A moral case for vegetarianism has been made by some philosophers and has become popular among a small group of people not noted for their reticence. The most influential of these philosophers is Peter Singer. Singer’s argument is that it’s immoral to cause suffering, that the suffering of non-human animals has equal weight with the suffering of humans, that you can’t eat meat without patronizing and encouraging the inflicting of suffering on animals, and that therefore it must be immoral to eat meat, except in cases of dire necessity.

I think this argument is mistaken, and I will now give you my chief counter-argument. My counter-argument contains a lemma—an intermediate conclusion that I can then use as a premiss for my final argument. To keep things short and simple, I’m not going to argue here for the lemma (though I am going to briefly explain the point of it), since I believe that most people, if they think about it even briefly will agree with it. I’m just going to state the lemma and then move on from there. (Although I say “my” counter-argument, I don’t mean to imply that there’s anything original about this. I’ve heard something similar to this before, though I have no idea who first came up with it. After all, it’s pretty obvious.)

Lemma: We’re not under any moral obligation to act so as to reduce the total amount of animal suffering below what it is in the wild, or below what it would be if humans didn’t exist. In other words, if the immorality of eating meat is dependent on humans causing animals to suffer, then it can’t be immoral to eat meat if the production of meat for human consumption does not increase the suffering of animals above what it would have been in the absence of any human intervention.

Explanation of the lemma: In the absence of human intervention, animals like deer and oxen would be eaten by non-human predators. When humans eat meat, they’re competing with other meat-eating animals, such as lions and wolves. If the predators disappear, this may lead to overpopulation of the former prey animals and consequent unwelcome environmental effects such as deforestation followed by soil erosion. The situation is not changed in principle if we move from hunting to the raising of livestock: the morally relevant issue is whether the cows or sheep we’re raising would suffer more, or less, or the same, if they were in the wild and being eaten by lions or wolves.

The lemma allows the possibility that some ways of treating animals may be immoral, but the lemma rules out the presumptive immorality of all cases of treating animals in such a way that their situation is no worse than they would face in the wild. In the case of hunting, this is clear enough. Anyone who knows cats knows that they love to keep their prey alive and toy with it before finally killing it, and this causes more suffering than would be caused by a quick kill with an arrow or a bullet. So human hunting causes less suffering than hunting by at least some other predators.

Could it be argued that by hunting deer, humans are causing suffering to lions and wolves by taking away their prey? This doesn’t look like a promising line of argument. Humans are hunters by nature, and it’s not clear why we would feel obliged to let other species of hunters have prey that we could have. A lion whose potential prey is killed by a human is no worse off than a lion whose potential prey is killed by another lion, and in either case the total lion population adjusts to the availability of prey for lions, with marginal lions always dying or otherwise failing to reproduce because of competition.

As we move from hunting to raising livestock, no important new issues of principle arise. Do farm animals suffer more or less than animals in the wild? It’s not clear that they suffer any more, and it seems likely that they suffer a lot less. The day-to-day life of a cow munching the grass and chewing the cud has less excitement than that of the wild ox, continually fearful of sudden attack by a predator, but I doubt that the cow would get a thrill from dangerous adventures the way some humans do. When death comes to the cow, it does not seem to cause any more suffering than death in the wild—and if we ever found out that it did, we could adjust our techniques of slaughter, without abandoning the practice of killing animals for food. My argument is not that all and any ways of raising and killing animals for food are morally acceptable, but merely that some feasible ways are morally acceptable, and therefore morality does not require vegetarianism.

Some people may feel that the life of an animal in the wild is in some way better than that of a farm animal, even though the farm animal experiences less actual pain and fear. Well, we observe, as real incomes rise, that there is a growing interest in both recreational hunting and in the demand for game animals, animals killed in the wild, in preference to farm-raised animals. The meat of game animals is leaner and tastes better. This trend is merely the tip of a broader movement towards free-range raising of animals. Suppliers of meat can charge more for meat that has been produced in a ‘more natural’ way, partly because of superior taste and partly because consumers feel better knowing that what they were eating was produced in a more natural way. As our incomes rise, we spontaneously move away from factory farming toward free-range farming, and then ultimately to preferring meat from animals that have been hunted in the wild.

If we accept the lemma, then the mere fact that some suffering occurs to animals when they’re raised for meat production is not enough to show that this is immoral. Instead, we have to show that they necessarily suffer more than they (or corresponding animals, which might be a bit different in a hypothetical alternative world) would suffer, if the human population were much smaller and the populations of lions and wolves much bigger.

Although I’m not offering arguments for the lemma, I do want to look at three possible ways of rejecting it. Someone could maintain that our obligation is simply to stop suffering wherever we can. One way to stop the suffering that comes from animals being harvested as prey would be to wipe out those animals. Thus, we could kill all oxen (including beef cows). At the same time, we would wipe out all the predators, the animals that would have eaten the oxen. This would mean wiping out virtually all animal species, including insects, birds, and fish, for all these animals are either predators or likely prey. Some folks would feel sad that all these species had disappeared, but they could console themselves with the thought that being extinct means you never have to suffer, whereas being extant means you do have to suffer.

Consistently, we should extend this to humans: they should be killed off, and then no human would ever suffer again. (Just to keep an eye on things and make sure everyone follows the rules, I’ll be the last one to go.) If allowing suffering is decisively immoral then every sentient living thing, including humans, should be made extinct, because this and only this guarantees no more suffering.

Another person might, however, approach the issue a bit differently. Instead of killing all animals, we could take over and manage the entire animal kingdom, transforming it into something very different from the way it has evolved, intervening with birth control drugs, factory-produced food, analgesics, and anesthetics. The former predators could be fed substitute foods made in factories from soybeans, or even directly from industrial chemicals. Since they would suffer somewhat from not being able to hunt, we would have to provide them with robotic imitation-prey, so that they could continue to experience the activity of hunting. Herbivores could be left to graze the wilderness, but fed fertility-reducing drugs to keep their populations stable. There would still be some suffering: accidents do happen, and every animal has to die, though we could try to limit this suffering by infiltrating the natural world with robots using analgesic and anesthetic dart guns, watching all the while for any impending pain or anxiety.

There are various aspects of this scenario which may not be very appealing. Be that as it may, it is not feasible right now, and won’t be feasible without a huge investment over many decades, if not centuries (think about the difficulty in ensuring that every fish in the oceans is guaranteed never to be eaten). So, even assuming that this ambitious intervention is morally required, we’re stuck for a while with the choice between a certain amount of suffering in the wild and a certain amount of suffering (probably the same or a bit less) down on the farm. And therefore, if we accept the lemma, we must reject the case for vegetarianism on grounds of the suffering caused by meat-eating.

Of course, most vegetarians will reject those two approaches and go for a third approach: simply have humans abstain from meat-eating. But what the lemma helps to bring out is that this option has an arbitrary quality. Turning humans into herbivores means excluding other herbivores from a large area of land, reducing the world’s populations of non-human herbivores. So the third approach is a kind of partial and inconsistent version of the first approach. Either we have an obligation to reduce animal suffering every chance we get, or we don’t have such an obligation. Eschewing the first two approaches means admitting that we have no such obligation.

We can kill animals for food without adding to the total net suffering in the animal kingdom, and this is morally okay.