(P1)
Pain counts against God's existence only if it is gratuitous.
+
Almost nobody thinks pain by itself is evidence against God. The serious versions of the objection say so outright, and they name the extra ingredient they need.
(1) Pain with a point is not evidence against a good God.
- Picture a parent pinning a screaming two-year-old still for a vaccination, or a surgeon breaking a badly set bone so that it can heal straight. Neither one is a monster. The pain is real, and it is doing something.
- So the question is never whether there is pain. It is whether there is pain that accomplishes nothing whatsoever.
(2) The word for pain that accomplishes nothing is "gratuitous."
- Gratuitous pain is pain God could have prevented without losing some greater good, and without having to permit some evil equally bad or worse. That is the objector's own definition and the standard one in the literature, and it covers pain of every kind, felt by any creature capable of feeling it.
- In plainer words, it is pain that does no work. A dentist's drill hurts and earns its keep. The same drill run on a healthy tooth for ten minutes, for no reason anyone could name, would be gratuitous.
- Notice how much that definition gives away before the argument even starts. It admits that suffering can be justified, and then rests the whole objection on the claim that this suffering is not.
(3) The reasoning runs by a form called modus tollens.
- Modus tollens goes like this: if the first thing were true, the second would have to be true; the second is not; so the first is not either. If it rained last night, the driveway would be wet. The driveway is bone dry. It did not rain.
- Applied here: if animal pain counts against God, that pain has to be gratuitous. Show that it is not gratuitous, and it does not count against him.
- A deductive argument works like a chain, and if every link holds, the conclusion cannot slip out. The links do not have to be certainties, only more plausible than their denials, which is a fairer bar than proof and still a real one.
(4) The two sides carry different burdens.
- No Christian has to name God's reason for any particular animal's pain. Naming it was never part of the claim, and demanding it loads the question.
- The objector does not have to prove God impossible either, and the serious versions never try. The claim is that after careful looking, no justifying reason turns up, and that this is good evidence there is none. That is an inference rather than a proof, and in plenty of settings a perfectly good one.
- Everything turns, then, on whether this is one of those settings. Think of a mechanic who cannot say why the engineer ran a wire the long way round. With the full schematic spread on the bench in front of him, his failure to find a reason tells him something. Without it, it tells him almost nothing.
(5) A second version of the objection asks about the pattern of pain rather than its point.
- It grants that some pain does work, and presses the arrangement instead: pain in nature tracks whatever keeps a creature alive and breeding, and a world with nobody steering predicts that pattern better than a world with a good God in charge.
- That is a comparison of likelihoods rather than a charge of pointlessness, so the premise that answers it is (P4) rather than (P1). Both versions do need the same thing in the end: an estimate of what a good God would probably have arranged, and an estimate that comes out low.
Therefore, the question the animal world raises is not whether animals hurt, but whether their hurting is pointless.
(P2)
Most animals may react to injury without feeling anything.
+
The writhing, the fleeing, the crying out: all of it is behavior a nervous system can produce whether or not anyone is home behind it. Telling behavior apart from feeling is the hardest question in the whole subject. It is also the least settled, and it is unsettled for the overwhelming majority of animals alive.
(1) Detecting damage and feeling damage are two different processes.
- Nociception is a nervous system's detection of injury and the response it triggers. Your own hand comes off a hot pan before you feel a thing. The spinal cord ran that move by itself, and the sensation arrives a beat later. The withdrawal and the hurt are separate events, and the withdrawal came first.
- The felt side has a name too. Philosophers call it phenomenal consciousness, meaning that there is something it is like to be the creature in question. There is something it is like to be you right now. There is something it is like to be a dog waiting at a door. There is plausibly nothing at all it is like to be a thermostat.
- Think of it this way: a robot vacuum backs away from a staircase every single time. It has sensors, it has a goal, it changes course. Nobody thinks it is frightened of the stairs.
(2) Very simple nervous systems produce surprisingly clever avoidance.
- The roundworm Caenorhabditis elegans is about a millimeter long and has exactly 302 neurons. It learns to avoid the smell of bacteria that once made it sick. A human brain has something on the order of 86 billion neurons, roughly 285 million times as many, and the worm still manages the trick.
- Rats whose spinal cords have been surgically cut off from the brain will still learn to hold a hind leg clear of a shock. That learning is happening in a length of cord with no brain attached to it, and nobody on any side of this debate thinks a length of cord is a subject of experience.
(3) Where the line falls between the two is genuinely unsettled.
- Researchers who work on invertebrate consciousness are candid that there is no agreed method for settling whether an insect feels anything, and no consensus on the answer. That is simply the state of the field, and over the past two decades it has moved toward wider attributions of feeling rather than narrower ones.
- A minority of philosophers of mind go further. They argue that the question is not merely open but unanswerable, because our concept of felt experience has no settled application outside the human case.
- Honesty requires the point to cut both ways. It is a good reason not to be confident that a wasp suffers. It is not a proof that a wasp does not, and treating it as one is the same mistake pointing the other direction.
(4) The arithmetic of the objection depends entirely on where that line falls.
- Of roughly 1.5 million described animal species, well over ninety percent are invertebrates, meaning animals with no backbone, and most of those are insects. Count individual creatures rather than species and the imbalance is wider still.
- So the staggering total the objection adds up is made overwhelmingly of beetles, mites, worms, and flies. Whether any of them feels anything is precisely what has not been established, in either direction.
Therefore, the size of the animal world's suffering cannot be read off the number of animals in it, because most of those animals may be reacting without feeling anything at all.
(P3)
The pain animals do feel is not the self-aware suffering humans know.
+
Vertebrates feel pain, and no honest treatment says otherwise. What they appear to lack is the thing that turns pain into the particular horror it is for us.
(1) Human suffering is made largely of knowing.
- Ask anyone who has waited on a biopsy result. The worst stretch is often the week before treatment starts, when nothing hurts yet at all. What hurts is knowing.
- The distinction is between feeling pain and grasping that you are the one in pain, that it was you yesterday, and that it will still be you tomorrow. Call that second thing self-aware suffering: pain taken up into a creature who knows that the life being damaged is its own. Strip it away and what is left is bad, but it is not the same thing.
(2) A creature can hurt without being able to think of itself as hurting.
- A dog with a cut paw yelps and limps and licks the wound. It does not lie awake the night before a veterinary appointment. It does not spend the drive over rehearsing what is about to happen.
- The deer does not know that it will die. The fish on the line is not contemplating its own mortality. Their pain is real, and it is confined to the moment it occupies.
(3) Some animals plan and remember far more than was once thought.
- Scrub jays that hide food go back for the perishable items first and leave the durable ones, as though they keep track of what they stored, where, and how long ago. That is a genuine result, it has held up, and it should make anyone cautious about drawing this line too confidently.
- Chimpanzees and orangutans have been recorded setting a tool aside for use hours later. One captive chimpanzee gathered stones in the quiet of the morning to throw at visitors in the afternoon.
(4) Even passing the mirror test is not holding your life together as a story.
- In the mirror-mark test, a mark is placed where an animal can only see it in a mirror. If the animal touches the mark on its own body rather than on the glass, it has connected the reflection to itself. Chimpanzees, dolphins, elephants, magpies, and a small reef fish called the cleaner wrasse have all passed some version of it.
- The wrasse result is disputed. Critics argue the fish was scraping at what looked like a parasite on another fish rather than recognizing itself, and that argument is not over.
- What none of it yet shows, the mirror results and the scrub jays alike, is a creature holding its own life together as a single story, with an ending it can see coming and dread. That is the capacity (P3) turns on. Researchers who argue for wide animal consciousness say so themselves: the richer kind of self-awareness is poorly evidenced outside our own species.
(5) Saying animal pain is different is not saying animals do not matter.
- Scripture is not squeamish about the difference. Jesus tells frightened disciples that not one sparrow falls to the ground apart from the Father, and then draws the conclusion they need: "you are of more value than many sparrows" (Matthew 10:29-31). The comparison only works if the sparrow is genuinely noticed. He gives them a graded picture rather than a flat one, which is exactly what (P3) claims.
- The righteous man cares for the life of his animal (Proverbs 12:10). Israel's law rested the ox on the Sabbath along with everyone else (Exodus 23:12) and forbade muzzling it while it worked (Deuteronomy 25:4).
- If anything, the picture is a merciful one. Whatever else follows, an animal is spared the dread, the anticipation, and the knowledge of its own death that make human dying what it is.
Therefore, animal pain is real, and it is not the self-aware suffering the objection asks us to picture when it invites us into the animal's place.
(C1)
Therefore, animal pain carries far less moral weight than it appears to carry.
+
(P2) and (P3) come at the same thing from two directions: how many creatures are suffering, and how much any one of them suffers.
(1) The tally shrinks in two ways at once.
- Start with how many creatures belong in the tally at all. Reacting to injury is not the same as feeling it, and most of the animals being counted are ones nobody can yet say feel anything (P2).
- Then take the ones that plainly do feel. What they lack is the self-aware suffering that carries the greatest moral weight (P3).
- What has not shrunk is whether animals matter. They do, and the texts quoted in (P3) say so.
(2) None of this licenses cruelty.
- A person who concluded from (P2) or (P3) that he may treat a dog however he likes has drawn exactly the wrong conclusion, and has broken a command besides.
- Level of awareness bears on how bad a given episode of pain is. It does not hand anyone a permission slip to cause it.
(3) What survives still needs an answer.
- Millions of vertebrates are torn apart, starve, and drown every day, and they feel it. Nothing in (P2) or (P3) explains that away.
- (P4) through (P6) do not try to explain it away either. They answer it.
Therefore, the weight of the animal world's pain is a great deal less than it looks, and what remains of it is genuine.
(P4)
Stable natural law is required for embodied creatures to act at all.
+
A world where creatures do anything at all has to be a world where doing a thing has predictable results. That requirement turns out to be expensive.
(1) Free action requires a world that behaves the same way twice.
- You can throw a ball to a child because gravity is boring. In a world where fire warmed you on Monday and froze you on Tuesday, no creature could learn anything, plan anything, or choose anything, because nothing it did would connect reliably to any outcome.
- The name for this reasoning is a natural order theodicy.
• A theodicy is an attempt to say what reason God might actually have for permitting some evil. It aims higher than a mere defense, which only has to show that some reason is possible. The natural order theodicy says the reason is the stability itself.
(2) The same physics that makes a body livable makes it breakable.
- A body soft enough to feel a hand laid on it is soft enough to be torn. A digestive system that can pull nourishment out of food can also absorb poison. Nobody has described a way to build the first without the second.
- Rock hard enough to hold a mountain up is hard enough to crush what it lands on. The property is one property.
(3) Predation is the other side of speed, hunger, and sight.
- The eyes and legs that let a deer outrun a wolf are the eyes and legs that let the wolf run down the deer. It is the same equipment, issued to both.
- A creature that has to eat is a creature that can starve, and in a world of finite food, some of the eating is going to be of each other.
(4) Suspending the laws case by case would eat the regularity it interrupts.
- The obvious repair is to leave the laws alone and have God step in whenever a creature is about to be hurt for nothing. But the stepping in is not free. A world where fire burns except when burning would injure something is a world where nobody can find out what fire does.
- Rare exceptions are a different matter, and Christianity has always held that God works some. What the repair needs is exceptions as a standing policy, and a standing policy of exceptions is not an exception at all. It is the rule, wearing a disguise.
- Think of a referee who reverses every call that would cost a player the game. After the third reversal there is no longer a game being played, only a series of decisions.
(5) The bill for a stable world is real.
- A law-governed world does not come free. It costs a great deal, and no theodicy makes that bill disappear.
- What can be said is that the cost was plausibly worth paying, because the alternative on offer is not a gentler world. It is a world in which no embodied creature could act at all.
- Think of a city that wants its people to be able to get places. It builds roads, and roads produce collisions. Nobody excuses a collision by pointing at the road. The point is that there is no version of getting around town that leaves the roads out.
- That is also the limit of what (P4) claims. It says why the possibility of animal pain is built into a world like this one. It does not say why this creature was burned in this fire, and it never pretended to.
Therefore, the regularities that make animal pain possible are the same regularities that make animal life possible, and they cannot be separated by any means we can describe.
(P5)
Removing predators would replace faster deaths with slower ones.
+
Every animal dies. Take the predators away and the deaths do not stop; they only take longer.
(1) The alternative to being eaten is not living forever.
- The objection tends to compare predation with nothing at all. The real comparison is predation against the other ways an animal in the wild comes to an end: infection, injury, parasites, cold, and the long decline of not getting enough to eat.
- Measured against those, being killed by something faster than you is the short version.
(2) A population with no predators eats through its food and then collapses.
- In 1944 the United States Coast Guard put twenty-nine reindeer on St. Matthew Island in the Bering Sea, a place with no wolves and no bears. Nineteen years later there were about six thousand of them, a two-hundredfold increase, and the lichen they lived on was essentially gone.
- Then came one exceptionally hard winter. By the next census there were forty-two animals left, forty-one of them female and one male, with no calves anywhere on the island. Nothing had hunted them. That was the entire problem, and the herd never recovered.
- A predator-free world is not a world without mass death. It is a world where mass death arrives as famine.
(3) Not every kill is quick.
- Some deaths by predator are appalling and slow. A pack can start feeding on an animal that is still alive, and Darwin's own example of pointless cruelty in nature was a cat playing with a mouse. Nobody who has watched a wildlife film needs to be told that predation is not gentle.
- What being caught actually feels like is another matter. Seized by a lion in 1844, the missionary David Livingstone described "a sort of dreaminess in which there was no sense of pain nor feeling of terror, though quite conscious of all that was happening." He compared it to a mouse going limp after the first shake of a cat, and guessed it was a mercy given to every animal taken by a carnivore. Physiology now has a name for it: stress-induced analgesia, the blunting of pain under extreme threat.
- (P5) is about populations rather than individual episodes. Across a whole population, the deaths predators cause run shorter than the deaths that replace them when the predators are gone, which is a comparison between two bad arrangements rather than a defense of one good one.
(4) Removing predators is something wildlife managers now spend money undoing.
- Wolves were returned to Yellowstone in 1995, some seventy years after the last of them had been killed off, into a park where elk had browsed the streamside willows down hard. In parts of the park the willows, the beaver, and the songbirds recovered afterward.
- Ecologists genuinely disagree about how much of that recovery the wolves caused, since drought, bears, hunting, and elk simply moving elsewhere all played a part. The case should not be pushed harder than the data allows. What is not in dispute is that the predator-free arrangement was failing on its own terms.
(5) A geologist made this case in 1836.
- More than twenty years before Darwin published, the geologist William Buckland argued that death by a predator is on balance a mercy. It subtracts from the total pain of dying, it cuts short disease and slow decay, and it holds numbers within what the land can feed.
- He was writing natural theology, the attempt to argue for God from the study of nature, and that genre has gone out of fashion. The reasoning does not depend on it. The reindeer on St. Matthew Island had never heard of Buckland, and it made no difference to how they died.
Therefore, the predation the objection finds hardest to accept is, inside a world of finite food, on the whole the gentler of the deaths actually on offer.
(P6)
The present state of creation is not its final state.
+
Christianity has never taught that the world as we find it is the world as it will be. The claim is old, it comes straight out of the text, and it changes what the fossil record is evidence of.
(1) Paul says creation is not yet what it will be.
- Creation "was subjected to futility, not willingly, but because of him who subjected it, in hope" (Romans 8:20). The Greek word behind "futility" is mataiotes, meaning emptiness or frustration: a thing not reaching what it is for, like a field that will not hold seed.
- The two details usually skipped are the important ones. Creation "was subjected," which is passive, so it did not drift into this condition; somebody put it there. And it was subjected "in hope," which means the condition was imposed with an end in view rather than suffered as a breakdown.
(2) The prophets' picture of the end has no predation in it.
- The wolf dwells with the lamb, the leopard lies down with the young goat, the lion eats straw like the ox, and "they shall not hurt or destroy in all my holy mountain" (Isaiah 11:6-9).
- Whatever that language finally amounts to, and Christians read it in more than one way, it is offered as the future rather than as a lost past. The direction of the hope is forward.
(3) Christians who have thought hardest about this have admitted they were guessing.
