Addendum: The Predation Problem and the Limits of "Rigorous" Veganism

A note before starting: this piece is only tangentially related to the earlier parts of the series — it stands largely on its own. But the central thesis of Part 1, that humans differ in kind because of a deeper telos, turns out to be relevant here too — this time in defense of the carnivore. That application is not mentioned explicitly in the essay.

The setup: why lions get a pass

The standard vegetarian/animal-rights position doesn’t claim a lion is permitted to kill — it claims the lion is a moral patient but not a moral agent.¹ The hinge is moral agency: only beings that can reason, recognize rights, and choose otherwise can bear duties or be held responsible.² The lion can’t deliberate, so its killing carries no moral weight — it’s a physical event, not a moral one. The animal isn’t outside moral concern (as a patient it can still be wronged by us); it simply can’t do wrong, so its predation poses no moral problem to be solved.³ The asymmetry with humans isn’t “lions are excused” — it’s that the category of moral responsibility doesn’t apply to lions at all. Humans, who can hear and act on moral arguments, are the only beings the framework holds answerable.

This already defeats the lazy “but lions do it” defense of meat-eating: we never derive human norms from animal behavior. We don’t condone infanticide because chimps commit it.

The trap

But “lions do it” cuts both ways, and that exposes a real inconsistency:

1. A vegan would intervene to stop a lion eating a human.
2. Their reason: the victim’s interests matter and deserve protection.
3. The deer’s interests also matter — that’s the entire basis of veganism.
4. So by the same logic, they must intervene to stop a lion eating a deer.
5. But they don’t.

So either the vegan is inconsistent, or the deer’s interests secretly matter less than the human’s — which collapses the anti-speciesism that grounds the whole position.

The usual escapes all fail against this:

• Positive vs. negative duty / “scale” — fails, because a lion eating a specific deer in front of you has identical proximity and scale to the lion-eating-human case.⁴
• Singer’s utilitarian weighting (humans have richer preferences) — fails, because it reintroduces a hierarchy of beings, so the framework is no longer “equal consideration.” A sliding scale, not a bright line.⁵
• Regan’s rights view (the lion has no duty, but we do) — explains why the lion isn’t culpable, but not why we needn’t intervene on the deer’s behalf when we could.⁶
• Ecosystem integrity — fails, because it’s an anti-individualist move that animal-rights theory rejects everywhere else (it’s the same logic used to excuse factory farming: “the species is fine, only the individual suffers”).⁷

The honest answer

The deepest tension: animal-rights philosophy is individualist (worth lives in the individual animal’s experience), but predation is systemic (only evaluable at the level of the whole system). An individualist framework has no principled way to say “this deer’s suffering matters when a human causes it but not when a lion causes it” without conceding that the cause’s moral status infects the victim’s worth — which is incoherent.

The only defensible vegan landing spot:

“We’re not claiming the deer matters less when the lion eats it. We’re claiming the universe of beings we can morally address is limited to moral agents — and you are one. The lion-eats-deer case is therefore genuinely tragic and unaddressable, not ‘fine.’”

That’s a real, consistent distinction (we hold humans and not lions accountable across the board, not selectively). But it’s a far more uncomfortable position than casual veganism usually occupies — and it’s worth spelling out exactly why, because the discomfort is the whole point.

Why this is uncomfortable: the bright line is gone

Casual veganism runs on a black-and-white claim: killing animals is wrong, full stop. It’s clean, categorical, and — crucially — it lets you never mention lions. Nature is off the table; the claim is about what you do, and the answer is a flat “don’t.”

The lion-deer trap destroys that flatness. You can no longer say “killing is wrong, period,” because the lion kills and you don’t want to condemn the lion or send yourself out to save the deer. So you must say something more specific about why your killing is wrong and the lion’s isn’t.

First, the dead end

The tempting move is to grade the deer’s worth down: say the deer’s interests count for less than a human’s because its mental life is simpler, fewer forward-looking projects, less lost in death. (This grading-by-cognitive-richness is broadly Singer’s preference-utilitarian view — the same one invoked earlier as a candidate escape — though Singer deploys it to weigh suffering, not to license predation; pressed into service as a lion-excuse it backfires immediately.) But this doesn’t even work as a lion-excuse — and worse, it detonates in your hand. If the deer is worth less because it’s a deer, then you can eat it too. Lowering the deer’s worth doesn’t explain why the lion may kill it and you may not; it just removes the reason anyone shouldn’t. A vegan cannot take this route. It defeats veganism in one step. So worth-grading is off the table — which means there is really only one route left.

The only real route — and it pays twice

The move that actually excuses the lion is the one from the top of this essay: grade responsibility, not worth. The deer keeps its full, equal worth; what the lion lacks is moral agency — the capacity to grasp and act on moral reasons. The lion can’t be blamed because it can’t be answerable; you can, because you can. This is the “humans have a higher moral capacity” line. Notice it says nothing about the deer being worth less — it’s entirely about the killer’s capacity for responsibility.

This is the only coherent escape. But it doesn’t dodge a cost by being clean — it gets hit from two directions at once:

Cost 1 — the intervention problem. Because this route keeps the deer a full, equal patient, the deer still has a real claim not to be killed. So the original trap walks right back in: if the deer is an equal rights-holder you could shield from the lion at low cost, why don’t you? Excusing the lion (“it’s not culpable”) does nothing to discharge your duty to assist a rights-holder in peril. You answer “why isn’t the lion guilty?” and reopen “why don’t you intervene?”

