Free Will, Animal Suffering, and the Contingency of Natural Laws

Imagine a world with no reliable cause and effect relationships. There are no regular connections between events of type A, events of type B, or C, etc. Imagine further that because you live in this world, there is no way to predict what will result from your intentions to act. For example, you may intend to shake my hand but end up punching me in the face instead! Or upon a second attempt at shaking my hand, bullets shoot out of your fingers and kill me! