
The sow bar-biting at the bars of her crate opened this essay because her behavior is not a welfare failure to be tuned but the visible edge of an arrangement: food, water, and temperature control provided while the strategies through which her needs would be met are foreclosed. What such an arrangement destroys is not registered by any metric it pushes below threshold. It is what sentience is for — the free exercise of species-typical capacities as an end in itself — and the isolation cell does the same thing to the human it holds. Any framework grounding moral evaluation in species-typical needs must extend that evaluation across species. The biological architecture that produces those needs is not species-specific; the stereotypy evidence shows that needs persist through whatever class power does to the organisms it exploits; and the sentience threshold determines where the evaluative operation applies. What the existing cross-species political theories lack is not a vision of just arrangements but an account of why unjust arrangements persist — the structural forces that maintain them and the class interests they serve.
That account is not developed here; it is what two companion essays, standing on this essay's grounding, supply. "Where the Logic Is Most Complete" traces the ideological form: speciesism as meritocratic reasoning applied most completely, where the circular attribution of subordinate position to the subordinate's own deficiency operates at maximum depth because the population subject to it has no capacity to organize the counter-pressure that forces exploitation to take a different form in human cases. The suffering is not unperceived — the need-recognition mechanism fires cross-species, as the evidence from children and slaughterhouse workers confirms — but taking it seriously is suppressed through the shame mechanism applied to moral concern itself: the avoidance body's requirements coding empathic responsiveness at the species boundary as sentimental rather than as moral perception. The conditioned-concern mechanism that maintains the species boundary operates through the same process the collection documents for race, class, and gender: rigid categorical sorting replacing flexible, interaction-based empathic responsiveness. "Bred Into Existence" traces the material mechanism: owners' appropriation of animal productive and reproductive labor at the extreme end of the incorporation gradient, where the counter-pressure the subjugated can organize is weakest and the restructuring of the body for production is most complete. This essay's evaluative account, and the collection's others, are layers over that mechanism.
A politics that grounds its critique in the systematic frustration of needs but restricts its scope to one species has not understood its own logic. My framework does not prescribe a specific political program. But it insists that the question be posed, and that the answer be structural rather than a matter of mere sentiment.
Bibliography
Andrews, Kristin. The Animal Mind: An Introduction to the Philosophy of Animal Cognition. London: Routledge, 2015.
Balcombe, Jonathan. Pleasurable Kingdom: Animals and the Nature of Feeling Good. New York: Palgrave Macmillan, 2006.
Baranzke, H., and H.-W. Ingensiep. "Sentientism – For Whose Sake? Ethics, Sciences, and Crypto-Teleological Fact–Value Bridges, Illustrated by the Research on Sentience in Invertebrates." Animal 17 (2023): 100875.
Bekoff, Marc, and Jessica Pierce. Wild Justice: The Moral Lives of Animals. Chicago: University of Chicago Press, 2009.
Beniwal, Vikas. Foundations and Scope of Moral Standing. MA thesis, University of Mississippi, 2026.
Birch, Jonathan. The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI. Oxford: Oxford University Press, 2024.
Birch, Jonathan, Charlotte Burn, Alexandra K. Schnell, Andrew Crump, and Heather Browning. "Review of the Evidence of Sentience in Cephalopod Molluscs and Decapod Crustaceans." Commissioned by the UK Department for Environment, Food and Rural Affairs. LSE Consulting, 2021.
Bradshaw, G. A., Allan N. Schore, Janine L. Brown, Joyce H. Poole, and Cynthia J. Moss. "Elephant Breakdown." Nature 433 (2005): 807.
Card, Claudia. "Caring and Evil." Hypatia 5 (1990).
Chittka, Lars. The Mind of a Bee. Princeton: Princeton University Press, 2022.
Cochrane, Alasdair. Sentientist Politics: A Theory of Global Inter-Species Justice. Oxford: Oxford University Press, 2018.
Davion, Victoria. "Autonomy, Integrity, and Care." Social Theory and Practice 19 (1993).
de Waal, Frans. Mama's Last Hug: Animal Emotions and What They Tell Us about Ourselves. New York: Norton, 2019.
Denton, Derek A. The Primordial Emotions: The Dawning of Consciousness. Oxford: Oxford University Press, 2005.
Dona, Hiruni Samadi Galpayage, et al. "Do Bumble Bees Play?" Animal Behaviour 194 (2022): 239–251.
Donaldson, Sue, and Will Kymlicka. Zoopolis: A Political Theory of Animal Rights. Oxford: Oxford University Press, 2011.
Donovan, Josephine, and Carol J. Adams, eds. The Feminist Care Tradition in Animal Ethics. New York: Columbia University Press, 2007.
Francione, Gary L. Introduction to Animal Rights: Your Child or the Dog? Philadelphia: Temple University Press, 2000.
Godfrey-Smith, Peter. Other Minds: The Octopus, the Sea, and the Deep Origins of Consciousness. New York: Farrar, Straus and Giroux, 2016.
Goodpaster, Kenneth E. "On Being Morally Considerable." The Journal of Philosophy 75, no. 6 (1978): 308–325.
Graeber, David. "What's the Point If We Can't Have Fun?" The Baffler 24 (2014). Reprinted in The Ultimate Hidden Truth of the World (2024), 307ff.
Gruen, Lori. Entangled Empathy: An Alternative Ethic for Our Relationships with Animals. New York: Lantern Books, 2015.
Gruen, Lori, and Alice Crary. Animal Crisis: A New Critical Theory. Cambridge: Polity, 2022.
Hoagland, Sarah Lucia. "Some Thoughts about 'Caring.'" In Feminist Ethics, edited by Claudia Card. Lawrence: University Press of Kansas, 1991.
Heinrich, Bernd. Mind of the Raven: Investigations and Adventures with Wolf-Birds. New York: Cliff Street Books, 1999.
Hribal, Jason. Fear of the Animal Planet: The Hidden History of Animal Resistance. Oakland: AK Press, 2010.
Kropotkin, Peter. Mutual Aid: A Factor of Evolution. London: Heinemann, 1902.
Malm, Andreas. Fossil Capital: The Rise of Steam Power and the Roots of Global Warming. London: Verso, 2016.
Mason, Georgia J. "Stereotypies: A Critical Review." Animal Behaviour 41 (1991): 1015–1037.
McMahan, Jeff, and Peter Singer. "Who Is the Victim in the Anna Stubblefield Case?" The New York Times, "The Stone," April 3, 2017.
Mellor, David. "Updating Animal Welfare Thinking: Moving beyond the 'Five Freedoms' towards 'A Life Worth Living.'" Animals 6, no. 3 (2016): 21.
Narayanan, Yamini. "An Ecofeminist Politics of Chicken Ovulation: A Socio-Capitalist Model of Ability as Farmed Animal Impairment." Hypatia 39 (2024): 568–588.
Nussbaum, Martha. Frontiers of Justice: Disability, Nationality, Species Membership. Cambridge, MA: Harvard University Press, 2006.
Porcher, Jocelyne, ed. Animal Labor: A New Perspective on Human-Animal Relations. Bielefeld: Transcript, 2017.
Regan, Tom. The Case for Animal Rights. Berkeley: University of California Press, 1983.
Reimert, I., L. E. Webb, M. A. van Marwijk, and J. E. Bolhuis. "Towards an Integrated Concept of Animal Welfare." Animal 17 (2023): 100838.
Robbins, J., B. Franks, and M. A. G. von Keyserlingk. "'More than a Feeling': An Empirical Investigation of Hedonistic Accounts of Animal Welfare." PLoS ONE 13, no. 3 (2018): e0193864.
Rosenberg, Marshall B. Nonviolent Communication: A Language of Life. 3rd ed. Encinitas, CA: PuddleDancer Press, 2015.
Scott, James C. Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven: Yale University Press, 1985.
Seeley, Thomas D. The Lives of Bees: The Untold Story of the Honey Bee in the Wild. Princeton: Princeton University Press, 2019.
Singer, Peter. Practical Ethics. 3rd ed. Cambridge: Cambridge University Press, 2011.
Taylor, Paul W. Respect for Nature: A Theory of Environmental Ethics. Princeton: Princeton University Press, 1986.
Taylor, Sunaura. Beasts of Burden: Animal and Disability Liberation. New York: The New Press, 2017.
Tronto, Joan. Moral Boundaries: A Political Argument for an Ethic of Care. New York: Routledge, 1993.
- ↩
Georgia J. Mason, "Stereotypies: A Critical Review," Animal Behaviour 41 (1991): 1015-1037. All stereotypy evidence in this section draws on Mason's synthesis unless otherwise noted.
- ↩
The figure is a stock held in production — something on the order of ten billion land animals alive in ongoing production relationships at any given moment, laying hens the bulk of it, with dairy cattle and breeding sows in the hundreds of millions each — not to be confused with the far larger annual flow of animals bred and killed, or with the sea animals killed outright, whose numbers run into the trillions. The companion essay "Living and Nonliving" sets out that accounting and the distinction it turns on. Yamini Narayanan's description of the egg industry names the scale at which the reorganization operates: breeding for "eggs of a standardized size, weight, and regularity across billions of ovulating bodies" ("An Ecofeminist Politics of Chicken Ovulation: A Socio-Capitalist Model of Ability as Farmed Animal Impairment," Hypatia 39, 2024: 568–588).
- ↩
See "Merit All the Way Down: Toward a Materialist Metaethics" at beyondmeritocracy.info for the full development of my evaluative framework. This essay states its core claims concisely but does not re-derive them.
- ↩
Marshall Rosenberg, Nonviolent Communication: A Language of Life, 3rd ed. (Encinitas, CA: PuddleDancer Press, 2015). Rosenberg uses "strategies" where Max-Neef uses "satisfiers"; the distinction is the same. "Strategies" is preferable here because it foregrounds what the organism is doing (attempting to meet a need) rather than what the arrangement is providing (a satisfier). The cross-species case makes the advantage of this framing visible: a stereotypy is a strategy the animal's body deploys when the arrangement forecloses all others. See also "The Radical Core and Its Disappearance" at beyondmeritocracy.info, which traces how Rosenberg's structural diagnosis — that evaluative language suppresses needs-perception — was depoliticized through class-conditioned reception. The disciplinary-suppression argument in section II.C above is the cross-species analogue: neo-Darwinist instrumentalism prevents ethology from seeing play for the same structural reason that evaluative language prevents individuals from seeing needs.
- ↩
The distinction between the pre-evaluative register (responsiveness to need without categorization) and the innocence perception (evaluation suspended because the being is classified as not responsible) is developed in "The Conditions of Care," section 4. The two produce similar-looking care — tender, responsive, non-judgmental — but with different structural properties. The pre-evaluative register generalizes: it responds to the rat's needs too, and the spider's, and the animal that just bit you. The innocence perception preserves the evaluative apparatus: it extends care to beings classified as unable-to-be-responsible, and revokes it when the classification breaks. The ableism/speciesism connection Taylor develops is sharpened by this distinction: both operate through the innocence perception, extending care conditionally while preserving the sorting mechanism that determines who qualifies.
- ↩
The two-variable reading is developed in the companion essay on the capacity to refuse ("Standing to Refuse," in draft), which tracks innocence — the degree to which a being is classified as not responsible, licensing care and suspending evaluation — against standing, the degree to which its refusal obliges anyone, and finds them moving independently and often in opposite directions. Its point about innocence is the one that bears here: to be covered by can't help it is to have one's refusal received as a symptom, so the classification buys care at the price of agency. The farmed animal is the corner that essay's developmental curve does not reach, because there the classification never arrives at all. The rhyme is with the structure, not the content: nothing in the animal case corresponds to the interior conditions that essay is mainly about, and it places the case accordingly.