Cost 2 — the marginal-cases cut. This is where the Argument from Marginal Cases bites — and it’s crucial to see that AMC attacks the responsibility criterion, not just a worth criterion.⁸ AMC is normally a weapon vegans use against meat-eaters: the carnivore says “humans matter and animals don’t because humans are rational/responsible,” and the vegan replies that infants, the severely impaired, and the comatose also lack rationality and moral responsibility, yet plainly still matter — so that capacity can’t be what does the work. The trap is that this route forces the vegan to pick up the very criterion AMC exists to puncture. Once “killing is wrong for humans because humans have the capacity for moral responsibility” is doing the load-bearing work, AMC turns it around as a reductio: infants and the profoundly impaired lack that capacity too, so either the criterion absurdly withdraws their protection, or — the obvious reading — the criterion was never what made killing wrong in the first place. It breaks on the infant. And the wreckage folds straight back into Cost 1: the only thing left that can ground the infant’s protection is the infant’s worth as a patient, independent of any responsibility-capacity — but the deer is a full patient too, so the same worth that obligates you to save the infant from the lion obligates you to save the deer. The criterion you used to mark humans off from the lion fails to mark some humans off from animals, and the only repair reinstates the intervention duty.

So it isn’t a fork where you choose which price to pay. The single viable route inherits both bills. That’s a stronger result than a fork: the one move that excuses the lion is squeezed from both sides simultaneously.

The one escape from Cost 2 — straight into Cost 1

There’s a sharp reply available to the marginal-cases cut, and it’s worth following because of where it lands. The vegan can say: “Responsibility governs only blame, not killability. The infant and the lion are alike in being non-culpable — not in being permissibly killed. I never said low-responsibility beings may be killed; I said they can’t be held answerable.” That’s correct and coherent.

But look at what it costs. If the responsibility criterion now governs only blame and not the wrongness of killing, then it has stopped explaining why your killing is wrong. It explains why the lion isn’t guilty — full stop. The wrongness of you killing the deer now needs a separate ground, and the only ground left is the deer’s worth as a patient… which is exactly the equal-worth claim that generates Cost 1. Escaping the marginal-cases cut by confining responsibility to blame throws you straight back onto the intervention problem. The two costs aren’t just both attached to the route — they’re welded together: every attempt to slip one tightens the other.

(There are standard patches for the marginal-cases cut — the species-norm reply, treating individuals by what’s normal for their kind, and the slippery-slope reply, holding all humans equal because grading them would corrode how we treat everyone.⁹ Both work as defenses, but both are species-typical or pragmatic — they reintroduce exactly the species-based, non-individualist reasoning anti-speciesist veganism set out to reject. The patch holds, but it’s bought with the framework’s founding premise.)

None of this makes the deer’s intrinsic worth rigorously demonstrable. The framework still can’t generate clean universal obligations from the “intrinsic equal worth” premise — it can only choose which discomfort to carry, and the viable route carries two.

What this proves — and what it doesn’t

The reductio does defeat one specific target: the version of veganism that claims its conclusions follow by pure logical force, such that the meat-eater is committing a rational error. Selective application of a premise is the signature of preference dressed as logic, and the vegan who condemns the BBQer while shrugging at lion-deer predation is applying the premise selectively.

But the reductio overreaches if it concludes “veganism is mere preference a BBQer can dismiss.” Almost no substantive ethical position clears that rigor bar. Run the identical move on the carnivore:

• You’d stop a man torturing a dog for fun — but cats play with mice before killing. Stop the cat, or admit your objection is “mere preference”?
• You’d stop a man killing a toddler — but chimps commit infanticide. Same trap.

The lion-deer problem isn’t a special weakness of veganism. It’s an instance of the general truth that “nature does X” never settles whether “humans ought to do X,” and that every moral framework carries obligations it can’t fully discharge.

The conclusion: underdetermined, not arbitrary

“Humans have a higher moral calling” is the premise that’s supposed to ground veganism — but it’s bivalent. You can route it toward:

• We rise above nature’s indifference → minimize all suffering, including the deer’s (veganism), or
• We are a morally distinct kind of being, stewards standing outside the animal moral community (dominion / use).

Both are coherent developments of the same premise. The premise underdetermines the conclusion. It is not deterministic.

So the disagreement isn’t about a fact or a logical step. Both sides concede animal suffering is real, and both concede humans are morally distinctive. The disagreement is a weighting: how much should the reality of that suffering constrain me, versus how much does human distinctiveness license using animals? You have to put in a weight — and the weight is where the values live, not the logic.

This is not relativism. The situation is specific and rare: two coherent settlements of a genuine tension — an equation with more than one root. Neither is forced, neither is absurd. That’s precisely the condition under which mutual respect is rationally mandated rather than merely polite.

• The vegan is owed respect because their position is a competent, non-arbitrary settlement of a real moral tension — and they’ve often thought harder about the suffering than the BBQer has.
• The BBQer is owed respect because “humans stand in a different relation to animals than to each other” is most of human moral tradition, religious and secular — not a dumb premise.

What neither gets to do is call the other irrational. The vegan who calls the BBQer a moral idiot, and the BBQer who calls the vegan a sentimental fool, are both mistaking “I weighted this differently” for “you reasoned badly.”

Not right vs. wrong — two valid roots of the same equation. Which is exactly why respect is the rigorous answer, not just the courteous one.