- Nobody in the tradition claims to know what becomes of a wren. The most widely read Christian book on the problem of pain, published in 1940, gave animals a whole chapter and then admitted inside that chapter that the whole of it was guesswork with no authority behind it.
- The honest position is that Scripture says creation will be liberated and does not say how far down the ladder that reaches.
(4) Hope is not an explanation.
- A promise about the end does not explain why there was pain at the beginning. That work belongs to the natural order and to the shape of animal death (P4, P5), and if those two fail, no promise will rescue them.
- Making something up to a creature afterward is not the same as having had a reason for it, and the difference is worth keeping straight.
- There is a third possibility, which philosophers call defeat: not paying a creature back for what it went through, but taking the suffering up into a larger good so that it becomes part of what that good is. Two friends fall out badly and are reconciled years later. The friendship afterward is not the friendship that would have existed had nothing gone wrong, and the injury is part of what it now is. Whether the promise in Romans 8 reaches animals in that way is exactly what nobody knows.
- What (P6) denies is that the fossil record is the last word about the creatures in it. An objection that treats the present arrangement as final is arguing against a Christianity that does not exist.
Therefore, the state of nature the objection points to is, on Christian teaching, a stage rather than a verdict.
(C2)
Therefore, animal suffering does not make God's existence improbable.
+
The animal world does not give anyone good reason to think God is not there. That is the whole claim, and it is narrower than the one people usually hear.
(1) The pain that remains is not gratuitous.
- Pain counts against God only if it is gratuitous (P1). Three things stand in the way of calling what remains gratuitous:
• A stable order whose regularities cannot be suspended case by case without destroying the regularity itself (P4).
• A food chain that shortens dying rather than lengthening it (P5).
• A creation ordered toward an end it has not yet reached (P6).
- On top of those three, a large share of what the tally counted was never shown to be suffering at all (P2, C1).
- Take them together and the "if" in (P1) has nothing left to fire on.
- The step from the goods those premises name to the flat words "not gratuitous" is the hardest step in the whole case, and it is not self-evident. Pretending otherwise would be dishonest.
(2) The case stands on two legs.
- The first leg gives positive reasons why animal pain in a world like ours would serve a purpose (P4, P5, P6). The second denies that our failure to see a purpose in some particular case is good evidence that there is none.
- Neither leg is asked to do the other's work. The positive reasons cover the arrangement. The limit on arguing from our own ignorance covers the individual episode that no arrangement was ever going to explain.
- Think of a defense lawyer who both produces an alibi and shows that the eyewitness could not have seen the street from where he was standing. Either would help on its own. Together they cover different ground.
(3) None of this is evidence that God exists.
- Nothing in (P1) through (P6) counts toward theism, the claim that God is real. The work is defensive from beginning to end, and an objection that fails to make theism improbable has not thereby made it probable.
- Whether God exists is settled, if at all, by everything there is. The arguments for theism have to be weighed against the difficulties, and the animal world is one item on a long list.
- Nor does any of it claim to know God's reason for any particular animal's pain. An argument that claimed that would deserve the suspicion it got.
- Nor does any of it say animals do not matter. The texts in (P3) rule that reading out.
(4) Better evidence about animal minds would weaken the case considerably.
- If the science settled that most animals possess rich, self-aware suffering that persists across time, (P2) would be false and (P3) would lose most of its force. Everything would then rest on (P4) through (P6) alone, and it would be a good deal weaker for it.
- That is a real vulnerability, and it is a moving one, since work on animal consciousness has been expanding the circle rather than shrinking it. Saying so is not a concession made under pressure. It is what the state of the evidence actually supports.
Therefore, the animal world is a real difficulty for Christian belief, and it is not a disproof of it.
See also:
CO / PoE: Intellectual PoE: Probabilistic
Natural Theology: The Moral Argument
On animal pain and theodicy:
Adams, Marilyn McCord. Horrendous Evils and the Goodness of God. Ithaca, NY: Cornell University Press, 1999.
Alston, William P. "The Inductive Argument from Evil and the Human Cognitive Condition." Philosophical Perspectives 5 (1991): 29-67.
Bergmann, Michael. "Skeptical Theism and Rowe's New Evidential Argument from Evil." Noûs 35, no. 2 (2001): 278-296.
Bergmann, Michael, and Michael C. Rea. "In Defence of Sceptical Theism: A Reply to Almeida and Oppy." Australasian Journal of Philosophy 83, no. 2 (2005): 241-251.
Craig, William Lane. Systematic Philosophical Theology, Volume IIb: On God: Excursus on Natural Theology, On the Trinity. Hoboken: Wiley-Blackwell, 2026.
Dougherty, Trent. The Problem of Animal Pain: A Theodicy for All Creatures Great and Small. Basingstoke: Palgrave Macmillan, 2014.
Hasker, William. "The Open Theist View." In God and the Problem of Evil: Five Views, edited by Chad Meister and James K. Dew, Jr. Downers Grove, IL: IVP Academic, 2017.
Hasker, William. The Triumph of God over Evil: Theodicy for a World of Suffering. Downers Grove, IL: IVP Academic, 2008.
Lewis, C. S. The Problem of Pain. London: Geoffrey Bles, 1940.
Moreland, J. P. The Recalcitrant Imago Dei: Human Persons and the Failure of Naturalism. London: SCM Press, 2009.
Moreland, J. P., and Scott B. Rae. Body and Soul: Human Nature and the Crisis in Ethics. Downers Grove, IL: InterVarsity Press, 2000.
Murray, Michael J. Nature Red in Tooth and Claw: Theism and the Problem of Animal Suffering. Oxford: Oxford University Press, 2008.
Plantinga, Alvin. "On Being Evidentially Challenged." In The Evidential Argument from Evil, edited by Daniel Howard-Snyder. Bloomington: Indiana University Press, 1996.
Swinburne, Richard. Providence and the Problem of Evil. Oxford: Clarendon Press, 1998.
Van Inwagen, Peter. The Problem of Evil. Oxford: Oxford University Press, 2006.
Wykstra, Stephen J. "The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of 'Appearance.'" International Journal for Philosophy of Religion 16, no. 2 (1984): 73-93.
On animal consciousness and ecology:
Berdoy, Manuel, Joanne P. Webster, and David W. Macdonald. "Fatal Attraction in Rats Infected with Toxoplasma gondii." Proceedings of the Royal Society B 267, no. 1452 (2000): 1591-1594.
Berthier, Marcelo, Sergio Starkstein, and Ramón Leiguarda. "Asymbolia for Pain: A Sensory-Limbic Disconnection Syndrome." Annals of Neurology 24, no. 1 (1988): 41-49.
Birch, Jonathan. The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI. Oxford: Oxford University Press, 2024.
Birch, Jonathan. "The Search for Invertebrate Consciousness." Noûs 56, no. 1 (2022): 133-153.
Birch, Jonathan, Alexandra K. Schnell, and Nicola S. Clayton. "Dimensions of Animal Consciousness." Trends in Cognitive Sciences 24, no. 10 (2020): 789-801.
"The Cambridge Declaration on Consciousness." Francis Crick Memorial Conference, Cambridge, July 7, 2012.
Carruthers, Peter. Human and Animal Minds: The Consciousness Questions Laid to Rest. Oxford: Oxford University Press, 2019.
Clayton, Nicola S., and Anthony Dickinson. "Episodic-like Memory during Cache Recovery by Scrub Jays." Nature 395 (1998): 272-274.
Cox, James J., et al. "An SCN9A Channelopathy Causes Congenital Inability to Experience Pain." Nature 444, no. 7121 (2006): 894-898.
Grau, James W., Douglas G. Barstow, and Robin L. Joynes. "Instrumental Learning within the Spinal Cord: I. Behavioral Properties." Behavioral Neuroscience 112, no. 6 (1998): 1366-1386.
Klein, David R. "The Introduction, Increase, and Crash of Reindeer on St. Matthew Island." Journal of Wildlife Management 32, no. 2 (1968): 350-367.
Kohda, Masanori, et al. "If a Fish Can Pass the Mark Test, What Are the Implications for Consciousness and Self-Awareness Testing in Animals?" PLOS Biology 17, no. 2 (2019): e3000021.
Middleton, Arthur D. "Is the Wolf a Real American Hero?" New York Times, March 9, 2014.
Mulcahy, Nicholas J., and Josep Call. "Apes Save Tools for Future Use." Science 312, no. 5776 (2006): 1038-1040.
"The New York Declaration on Animal Consciousness." New York University, April 19, 2024.
Osvath, Mathias. "Spontaneous Planning for Future Stone Throwing by a Male Chimpanzee." Current Biology 19, no. 5 (2009): R190-R191.
Ripple, William J., and Robert L. Beschta. "Trophic Cascades in Yellowstone: The First 15 Years after Wolf Reintroduction." Biological Conservation 145, no. 1 (2012): 205-213.
Critic sources engaged:
Almeida, Michael J., and Graham Oppy. "Sceptical Theism and Evidential Arguments from Evil." Australasian Journal of Philosophy 81, no. 4 (2003): 496-516.
Ayala, Francisco J. Darwin's Gift to Science and Religion. Washington, DC: Joseph Henry Press, 2007.
Dawkins, Richard. River Out of Eden: A Darwinian View of Life. New York: Basic Books, 1995.
Dawkins, Richard. The Greatest Show on Earth: The Evidence for Evolution. New York: Free Press, 2009.
Draper, Paul. "Pain and Pleasure: An Evidential Problem for Theists." Noûs 23, no. 3 (1989): 331-350.
Gould, Stephen Jay. "Nonmoral Nature." In Hen's Teeth and Horse's Toes. New York: W. W. Norton, 1983.
Rowe, William L. "The Problem of Evil and Some Varieties of Atheism." American Philosophical Quarterly 16, no. 4 (1979): 335-341.
Rowe, William L. "Evil and the Theistic Hypothesis: A Response to Wykstra." International Journal for Philosophy of Religion 16, no. 2 (1984): 95-100.
Rowe, William L. "The Evidential Argument from Evil: A Second Look." In The Evidential Argument from Evil, edited by Daniel Howard-Snyder. Bloomington: Indiana University Press, 1996.
Singer, Peter. Animal Liberation. New York: New York Review, 1975.
Historical sources:
Buckland, William. Geology and Mineralogy Considered with Reference to Natural Theology. 2 vols. Bridgewater Treatises 6. London: William Pickering, 1836.
Darwin, Charles. "Letter to Asa Gray, 22 May 1860." In The Correspondence of Charles Darwin, vol. 8, edited by Frederick Burkhardt et al. Cambridge: Cambridge University Press, 1993.
Hume, David. Dialogues Concerning Natural Religion. London, 1779. Part XI.
Livingstone, David. Missionary Travels and Researches in South Africa. London: John Murray, 1857.
Biblical citations:
Genesis 1:29-30; Exodus 23:12; Deuteronomy 25:4; Proverbs 12:10; Isaiah 11:6-9; Matthew 10:29-31; Romans 5:12; Romans 8:19-23; 1 Corinthians 9:9
See also:
- CO / PoE: Intellectual PoE: Logical
- CO / PoE: Intellectual PoE: Probabilistic
- CO / PoE: Intellectual PoE: External
- CO / PoE: Emotional PoE
- CO / PoE: Proof of God from Evil
+
A fawn is trapped in a forest fire, badly burned, and lies in agony for several days before it dies, and nobody ever sees it. An all-powerful God could have prevented that with no loss to anything, and did not.
The philosopher William Rowe put this case at the center of the modern debate in 1979, and it deserves to be taken in his own terms rather than in a softened version. His argument runs in three steps:
- There exist cases of intense suffering that an all-powerful, all-knowing being could have prevented without thereby losing some greater good, and without having to permit some evil equally bad or worse.
- An all-knowing, wholly good being would prevent any intense suffering it could, unless it could not do so without losing some greater good or permitting some evil equally bad or worse.
- So there is no all-powerful, all-knowing, wholly good being.
The fawn is his example precisely because it strips out every human-centered excuse. No free will is involved, no lesson is being learned, no soul is being made, and there is nobody watching to be moved by it.
1. Something very like Rowe's fawn happens somewhere every day.
It has happened for as long as there have been forests and lightning. That pain is real, it is not small, and no Christian should be comfortable about it.
Rowe himself deserves credit here. He called his position "friendly atheism," by which he meant that a believer could be perfectly rational in rejecting his conclusion. That is a rare and admirable thing to say about one's own argument, and it should set the tone of any reply.
2. The weak link is the step from "we see no good reason" to "there is no good reason."
This move has a name in the literature: a noseeum inference, from "I no see 'um, so it ain't there." Sometimes that inference is excellent and sometimes it is worthless. What settles which is whether we would expect to see the thing if it were there.
- Look across an empty garage and see no rottweiler, and you may safely conclude there is no rottweiler. A rottweiler is the sort of thing you would notice.
- Look across the same garage and see no fleas, and you have learned nothing at all about fleas.
- So the question is whether the reasons of a mind that holds the whole causal history of a universe together are more like rottweilers or more like fleas. Rowe's argument requires that they be rottweilers, and that is exactly the claim that has never been established.
Rowe answered this in print, and his answer is a good one. Nobody, he said, is being asked to survey God's mind. In ordinary cases we can see well enough what a suffering was for, and our repeated failure to see it in cases like the fawn's is itself evidence.
That is a fair point about method, and it puts the question exactly where it belongs. Everything now turns on one thing: are the goods we can already see a fair sample of the goods there are?
Think of a juror handed three pages out of a thousand-page file. He may report quite honestly that he found nothing in those pages about where the defendant was that night, and still have learned nothing whatever about where the defendant was. The goods we can see sit inside a human-sized frame: a surgery, a lesson, a rescue. The goods at issue would have to run through the whole causal order of a world and the whole of its future. Nobody has shown that our three pages are the representative ones.
3. Rowe's first step needs a minimum that may not exist.
Grant that God could have spared this fawn without losing anything. The same will be true of the next fawn, and the one after that, all the way down, because for a great many goods there is no least quantity of suffering that secures them, in the way that there is no smallest sum of money that counts as generous.
- Where there is no minimum, any amount God permits is an amount some smaller amount would have matched. So the fact that this particular case could have been spared shows nothing at all about whether it was pointless. It shows only that a line had to be drawn somewhere.
- An objector can answer that he needs no minimum, only an excess. That is fair, and it hands him the harder task of saying where the excess begins. Nobody has been able to make that measurement, which is the same wall the noseeum inference runs into.
4. The fawn hurts without knowing that it is dying.
A fawn has a backbone and a nervous system, and it hurts (P3). What it does not do is count the days, or grieve in advance for a life it will not have. Much of the horror the example produces in a reader comes from imagining a human consciousness inside that body over those days.
Removing that projection does not make the case go away, and nobody should pretend it does. It does mean the quantity being weighed is smaller than "several days of agony" sounds to a creature that can dread.
5. Christianity does not need to name the good that fawn's death served.
What it needs is that the pain not be gratuitous, and (P4) through (P6) give three general reasons why animal pain in a world like ours would not be:
- Natural law is stable, which is what makes animal life possible at all (P4).
- Fire and predation shorten deaths that would otherwise come slowly (P5).
- Creation is ordered toward an end it has not yet reached (P6).
None of those is a story about that particular fawn. None of them needs to be.
6. Rowe's own strongest version does not stay with animals.
In his later statements of the argument he set a human case beside the fawn: a child murdered where nobody could see it. He did that precisely to shut off replies that turn on what animals can and cannot experience, and it was the right instinct on his part.
He kept the fawn as well, and for an equally good reason, since the fawn shuts off the replies that turn on free will and on souls being made. Each case was chosen to close a door the other leaves open, and (P2) and (P3) close only one of them. The human case is a different argument and it needs a different answer.
See also:
CO / PoE: Intellectual PoE: Probabilistic
+
Pain and pleasure in nature track what helps a creature survive and reproduce, not what any creature deserves. That is exactly the pattern you would expect if nobody is in charge, and not what you would expect from a loving God.
The strongest published form of this objection is Paul Draper's, from 1989, and it is not the crude complaint it is often mistaken for. Draper is not asking why God allows pain. He is running a comparison between two rival hypotheses and asking which one better predicts what we actually find:
- Theism: a good and powerful God stands behind the living world.
- What Draper calls the hypothesis of indifference: neither the nature nor the condition of living things on earth is the result of anything done by non-human persons, kindly or cruel. In plain terms, nobody is steering.
He then sorts pain and pleasure into three kinds: the sort that is biologically useful to the creature that has it, the sort that is useful to nobody, and the sort whose usefulness is unknown. His case is that the whole pattern, and the second kind especially, is far more surprising on theism than on indifference. Richard Dawkins gave the popular form of the same thought: a universe of blind, pitiless indifference is precisely what we should expect to see.
1. The observation is correct.
Pain in the animal world does track biological usefulness. Creatures hurt where hurting keeps them alive, and feel good where feeling good gets them fed and bred. There is no use pretending the correlation is not there, and a reply that starts by disputing the data is a reply that has not read the paper.
2. Useful pain is exactly what a law-governed world produces.
Draper's argument is comparative, so what theism actually predicts matters enormously. Christian theism is not the hypothesis that God arranges each sensation by hand. It is the hypothesis that God made a world running on stable regularities in which embodied creatures can act (P4). In such a world, pain systems get built the way everything else gets built, by the ordinary operation of those regularities, which means they will track survival.
Build a world in which a hot stove teaches a child something, and you have built a world in which a hot stove hurts. The lesson and the burn are one event. So the correlation Draper identifies is not an embarrassment to the natural order theodicy (P4). It is one of its predictions.
That matters more than it looks. His comparison needs one particular number to come out low: the probability of the observed pattern given theism. Nobody has measured that number and nobody could. It is an estimate of what a good God would have been likely to arrange, and once theism is filled out with a world of stable regularities, the estimate stops being low.
3. The hard part of his case is the pain that is useful to nobody.
Draper's sharpest examples are not the toothache that keeps you off a bad tooth. They are the pains that do nothing for anyone: the agony of an animal already dying, chronic pain that goes on long after the injury it once signaled has healed, the burned fawn that will not recover whatever it feels. If pain were designed, why would it keep firing when it has nothing left to protect?
Three things can be said, and none of them is a knockdown blow:
- A warning system tuned to be reliable in the ordinary case will misfire in the extraordinary one. A smoke alarm sensitive enough to catch a real fire also goes off over burnt toast, and the alternative is an alarm that sleeps through fires.
- Building a system that works in general but falls silent in the hopeless case means deciding, case by case, which cases are hopeless. That is the standing policy of exceptions that (P4) says a stable order cannot survive.
- How much pain actually falls into this category is not known, and Draper does not claim to have measured it. The category is real. Its size is an estimate, and an estimate is what the argument's weight is resting on.
4. Draper's comparison is with bare theism.
Bare theism is theism stripped to its minimum: some powerful, good God exists, and nothing further is assumed. Draper says so himself, and he is careful to add that further background knowledge can change the verdict. Christian theism is not bare theism. It carries content that bears directly on the animal world:
- A world built to run on stable regularities, which is why pain systems would track biology in the first place (P4).
- A moral weight that depends on the kind of awareness the sufferer has (P3).
- A creation subjected to futility and awaiting liberation (P6).
Whether that additional content is independently credible is a fair question, and it is answered by the case for Christian theism as a whole rather than by anything said about animals. What is not fair is scoring bare theism and then announcing the result for Christianity.
5. The pleasure has to be counted too.
Draper's subject is the distribution of pain and pleasure, so the good half of the ledger is his business as well. He can point out, quite fairly, that biologically useful pleasure is exactly what indifference plus evolution would produce, so a theist cannot simply gesture at happy animals and call it a rebuttal.
What is harder for him is his own second category running the other way. A great deal of pleasure in the animal world is not obviously doing any biological work:
- Play that carries on long after it has finished training anything.
- Curiosity that costs energy and returns nothing in particular.
- Birdsong well past what holding a territory requires.
A biologist can propose a function for each of these, and often does. A biologist can also propose a function for most pain. The ledger has two columns, and the objection tends to audit only one.
6. The dispute is live.
Draper's argument has not been refuted to the satisfaction of the philosophers who work on it, and no reply anyone has offered has gone unchallenged in print. Calling the matter settled in either direction would be false. What can be said is narrower and still worth something: at most, one strand of evidence favors indifference, and the question of God's existence is not decided by one strand.
+
Ichneumon wasps lay their eggs inside living caterpillars, and the larvae eat the host from the inside while keeping it alive as long as possible. Darwin said he could not persuade himself that a good and all-powerful God designed that.
Darwin wrote that to the botanist Asa Gray in May of 1860, and Stephen Jay Gould later built an entire essay around the horror those wasps caused the nineteenth-century writers who argued for God from nature. Nature supplies the wasps with plenty of company:
- The cuckoo chick hatches first and heaves the host bird's own eggs and chicks over the side of the nest.
- The tongue-eating louse Cymothoa exigua consumes a fish's tongue and then takes its place as a working organ.
- The parasite Toxoplasma gondii alters an infected rat's brain until the rat stops avoiding the smell of cats.
1. Nearly every creature on that list is an invertebrate (P2).
The wasps, the caterpillars, and the lice have nervous systems that may well not support any experience at all, and nobody has a method that would settle it. A caterpillar being eaten from within squirms and tries to get away, and squirming is exactly what a nervous system with no experiencer would produce. The cuckoo chick heaving eggs over the side is not making a choice; it is running hardwired behavior, no more open to moral assessment than a Venus flytrap closing.
The horror a reader feels watching the caterpillar comes largely from imagining being the caterpillar. Whether there is anyone in there to be is precisely the open question, and an objection that helps itself to the answer has assumed the thing it needed to show.
2. What the parasite does to the rat is remove its fear.
Toxoplasma gondii is the one item on the list involving a mammal, and mammals feel things (P3). The infected rat loses its aversion to the smell of cats, and more recent work suggests the effect is a general loss of fear rather than a targeted attraction to the animal that will eat it. Either way, the rewired rat is less frightened than an ordinary rat, not more. Whatever else that is, it is not the infliction of suffering.
3. Parasitism does the work that predation does, on a smaller scale.
Ichneumon wasps are among the most effective natural controls on caterpillar populations, which is why agriculture takes an interest in them. Without them, caterpillar numbers rise, strip the vegetation they feed on, and then fall by starvation, which is the St. Matthew Island pattern at a different scale (P5). Darwin's revulsion was real and it was decent. It did not take the ecological arithmetic into account, because in 1860 nobody had done it.
4. Darwin reached for the same answer in the same letter.
A few lines from the sentence about the wasps, Darwin says he is inclined to see everything as the result of designed laws with the details left to work themselves out. Then he admits at once that the notion does not satisfy him, that the whole subject is too deep for a human mind, and that a dog might as well speculate about Newton.
He was not writing a case against God. He was writing that he was bewildered, and the handle he reached for is the same one (P4) takes hold of.
5. The protest borrows a moral standard.
Darwin was not describing the wasps; he was condemning them. A careful objector will say straight away that he does not have to own the standard he is using. He is arguing from the inside, borrowing the Christian's own conviction that cruelty is wrong and showing that the world does not look like the work of a God who shares it. That is a legitimate way to argue and it should be granted without a fuss.
What it costs him is the rhetoric, and the freedom to pick and choose. Run from the inside, the objection can no longer say that nature is objectively cruel, only that a Christian ought to find it hard to square. And an argument run from inside a system has to take the whole of that system, which here includes a stable order nobody has shown could be had on other terms (P4) and a creation that has not arrived at what it is for (P6).
Meanwhile the version most readers actually meet does say nature is objectively cruel, with the wasp standing as a real indictment of a real moral order. That version needs a standard, and it is the version least able to supply one. If physical nature is all there is, and nothing stands behind it, then there is no cruelty in nature at all. There is only what happens.
See also:
Natural Theology: The Moral Argument
+
Forget the insects. The fox in the leg-hold trap, the beached whale, and the deer dying slowly on a roadside all genuinely feel pain, and telling us they lack self-awareness is cold comfort.
This is the right place to press, and the strongest form of the objection presses exactly here rather than on the wasps. An objector can concede every word of (P2) about invertebrates and lose nothing, because he never needed them. Dogs, cats, deer, horses, whales, and millions of other vertebrates have the neural equipment for felt pain and they use it. They cry out. They favor the injured limb. They flee.