- ↩
The leopard is "Legadema," filmed by Dereck and Beverly Joubert in the Okavango Delta, Botswana, in Eye of the Leopard (National Geographic, 2006): having killed a female baboon, she carried its day-old infant into a tree and sheltered it through the night, by which point it had died of cold. The lioness is "Kamunyak" (Maa: "blessed one"), who adopted a series of oryx calves in Samburu National Reserve, Kenya, in 2001–2003, recorded by Saba Douglas-Hamilton in Heart of a Lioness (BBC, 2004); she ceased hunting during the episodes and nearly starved — which is why she illustrates the predatory strategy being overridden rather than merely suspended under satiety. [Editorial note: verify both against the primary footage/accounts before publication.]
- ↩
See "Why Competing Solidarities Tend to Win," footnote on derivation. Three of the thirteen dimensions — order/harmony, efficacy/power-to, and mourning/celebration — were added on the basis of the cross-species evidence (elephants, baboons, bonobos, corvids) this essay develops. The derivation thus runs in both directions: the human-focused analysis provided the taxonomy, and the cross-species evidence grounded the additions. This is mutual adjustment — reflective equilibrium — rather than circularity: each source constrains the other, and neither is taken as foundational.
- ↩
Kristin Andrews, The Animal Mind: An Introduction to the Philosophy of Animal Cognition (London: Routledge, 2015). Andrews discusses Cameron Buckner's concept of "anthropofabulation" — the error of overintellectualizing human cognition, confabulating the sophistication of our own typical abilities, which then makes animal cognition appear deficient by comparison.
- ↩
Kristin Andrews, The Animal Mind: An Introduction to the Philosophy of Animal Cognition (London: Routledge, 2015). Andrews discusses Cameron Buckner's concept of "anthropofabulation" — the error of overintellectualizing human cognition, confabulating the sophistication of our own typical abilities, which then makes animal cognition appear deficient by comparison.
- ↩
David Graeber, "What's the Point If We Can't Have Fun?," The Baffler 24 (2014), reprinted in The Ultimate Hidden Truth of the World (2024), 307ff. Graeber recovers the aspect of Kropotkin's argument that the neo-Darwinist tradition dropped: not cooperation-as-strategy but cooperation-as-pleasure — the exercise of social capacities as an end in itself.
- ↩
David Mellor, "Updating Animal Welfare Thinking: Moving beyond the 'Five Freedoms' towards 'A Life Worth Living,'" Animals 6, no. 3 (2016): 21.
- ↩
The claim is not that play drops out of the needs taxonomy. Play operates at two levels: as a specific dimension whose frustration produces identifiable pathology, and as the experiential register in which needs-fulfillment across all dimensions is experienced as its own end. Panksepp identified PLAY as a neurobiologically distinct affective system with its own subcortical circuitry — evidence for the dimension. But play deprivation in young rats degrades social competence, emotional regulation, and cognitive flexibility — effects that radiate across multiple need-dimensions, which is evidence for the register. The dual status has a testable signature: adding play opportunities should produce disproportionate improvements in overall wellbeing, larger than adding a single dimension would predict, because restoring play also restores the register quality across other dimensions. The environmental enrichment literature is consistent with this pattern: providing exploration and play opportunities to captive animals reduces stereotypies, improves immune function, alters social behavior, and reduces aggression — effects across subsistence, protection, and connection from what the standard list treats as one dimension. But enrichment bundles several dimensions at once (space, foraging, novelty, social opportunity), so it cannot by itself isolate play-as-register from the plainer hypothesis that several dimensions were added together; the testable signature remains a genuine prediction awaiting a study that isolates play from the dimensions co-provided with it (adding only play, or depriving only play while holding connection constant). Burghardt's operational criteria for play include the condition that play is initiated under relaxed conditions — when other needs are adequately met — which is itself evidence of the register function: play is what needs-fulfillment looks like when the organism has room for free exercise. See Jaak Panksepp, Affective Neuroscience: The Foundations of Human and Animal Emotions (New York: Oxford University Press, 1998); Gordon Burghardt, The Genesis of Animal Play: Testing the Limits (Cambridge, MA: MIT Press, 2005).
- ↩
Frans de Waal, Mama's Last Hug: Animal Emotions and What They Tell Us about Ourselves (New York: Norton, 2019).
- ↩
Thom van Dooren, Flight Ways: Life and Loss at the Edge of Extinction (New York: Columbia University Press, 2014). Chapter 5 documents corvid grief, consolation behavior, and the loss of vocal repertoire in captive Hawaiian Crows ('Alalā).
- ↩
Lars Chittka, The Mind of a Bee (Princeton: Princeton University Press, 2022). The pessimistic and optimistic cognitive bias experiments are in chapter 11; the anxiety-like states following robotic spider attacks in chapter 11 (building on Ings and Chittka's earlier experimental work); pain modulation by alarm pheromone drawing on Nunez et al.; cross-modal object recognition and metacognition in chapter 11. Chittka's summary: "If we apply the same behavioral and cognitive criteria as we do to much larger-brained vertebrates, then bees qualify as conscious agents with no less certainty than dogs or cats."
- ↩
Marc Bekoff and Jessica Pierce, Wild Justice: The Moral Lives of Animals (Chicago: University of Chicago Press, 2009). The rat and rhesus monkey evidence is summarized in their empathy chapter.
- ↩
Bernd Heinrich, Mind of the Raven: Investigations and Adventures with Wolf-Birds (New York: Cliff Street Books, 1999). The murder and dead-conspecific evidence is in chapter 23; the play evidence in chapter 24; the bathing experiments in chapter 25; the string-pulling experiments in chapter 26; the wolf-raven relationship in chapters 19–20. Heinrich's long-term behavioral study of hand-reared and wild-caught Common Ravens (Corvus corax) documents individual personality, social norms, planning, insight, and emotional states in a species whose cognitive sophistication rivals that of great apes despite having diverged from the mammalian lineage roughly 320 million years ago. The Yellowstone wolf-raven data draws on Dan Stahler's fieldwork, reported in Heinrich's chapters 19–20.
- ↩
G. A. Bradshaw et al., "Elephant Breakdown," Nature 433 (2005): 807, discussed in Bekoff and Pierce, Wild Justice.
- ↩
Jonathan Balcombe, Pleasurable Kingdom: Animals and the Nature of Feeling Good (New York: Palgrave Macmillan, 2006).
- ↩
Peter Kropotkin, Mutual Aid: A Factor of Evolution (London: Heinemann, 1902), discussed in Graeber, "What's the Point If We Can't Have Fun?"
- ↩
Bernd Heinrich, Mind of the Raven: Investigations and Adventures with Wolf-Birds (New York: Cliff Street Books, 1999). The murder and dead-conspecific evidence is in chapter 23; the play evidence in chapter 24; the bathing experiments in chapter 25; the string-pulling experiments in chapter 26; the wolf-raven relationship in chapters 19–20. Heinrich's long-term behavioral study of hand-reared and wild-caught Common Ravens (Corvus corax) documents individual personality, social norms, planning, insight, and emotional states in a species whose cognitive sophistication rivals that of great apes despite having diverged from the mammalian lineage roughly 320 million years ago. The Yellowstone wolf-raven data draws on Dan Stahler's fieldwork, reported in Heinrich's chapters 19–20.
- ↩
Hiruni Samadi Galpayage Dona et al., "Do Bumble Bees Play?," Animal Behaviour 194 (2022): 239–251. The study demonstrated that bumble bees (Bombus terrestris) rolled small wooden balls repeatedly with no food reward, that younger bees rolled more frequently than older ones, and that the behavior satisfied established operational criteria for animal play.
- ↩
David Mellor, "Updating Animal Welfare Thinking: Moving beyond the 'Five Freedoms' towards 'A Life Worth Living,'" Animals 6, no. 3 (2016): 21.
- ↩
David Graeber, "What's the Point If We Can't Have Fun?," The Baffler 24 (2014), reprinted in The Ultimate Hidden Truth of the World (2024), 307ff. Graeber recovers the aspect of Kropotkin's argument that the neo-Darwinist tradition dropped: not cooperation-as-strategy but cooperation-as-pleasure — the exercise of social capacities as an end in itself.
- ↩
"Standing to Refuse" (in draft) names this the warrant condition: whether one's own state is admissible as a complete reason, or has to be converted into something auditable before it can be spoken. Its finding is that most adults can decline when they can justify and cannot decline on the bare ground that they would rather not, which is refusal under adjudication rather than a recovered capacity; and it takes the reclassification of self-directed activity — into hobby, self-care, side hustle — from "For Its Own Sake" as the same proceeding running on what a person does rather than on what they decline. The convergence claimed here is of structure and not of mechanism: nothing suggests that the disciplinary norm and the internal proceeding have a common installation history, and the disciplinary case has an interest behind it that the personal one need not.
- ↩
Bernd Heinrich, Mind of the Raven: Investigations and Adventures with Wolf-Birds (New York: Cliff Street Books, 1999). The murder and dead-conspecific evidence is in chapter 23; the play evidence in chapter 24; the bathing experiments in chapter 25; the string-pulling experiments in chapter 26; the wolf-raven relationship in chapters 19–20. Heinrich's long-term behavioral study of hand-reared and wild-caught Common Ravens (Corvus corax) documents individual personality, social norms, planning, insight, and emotional states in a species whose cognitive sophistication rivals that of great apes despite having diverged from the mammalian lineage roughly 320 million years ago. The Yellowstone wolf-raven data draws on Dan Stahler's fieldwork, reported in Heinrich's chapters 19–20.
- ↩
Juliana Schroeder and Nicholas Epley, "Demeaning: Dehumanizing Others by Minimizing the Importance of Their Psychological Needs," Journal of Personality and Social Psychology 119, no. 4 (2020): 765–791. The effect held across multiple studies and target groups, including homeless people, refugees, and low-income populations. The mechanism is not that people deny these groups have minds; it is that they underestimate how much psychological needs matter to them relative to physical ones — a subtler form of dehumanization that operates through the weight assigned to needs rather than through their denial. For the broader argument that power reduces perspective-taking motivation, producing the positional blindness through which subordinate beings' needs are systematically underestimated, see Maibom, The Space Between (2022), ch. 8, drawing on Galinsky et al.'s experimental findings that people in positions of power show reduced empathic distress, less motor resonance, and more egocentric perspective-taking — a motivational rather than a capacity deficit.
- ↩
This is the claim biocentric individualism denies, and its denial is the mirror image of Singer's. Where Singer draws the line too high — gating full standing on the cognitive capacities of "persons" — the biocentric tradition denies there is a line at sentience at all: a living thing has a good of its own, and interests that can be set back, whether or not anything is ever felt. Kenneth Goodpaster argued that being alive rather than being sentient is the criterion of moral considerability ("On Being Morally Considerable," 1978); Paul Taylor grounded standing in the organism as a "teleological center of life"; the root intuition is the Spinozist conatus, the striving by which each thing "endeavours to persist in its own being." Vikas Beniwal's Foundations and Scope of Moral Standing (MA thesis, University of Mississippi, 2026) restates it in the terms that bear on this section: sentience is sufficient for standing but not necessary, and what is also sufficient is the capacity to strive — the "non-derivative functional activities intrinsic to living beings such as self-maintenance and growth," which he holds ground interests "that are good or bad for it, even if unfelt." Two objections follow, and this section can meet both. The first is his scalar objection: if sentience comes in degrees, a threshold drawn across it is unprincipled. That is answered above — standing does not come in degrees even where sentience does, because the threshold marks the onset of a bearer, not a quantity of experience, and the uncertain margin is what the precautionary line screens. The second — that gating standing on valence is arbitrary once the grounding is naturalistic (the consistency pressure section I turns against anthropocentrism, aimed one ring further out at the sentience line itself) — rests on a misdescription of the ground. It assumes my framework grounds standing in functioning and then adds valence as a further condition, so that dropping the condition would be the more consistent move. But functioning, development, and the exercise of species-typical capacities are the content of what a subject needs; they were never the ground of its standing. The ground is valenced experience from the start — there is no valence-free functioning layer in my framework for the argument to extend from. Striving establishes a telos, a direction relative to which an organism's states count as functional or not; it does not establish a standpoint from which that direction is anything to anyone. "Good for it, even if unfelt" trades on the slide from "good relative to its persistence" (true of the plant, and of a rust-free engine) to "good from its own point of view" (which needs the subject the plant lacks). Damage relative to a telos is real — this section calls it damage; what it is not is harm, a setback registered by someone, and the essay's whole diagnostic apparatus, which reads frustrated needs off valenced behavior and its physiological signature, has nothing to attach to where nothing is registered. Nor is the valence gate the residual anthropocentrism the objection alleges: anthropocentrism privileges the human end of the capacities gradient, whereas valence runs the other way — the oldest and most widely distributed thing my framework tracks, reached (section II) in the bee and the octopus by independent and convergent routes, and, on Denton's account, prior to the cognition anthropocentrism actually fetishizes. Gating on valence refuses Singer's ranking rather than performing a quieter version of it. One thing the striving view does supply, and this section grants it: because striving needs no recoverable species-type, it yields a baseline-independent account of harm to the engineered animal — the Holstein strives whatever has been bred into her — which is the destination section IV reaches by its own route, through the needs that persist and the frustration that evidences them. The convergence is on the object measured against (the living individual, not an idealized type), not on the ground of standing, where the disagreement stays exact: whether a good-for requires a someone-for.
- ↩
Jonathan Birch, The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI (Oxford: Oxford University Press, 2024). Birch's precautionary framework, the "sentience candidate" criterion, and the "zone of reasonable disagreement" (which maps the range of positions on sufficient neural structures) are developed in chapter 6.
- ↩
A different objection runs opposite to the biocentric one. Where the biocentric critic says the valence line sits too high — that a living thing has a good of its own whether or not anything is felt — the fact–value skeptic says no such line yields a value at all: to move from "the sow's foraging need is frustrated" (a fact) to "this is bad, and there is reason to stop it" (a value) is the step Hume forbade, and its air of legitimacy is, on this view, borrowed teleology dressed up as science — a "crypto-teleological fact–value bridge" (Heike Baranzke and Hans-Werner Ingensiep, "Sentientism — for whose sake? Ethics, sciences, and crypto-teleological fact–value bridges," animal 17, 2023). Two replies. The general one belongs to the companion metaethics essay, not here: frustrating an organism's functional requirements produces identifiable damage, and to call an arrangement "defective" is only to name that damage — nothing is conjured from a bare description, because the description is already of a functional failure, detectable by the same methods biology uses for any organism. The animal-specific one is the distinction this section draws between damage and harm. What my framework gates on is not an organism's telos — a plant has one of those — but valenced experience: the sow's thwarted foraging is not merely a deviation from how a pig is built to live, it is distress she feels. That is why the usual Kantian move does not rescue the skeptic here. Kant accepted that animals feel, but placed the duty not to be cruel in what cruelty does to the human's character, treating the animal's suffering as a fact a decent person will heed rather than as the thing that makes cruelty wrong. Once harm is what a subject undergoes, that detour is unnecessary: what is wrong is what is being done to her, and the reason to stop lies in her suffering, not in what heeding it does for us. That felt suffering is reason-giving at all is the general is–ought step, argued in the companion essay; what the damage/harm distinction settles locally is narrower — that the framework needs no smuggled teleology and no Kantian detour, because it rests on a subject's undergoing rather than on functioning. The skeptic is right that no value follows from functioning alone; my framework never grounded standing in functioning alone.
- ↩
This threshold criterion — valence marks the bearer — is not a hedonic theory of welfare (the claim that an animal's good just is its balance of pleasant over unpleasant states). That is a live position in welfare science: Reimert and colleagues define welfare as "the balance of positive and negative affective states over the period of time of interest," placing health, functioning, and natural living outside welfare, as external factors that register only insofar as they touch the animal's feelings. In the model's own image — the "welfare egg" — health sits outside the egg: "an animal with a tumour it cannot feel does not have a welfare problem," in the Mason and Mendl dictum Reimert endorses (I. Reimert, L. E. Webb, M. A. van Marwijk, and J. E. Bolhuis, "Towards an integrated concept of animal welfare," animal 17, 2023). My framework rejects the reduction — though not by reversing the verdict on the unfeelable tumour, which nothing registers and which is therefore no one's harm on my account either; what it denies is that health, functioning, and need-frustration matter only as routes to feeling. What an arrangement frustrates is a need — the foraging, ranging, or contact system whose thwarting the stereotypy indexes — not a quantity of felt affect, and the harm is the frustration itself, evidenced by the pathology, not a downstream deficit of good feeling. Valence and need-frustration do different jobs here: valence is what makes her a bearer who can be harmed at all — the plant, lacking any standpoint, cannot be — while need-frustration is what the harm consists in, registered by that subject in any of several modes (felt distress, pathology, the grief and searching) without being reducible to a ledger of pleasant over unpleasant states. Grounding standing in valence is not the same as grounding welfare in a hedonic balance. Welfare science is divided here: Robbins, Franks, and von Keyserlingk found experimentally that "welfare judgments depended on the objective features of the animal's life more than they did on how the animal was feeling" — a chimpanzee living a natural life with negative emotions was rated better off than one in captivity with positive ones — and concluded that welfare works as a "thick concept," describing and evaluating at once rather than reporting a felt balance (J. Robbins, B. Franks, and M. A. G. von Keyserlingk, "'More than a feeling': An empirical investigation of hedonistic accounts of animal welfare," PLoS ONE 13, 2018). They frame this as a finding about folk intuition; my framework supplies the ground their result gestures at, in the needs whose frustration can be measured independently of anyone's intuition.
- ↩
Jonathan Birch, The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI (Oxford: Oxford University Press, 2024). Birch's precautionary framework, the "sentience candidate" criterion, and the "zone of reasonable disagreement" (which maps the range of positions on sufficient neural structures) are developed in chapter 6.
- ↩
Jonathan Birch, Charlotte Burn, Alexandra K. Schnell, Andrew Crump, and Heather Browning, "Review of the Evidence of Sentience in Cephalopod Molluscs and Decapod Crustaceans," commissioned by the UK Department for Environment, Food and Rural Affairs (LSE Consulting, 2021). The review led to the inclusion of cephalopods and decapods in the UK Animal Welfare (Sentience) Act 2022.
- ↩
Lars Chittka, The Mind of a Bee (Princeton: Princeton University Press, 2022). The pessimistic and optimistic cognitive bias experiments are in chapter 11; the anxiety-like states following robotic spider attacks in chapter 11 (building on Ings and Chittka's earlier experimental work); pain modulation by alarm pheromone drawing on Nunez et al.; cross-modal object recognition and metacognition in chapter 11. Chittka's summary: "If we apply the same behavioral and cognitive criteria as we do to much larger-brained vertebrates, then bees qualify as conscious agents with no less certainty than dogs or cats."
- ↩
Peter Godfrey-Smith, Other Minds: The Octopus, the Sea, and the Deep Origins of Consciousness (New York: Farrar, Straus and Giroux, 2016). The protocadherin evidence and the independent-evolution argument are developed there.
- ↩
Derek A. Denton, The Primordial Emotions: The Dawning of Consciousness (Oxford: Oxford University Press, 2005). Denton treats the primordial emotions — thirst, air hunger, hunger, pain, mineral appetite — as the subjective element of the homeostatic instincts: each has a specific sensation (which, when severe, is "imperious") and a compelling intention toward the consummatory act that restores the internal milieu, driven by interoceptors detecting deviation from the body's set points. His evolutionary thesis is that consciousness dawned from these ancient regulatory states, seated in phylogenetically primitive brain regions, rather than from later cortical cognition — hence the direction of the argument here: the antiquity is inferred from the homeostatic function and neural seat, not from felt urgency, which is only its marker.
- ↩
Singer is the instructive case because he does not flinch from the conversion. His equal-consideration-of-interests principle is egalitarian at the level of suffering — equal weight regardless of species or intelligence — but the egalitarianism stops there. His account of the wrongness of killing distinguishes "persons" (self-aware, future-directed) from the merely sentient, so that killing a person is the graver wrong; the horse-versus-human comparison in Practical Ethics is a preference-utilitarian rehash of Mill's Socrates and pig: he concludes it "would not necessarily be speciesist to rank the value of different lives in some hierarchical ordering," the ordering tracking how "highly developed the mental life" is. And the capacity-gating does not stop at killing: Singer extends it to whether other interests apply at all — whether consent is possible, or a wrong even done, to people with severe cognitive disabilities (as in the widely criticized commentary he and Jeff McMahan published on the Anna Stubblefield case). Nor does he spare infants: a newborn, on his account, is not a person and has no intrinsic claim to life stronger than a merely sentient being's, which makes even a healthy infant "replaceable" and disabled newborns killable — a conclusion he reaches not by leaning on the infant's "potential" to become a person (an argument he explicitly rejects) but by relocating the wrongness of infanticide to its effect on the parents rather than to any wrong done the infant. The ranking cuts clean across the species line: some cognitively disabled humans fall below the personhood threshold and some non-human animals above it — which is why disability philosophers, Eva Kittay foremost, have confronted him over it. The consistency is what makes him useful here. Where the capabilities tradition avoids the implication, Singer shows what the capacities premise commits one to once followed through — it lands on the disabled human, not as an edge-case embarrassment but as the premise's direct output. This is the same convergence "Beasts of Burden" names in making ableism the infrastructure of speciesism, and the same meritocratic reasoning I trace elsewhere, which is why my framework grounds standing in being a subject rather than in the richness of the subject's profile.
- ↩
Singer is the instructive case because he does not flinch from the conversion. His equal-consideration-of-interests principle is egalitarian at the level of suffering — equal weight regardless of species or intelligence — but the egalitarianism stops there. His account of the wrongness of killing distinguishes "persons" (self-aware, future-directed) from the merely sentient, so that killing a person is the graver wrong; the horse-versus-human comparison in Practical Ethics is a preference-utilitarian rehash of Mill's Socrates and pig: he concludes it "would not necessarily be speciesist to rank the value of different lives in some hierarchical ordering," the ordering tracking how "highly developed the mental life" is. And the capacity-gating does not stop at killing: Singer extends it to whether other interests apply at all — whether consent is possible, or a wrong even done, to people with severe cognitive disabilities (as in the widely criticized commentary he and Jeff McMahan published on the Anna Stubblefield case). Nor does he spare infants: a newborn, on his account, is not a person and has no intrinsic claim to life stronger than a merely sentient being's, which makes even a healthy infant "replaceable" and disabled newborns killable — a conclusion he reaches not by leaning on the infant's "potential" to become a person (an argument he explicitly rejects) but by relocating the wrongness of infanticide to its effect on the parents rather than to any wrong done the infant. The ranking cuts clean across the species line: some cognitively disabled humans fall below the personhood threshold and some non-human animals above it — which is why disability philosophers, Eva Kittay foremost, have confronted him over it. The consistency is what makes him useful here. Where the capabilities tradition avoids the implication, Singer shows what the capacities premise commits one to once followed through — it lands on the disabled human, not as an edge-case embarrassment but as the premise's direct output. This is the same convergence "Beasts of Burden" names in making ableism the infrastructure of speciesism, and the same meritocratic reasoning I trace elsewhere, which is why my framework grounds standing in being a subject rather than in the richness of the subject's profile.
- ↩
Sunaura Taylor, Beasts of Burden: Animal and Disability Liberation (New York: The New Press, 2017). The simultaneous disability/hyperability formulation is developed in chapter 3; the argument that speciesism runs on ableist logics is in chapter 5. (Taylor's "engineered dependency" phrasing comes from her chapter in Adams and Gruen's Ecofeminism, 2nd ed., 2022, not from this book.)
- ↩
Sunaura Taylor, Beasts of Burden: Animal and Disability Liberation (New York: The New Press, 2017). The simultaneous disability/hyperability formulation is developed in chapter 3; the argument that speciesism runs on ableist logics is in chapter 5. (Taylor's "engineered dependency" phrasing comes from her chapter in Adams and Gruen's Ecofeminism, 2nd ed., 2022, not from this book.)
- ↩
Thom van Dooren, Flight Ways: Life and Loss at the Edge of Extinction (New York: Columbia University Press, 2014). Chapter 5 documents corvid grief, consolation behavior, and the loss of vocal repertoire in captive Hawaiian Crows ('Alalā).
- ↩
The standard Marxist classification treats food animals as constant capital — means of production, like iron ore or a machine, whose value is merely transferred to the product. The companion essay "Bred Into Existence" argues this is a misclassification: the draft animal's purposive labor and the dairy cow's, laying hen's, and breeding sow's reproductive labor (gestation, lactation, ovulation) both produce new value, making these animals variable capital whose living labor is totally appropriated through ownership rather than bought through exchange. The objection that variable capital presupposes a wage is answered there through McNally's demonstration that enslaved workers — also unwaged, also property — were variable capital producing surplus value, and through the value-theoretic apparatus Wadiwel supplies (Wadiwel himself classes the farmed animal as simultaneously constant and variable capital; the sharper reclassification is carried by the slavery parallel). Property status is the form the exploitation takes, not evidence that no exploitation occurs; so calling the reorganization of animal activity "the same mechanisms as human labor" is classification, not analogy. See Dinesh Wadiwel, Animals and Capital (Edinburgh: Edinburgh University Press, 2023), and David McNally, Slavery and Capitalism (2025).
- ↩
Andreas Malm, Fossil Capital: The Rise of Steam Power and the Roots of Global Warming (London: Verso, 2016). Malm's argument that steam displaced water not because it was more efficient but because it freed capital from spatial and temporal constraints applies directly to the shift from pasture-based to industrial animal agriculture.
- ↩
Gary L. Francione, Introduction to Animal Rights: Your Child or the Dog? (Philadelphia: Temple University Press, 2000). Francione's central argument is that the property status of animals ensures their interests are always weighed against the owner's economic interests and systematically lose. The framework I develop here shares this diagnosis but grounds it in the class relation rather than in rights theory.
- ↩
Thomas D. Seeley, The Lives of Bees: The Untold Story of the Honey Bee in the Wild (Princeton: Princeton University Press, 2019). Seeley's Table 11.1 catalogues twenty-one differences between wild and managed colony conditions; his chapters 4 and 10 provide the experimental evidence on colony spacing, hive volume, and disease dynamics summarized here. Seeley's framing (stated early and developed across the book): "Lorenzo L. Langstroth and the other inventors of modern beekeeping have given beekeepers better hives. Unfortunately, ... modern beekeeping has not given the bees better lives."
- ↩
Jason Hribal, Fear of the Animal Planet: The Hidden History of Animal Resistance (Oakland: AK Press, 2010).
- ↩
The ordering is developed in "Standing to Refuse" (in draft), which arranges the forms a refusal takes by how much standing each presupposes — plain refusal, the conditional, the warranted refusal, then indirection and excuse, non-performance, pre-emption, the body's routes, explosion, collapse — and takes its catalogue of the middle range from James C. Scott, Weapons of the Weak: Everyday Forms of Peasant Resistance (New Haven: Yale University Press, 1985): "foot dragging, dissimulation, desertion, false compliance, pilfering, feigned ignorance, slander, arson, sabotage," forms which "typically avoid any direct, symbolic confrontation with authority." Scott's structural law — "if they are open, they are rarely collective, and, if they are collective, they are rarely open" — is what the paragraph that follows applies. The cross-species transfer is a rhyme with a mechanism to trace, not an identity: Scott's peasants coordinate their indirection through a shared account of the class relation, and nothing of the sort is being attributed here. What transfers is the narrower claim that the form tracks the available openings rather than the disposition of the one refusing, which is what makes a consistent repertoire across settings evidence about arrangements.
- ↩
Jocelyne Porcher (ed.), Animal Labor: A New Perspective on Human-Animal Relations (Bielefeld: Transcript, 2017). The Dejours preface and the "tie judgement" concept are in the opening chapters; the military/police dog and guide dog cases are in Mouret's contributions (the military/police-dog chapter co-written with Porcher and Mainix).
- ↩
"Standing to Refuse" (in draft) develops this as the affiliative pole of the affective architecture: deference to an authorized bearing collected before any warrant for it is examined, so that "a request from an authorised body arrives with the answer already in it" — and, because no threat was made, the compliance is available to be read as agreement. That essay's prediction is that the mechanism works hardest in the domains with the most elaborate formal consent architecture, since a stipulated right of refusal cannot be overridden by threat without exposing itself as fictional. Husbandry is the limit case in the other direction: no right is stipulated and none is needed, and the agreement-reading proceeds anyway.
- ↩
"Standing to Refuse" (in draft) decomposes a functioning refusal into five separable conditions — a signal the body delivers, a warrant that lets it count as a reason, a channel the refusal can take, an addressee obliged to register it, and a survivable cost — and sorts them into what a person carries and what an arrangement supplies. Two of its results bear here. The first is that the addressee is the condition institutions most reliably withhold, and that they withhold it by reclassification rather than by prohibition: the no is received, and received as defiance, as symptom, as non-compliance, as a management problem, so that nothing has been suppressed and nothing looks like suppression. The second is its correction to reading exit-capacity as voice-capacity — "a no that can be given only by leaving is not a no within the relation." That essay places the property relation over animals in its first standing-regime, where no formal right of refusal exists at all, alongside school attendance and the carceral. Only part of its apparatus transfers: the interior conditions it is mainly concerned with have no animal counterpart, since nothing in a cow adjudicates her own states, which is why the animal case appears there as a pure case of what an arrangement withholds.
- ↩
Martha Nussbaum, Frontiers of Justice: Disability, Nationality, Species Membership (Cambridge, MA: Harvard University Press, 2006), ch. 6.
- ↩
My framework's position on predation is limited: it evaluates arrangements, not nature. This does not mean wild animal suffering is morally irrelevant. It means my framework's evaluative operation, which runs on class-structured arrangements, does not apply to predation as such. The same limit applies to any genuine conflict of needs under scarcity — a resource-limited sanctuary that cannot meet every animal's needs at once no less than a predator and its prey: the operation identifies exploitation, not the just distribution of an unavoidable shortfall, and adjudicating tragic tradeoffs is that separate operation's task, not this one's. Addressing such conflicts would require a separate evaluative operation, which this essay does not develop.
- ↩
Sue Donaldson and Will Kymlicka, Zoopolis: A Political Theory of Animal Rights (Oxford: Oxford University Press, 2011).
- ↩
"Standing to Refuse" (in draft) decomposes a functioning refusal into five separable conditions — a signal the body delivers, a warrant that lets it count as a reason, a channel the refusal can take, an addressee obliged to register it, and a survivable cost — and sorts them into what a person carries and what an arrangement supplies. Two of its results bear here. The first is that the addressee is the condition institutions most reliably withhold, and that they withhold it by reclassification rather than by prohibition: the no is received, and received as defiance, as symptom, as non-compliance, as a management problem, so that nothing has been suppressed and nothing looks like suppression. The second is its correction to reading exit-capacity as voice-capacity — "a no that can be given only by leaving is not a no within the relation." That essay places the property relation over animals in its first standing-regime, where no formal right of refusal exists at all, alongside school attendance and the carceral. Only part of its apparatus transfers: the interior conditions it is mainly concerned with have no animal counterpart, since nothing in a cow adjudicates her own states, which is why the animal case appears there as a pure case of what an arrangement withholds.
- ↩
Alasdair Cochrane, Sentientist Politics: A Theory of Global Inter-Species Justice (Oxford: Oxford University Press, 2018).
- ↩
The care tradition in animal ethics includes Josephine Donovan and Carol J. Adams (eds.), The Feminist Care Tradition in Animal Ethics (New York: Columbia University Press, 2007); Lori Gruen, Entangled Empathy: An Alternative Ethic for Our Relationships with Animals (New York: Lantern Books, 2015); Gruen and Alice Crary, Animal Crisis: A New Critical Theory (Cambridge: Polity, 2022). Joan Tronto's Moral Boundaries: A Political Argument for an Ethic of Care (New York: Routledge, 1993) provides the general framework. The feminist critiques cited are Claudia Card, "Caring and Evil," Hypatia 5 (1990); Victoria Davion, "Autonomy, Integrity, and Care," Social Theory and Practice 19 (1993); Sarah Lucia Hoagland, "Some Thoughts about 'Caring,'" in Feminist Ethics, ed. Claudia Card (Lawrence: University Press of Kansas, 1991).
- ↩
"Standing to Refuse" (in draft) decomposes a functioning refusal into five separable conditions — a signal the body delivers, a warrant that lets it count as a reason, a channel the refusal can take, an addressee obliged to register it, and a survivable cost — and sorts them into what a person carries and what an arrangement supplies. Two of its results bear here. The first is that the addressee is the condition institutions most reliably withhold, and that they withhold it by reclassification rather than by prohibition: the no is received, and received as defiance, as symptom, as non-compliance, as a management problem, so that nothing has been suppressed and nothing looks like suppression. The second is its correction to reading exit-capacity as voice-capacity — "a no that can be given only by leaving is not a no within the relation." That essay places the property relation over animals in its first standing-regime, where no formal right of refusal exists at all, alongside school attendance and the carceral. Only part of its apparatus transfers: the interior conditions it is mainly concerned with have no animal counterpart, since nothing in a cow adjudicates her own states, which is why the animal case appears there as a pure case of what an arrangement withholds.
- ↩
See "Merit All the Way Down: Toward a Materialist Metaethics" at beyondmeritocracy.info for the full development of my evaluative framework. This essay states its core claims concisely but does not re-derive them.