The problem of animal suffering has become an area of some focus within the philosophy of religion in recent years. Michael J. Murray's Nature Red in Tooth and Claw (Reference Murray2008) proposes a number of different theistic responses, while Trent Dougherty's The Problem of Animal Pain (Reference Dougherty2014) argues that animals, by which I shall mean non-human animals throughout this article, are capable of morally relevant spiritual development through suffering. In the theological literature, Christopher Southgate's The Groaning of Creation (Reference Southgate2008) suggests that evolution, with the attendant suffering, is the only way God could have brought about the beauty and diversity of the natural world, while Nicola Hoggard Creegan's Animal Suffering and the Problem of Evil (Reference Creegan2013) uses a biblical-based analogy to defend a similar position to Southgate's. However, I believe one important line of response has not been touched upon in the contemporary literature. In this article, I present a new defence of classical theism against the problem of animal suffering. Specifically, I argue that cases like William Rowe's celebrated ‘Bambi’ case (Rowe (Reference Rowe1979), 337) do not provide evidential support of the kind required to render the evidential problem of evil a compelling argument against classical theism, in brief, because the rational theist can simply deny that such cases ever occur.
The evidential problem of evil and the Bambi case
William Rowe (Reference Rowe1979, 336) famously argues as follows:
Argument 1
P1.1 There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse.
P1.2 An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse.
From this, we may conclude that there is no omnipotent, omniscient, and wholly good being. For what follows, I will take the intended target of the problem to be a generic Abrahamic monotheism, a notion of God abstracted from the religious traditions of Judaism, Christianity, and Islam. That is, a God who is omnipotent, omniscient, omni-benevolent, eternal, personal, and interventionist. I shall throughout refer to such a being simply as ‘God’ and to the thesis that such a God exists simply as ‘theism’.
P1.2 is largely accepted by all sides. P1.1 is the controversial premise, which is usually taken to require evidential support. Rowe appeals to a very cleverly chosen example of apparent evil in order to support P1.1.
Suppose in some distant forest lightning strikes a dead tree, resulting in a forest fire. In the fire a fawn is trapped, horribly burned, and lies in agony for several days before death relieves its suffering. So far as we can see, the fawn's intense suffering is pointless. For there does not appear to be any greater good such that the prevention of the fawn's suffering would require either the loss of that good or the occurrence of an evil equally bad or worse. Nor does there seem to be any equally bad or worse evil so connected to the fawn's suffering that it would have had to occur had the fawn's suffering been prevented. (ibid., 337)
The ingenuity of the fawn case (since baptized ‘Bambi’) lies in the fact that it seems, under very plausible assumptions, to escape the ready-made explanations provided by the free-will and soul-making theodicies.
The free-will theodicy holds that free will is a good so great that God wants humans to possess it, but it is impossible (even with the help of an omnipotent being) for humans to possess genuine free will without also possessing the capacity to cause evil. God, then, would reasonably allow evil to exist in the world, if it is the result of intentional human action.
The soul-making theodicy holds that God would want his creation to suffer in this world because only by so doing can they fully develop their eternal souls. I find the soul-making theodicy most plausible if we link it to virtue ethics. The best thing humans can aim at, on this view of morals, is to first attain and then maximize the virtues. When we reflect on the virtues, we note that there are some which are such that it is impossible (even with the help of an omnipotent being) for a human to have, or at least maximize, without enduring some evil. Plausible examples include fortitude, moral rectitude, perseverance, hope, bravery, and so on. The thought is that these virtues are not mere dispositions. Rather, necessarily, one only has a virtue like perseverance if one has actually persevered through something. The possessing of some such goods is so beneficial for humans, according to this version of the soul-making theodicy, that God justly allows the requisite amount of evil to befall humans in order that they might have the opportunity to gain these goods, or suitably high levels of these goods. The balance between suffering on one hand and the opportunity for achieving the virtues that come with suffering on the other hand (after we have factored out cases of moral evil resulting from free will, if the two theodicies are run together, as I think they should be) is optimal. Or so this version of the soul-making theodicy holds.
Bambi, however, does not have a soul of the kind which can improve through the achievement of virtues only accessible through suffering. At least it is plausible to think not, and this view is fairly common among the religious traditions associated with classical theism. Moreover, it is plausible to say that Bambi either does not manifest virtues or, if he does, they are not of sufficient moral significance to justify the terrible evil done to him. Finally, Bambi has no free will. So, it seems, the evil done to Bambi cannot be explained either as a necessary consequence of Bambi having free will or as the best way of giving Bambi the opportunity to manifest virtue.
It is noteworthy that the soul-making theodicy can, with some degree of plausibility, be extended to explain some instances of animal suffering, even under the assumption that animals cannot build a morally relevant character through suffering. Specifically, humans are capable of using suffering to develop good moral characters in two different ways. Humans can develop certain virtues (like perseverance or fortitude) by responding in certain ways to their own suffering, but additionally, humans can also develop virtues (like empathy, for example) by responding in certain ways to the suffering of others. Defenders of the soul-making theodicy might, then, explain animal suffering by the opportunity it provides humans to develop certain sorts of very important virtues through interaction with the animals that are suffering.Footnote 1
Although various objections might be raised against either of these defences, both have many defenders both historical and contemporary and the purpose of this article is to consider whether a problem which threatens to evade either of these responses can be answered. The rest of this article, therefore, will adopt the working assumption that free will and the soul-making, taken together, can account for all the evil that is visible to humans and all the evil, visible or invisible to humans, that results from intentional human actions.Footnote 2
Tollensing the ponens: miraculous animal non-suffering
There is plenty of compelling evidence of the unseen deaths of animals. Bones, fossils, and by-products of these (petroleum, for example) provide, I believe, good evidence that animals have been dying in natural occurrences for many hundreds of millennia, in fact well before the first humans developed. My objection to versions of the problem of evil that turn on cases involving unseen animals is not scepticism about the deaths of such animals but scepticism about the suffering of such animals.Footnote 3
Scepticism about animal suffering is not a new position. It is famously associated with René Descartes. Although Descartes's view that animals lack an internal mental life, or at least a mental life sufficiently developed to attribute morally relevant pleasures and pains to them, has long since gone out of fashion among philosophers, it has some contemporary defenders, notably Murray and Carruthers.Footnote 4 Nevertheless, I am not advocating neo-Cartesianism. I think we have plenty of evidence that animals have subjective mental states. Rather, I wish to defend the so-called ‘Moorean shift’ as a rationally justified response to cases of unseen animal suffering. So, whereas the proponent of the evidential problem of evil may reason as follows:
Argument 2
P2.1 If there is a God, then animals wouldn't suffer when there is nobody around to see them or cause their suffering.
P2.2 Animals (sometimes) suffer when there is nobody around to see them or cause their suffering.
C2 There is no God.
I hold that the theist may with equal justification turn this reasoning on its head:
Argument 3
P3.1 If there is a God, then animals wouldn't suffer when there is nobody around to see them or cause their suffering.
P3.2 There is a God.
C3 No animal suffers when there is nobody around to see it or cause its suffering.
The point is that P2.2 is no better evidenced than P3.2 and perhaps less well evidenced. Why wouldn't an unobserved animal suffer when it dies? In keeping with many attempted defences against the problem of evil, I will provide a ‘just-so story’ which can answer this question. My just-so story is simple: every time an animal is about to suffer a gratuitous pain, God intervenes with a local miracle which effects only the animal's internal mental life and prevents the suffering. I am not committed to the truth of the just-so story, but I am committed to its coherence, and that it is compatible with and does not undermine the prior probability of theism.
It might be objected that the probability of this being a world with a God who regularly intervenes to prevent animal suffering is very low because such a world would involve a high degree of irregularity. Why should an irregular world in the relevant sense have a low probability attached to it? Perhaps the obvious reason is that an irregular world is a less simple world, or so it seems, and simplicity, it is sometimes suggested, is a theoretical virtue precisely because simpler theories are more likely to be true.
There are two problems with this line of reasoning. First, it is by no means agreed that simpler theories are more likely to be true. Second, it is not at all clear that the possibility I have sketched in this article is any less simple than other alternatives, at least in any relevant sense of the term ‘simplicity’.
Nevertheless, one might be tempted to argue as follows: call the proposition that there exists an entity with the attributes of the God of classical theism ‘G’ and call the proposition that there exists an entity with the attributes of a God of classical theism, who intervenes in all cases to prevent unseen animal suffering, ‘G1 ’. We may think that G is ceteris paribus more probable than G1 on the grounds that, in general, adding properties to a posited object brings the probability of that object existing down.
So far, I think this is right, and I grant that the probability of G1 is lower than that of G. However, on its own, this fact is not enough to cause trouble for my proposal. Two possible conclusions would cause trouble for my proposal. First, my proposal would face a serious objection if it could be shown that G1 is less probable than the existence of a God who does not intervene or a God who intervenes less regularly than according to G1. However, these results cannot be derived from the fact that G is more probable than G1. G does not tell us anything about the number or nature of God's interventions. Attaching a higher probability to G than to G1 is perfectly compatible with attaching the same or a lower probability to the existence of a non-or-less-intervening God than to G1. Second, my proposal would face a serious objection if it could be shown that G1 is less probable than atheism or Paul Draper's (Reference Draper1989) Hypothesis of Indifference. Again, though, this does not follow from the fact that the probability of G1 is lower than that of G. Admittedly, it does follow, under very plausible assumptions about the ordering of probabilities, that if G1 is equally as probable as atheism, then G is more probable than atheism. It might be charged that, minus a convincing argument for the existence of God, this result involves a kind of question-begging against the non-theist. However, this result is also not entailed by the current proposal; this is because we need not hold that G1 is equally as probable to atheism. We might rather hold, with van Inwagen (Reference van Inwagen2006, 115–116) for example, that we simply do not have any reasonable basis for assigning relative probabilities to these alternatives. From this position, it does not follow that G is unfairly assigned a higher probability than atheism. With the failure of this line of attack, I do not see any other plausible strategy for moving from the higher probability of G than G1 to the probable falsity of the latter. It seems to me at this point that a compelling case has yet to be made out to the effect that G1 is inherently unlikely.
Evidence and animal suffering
I anticipate that this response to the problem of unseen animal suffering will be met with more than its fair share of incredulity. When we try to pin down the source of the misgivings, I think several worries present themselves. I will break down my discussion of objections over two sections. In this section, I will consider objections related to the general worry that my position slides into a form of Cartesianism. In the next section, I consider objections on the grounds that my response goes wrong in abandoning nomic regularity with respect to the mental lives of animals.
The slide into (sub-)Cartesianism
One worry is that this sort of reasoning is the beginning of a slippery slope that leads to Cartesianism. In response, I shall argue that unseen animals are a special case because we do not, generally, have compelling evidence for unseen animal suffering, and in those special circumstances which do allow us to have such evidence, those same circumstances also provide the makings of plausible appeals to the soul-making or free-will theodicies to explain why God would allow these particular instances of evil.
The worry is that, if we cannot conclude that unseen animals suffer because, for all we know, God might be intervening in their internal mental lives, then we can also not conclude that visible animals suffer for the same reason. The first thing to note here is that, even if this thought is right, the result is not Cartesianism. Cartesianism is the view that animals lack conscious mental states altogether. My position does not provide any reasons to doubt that animals, visible or otherwise, experience all pleasurable and neutral mental states. If my just-so story gives us as much reason to doubt visible animal suffering as invisible animal suffering, even so, Cartesianism would not be the result.
Call the view that animals experience only pleasurable and neutral mental states ‘sub-Cartesianism’. Sub-Cartesianism would still be a problematic consequence. For one thing, it would provide too simple a resolution to the difficult moral issues around killing animals for food. If theists were justified in reasoning that God would never allow animal suffering under any conditions, then this would seemingly provide justification for meat-eating on the grounds that God would ensure the animals did not suffer. The point of course would generalize to all deliberate attempts to harm animals. If the just-so story I have proposed has the consequence that killing and torturing animals is morally acceptable, this is a problem for the just-so story.
Nevertheless, I do not think that sub-Cartesianism follows from my view. The Moorean shift strategy that I propose depends on P3.2, that God exists, being no less well-motivated than P2.2, that animals suffer when there is nobody about to see them. If we do not have evidence for the suffering of unseen animals, and assuming we are not already weighting the probability of the existence of an interventionist God lower than the probability of the non-existence of such a God, then P3.2 is at least as well motivated as P2.2, or so I have suggested. But consider a parallel thesis to P2.2 with reference to unseen animals replaced by reference to visible animals.
P2.2* Animals (sometimes) suffer when there are humans around to see them and/or cause their suffering.
If one were to reject this thesis, using reasoning parallel to mine with respect to P2.2, it would have to be the case that P3.2, that God exists, is as well evidenced as P2.2*. Now, of course, many theists will hold that this is so. However, nothing I have said requires that I take this position. It seems to me clear enough that P2.2* is very well evidenced indeed, and only very strong evidence for the existence of God would serve to provide the makings of an argument for Cartesianism parallel to my argument 3.
Key to my response to this worry is that there is a difference between our evidential relation to P2.2 and to P2.2*. To justify this claim, we must briefly consider our evidential relation to the conscious states of others generally. I shall argue that the kind of evidence available supports the existence of visible animal conscious states in a way that it does not support the existence of unseen animal conscious states, with exceptions to be noted in the next sub-section.
The standard view is that our justification for believing in the existence of conscious states that are not our own is dependent on an inference, either analogical (Mill (Reference Mill1889 [2005]), Russell (Reference Russell1948), Ayer (Reference Ayer1956), for example), or abductive (Pargetter (Reference Pargetter1984) ), or a hybrid of both (Melnyk (Reference Melnyk1994) ), which takes as evidence the behaviour of the subject of the purported mental states. These are so-called ‘inferentialist’ accounts. According to other philosophers, our knowledge of the conscious states of other agents is, or at least sometimes is, immediate and non-inferential (Dretske (Reference Dretske1973), Cassam (Reference Cassam2007), McNeill (Reference McNeill2010) ). This view is the so-called ‘perceptual hypothesis’ according to which conscious mental states that are not our own are nevertheless evident to our senses; we can see that others are happy, angry, or in pain, for example. I maintain that on both of these accounts, we have evidence for P2.2* which we lack for P2.2.
It is easy to see this with respect to the perceptual hypothesis. All versions of the hypothesis depend on a direct sensory connection between the perceiver and the object experiencing the observed conscious states. That is the whole point of the perceptual hypothesis. Clearly, we cannot have this sort of observation of unobserved animals.
If we were to have evidence of the perceptual states of unobserved animals, it would have to be inferential. However, turning back to inferential justifications for beliefs in external conscious states, the empirical literature on animal pain provides good examples of the structure of these kinds of arguments. A starting point for the contemporary research is the distinction between evidence for nociception, the capacity for sensing noxious stimuli, and evidence for the conscious experience of pain. Evidence for the first is not taken as sufficient evidence for the second (Melzack & Wall (Reference Melzack and Wall1965) ). It is therefore not sufficient to show that animals have a certain biological make-up, or that they are being subjected to a noxious stimulus in order to conclude that animals experience pain.Footnote 5 Empirical arguments for the presence of pain in animals typically involve comparing the behaviour of a sample of animals in contact with a noxious stimulus to the behaviour of a sample of animals which are not in contact with that stimulus. (See, for example, Lynne Sneddon's experimental work on fish pain: Sneddon (Reference Sneddon2003), Sneddon, Braithwaite, & Gentle (Reference Sneddon, Braithwaite and Gentle2003); see also Chandroo, Yue, & Moccia (Reference Chandroo, Yue and Moccia2004) and Sherwin (Reference Sherwin2001). Similarly see Elwood and Appel's (Reference Elwood and Appel2009) experiments with hermit crabs.Footnote 6)
What can experimental work tell us about animal suffering? I think it provides inductive evidence from which we can justifiably infer that the observed animals suffer during the experiments. On the perceptual hypothesis, we can justifiably believe that we are seeing animal suffering. On inferentialist accounts, we have good, though defeasible, evidence for inferring that an animal was suffering. I think we might even defeasibly infer that there is a law of nature operating according to which, in general, certain stimuli cause suffering in animals. However, I do not think we are rationally compelled to infer that every time an animal meets these stimuli, it suffers. For one thing, any number of intervening physical phenomena might intervene to block the usual cause and effect, for another, the rational theist will be open-minded about the possibility of miraculous exceptions to the operation of these laws of nature. In short, I am suggesting that while experimental work can provide good evidence for P2.2*, that visible animal suffering sometimes occurs, it does not provide compelling evidence P2.2, that unseen animal suffering occurs. For this reason, my view does not involve a slide to even sub-Cartesianism.
The asymmetry of divine intervention
In the previous section, I argued that my position does not slide into sub-Cartesianism because there is an asymmetry in our epistemic relations to visible and invisible animal suffering. That is, we have a reason to believe that God is not always intervening in cases of visible animal suffering which is not available to us in cases of invisible animal suffering. It has been suggested to me that a worry remains in this area: if we are justified in believing that God often allows visible animal suffering while intervening to prevent invisible animal suffering, the rational theist who accepts this view must provide some explanation for why God allows visible animal suffering. I have already provided an answer to this objection: I assume that the popular free-will and soul-making theodicies can account for all visible animal suffering. Nevertheless, the objection might be pressed that if God could have intervened to remove the suffering of animals while keeping external appearances as they are, and if God is in fact doing just this in the case of invisible animals, then surely the cases of visible animal suffering are gratuitous.
My main concern in this article is of course invisible animal suffering, not visible animal suffering, and I am not convinced that this line of objection is any more serious for my view than for the traditional free-will and soul-making defences. Nevertheless, I will propose three possible explanations (once again, these are ‘just-so’ stories, to the truth of which I am not committed) for the apparent asymmetry of God's interventions, i.e. intervening to protect invisible animals, while permitting visible animals to suffer.
The first reason is that, perhaps, if God were frequently to allow the visible manifestations of suffering while eliminating the actual suffering, this would constitute a systematic form of deception incompatible with God's nature.Footnote 7
The second reason is that, if God were always to intervene to prevent animal suffering, this would involve a diminution of our freedom. Part of the great good of being free that God has given human creation is that we can act freely in the world. A world in which we can choose freely and act freely is better than a world in which we can choose freely but act only in accordance with God's wishes. Acting freely involves being able to realize the intended consequences of our choices, or so I think. This is so even when the results look the same to us regardless of whether or not God intervened. In short, the great gift of free action entails that humans be able to make animals suffer when they choose to act in such a way that animal suffering would naturally result.Footnote 8
The third reason God might have for allowing visible animal suffering is that, if God did always intervene in these cases, that would provide humans with the chance to thwart an important part of the divine plan. It could be that God attaches great significance to the ethical decisions that humans make with respect to their treatment of animals (e.g. ‘should I eat meat?’, ‘should I have a pet?’, ‘should I attend the circus?’). If God always did intervene, then it would be possible for humans to come to realize that God always intervenes (they could employ a version of my Argument 3, for example, which concludes that no animal ever suffers). However, if humans were able to justifiably reason in this way, then the ethical decisions would no longer arise for them. They could eat meat, keep pets, attend circuses, and so on, safe in the knowledge that God was not allowing the animals to suffer in any case. If God does attach value to the moral decision we make with respect to our treatment of animals, then plausibly God would not allow us to escape these moral conundrums quite so easily as this.
Naturally more can be said about each of these reasons for thinking that God sometimes allows visible animal suffering to occur. However, I will have to leave a more thorough discussion for future research.
Positive evidence of unseen animal suffering
It might be objected at this point that in some cases we can justify our belief in conscious states that are not our own on some other grounds than observation. If such a justification does hold equally well for observed and unseen animal suffering, then again my position is threatened by the dilemma of endorsing sub-Cartesianism or accepting some gratuitous animal suffering.
I can think of three plausible ways in which we might justifiably arrive at beliefs about the behaviour of unseen animals. First, our beliefs about the behaviour of unseen animals might be justified on the basis of testimony. Second, they might be justified on the basis of immediate perception of the results of animal behaviour, without perception of the behaviour itself. Third, they might be justified on the basis of the results of intermediary devices which report the behaviour of animals to which no human has immediate sensory access. I shall take each of these cases in turn, and argue that none of these possibilities raises insurmountable problems for my view.
It is quite clearly the case that our beliefs about the behaviour of others are often based solely on testimony. I think that testimony is often quite sufficient evidence to justify our beliefs about the behaviour of others. This goes for animals as well. I can quite justifiably believe that Tibbles the cat sat on the mat, simply because you told me she did, and you were there the whole time watching her (assuming I justifiably believe you to be generally trustworthy and so on). Indeed, you might not have watched her; perhaps you heard it from another trustworthy source. However, on reflection it is also clear that the chain of testimony must bottom out somewhere in some sort of evidential relation other than testimony, if the final resulting belief is to be justified. Somewhere, someone must have seen Tibbles on the mat, or perhaps there was something about the mat that pointed definitely to Tibbles having sat there, or perhaps the cat monitor picked up Tibbles's every movement, including her sitting on the mat. In any case, as the justification for the resulting belief will ultimately depend on some relation to Tibbles other than simple testimony, we can leave Tibbles, and consider these other alternatives.
We can turn then to the second possibility, that is, cases where we find evidence of animals having suffered after that suffering has ended. Let us take the following case as an example: I am hiking in the countryside and I come across a recently deceased fawn, the look of terror and pain still on its face. The fawn is, however, dead and has been for some time before I arrived to bear witness. I did not see the animal suffering, but I did have perceptual evidence that the suffering had occurred.Footnote 9
However, cases like this one would require that the animal's remains are sufficiently intact to provide evidence not just of the animal's death, but of the animal's conscious state immediately prior to death. On the perceptual hypothesis, I would need to be able to see the suffering. On an inferentialist account, I would need to be able to see clear evidence of the pain behaviour the animal engaged in. In such a case, I suggest, the same appeal to the soul-making theodicy could be made here as in the case of visible animal suffering. The soul-making theodicy, as I have set it out, posits the opportunity to manifest particular virtues as reasons for God's allowing evil. I have suggested that, in cases of visible natural animal suffering, the theist can appeal to the opportunity for manifesting virtues like empathy as the reason God allows the animal's suffering. In cases where the remains of an animal tell an unambiguous story of suffering, it seems to me that such an appeal can still be made.Footnote 10 I think, then, that this case is not decisive. I grant, however, that cases of this sort, and the sort we turn to next, do force me to admit that the conclusion of Argument 3, which I have defended, namely that ‘no animal suffers when there is nobody around to see it or cause its suffering’ cannot be read quite literally. By ‘see’, in this case, I mean ‘have some sensory relation to sufficient for providing an opportunity for soul-making’. If I am permitted to interpret my own words in this slightly liberal way, I think my position remains intact.
We turn, finally, to our third case, in which a machine tells us that an animal is displaying pain behaviour, but we are not actually looking at the animal. In some respects this is the hardest case for my account to deal with, but I do not think it is fatal to the proposal. First let me note that my thesis does not involve a modal claim. I am not arguing that we could never have evidence of suffering in animals who are sufficiently distant from us that their suffering does not provide the opportunity for us to manifest any of the virtues, if such suffering exists. Such mediated evidence may well be possible. Indeed, Elwood and Appel's (Reference Elwood and Appel2009) results involving crabs leaving their shells after receiving electrical shocks can presumably be replicated without humans being present at the time the shocks were applied. Moreover, I think it would be wrong to suggest that this information provides a genuine opportunity to manifest any particular virtue, in the way that directly experiencing the death throes of a sentient creature may induce emotions of a life-changing character. However, it is worth noting a few conditions that would have to be met for such cases to provide evidence of gratuitous suffering (suffering escaping the explanatory capacities of the free-will and soul-making theodicies, that is). The behavioural reports that are generated by any mediating machine would have to be sufficiently lacking in detail that the humans tabulating the results are not provided with the opportunity to manifest high degrees of empathy, while simultaneously being sufficiently detailed to glean positive evidence of animal suffering (a very difficult matter, as the literature on what counts as evidence for animal suffering attests). Moreover, if the purported evidence is the result of an experiment in which an animal is actively put into a position where its pain behaviour is being registered, then it would be plausible to maintain that an act of free will was responsible for the animal's suffering. This would provide an alternative explanation for the suffering, to which the theist could appeal. This would seem to provide a blanket response to all relevant scientific studies of animal pain, including Elwood and Appel's and Sneddon's. Reasonably preventable animal suffering would also have to be ruled out on similar grounds. The mediated evidence would, in short, have to involve unseen and unplanned animal suffering of such a kind that we can justifiably conclude that animal suffering exists, but to which we are not so closely connected as to justify an appeal to the soul-making theodicy. It is not at all clear to me that we have this kind of evidence.
The complexity of animal suffering demands too much divine intervention
Another related worry is that the just-so story that God intervenes miraculously to spare unseen animals suffering loses its plausibility when we consider cases of complex animal suffering, for example, suffering related to conditions like depression, post-traumatic stress disorder, anxiety, and other similar conditions which might be expected to lead to kinds of suffering which are drawn out through time rather than being episodic in nature.
Consider the following example,Footnote 11
An experienced researcher of primate behaviour leaves the orang-utan clan she has been studying and subsequently returns to find all the orang-utans dead. All the evidence points to death by slow starvation over months. She has studied many other orang-utan clans in periods of drought and has seen how they are impacted by gradual starvation. The final months of their lives are characterized by physical pain but just as much by persistent and worsening emotional distress. For months, there is not a waking moment in the life of any of the orang-utans which is not rendered visibly painful and unhappy by the combination of hunger and of the deaths of their clan-mates.Footnote 12
If we respond to this case by appealing to our just-so story that God intervenes to spare unnecessary suffering, we face the worry that we have allowed too many divine interventions. Consider: mental suffering in some degree or other characterizes much of our own lives; the moments of our lives that are absolutely free of anxiety, stress, or worry are few. It might be thought plausible that the same should in general be assumed to apply at least to complex animals like primates. But, if we are to apply the just-so story to all cases of unseen animal suffering, including complex suffering like that experienced by the orang-utans, it seems that we might end up in a position of affirming that God is literally intervening in the lives of each unseen animal constantly, in order to spare them the unnecessary emotional stresses that are the daily fare of all complex animals. Why might this be thought to be a problem? It might even be thought to show that the just-so story is actually incoherent in that the story seems now to suggest that God miraculously intervenes in the conscious lives of unseen animals most of the time, and it might be thought that it is conceptually impossible for miracles to be taking place most of the time. More specifically, a miracle, we are often told, is a violation of the laws of nature. If the supposed laws of nature (in this case, the laws relating certain painful stimuli to their conscious affects) are being broken most of the time, then they are not the actual laws of nature.Footnote 13
This objection depends on deriving some problematic consequence from the thesis that God miraculously intervenes in the conscious lives of animals most of the time. However, I see no compelling reason for thinking that my defence involves this problematic thesis, even if we extend the just-so story to cases of sustained non-episodic unseen animal suffering. Even if the available scientific data can support the hypothesis that most conscious animal life has occurred beyond human contact and that, most of the time, this life involves morally significant levels of suffering, still this would not commit me to saying that God intervenes most of time in the conscious lives of animals. As I have noted before, my defence is intended to run hand-in-hand with the soul-making and free-will theodicies. So, on my view, it is not the case that God intervenes in every case of unseen animal suffering. God only intervenes in the gratuitous cases. There are undoubtedly cases of unseen animal suffering which are not gratuitous because they are the results of human free will. For example, the fawns that have died in deliberately set forest fires, the bats that died in the Hiroshima bombings, the big cats that died when their natural prey was hunted to extinction, and the polar bears that drowned as a result of global warming, to say nothing of the millions of creatures who suffer as a result of the meat, fur, and dairy industries. For all I know (I strongly suspect, for all anyone knows), most contemporary animal suffering is the result of human activity. All of these cases would need to be factored out. As for pre-human animal life, it would not be sufficient simply to show that most complex animal life pre-existed humans. We would also need quite detailed evidence about the cognitive capacities and the daily lives of these creatures: how their brains had developed, whether they were capable of experiencing complex suffering or only episodic suffering, how much of the lives of how many of these creatures might probably have been characterized by suffering, and so on. It seems to me that, with complications like this, we do not have sufficient evidence to infer that if God intervenes to prevent gratuitous animal suffering, God intervenes most of the time.
My just-so story is only committed to the claim that God intervenes miraculously in the conscious lives of unseen animals some of the time. So, if it is combined with an account of laws of nature like that of Swinburne (Reference Swinburne1968), according to which they allow for exceptions, I do not think there is anything incoherent about it.