1. The fox in the trap is genuinely in pain.
A Christian who says otherwise has said something false and should stop. The fox is in pain, the pain is bad, and a good person seeing it would end it. That is where an honest answer starts, not a concession extracted from it later.
2. That pain lacks the structure that makes human suffering what it is.
The terrible thing about a human being in long confinement is not the bare physical sensation. It is the knowledge of yesterday, tomorrow, and forever: the anticipation, the humiliation, the sense of a life being taken piece by piece. The fox has the sensation. On the available evidence it does not have the rest, and the rest is where most of the moral weight of human suffering sits.
That is a claim about how much, not about what kind of thing. The fox's pain counts for less than a human's would in the same trap. It does not count for nothing.
3. The examples come from the slow end of a range.
Choosing the worst cases is fair enough in an objection, and nobody should pretend a leg-hold trap is typical. Some of the sharpest examples are ones people caused: the trap, the car, the puppy mill. Scripture indicts those and so should we (Proverbs 12:10), and they are evidence about human beings rather than about the world God made.
The rest is not so easily set aside, and (P5) does not try to. Wild animals do starve slowly, do die of infection, and do linger with injuries nothing will heal. What (P5) claims is comparative: across a population, the deaths predators cause run shorter than the deaths that replace them when the predators are gone.
4. The pain that remains is what (P4) through (P6) were built for.
Grant all of it. The vertebrate pain in the world is still pain inside a stable natural order that embodied creatures require (P4), inside a food chain that shortens dying rather than lengthening it (P5), in a creation that Christianity says is not yet finished (P6). The objection does not slip past those three. It lands squarely on them, which is where they were placed to take the weight.
+
If pain matters morally only when the sufferer knows it is happening to them, then the pain of newborns and of people with severe cognitive disability would not matter either. That conclusion is monstrous, so the premise producing it is false.
This objection deserves to be taken as seriously as any raised about animal suffering, and it is not a debating point. If a Christian answer to animal suffering works by making moral status depend on how self-aware a creature currently is, it has bought that answer at a price no Christian can pay, because the same move would put infants, the senile, the comatose, and the profoundly disabled outside the circle of moral concern. Any argument with that consequence should be abandoned, whatever it does for the problem of evil.
1. (P3) says self-awareness makes pain worse, not that it makes pain matter.
Those are different claims, and only the second has the monstrous consequence. A newborn's pain matters completely. The reason it is not identical in weight to an adult's fear of a diagnosis is that the newborn is not doing the dreading, and that is a statement about what the experience contains, not about what the child is worth.
2. Moral status comes from what a thing is, not from what it can do today.
Human beings bear the image of God. That is a fact about the kind of thing a human is, and it does not switch off during sleep, anesthesia, infancy, or dementia. A sleeping adult has no self-awareness at all for hours at a time and does not thereby become available for experimentation.
- The distinction philosophers use here is between a capacity and its exercise. A pianist who is asleep still has the capacity to play; she is simply not exercising it tonight. Push it one step down and there is a deeper capacity underneath: a child who has never touched a keyboard has, by being the kind of creature she is, the capacity to learn. That deeper one is grounded in what she is rather than in what she has so far done, which is why sleep, injury, and immaturity cannot take it away.
- An infant is a being of the kind that grasps its own life. A wasp is not. That is a difference in what they are, and it is exactly the difference the objection's counterexamples do not have.
3. Calling the appeal to kinds "speciesism" names the move rather than defeating it.
The comeback, pressed hardest by Peter Singer, is that grounding moral status in species membership is itself the arbitrary step. It picks out a biological category and then treats it as morally decisive, which is the shape of prejudices we have learned to distrust.
- The claim is not about biology. Bearing the image of God is a claim about what a human being is, not about a set of chromosomes, and whether that claim is true is settled by the case for Christianity as a whole.
- Singer takes the consistent alternative and grades by present capacity all the way down, accepting that some animals then outrank some humans. Almost nobody follows him there. This objection cannot follow him at all, since its entire force came from the monstrousness of downgrading the newborn.
4. The objector has to grade suffering too.
Almost nobody thinks a mosquito's death and a child's death are morally equivalent, which means almost everybody is grading somehow. The question is whether the grading goes by the kind of being involved or by a measurement of what the individual can do right now. It is the second that produces the monstrous results. That second option is the one the objection rightly rejects, and it is the one Christian teaching has never held.
5. Cruelty to animals is a sin.
On the practical question the tradition has never been ambiguous. Israel's law rested working animals (Exodus 23:12), forbade muzzling the ox that treads the grain (Deuteronomy 25:4), and made care for an animal's life a mark of righteousness (Proverbs 12:10). Jesus said the Father attends to the death of a sparrow (Matthew 10:29).
One text is worth heading off, because at first glance it looks like a counterexample. Paul quotes the ox law and asks whether it is for oxen that God is concerned (1 Corinthians 9:9), which sounds like a denial that the law protected the animal. Read that way it would contradict the law's own plain sense. Read as the standard Jewish argument from the lesser case to the greater, which is what Paul is plainly doing, it says the opposite: if God writes legislation for an ox, how much more for the people doing the work.
Nothing in (P2) or (P3) loosens any of that, and a reader who came away with permission to be careless with an animal has read it backwards.
See also:
Natural Theology: The Moral Argument
+
For hundreds of millions of years before humans existed, animals that could feel were hunted, starved, and eaten in their billions. Even if each episode is small, the total is staggering, and God could have made humans without that prologue.
The scale is worth stating exactly, because it is the objection's whole force:
- Complex predator-prey relationships appear in the fossil record by the Cambrian, roughly 540 million years ago.
- The end-Permian extinction about 252 million years ago killed the large majority of marine species.
- The end-Cretaceous extinction about 66 million years ago ended every dinosaur except the birds.
- Billions of animals still die daily.
God could have created humans directly, or run an evolutionary process without predation and parasitism, and did neither.
1. Suffering does not obviously add up across creatures the way the objection needs.
The number only does its work if pains in separate creatures sum into one quantity of badness. Aggregating across individuals is standard in a good deal of moral philosophy, and refusing to do it carries real costs, so this is not a cheap escape and should not be offered as one.
Still, it is far from obvious that a million brief episodes in creatures that cannot think about themselves add up to the weight of one person's long conscious ordeal, any more than a million paper cuts spread across a million strangers add up to a murder. Nobody undergoes the total. There is no sufferer of the sum.
Quantity is not irrelevant, and pretending otherwise would be dishonest. The point is narrower: the objection's most impressive move, the sheer size of the number, is doing less work than it appears to, and less again once most of the number turns out to be creatures nobody can yet say feel anything (P2).
2. The long ages were not empty, even on a purely physical accounting.
Those hundreds of millions of years produced the ecosystems that support complex life, the atmosphere with the oxygen in it, the soil, and the fossil carbon that human civilization ran on for two centuries. They produced brains capable of language and of asking this question. A world with an actual history behind it is not obviously worse than a world manufactured last Tuesday with the appearance of one.
3. The alternative on offer is not obviously better.
The objection does not need a recent creation and should not be saddled with one. God could have made an old world in which nothing felt anything until human beings arrived, or a world with a wholly different biology. Grant all of that. The objection still needs the alternative to be better all things considered, not merely available, and that is a claim about how an entire world would have run.
On the physics and biology we actually have, a world under stable law reaches creatures like us through populations that vary, compete, and die. Removing that means either different laws throughout or continual intervention within them, which is the bill (P4) itemizes. Whether some third arrangement was open to God is exactly the sort of question about unrealized worlds that nobody is in a position to answer.
4. Judging that the process was excessive requires knowledge nobody has.
Whether God could have reached the same ends with less animal pain is a question about which whole worlds were available to him and how each of them would have turned out. Confident answers to that are not on offer to finite minds.
It is the noseeum problem again. Moving from "I see no good reason" to "there is no good reason" works only where we would expect to spot the reason if it were there, and hundreds of millions of years of causal history is the last place on earth to expect that. The same limit constrains every version of the argument that evil makes God unlikely rather than impossible.
See also:
CO / PoE: Intellectual PoE: Probabilistic
+
The 2012 Cambridge Declaration and the 2024 New York Declaration on Animal Consciousness say conscious experience is far more widespread than assumed, reaching birds, octopuses, fishes, and possibly insects. Sorting animals into levels of awareness is out of date.
The evidence behind these statements is genuine and worth listing rather than waving at:
- Octopuses solve puzzles, open jars, and appear to change color while sleeping.
- New Caledonian crows manufacture hooked tools and hang on to favored ones.
- Honeybees tell each other the direction and distance of food by dancing.
- Elephants return to the bones of dead herd members and handle them.
- The cleaner wrasse, a fish the size of a finger, passed a version of the mirror-mark test in 2019.
The New York Declaration goes further than the Cambridge one, holding that there is at least a realistic possibility of conscious experience in all vertebrates and in many invertebrates, including cephalopods (octopuses and squid), decapod crustaceans (crabs and lobsters), and insects.
1. The declarations answer a different question from the one (P3) asks.
Three questions get run together in this debate, and they need pulling apart:
- Does the creature react to damage? Nearly every animal does, and so does a robot vacuum.
- Does it feel anything while it reacts? This is the question both declarations address.
- Does it know that it is the one feeling it, that it was yesterday, and that it will be tomorrow? This is the question (P3) turns on, and neither declaration addresses it.
On that third question the declarations are silent rather than contrary.
2. The 2024 statement claims a "realistic possibility," not a consensus.
The New York Declaration is carefully worded: strong scientific support for conscious experience in mammals and birds, and a realistic possibility of it in the rest, insects included. Both documents are position statements signed by researchers who work in the area rather than surveys of a settled field, and that fact cuts in both directions. It means neither can be quoted as the verdict of biology. It also means nobody may quote biology as having ruled invertebrate feeling out, because it has not.
Jonathan Birch, one of the philosophers behind the 2024 statement and a leading researcher on invertebrate minds, has written plainly that there is no agreement on whether any invertebrate is conscious and no agreed method that could settle it. An objector who treats these documents as a settled consensus is quoting them for more than they say, and the researchers are the ones saying so.
3. Evidence for the top grade of self-awareness remains very thin.
Researchers who work on this treat self-awareness as several things rather than one, and the grades come apart:
- Telling self from other, the most basic version, which a great many animals manage.
- Knowing where your own body ends, which is what the mirror-mark test probes.
- Holding yourself together across time as the one who has been having your experiences all along, and expecting to go on being that one.
Various animals reach the second. Evidence for the third outside our own species is scarce, and the third is the grade (P3) is about.
4. If the circle widens, the answer widens with it.
Suppose the declarations are eventually vindicated all the way down to bees. Then (P2) is false and should be given up, because the tally it holds open would fill. What would be left is (P4) through (P6), and the bee's pain would sit inside the same stable order, the same food chain, and the same unfinished creation as the deer's. That is a weaker case than the one with (P2) in it, and saying so is the honest thing to do about a scientific question still in motion.
+
Animal death and predation ran for hundreds of millions of years before any human sinned. An appeal to a fallen creation cannot explain pain that came first, so Romans 8 does no work here.
The objection assumes the mainstream geological chronology, on which predation, parasitism, and mass extinction sit in the fossil record long before anything human. Christians divide over that chronology, and (P4) through (P6) do not need it settled. If the deep past runs the way the objection says, those premises still have to hold, and they do. What the objection is right about without qualification is that a great deal of popular Christian writing on animal suffering helps itself to a sequence it has never defended.
1. Nothing in (P4) or (P5) depends on when predation began.
(P4) and (P5) are claims about what a world of embodied creatures under stable law requires. They would hold in a world with a fall, in a world without one, and at whatever date the rocks turn out to support. A stable natural order and a food chain that shortens dying are features of the world as such.
2. Romans 8 is not obviously a claim about chronology.
Paul says creation "was subjected to futility, not willingly, but because of him who subjected it, in hope" (Romans 8:20). Most readers take the one who did the subjecting to be God, doing it deliberately and with an end in view; a minority take it to be Adam, whose sin brought the subjection about. Read either way, Paul's stress falls on where creation is going rather than on the date it started going there. Mataiotes, the word behind "futility," fits a creation that has not yet arrived at what it is for at least as naturally as it fits a machine that used to work and broke.
3. The verse usually set beside it is about human death.
Sin came into the world through one man, and death through sin, "and so death spread to all men because all sinned" (Romans 5:12). The sentence names its own subject. Paul is tracing the human race from Adam to Christ, and the death he says spread is the death that spread to all men. Whether animals died before that is not a question the verse takes up, which is why it cannot be made to answer it either way.
4. (P6) works on every Christian reading of these texts.
Christians who accept the same Scriptures disagree in-house about how to take them:
- Some hold that animal death began with human sin, and read the geological record in that light.
- Some hold that the fall's effects reach backward, so that God orders creation from the beginning in view of a rebellion he foresees.
- Some hold that creation was made incomplete from the start, always meant to be brought to a glory it did not begin with.
- Some take the passage to concern the human world's disorder spilling outward.
(P6) is deliberately worded to sit above all four. It claims only that the present state is not the final one, which every one of these readings affirms.
5. The objection defeats the fall as an explanation of animal pain.
If someone offers Adam's sin as that explanation, and offers it on a deep chronology, the objection defeats him, and it should. But (P6) was never that explanation. It offers hope about the end, and hope is not an explanation. The explanatory work sits in (P4) and (P5), which no chronology touches.
+
An all-powerful God could have made an ecosystem with no predation at all, and Genesis 1 suggests the original creation was vegetarian. If a predator-free world was possible, the appeal to ecological necessity collapses.
The objection has both a philosophical half and a scriptural half. Philosophically, it says that the actual world is not necessary but chosen, and that a better option was available. Scripturally, it points to Genesis 1:29-30, where God gives green plants for food to every beast of the earth, and reads that as a description of a creation without carnivores.
1. A vegetarian world is not the gentle world the objection pictures.
Herbivores in a world with no predators still meet a limit, and the limit is the food supply. The reindeer of St. Matthew Island ate nothing but lichen and died by the thousand anyway (P5). Take the carnivores out and the following remain:
- Starvation. Wildebeest die in droughts and deer die in hard winters, and dying that way takes weeks.
- Disease and parasites. Neither of them requires anyone to eat anyone.
- Competition. Animals fight savagely over water and forage, and the fighting gets worse as the forage runs out.
The comparison the objection needs is not "predation versus no predation" but "predation versus whatever replaces it," and the second comparison is nowhere near as one-sided as the first.
2. Whether Genesis 1 teaches an originally carnivore-free biology is genuinely contested among Christians.
Genesis 1:29-30 is read by some as a straightforward description of pre-fall biology, and by others as a statement about God's provision and the goodness of what he made, in a passage never written as a biology report. Both readings have careful defenders who take the text with equal seriousness, and neither is a dodge.
Whichever reading is right, (P4) and (P5) do not depend on it. And it is worth noticing where the vision of a predator-free world sits in Scripture itself: Isaiah puts the wolf with the lamb in the age to come (Isaiah 11:6-9), which is a hope about the end rather than a report about the beginning.
3. Showing an alternative was possible is not showing it was better.
That God could have created differently is true and not in dispute. For the objection to bite, the alternative has to be better all things considered: better at supporting embodied creatures who act, better at holding populations within what the land can feed, better at serving whatever God intends creation for. Establishing that means tracing consequences through a world nobody has seen, and no one on either side of the question can do it.
4. A world's pain can have a point even if another world was possible.
The claim was never that this world is the only one God could have made. It is that this world's pain is not gratuitous, and the two come apart completely. "Gratuitous" does not mean "avoidable by building something else instead." It means the pain serves nothing, which is a question about this world and what holds together inside it.
Think of a surgeon choosing between two workable operations. That the other one existed is not by itself a criticism of the one he performed, and it is certainly not evidence that the cutting he did was pointless.
+
Why does damage-avoidance have to hurt? A creature could be built to flinch from injury with no felt agony at all, which makes the feeling an unnecessary extra a good God would have left out.
This is the sharpest form the complaint takes, and it has been in print since 1779. Hume put it in Philo's mouth: grant that a world run by general laws will contain hazards, and still ask why pain had to be the signal, when animals might have been driven to eat, flee, and mate by nothing worse than a falling off of pleasure.
The objection has a second edge here, because (P2) hands it over. If a nervous system can detect damage and respond to it with nobody home behind it, then the feeling looks like precisely the part that could have been left out.
1. (P2) and (P4) do pull against each other.
(P2) says a nervous system can detect and avoid damage without anyone feeling anything. (P4) says the pain in the world is the price of a stable order. Put those side by side and the obvious question is why the price had to include the feeling. Nothing is gained by pretending the two premises sit comfortably together. They have to be reconciled, not stacked.
2. A feeling can be weighed against other things, and a reflex cannot.
A reflex handles one situation in one way, and it cannot weigh the injury against anything else.
- A gazelle with a torn flank has to keep running anyway, because the lion is a nearer problem than the wound. Something has to hold the damage in mind as one consideration among several and then let it be outweighed.
- A felt state does that and a reflex does not. It persists after the stimulus is gone, it competes for attention, it can be traded against hunger or fear, and it teaches the animal something it can carry into a situation it has never met before.
- That fits the pattern in the animals themselves. The simplest creatures, the ones with the least to decide, are the ones most plausibly running on reflex alone. The animals nobody seriously doubts are conscious are the ones with the most to decide.
3. Take the feeling out of a creature built around it and the result is a catastrophe.
People are born, rarely, unable to feel physical pain, through faults in single genes that build the nerve fibers and channels carrying the signal.
- They bite through their tongues as toddlers. They walk on fractures. Their joints are destroyed by injuries they never noticed, and many die young of wounds and infections nobody caught in time.
- A fair objection is that these people lack the signal as well as the feeling, so the case has not yet pulled the two apart. That is right, and there is a rarer condition that pulls them apart precisely.
- In asymbolia for pain, which follows damage around a fold of cortex called the insula, the sensation arrives intact and the patient reports it as painful, while the unpleasantness that normally rides along with it is simply absent. Those patients do not draw back from harm the way everyone else does.
- What the pair of conditions cannot settle is what some other designer might have managed from scratch. What they show is that in the one system we can examine from the inside, taking the hurt out of the signal does not leave the avoiding behind.
4. The "reduced pleasure" alternative is not obviously a different system.
Philo's proposal is that creatures be steered by degrees of pleasure rather than by pain. But whatever sits at the bottom of that scale, the state a creature is in when things are going worst for it, is the state it will work hardest to escape. Rename it and the question simply repeats: why does the bottom of the scale have to feel bad? A scale with no bad end has no urgency at the bad end, which is where urgency is needed.
5. Nobody has surveyed the alternatives.
None of this shows that God could not have built something better. It shows that the obvious alternative is not obviously better, which is all (P4) ever claimed. What a maximally good designer could have managed is a question about worlds nobody has ever seen, and on those the only honest answer is that we do not know.
+
If we cannot tell whether any suffering is pointless, we cannot tell whether we should prevent any of it either. The move that blocks the argument from evil wrecks ordinary moral reasoning along with it.
The move under attack has a name: skeptical theism, the position that our view of the goods a world might contain is too narrow to carry the leap from "I see no reason for this" to "there is no reason for this." The skepticism is about the range of our own sight, not about God.
Michael Almeida and Graham Oppy press the sharpest form of the objection. If our grasp of the goods available to an all-knowing mind is that poor, it is poor enough to block a great deal else, including the everyday reasoning by which we decide to step in and relieve someone's suffering, since for all we know God is permitting it for a good we cannot see. This is contested ground with able defenders on both sides, and it has not been settled in print.
1. Unrestricted skepticism is not what is claimed.
Grant the objection its target. Someone who said we can know nothing whatever about God's purposes would have no way left to call God good, no reason to trust a promise, and no way to tell a revelation from a fraud. That position should be refused.
The claim in play is narrow and specific: that we are poor judges of what goods are available across the whole causal history of a world, and of how things would have run had one event gone differently. That is a limit on one kind of survey, not a license for doubt about everything. You can be sure a surgeon is competent and kind while having no idea at all what she is doing with the instruments.
2. Our duties are set by what we know, not by what God knows.
The objection needs a premise it rarely states out loud: that I ought to step in only if I know no greater good will be lost by my stepping in. That is a consequentialist premise, meaning one that makes an act right or wrong purely by how the totals of good and bad come out afterward. Almost no moral theory accepts it in that form, and no Christian one does.
- A nurse gives morphine because the patient is in pain and she can relieve it. She does not withhold it on the theory that the pain might be serving a purpose she cannot see, and nothing about a limited view of God's purposes suggests she should.
- Obligations attach to agents in their situations. I know the child is drowning and I can reach him. That is the whole of my case, and whether an all-knowing being would have a reason to permit the drowning is not a fact I have access to, which is exactly why it is not a fact I can act on.
- Christianity makes this explicit rather than leaving it to inference. God commands mercy, and the command is not conditional on our being able to audit his providence.
3. God's goodness is not a conclusion drawn from surveying outcomes.
If Christians believed God was good because they had checked the world's ledger and found it balanced, this objection would land hard. That is not where the belief comes from. It rests on revelation and the character of Christ, and the moral argument runs the other way, from the reality of moral obligation back to its source. So a limited view of what good a particular fire serves does not undercut the ground of the belief. It only keeps that fire from confirming or refuting it.
4. The standard of access cuts at the argument from evil too.
The argument from evil needs us to be reliable judges of which goods a world could contain. If we are reliable judges, the Christian gets to use that competence as well, and the goods named in (P4) through (P6) are not especially hard to see. If we are not, the argument's key premise loses its support. No witness gets to claim expert eyesight on the questions that help his case and blindness on the ones that do not.
See also:
CO / PoE: Intellectual PoE: Probabilistic
Natural Theology: The Moral Argument
+
Uncertainty about insect minds rescues nothing. If there is even a modest chance a bee feels pain, the expected suffering across quintillions of insects is still overwhelming, and a good God would not gamble with it.
This is the best reply to a premise that rests on "nobody knows," and it is the reasoning behind the practical half of the 2024 declaration, which says that where there is a realistic possibility of conscious experience it is irresponsible to ignore it. Applied to the problem of evil the reasoning runs like this: multiply a modest probability of sentience, the capacity to feel anything at all, by an enormous number of creatures, and the product is enormous whatever the probability turns out to be.
1. The practical version is right.
Where it is uncertain whether a creature can suffer, caution is the right policy. That is how a decent person treats a lobster, and it follows from the animal-welfare texts rather than sitting awkwardly beside them (Proverbs 12:10). Nothing in (P2) argues for carelessness. (P2) says we do not know, and "we do not know" is a reason for care rather than against it.
2. The multiplication needs a number nobody has.
Expected-value reasoning takes a probability and multiplies it by a quantity. The trouble is that the probability here is not small; it is missing. The researchers who framed the "realistic possibility" standard chose that wording on purpose, holding that no responsible figure for insect sentience can be given at present, which is why what they offer is a threshold for acting rather than an estimate of the world. A precaution tells you what to do while you are in the dark. It does not tell you what the dark contains.
3. The evidential version also needs the pains to sum.
Reasoning of this kind requires that a faint pain in one insect and a faint pain in another combine into a larger quantity of badness held by nobody in particular. That assumption is doing a great deal of work, and it is not obvious. There is no bee who has the total.
4. The multiplication runs down both columns of the ledger.
If a modest probability of insect sentience multiplies into an enormous quantity of suffering, the same probability multiplies into an enormous quantity of ordinary insect well-being: warmth, feeding, flight, mating, the long stretches in which nothing bad is happening at all. The expected value of a quintillion faintly sentient lives is not obviously negative, and the objection needs it to be negative, not merely large.
5. The goods named are features of the arrangement, not payments to each creature.
(P4) and (P5) do not offer a compensating good to each animal one at a time. They say that a world of embodied creatures under stable law cannot be had on other terms, and that claim does not weaken as the number of creatures grows. A city's road system is not justified one driver at a time, and it does not stop being justified when the population doubles. If anything, a large, old, fully populated world is exactly the thing (P4) and (P5) were claims about in the first place.
See also:
CO / PoE: Intellectual PoE: Probabilistic