Should we rid the world of carnivores if we could?

by Alexandre Erler

In a provocative piece for the New York Times, Jeff McMahan remarks that cruelty pervades the natural world: he stresses the vast amount of suffering and the violent deaths inflicted by predators on their innocent victims. He then invites us to consider a daring way of preventing such suffering and deaths: “Suppose that we could arrange the gradual extinction of carnivorous species, replacing them with new herbivorous ones.  Or suppose that we could intervene genetically, so that currently carnivorous species would gradually evolve into herbivorous ones, thereby fulfilling Isaiah’s prophecy.  If we could bring about the end of predation by one or the other of these means at little cost to ourselves, ought we to do it?” McMahan’s conclusion, which he describes himself as “heretical”, is that we do have a moral reason to desire the extinction of carnivorous species, and that it would be good to bring about their extinction if this could be done “without ecological upheaval involving more harm than would be prevented by the end of predation”.

As McMahan had anticipated, his piece has attracted quite a number of outraged comments. However, few of these can be said to offer objections that (excuse the pun) have any meat to them. Many actually raise points that McMahan himself considers, and replies to, in his essay. Among the weakest of these are the idea that implementing his proposal – even with the specific provisos he gives – would mean “playing God”, or acting “against nature”. Such lines of argument have a host of absurd implications: for instance, one might just as well use them to condemn morally innocuous practices like contraception or vaccination. They are therefore quite easily dismissed.

A slightly more plausible line would question whether McMahan’s proposal would really achieve its purpose: would we really reduce the total amount of suffering among nonhuman animals by abolishing predation? Would death after years of decay due to ageing really involve less pain for the animals concerned than the admittedly violent but comparatively short-lived attack of a predator? One might hesitate to give an answer here, though it still seems likely, in view of the multiple years that would be added to the lives of potential preys by the disappearance of their predators, that bringing about such an outcome would overall be in the interest of the former.

Yet another, more plausible objection, which McMahan himself considers, would appeal to the loss, both in terms of intrinsic and instrumental value, that the extinction of carnivorous species would represent. Wouldn’t his proposal mean depriving us of the joys of keeping cats and dogs as pets? (Dogs can admittedly survive without meat, but it's not clear that they can be strong and healthy without it.) And independently of the impact on our own well-being, wouldn’t the natural world be greatly impoverished and “watered down” if we brought about the extinction of carnivorous species to benefit the more “peaceable” ones?

I think there is something to this line of argument, and that McMahan dismisses it too quickly when he argues that the criteria for individuating species are largely arbitrary and that this casts doubt on the idea that species can have intrinsic value. Surely, if beautiful animals such as tigers or lions were to disappear completely, this would be a significant loss for the planet. Still, it is true that the badness of such a loss would have to be weighed against the benefit derived from the disappearance of predatory acts. And we might well conclude
that the interests of their potential victims outweighed the value, both intrinsic and instrumental, of having lions and tigers around. However, I think this objection still shows that if we could choose between the two alternatives presented by McMahan at the beginning of his essay (which he doesn’t make much
use of later on), we should choose to genetically modify carnivorous species to make them evolve into herbivorous ones, rather than causing their extinction, as at least most of the value under discussion would thereby be preserved. Assuming the idea of a vegetarian lion or tiger is by definition impossible, we
could at least produce very similar animals (“vions” and “vigers”?) that would have the beauty and majesty of lions and tigers without the cruelty. (I am not a utilitarian, but it seems to me that utilitarians can agree with this
suggestion, if only for the sake of promoting the well-being of people who love tigers and lions.) Furthermore, the complete disappearance of carnivores might actually not be necessary to put an end to the suffering and death caused by predation. Indeed, in vitro meat, an ethical alternative to meat grown in the lab, might soon be a reality, which would at least allow us to keep some of our favorite meat-eating pets – Doberman Pinschers and Rottweilers included. More generally, we could still have carnivorous species around provided that they did not need to hunt other sentient creatures.

Obviously, as McMahan is well aware, the most formidable obstacle to realizing his idea lies in the huge ecological imbalance that this might produce. Were we to sterilize all predators (which would already pose a major technical challenge), the population of herbivores would likely soon explode unless we could embark on a
large-scale program of population control that would be very difficult to carry out to say the least. At the level of concrete action, it seems much more urgent and realistic to focus on reducing the huge amount of needless suffering and death that we are inflicting ourselves on animals. Nevertheless, McMahan’s essay is a thought-provoking one, and makes a surprisingly plausible case for the desirability, if not of the complete extinction of carnivorous species, at least of a drastic reduction in their numbers together with the “conversion” of most of them into herbivorous, or at least non-hunting alternatives. The reactions to McMahan’s piece also reveal how deeply rooted the prejudice that we shouldn’t go “against nature” seems to be – a prejudice that arguably thwarts progress in a number of different domains of research today, such as the fight against ageing.

  • Facebook
  • Twitter
  • Reddit

26 Responses to Should we rid the world of carnivores if we could?

  • ledge says:

    The thought occurs that this is a reductio ad absurdum against vegetarianism…

  • Matt says:

    I’m not sure we should be so quick to dismiss the hesitation to “go against nature” or “play God.” For one thing, these arguments (when made carefully) might not overgeneralize as much as you seem to think they do; for example, relatively few people think that vaccination is immoral, though a lot think that human cloning might not be moral.

    Further, consider a scenario in which we (humanely) replace all non-human animals with really accurate robot replicas. Assuming the robots we use as replacements cannot feel pain, we would have completely ended animal suffering, while preserving all the superficial beauty of the natural world. There would still be things that looked and acted like tigers and lions, for example. There are probably a number of reasons to think this is a bad idea, even if feasible, and we might think that one reason is that it’s an impermissible tampering with “nature,” or “playing God.” Just to be clear, I don’t mean to endorse this style of argument, I just want to emphasize that perhaps we cannot dismiss it so easily.

  • Nick says:

    Another reason would be that a world of robots would offer no pleasure among its robots.

  • Matt: if all the objector to your hypothetical scenario could argue is that it would mean “playing God” or “tampering with nature”, I think this would constitute a very weak objection indeed. What would be bad about your scenario, it seems to me, is that the “animals” that would succeed the extinct ones would not be real animals. One might perhaps say that they wouldn’t be “natural” animals, but I would find this very misleading. A leopard, say, that had been produced by cloning, but was otherwise exactly the same as its “natural” counterparts (let’s assume it had been raised among them by its clone’s mother, etc.) would presumably not count as a “natural” leopard, yet it seems to me that it would still have just the same value as the natural ones. Your scenario would lead to a loss of value in the world because it would involve the disappearance of real, conscious, flesh and blood animals in favour of (presumably unconscious) machines, not because it would mean tampering with nature.

  • Matt says:

    Alexandre, you’re right that there’s a lot going on here, and perhaps many reasons to think pursuing the scenario I suggested is inadvisable; like I said, there are many reasons to think this is a bad idea.

    If I may, let’s consider the question in a stark form: do you think that there is any sense at all to objecting to an action because it would go “against nature”? I don’t find it at all obvious (as you seem to) that there is no sense to this kind of objection. Here’s another thought experiment: suppose that scientists told us that in order to save a very rare primate species from extinction, the only option would be for some humans to volunteer to be surrogate mothers for this rare primate. That is, humans would give birth to a non-human primate. (Let’s assume here that all the consent would be genuine, that the volunteers would not disproportionately be from certain demographic groups, etc.) It’s not immediately clear to me what to think of this, morally. But if someone said to me that it should not be pursued because it was “against nature,” I certainly would not find it obvious that this was wrong. I would want to know a lot more about what was being claimed before I would say whether I agreed, of course, just like with any other ethical slogan.

    One further thought: My sense of your post is that you think it’s a particularly conservative maneuver to appeal to an action’s being ‘against nature.’ Hence your examples of vaccination and contraception, opposition to which is more common among conservatives than liberals. I would think that this appeal, in the sense relevant to McMahan’s piece– i.e. going against nature as a whole, rather than human nature in particular– is actually more common in certain liberal circles. Think of the Gaia movement, for example.

    This has been a really interesting discussion, thanks!

  • Anthony Drinkwater says:

    Interesting discussion !
    I’m inclined to think, however, that far from talking about “playing god”, we are actually discussing “playing man”.
    McMahan’s text is in fact completely anthropocentric. The species he mentions are ones that most resemble man : they run, they hunt, they play, they have body processes similar to our own, they exhibit an aesthetic of form, muscle and behaviour with which we (naturally?) empathise. They are an extension of the concept of domestic pets, equally cute but much more risky.
    Where do bacteria, viruses or slugs come in this desire to eliminate suffering ? To say nothing of the DNA that drives us.

    Why “playing man” ?
    Because man, having created god in his own idealised likeness would, if we implemented McMahan’s plan, be commiting the same category error in deciding that certain life forms which most resemble him are worthy of special attention or manipulation.
    Of course, man has always manipulated “nature”, but to justify manipulation for ethical reasons (“for their own good”, we might say) strikes me more as misplaced arrogance, if not a new form of colonianism.

  • Well, now, and I thought I’ve seen the worst of it at my last vegetarian vs. meat-eaters debate as far as fringe ideas go.

    A few quick points that I think fatally undermine the position of McMahan:

    (1) Yes, herbivorous are suffering, but no they cannot be said to be innocent victims, at least in the relevant respect. They are not being unjustly subjected to pain and death, for the very simple reason that considerations of justice do not extend to the interaction between predator and prey (when the predator is a non-human animal). At bottom there is no reasoned moral imperative to prevent the suffering of a deer being taken down by a pack of wolves. Sure, it elicits empathy, but that carries little weight without some substantive support from moral reasoning. Why, exactly is it wrong for a wolf to take down and eat a deer? Is it because the deer will suffer? Well, then suppose the wolf is out of the question. The deer continue to reproduce unchecked and inevitably devastate their feeding grounds. Orders of magnitude more deer die of starvation and suffer as a result. Is it not more “humane” to let loose the wolves and inflict some suffering in order to avoid far greater suffering?

    Furthermore, if its the suffering of the deer that truly motivates us (it is the decisive moral factor) should we not vaccinate the deer against parasites and diseases, clear and pave their migration routes to avoid injury, provide painkiller at birth, prevent mating rituals among males? The list goes on…

    Clearly it is not simply that the deer suffer that matters. As when we consider other human beings, it is whether or not the suffering is just or unjust that matters, and as mentioned above, considerations of justice do not seem to extend to interactions between predator and prey (at least where the predators are non-human animals).

    (2) In the overwhelming majority of cases predators take down prey for a very specific reason: to survive. Without getting into it, a quick argument will legitimately recast this as acting in self-defense or the preservation of oneself. Few, except for the most hard-core pacifists, reject the notion that self-defense or ensuring ones own preservation by the minimally effective means is a legitimate reason to harm another. Predators do not kill on a whim, or with abandon, or for the pleasure of it. They kill so they can eat and survive. If anything, wild predators are being courteous enough to allow the prey to escape, unlike humans slaughtering livestock.
    Even if we were to extend moral considerations to predator-prey interactions, we would have to agree, it seems to me, that the predators have a good and decisive reason for doing as they do.

    At bottom I see no moral force behind a proposition to get rid of predators.

  • Peter Wicks says:

    Surely the key question here is precisely to what extent we should think of moral considerations – whether we talk about them in terms of justice, rights, avoidance of pain or whatever – as extending to non-humans. It seems to me that McMahan’s proposal rests on the assumption that we should: otherwise, why should we care whether animals are experiencing pain? As far as I can see Alexandre doesn’t address the possible objection that we shouldn’t. By contrast, this seems (on a fairly cursory reading admittedly, so apologies if I’ve got this totally wrong) to underpin Dmitri’s first objection.

    As always, my own starting point is that there cannot be an absolute right or wrong answer to this, or any ethical question. You have to start from some kind of choice about what you consider morally relevant or not. In response to Dmitri, though, I don’t think “justice” is the only possible ground for taking moral positions. We could, for example, decide that we value animal welfare and disvalue animal suffering. Nor do I think such a point of view is as “fringe” as Dmitri suggests, although for sure it’s limited (mainly) to people living in urban areas with Western standards of living.

    In any case I think this is a fascinating discussion, and if not quite the “reductio ad absurdum” that ledge suggests, at least helps to put into sharp relief some quite fundamental issues in relation to vegetarianism, Jainism and non-human rights (with relevance also to abortion/infanticide debates).

  • Peter Wicks says:

    Actually I should nuance what I just said. The Jainist and (in less extreme form) Buddhist religions show that concern for animal welfare is NOT limited, and perhaps not even mainly limited, to people living in urban areas with Western standards of living.

  • Simon Rippon says:

    I think the first commenter has the wrong -ism in mind; McMahan’s article is a reductio of *utilitarianism* rather than *vegetarianism*. McMahan assumes that because suffering is bad for the animal that suffers, a world in which it occurs is all things being equal worse than a world in which it does not occur, and therefore we always have a moral reason to prevent animal suffering. But it isn’t obvious that his entailment holds.
    Retributionists should be able to see this clearly: suppose a possible world W1 contains only two people, A and B. And suppose A does horrible things to poor defenceless B, and gets a lot of pleasure out of it. Now, suppose that W2 is exactly like W1 except that after A tortures B, some horrible natural calamity happens to A. Many would call this “just desert”, and think that W2 is a better world than W1, though it contains additional suffering.
    You don’t need to agree with the retributionists that suffering per se can make a world better overall to deny McMahan’s argument. You might just think that some suffering is value neutral in the impersonal sense. That is, what is bad for some animal may not make the world any worse. The world is, arguably, no worse for being a place in which there are natural predators and prey (even though it is certainly bad for the prey that they are preyed upon). And therefore, we have no moral reason to do anything to change it.

  • Peter Wicks says:

    Simon: I’m not sure how seriously to take your suggestion that McMahan’s article is a reductio of utiliarianism. I certainly don’t agree with it! Utilitarianism, as I see it at least, is a worthy attempt (or rather class of attempts) to build an ethical framework that, if implemented, would lead to greater happiness/well-being for some class of being, with sentience generally playing a determinant role with regard to which beings belong to this class, and what weight we give to their welfare.

    As such, I don’t think utilitarianism has any problem handling this case. First we can question the extent to which animals that are typically preyed on by predators are sentient. Then we have all the other objections raised by Alexandre, notably the ecological imbalance concern, to which I would also add an objection along the lines of, “Don’t we have anything better to do?” like trying to avoid civilisational collapse. What is nevertheless left of McMahan’s position, from this utilitarian (and pragmatic) standpoint, is a recognition that, other things being equal, we would prefer a world without suffering, even the suffering of animals.

  • Matt: I don’t want to say that the “going against nature” objection simply has no meaning. What I want to say, however, is that it needs to be spelt out further if it is sometimes to be useful at all. If we were to chop down all the trees in a national park and cover all the area with residential blocks made of concrete, it would probably be correct to say that this is wrong because it is “against nature” – but a more accurate way of putting it would be to say that we have destroyed a natural reserve and that much value has been lost in the process. To just say that sth. is “against nature”, and to leave it at that, has no force as it seems to entail the assumption that nature is good and that whatever goes against it is bad, which fatally leads to the absurd implications I have mentioned.

    Dmitri: as I understand the notion, every being whose interests are harmed when (s)he has done nth. to deserve it is an innocent victim. E.g. people (or animals) who are killed by floods or earthquakes can therefore be innocent victims in this sense, even though no one may have acted unjustly to bring this about (I assume that floods and earthquakes are not agents and that they are often beyond human control). The example of natural catastrophes seems to me to raise problems for your idea that “it is whether or not the suffering is just or unjust that matters”, as it appears to imply that we should not prevent the suffering that such events cause if it has not been unjustly brought about (which I take to be the case much of the time at least). I don’t think you want to endorse that conclusion.

    Simon: some suffering might be value neutral in the impersonal sense, but it seems to me that the intense suffering and violent deaths experienced by preys at the hands of predators is a rather unlikely candidate for such a description. I assume that these animals are not moral agents; therefore, they cannot deserve to suffer and this cannot be a case of “just desert”. But surely suffering is prima facie bad, and in the case that interests us there doesn’t seem to be anything that defeats this prima facie badness. If so, how can one deny that the world is made overall worse by the presence of that suffering? If you want to say that it is indeed made overall worse when humans are the victims of accidents or other calamities, but deny that it is so when nonhuman animals appear to experience suffering of similar magnitude, then you open yourself to the accusation of speciesism.

  • Matt says:

    Alexandre: Thanks very much for your reply. I think perhaps we don’t disagree that strongly on the status of arguments from ‘nature.’ My concern was to emphasize that the relationship between man and nature is complex and may give rise to a variety of moral demands; dismissing the very common way of objecting to something as “against nature” seemed to me likely to obscure those demands.

  • Simon Rippon says:

    Peter: No, what’s left after the objections to MacMahan’s argument are accounted for isn’t a *preference*. What’s left is the claim that: If sentient animals could be relieved of *any* kind of suffering without generating equal suffering of another kind, or reducing pleasure by an equivalent amount, then we would have *moral reason* (or, an even stronger claim, we *morally ought*) to relieve them of it. I regard this conclusion as absurd, and therefore a problem for utilitarianism. My objection is not quite a “demandingness” objection if the claim is restricted to the moderate one that we have moral reason to intervene, rather than the stronger claim that we morally ought to. But I do think there is a massive over-generation of moral reasons here.

    Alex: I am claiming that not all animal suffering counts morally in the same way that human suffering does. If that’s what you mean by speciesism, then I plead guilty as charged. (I am not claiming, of course, that no animal suffering matters morally.) Note that utilitarians are themselves equally committed to “sentientism”: why, we might ask, should the lives of only sentient creatures matter morally? A potato(we may assume) doesn’t feel pain or pleasure, and so its being “preyed upon” simply doesn’t count morally, according to the utilitarian. The world, according to the utilitarian, is no worse for having one potato less in it. But wasn’t being “preyed upon” bad for the potato in much the same way (curtailing of the life it would have lived) as the way in which being preyed upon is bad for animals – unless we simply assume that only sentient lives count morally? What’s the moral justification for that assumption?

  • Andrew Watkins says:

    There is an underlying current of anthropocentrism in the suggestion that we should intervene to produce a world which is more just in our terms.

    Apex predators play an important role in maintaining the health of a population. It has already been pointed out that without predation numbers might expand, creating a more miserable death for more at some later time, but in a biological sense perhaps a more important function of the apex predator is to take out the weak, slow and sick before they do too much breeding. Predators tend to go for the easy targets, which has value in a population genetic sense and may indirectly increase the sum total of animal happiness/utility.

    The other issue is of human hubris. We see the world through our own values and needs and have a remarkably poor understanding of how it works. One only has to look at the history of introduction of feral pests such as the cat, fox, cane toad, rabbit and white man into the environment of Australia to see that we are rather more like the Sorcerer’s Apprentice than any more sentient being!

    Why not just leave it alone if we don’t understand it?!

  • Peter Wicks says:

    Andrew’s point is important and I have a lot of sympathy with it, but (i) I think it’s at least partly addressed by Alex’s original post, and (ii) to some extent it misses the point, namely not whether this is actually a practicable suggestion (it isn’t) but rather whether, as Simon puts it: “If sentient animals could be relieved of *any* kind of suffering without generating equal suffering of another kind, or reducing pleasure by an equivalent amount, then we would have *moral reason* (or, an even stronger claim, we *morally ought*) to relieve them of it.”

    Simon regards such a conclusion as “absurd”, but I don’t understand why, even in the stronger version (“morally ought”). If we could really do this “without generating equal suffering of another kind”, then why don’t we have a moral reason, or even a moral obligation, to do so? In other words, why shouldn’t we regard utilitarianism’s “sentientism” as a sensible compromise between pure speciesism (ignoring totally the suffering of other animals, even the most sentient of them) and full-scale Jainism or (a different version of extreme non-anthropocentrism) deep ecology?

  • Marco Antonio Oliveira de Azevedo says:

    “I regard this conclusion as absurd, and therefore a problem for utilitarianism.” I totally agree with Simon. I think we should take McMahan’s very stimulating, but also awkward, paper (probably contrary the author’s intentions) as a reduction ad absurdum argument. Admitting a place for intuitions in moral argumentation, we can take McMahan conclusions as intuitively absurd. My point then is that we should rethink the utilitarian conceptions of “duty”, “ought” and “moral obligation”. We, rational persons, actually don’t have any kind of general moral obligation of reducing the global suffering in the world. Considered only in terms of pain and pleasure, *suffering* doesn’t have a clear *moral sense*. In other words, it is still not a relevant moral category. Then, even “if we could really do this ‘without generating equal suffering of another kind’”, we still not have any weak or strong moral reason for action. Anyway, fortunately utilitarianism is not the only persuading moral theory.

  • Peter Wicks says:

    Marco: I agree that utiliatarianism is not the only persuasive moral theory, although it happens to be the one that I find most persuasive. There are other choices. What I’m struggling with is statements such as “we can take McMahan conclusions as intuitively absurd”, as if we all necessarily share the same intuitions. Personally, I don’t find McMahan’s conclusions absurd. I do find them flawed, but for the reasons I have given above, all of which are perfectly consistent with a basically utiliatarian view.

    Let’s perhaps pause here to recall what a reductio ad absurdum actually is. It is a logical proof that starts by assuming the opposite, and then demonstrates *logically* that this leads to a contradiction. Simply claiming that something is “intuitively absurd” does not constitute logical proof. In fact, it is just the same thinking that led people in a previous century to insist, ignoring the evidence, that the earth is stationary and the sun and stars revolve around it.

    To show that this case exposes a fatal problem for utilitarianism it would be necessary to show that a utilitarian perspective leads to conclusions that the utilitarians themselves would not be willing to follow in practice, were such a situation to arise. This does not appear to be the case, and I am not aware of any other case that utilitarianism, if applied appropriately, cannot address adequately. That some people don’t like utilitarianism as a moral framework, and/or find the conclusions it leads to absurd, and/or mistakenly believe that it leads to conclusions that they find absurd, proves only that when it comes to ethics there are choices to be made, and not everyone will necessarily make the same choice. Unlike in science, neither logic nor evidence imposes such a choice: it is one that we have to make ourselves.

  • Anthony Drinkwater says:

    I’m not sure that the McMahan proposition argues for or against utilitarianism, or any other ethical theory, whether it be explicative or normative.
    Doesn’t the central question centre rather on the extension, or universe of application, of a theory rather than its content ?
    From this point of view, the debate hinges on moral agents and the objects of their actions (or non-actions). McMahon’s argument seems to me to rest on an assumption that predators are moral agents and that man should somehow wish to control their behaviour. As if I should be obliged to stop my cats from chasing birds (which they do not do mainly in order to survive, Dimitri).
    I agree of course that I, as a moral agent, should not behave like my cats, but this is far from imposing my view on other species, nor does it imply that I should seek (genetically or otherwise) to modify their behaviour.

  • Peter Wicks says:

    Anthony: I don’t quite agree that McMahan’s argument rests on the assumption that predators are moral agents. I do agree that his proposition doesn’t argue (at least not convincingly) for or against utilitarianism or any other ethical theory. This being the case, we could choose an ethical standpoint (I prefer this term to “theory” since ethics, unlike science, is not a search for the truth) in which we simply want to reduce the suffering of animals, irrespective of whether we regard them – or their predators – as “moral agents”.

    I’d also like to come back to the comment Simon made earlier in which he emphasised the distinction between *preferences* and *moral reasons*. Somehow I think this goes to the heart of our different positions on this issue. If ethics is, as I claim, a matter of choice, then ultimately it is also a matter of preference. That doesn’t mean all preferences are moral preferences (vanilla vs chocolate clearly isn’t), and we may choose ethical frameworks that give us “moral reasons”…but these are based ultimately on choices – and therefore preferences – regarding the underlying framework. All this is a bit abstract I know but the idea that ethics involves choice rather than truth seems to me to be rather important.

  • Marco Antonio Oliveira de Azevedo says:

    Peter. I must acknowledge that you are right concerning “reductio”. Hence I can’t say that my suspicion that McMahan’s conclusion is absurd can be used to reduce McMahan’s assumptions to a falsity. Then, let me try other direction: let’s take McMahan’s suggestion seriously.

    Well, his first question was: “If we could bring about the end of predation by one or the other of these means at little cost to ourselves, ought we to do it?”. It should be noted here that we are not talking about a real practical decision. McMahan himself actually said: “I concede, of course, that it would be unwise to attempt any such change given the current state of our scientific understanding. Our ignorance of the potential ramifications of our interventions in the natural world remains profound. Efforts to eliminate certain species and create new ones would have many unforeseeable and potentially catastrophic effects.” Hence, we shouldn’t make efforts to eliminate predators creating new “benign” species now ONLY by precautionary reasons. But it doesn’t mean that we also cannot stimulate researches. Hence, since we still don’t have any safe technologies, why not stimulate scientific researches? Why don’t encourage private or even public financial support to those experiments? It is plainly possible to design safe experiments in genetics within the limits of sensible precautionary principles. Let’s assume that it is possible. Would it be a wise decision? Ought we to defend public financial support to those experiments? If we conclude that we have other public priorities, it doesn’t mean that it cannot be considered as a wise police for some private groups. Anyway, my suspicion still is that there is something wrong with that, not because it would be wrong to develop technologies for creating new non-predatory species. It is not a matter of *absolutely wrongness*. Creating new predatory species is wrong not because there is a moral principle against it. It is wrong, first, besides the fact that we don’t have any private or public reason for doing that, also, and second, because it is based on a misunderstanding on what counts as goodness for wild animals. My main disagreement, then, is on the very conclusion that “if we could bring about the end of predation at little cost to ourselves, we ought to do it”, for here we are simply assuming that the end of predation would be GOOD for predated animals.

    Anthony is right. McMahan’s utilitarianism rest on an anthropomorphization of animal needs and well-being (Anthony has called it an “anthropocentric” view, but I think the flaw is not in “centralizing” morality in us, humans; it is a problem on what really matters to animals). Hence, my reference to an intuition that McMahan’s conclusion is an absurd (it was my interpretation of Simon’s comment) must be read as the conclusion that McMahon’s suggestion cannot be taken seriously, since we are not thinking appropriately on animals’ well-being. Animals’ well-being (I’m referring obviously to wild animals) depends on natural conditions, including ecological ones. In the natural environment, what is good for an animal cannot be dissociated from what is good for his species. And what is good for natural species is a resultant of natural selection, including “ecological balance”. Animal goodness after all is very different from human goodness (See here the very interesting thoughts of Philippa Foot in her “Natural Goodness”).

    PS: On the “aesthetic loss argument” (one of Alexandre’s point). Alexandre said that: “replacement of all non-human animals with really accurate robot replicas” could be thought to someone as a solution to the “aesthetic loss” problem. If goodness and badness were only a human subjective problem, then the fact that animals are actually ORGAs, not MECHAs, would not be a problem. But if goodness and badness are instead OBJECTIVE states (as I think they are, for I am an objectivist), then the problem of animal suffering obviously cannot be solved by substituting animals for robots (for in this case we would simply extinguishing the subjects and substituting them by other kind of beings only for OUR *pleasure*). Animal suffering, if it is a real problem at all, is not a problem analogous to persons’ and animals’ predicament of suffering of incurable diseases or deficiencies. Even if it would be feasible, substituting animals by robots would not be a solution FOR animals. Alexandre is right: the “animals” that would succeed the extinct ones would not be REAL animals (they would be *real entities*, although of another kind of *species*).

    On Dimitri’s point if what matters is whether or not the suffering is just or unjust. I don’t think that McMahan’s point is one of justice. Even concerning humans, we can have a lot of reasons for action outside the field of human justice. We can have reasons for doing good for animals without being fair or unfair to them. It is not justice here what matters. Anyway, we could develop an approach where justice would matter. Our question would be: Would deer have, for example, some right against us, humans, of being protected from their predators? Would that be a *natural right*? Anyway, nothing prevent us of giving them that claim (of course, we should be their surrogates in that case). But in this case we still would be embarrassed with the same question: would “being protected from their predators” something that promotes the natural well-being of deer? And what about the goodness of predators, for we would be creating rights to deer imposing harms on them. Of course, we cannot give rights to deer by imposing duties on predators (they are not a kind of subject that can bear duties – only rational beings can bear duties). Nevertheless, predators in that case would be innocently harmed simply because we prefer to protect the deer following our own idiosyncratic feelings.

    Anyway, my last point is this: even if it is true that choice and preferences have a place in ethical decisions, ethics is not ultimately a simple matter of choice or preference; ethics is at least partially a matter of reason and truth. And the problem of what is good or bad for some entity, natural or artificial, is primarily a matter of what is or is not the case.

  • Alcino Eduardo Bonella says:

    For me it seems that McMahan have forgotten completely that many predators could be “persons” in the tecnical or philosophical sense of the word. Even do, if predator would be mere “changed” in certain aspects, dietry aspects for example, the “person problem” (the inerent value of persons and the no substitution argument) does not apply. But it applies to extinction program, isn’t?

  • Peter Wicks says:

    Marco, thanks for this comprehensive reply. I’d like to make four points.

    1. The issue of what is “good” for a living organism is an interesting one. I tend to take the view that, prior to the development of human culture, the teleological nature of a living organism lies in our tendency to try to pass on our genes to future generations. So what is “good” for the organism is just what helps it to pass on its genes. Once language and culture emerge, the teleological nature of (in this case) humans changes: I am not a slave to my genes; what is good for me is something that I must decide.

    2. “Animals’ well-being (I’m referring obviously to wild animals) depends on natural conditions, including ecological ones.” If we take the above definition of what is “good” for a living organism, the animals’ well-being really depends on natural conditions to the extent that those conditions help the animal pass on its genes. What is good for the individual is not necessarily what is good for the species, except where the individuals concerned are genetically identical.

    3. “Ethics is at least partially a matter of reason and truth.” I would not disagree with this. It is possible (indeed usual) to have ethical positions that are logically inconsistent and/or based on false assumptions about reality, and precisely (I would suggest) it is the role of ethics to expose such imperfections and remove them, to the extent possible. My claim is only that reason and truth *alone* cannot determine our ethical choices.

    4. This being the case, it is not clear to me that the charge of “anthopomophism” renders an ethical system (such as utilitarianism) illegitimate (or “absurd”). We may decide, basically as a (human) cultural value, that we want to maximise pleasure and minimise pain in sentient creatures, where sentience is taken to depend on some measure of sophistication of brain function. That is a choice, and not (to my mind) an absurd one, even if we finally decide against that choice.

  • Sasha says:

    Hey Alex, as you might imagine I’m sympathetic to this argument. But I’m also concerned that it’s too far off becoming a possibility for it to be worth investing time in. Most of the critique on the NYT piece is angry variations on ‘it’s just impossible’, which suggests that a much better way than philosophical discussion to win support for the idea might just be proof-of-concept.

    It occurred to me that all violent death is not equal, though. Consider going after just the critters, for eg:

    Ichneumonid Wasps: ‘When the wasp hatches from the egg it starts to eat the body of the caddis fly larva. However, it does no eat the nervous system, as it does not want to kill the caddis larva. Only once the caddis larva has fixed its case to a stone to prepare for pupation will the wasp kill it by eating all the caddis larva.’

    Sacculina Carcini: ‘The female Sacculina punches numerous holes in the neutered crab’s body so that male barnacles can get in and fertilize its eggs. Once that’s accomplished, the tendrils take over the crab’s nervous system, making it autonomously guard, care for and clean the parasitic egg sac as if it were its own. And once they hatch, a repeat performance is in order.’

    Killer Whales: ‘As much as six hours may pass from initial attack to kill with ramming, biting, pulling on the pectoral fins, and attempts to separate the mother from the calf.’

    (this one might be a less popular, but it meets the criteria) Domestic cats: ‘We admire their magnificent beauty, strength and agility. But we would regard their notional human counterparts as wanton psychopaths of the worst kind.’

    In theory, it might be possible to start by removing just the most torturous and inefficient hunters from the ecosystem and allow the populations of quicker killers to expand correspondingly. It would still have profound and unpredictable effects on the local ecology, but much less so than removing *all* the predators. It would also reduce the associated risk of unchecked herbivore populations wiping out all the local flora.

  • Peter Wicks says:

    One thing I find interesting about Sasha’s comment is that it makes the whole thing seem so much more real. While McMahan’s initial idea seems so remote that it hardly engages the emotions, Sasha reminds us that there are actually steps we can take in that direction. I shan’t comment further now on whether that would be a good idea or not, except to mention that from the “sentientist” point of view you’d want to take account of the neurological sophistication of the prey and not only the efficiency of the kill.

  • Sasha: the cat example does indeed show what sacrifice would be involved if we could implement McMahan’s proposal. As a cat lover, I would find it terrible to ensure that cats go extinct, and I would hope we could rather find a way of removing their natural urge to hunt other sentient creatures. The only alternative I’m aware of to ensure that cats don’t hunt is to confine them indoors, which seems questionable from an ethical perspective.

    Regarding your other examples, as Peter says we will surely want to take into account whether or not the predator’s victims have the capacity to feel pain. As far as I can tell, this does not yet commit us to “sentientism”, as Simon calls it, if the term means that only the lives of sentient creatures matter morally. The view I hold (I don’t know what utilitarians will say on this issue) is that the lives of non-sentient creatures also have some degree of moral significance, which is why I would regard it as wrong of someone to go around killing fish (assuming they are not sentient) or even insects just for the fun of it. However, it seems plausible to think that the capacity for sentience gives us an additional moral reason not to harm the animal in question, and thus that the interests of sentient creatures matter more morally than those of non-sentient ones. If so, we have a stronger reason to prevent harm to sentient than to non-sentient animals when we can.

Recent Comments

Authors

Affiliations