These considerations also provide an answer to another closely related objection to my view that I have encountered. According to this latest objection, my view involves the problematic consequence that animals develop in such a way that they are capable of experiencing pain when they only do actually experience this pain in a comparatively small number of cases. There are two worries in this area. One of these concerns how evolution might allow for this sort of development. I will answer this objection in due course. The other concerns why God would allow animals to develop in this way. My answer to this second worry should be clear from the above considerations. God allows animals to develop the ability to feel pain in order to provide humans with important (perhaps unique) opportunities for free will and soul-making. Moreover, it is not clear that this is involves God in allowing the development of biological capacities that are very rarely used by animals (i.e. pain sensitivity). Finally, even if animals do comparatively rarely feel pain due to regular divine interventions, it is worth keeping in mind that with the expansion of the human population and the even faster expansion in the ability of humans to affect their environment, it is very likely that whatever might have been the case in the past, humans will at some point in the near future be responsible for most animal suffering, which might involve God in fewer future interventions.
I now take myself to have gone at least some way to answering the various objections in this area; nevertheless, some philosophers will think it a disagreeable story. I consider their views next.
Nomic regularity
The just-so story that I have appealed to in this article involves regular and systematic divine interventions in the world. There are those, including Swinburne (Reference Swinburne1998) and van Inwagen (Reference van Inwagen2006), who think that God would not want to design a nomically irregular world. Relatedly, there is at least one influential argument for the existence of animal suffering, from Peter Singer (Reference Singer1990), which depends on animal pain having been distributed evenly over time and regardless of the presence or absence of humans. If this argument is successful, then it would support a sort of regularity with respect to animal suffering that is at odds with my proposed solution to the problem of animal suffering. I shall consider both of these worries in turn.
Nomic irregularity is a defect
The objection is that God is less likely to create an irregular world than a regular one because the latter is in some way superior to the former. Ironically, this claim is often defended by those who are engaged in the same project I am engaged in now: trying to show that animal suffering does not make the existence of God less likely. In fact, an appeal to the supposed divine preference for nomic regularity provides one of the most common responses to the problem of animal suffering. The idea can be found in the work of Swinburne (Reference Swinburne1998), van Inwagen (Reference van Inwagen2006), and Hasker (Reference Hasker2008) and is challenged by Murray (Reference Murray2008) among others. Van Inwagen (Reference van Inwagen2006, 120) expresses the thought as follows: ‘[b]eing massively irregular is a defect in a world, a defect at least as great as the defect of patterns of suffering morally equivalent to these contained in the actual world’. An argument can be given as follows: sentient animals could exist without experiencing pain only if there exists massive irregularity in the world. It is better that animals regularly experience pain than that there exists massive irregularity in the world. From this it follows, under plausible assumptions, that if the value of the existence of sentient animals outweighs the evil of the suffering they experience, then God would choose to create sentient animals, pain and all, other considerations being equal. Swinburne (Reference Swinburne1998, 176–192) has argued that irregularity is defective for extrinsic reasons, in that it makes the exercise of morally relevant free wills impossible. Van Inwagen (Reference van Inwagen2006, 113–134) argues that irregularity is, or at least might be, intrinsically defective. Naturally, both of these views are incompatible with mine, and I will respond to each in turn.
Swinburne's position is easily dealt with. Swinburne intends to explain the natural evil by, in effect, extending the free-will theodicy. His position is that the morally relevant exercise of free will requires that humans have sufficient knowledge of the likely outcomes of their freely chosen actions. If humans were wholly unaware of what would be likely to result from their free choices, then those choices would not be evaluable as morally good or evil, even by God. The great good of morally relevant free will therefore requires a certain amount of knowledge of causes and effects, and, thinks Swinburne, this knowledge is only attainable in a world where the same sorts of causes lead to the same sorts of effects with lawlike consistency. Setting aside the objection that God could have simply implanted the relevant knowledge in the mind of the agent immediately prior to the free choice, without building into the world a nomic regularity that would lead to the deaths and sufferings of billions of animals, not to mention many innocent humans, nevertheless even if Swinburne's central claim is true, it does not undermine my position. Swinburne's defence, it seems to me, can justify only such nomic regularity as is required for the operation of morally relevant human free will. My defence concerns failures of nomic regularity which only occur beyond human sensory contact. It seems to me that these positions are perfectly compatible: perhaps God does allow the laws of nature to operate with absolute or near absolute consistency when humans are in a position to confirm their operation, but allows exceptions (of the kind required for the defence outlined above) when humans are not in a position to confirm their operation. If this is a logical possibility, and it seems to me that it is, this would suggest that, at most, God would not bring about a world which is insufficiently regular for the purposes of morally relevant human decision-making, and this is not a problem for my position.
I turn, then, to the stronger claim, that God would not bring about a nomically irregular world because nomic irregularity is intrinsically defective. We should note at the outset that the best contemporary exposition of this position, van Inwagen's, does not in fact involve a defence of its truth. As van Inwagen explains his view:
we should recall that a defence is not a theodicy, and that I am not required to show that it is plausible to suppose that massive irregularity is a defect in the world, a defect so grave that creating a world containing animal suffering morally equivalent to the animal suffering of the actual world is a reasonable price to pay to avoid it. I am required to show only that for all anyone knows this judgement is correct. (van Inwagen (Reference van Inwagen2006), 120)
Once again, this position is perfectly compatible with the defence I have sketched. I can accept that, for all anyone knows, massive irregularity is a defect in the world and, for all anyone knows, it isn't. So long as, for all anyone knows, massive irregularity is not a defect in the world, then the objection that God would not create a world in which unnecessary animal suffering was avoided due to massive irregularity does not get off the ground.
Nevertheless, van Inwagen provides some considerations which might be taken as evidencing the hypothesis that massive irregularity is a defect, and I must consider these considerations in order to justify my position.
Van Inwagen's first consideration is this:
One minor point in favour of this thesis is the witness of deists and other thinkers who have deprecated the miraculous on the ground that any degree of irregularity in a world is a defect, a sort of unlovely jury-rigging of things that is altogether unworthy of the power and wisdom of God. Presumably such thinkers would regard massive irregularity as a very grave defect indeed. And perhaps there is something to this reaction. It does seem that there is something right about the idea that God would include no more irregularity than was necessary in his creation. (ibid.)
The claim is an appeal to intuition, and it is not an intuition I share. As a self-report on my own aesthetic sense, I do not find the notion of a miracle unlovely in any way. Moreover I do not see why we should think God would create the most regular world possible. Finally, if it turns out that the other responses to the problem of unseen animal suffering fail, it might turn out that systematic irregularity with respect to the suffering of unseen animals is a necessary feature of the best possible world. I do not think, then, that this consideration fatally undermines the thesis of this article.
Van Inwagen's (ibid., 120–121) second consideration is that a massively irregular world is a deceptive world, and, as Descartes famously argues, God would not engage in systematic deception. There are two possible responses. One is to bite the bullet and accept that God is a deceiver, but hold that deception in this case is nevertheless the best possible course of action for God to take, given the alternatives. The other is to deny that massive irregularity of the kind involved requires divine deception. I prefer the second response, because I do believe that God does not deceive and have myself appealed to this very fact earlier in this article. On my view, God does not deceive us into thinking that animals have suffered beyond our immediate experience when they actually have not. Rather, through the fault of our own intellects, we systematically draw conclusions for which there is no good evidence. This is an intellectual failing in us, not a moral failing in God. I think some other reason must be given for thinking that God would not want to bring about an irregular world and I can think of none.
An abductive argument for evenly spread animal suffering
One final objection to my view is derived from Peter Singer's (Reference Singer1990) influential evolutionary argument for animal suffering. The idea is that the survival of animals over the millennia is better explained by their having developed the disposition to avoid painful stimuli than by their not having developed so useful a disposition. But if we take it that they do have this disposition, a view which is backed up by our observations of animal behaviour, this would be better explained by their actually feeling pain regularly and predictably than by their not feeling pain regularly and predictably. Singer's argument, then, depends on two abductive inferences, both of which have prima facie plausibility: first from the fact that animals have survived for millennia to the conclusion that animals have the disposition to avoid painful stimuli and second from this to the further conclusion that animals have been experiencing pain for millennia.
It seems quite plausible of course that an omnipotent God could have set up the world in such a way that animals would survive even without having evolved a disposition towards active pain-avoidance behaviour, or that God could have instilled a disposition for pain-avoidance behaviour in animals even while withholding the experience of pain from them. However, given the abductive nature of Singer's argument, the mere logical possibility of these alternative stories is not enough to blunt the force of the objection. What is needed is some alternative explanatory story which fits the agreed facts just as well as Singer's proposed explanation. If either of Singer's abductive inferences can be blocked by the availability of an equally well-fitting alternative explanation, the objection can be answered.
I do, however, think the second inference of Singer's argument, i.e. from the fact that animals display a disposition for pain-avoidance behaviour to the conclusion that they have been feeling pain for millennia, can be blocked by the alternative explanation that God might be intervening to bring about the behavioural disposition without the pain. More particularly, I suggest the following. Assuming that Singer's first abductive move is unchallenged and that it is an agreed fact that animals have a disposition to try to avoid pain (and have had this disposition for millennia), there are two equally good explanations for why this is so. Singer quite reasonably holds that this is so because animals have developed this disposition evolutionarily as a result of feeling pain. The rational theist may hold, with equal justification, that animals have this disposition because God gave it to them miraculously as a result of God's willing for animals to survive through the ages. Various reasons can be found in the literature for why God would want animals to survive through the ages: for example, for the value of diversity, or because they provide opportunities for soul-making as described above. In any case, I think it is very plausible to suppose God would want animals to survive. Moreover, I have already argued that theistic theories involving divine interventions are not necessarily less probable than atheism. Given this, the claim that God would bring about the survival of animals by miraculously gifting them a particular disposition is not one that brings the theist's explanation into disrepute.
I conclude from this that the theist has a perfectly good explanation, an alternative to Singer's, for why animals have the disposition to try to avoid pain. With this explanation in hand, Singer's argument does not show that animals have been experiencing pain for millennia.
Conclusion
In this article, I have proposed a new defence of theism against the problem of animal suffering. The defence employs a Moorean shift response, to conclude that animals for whom suffering is not required for the free will or soul-making of humans never do suffer.
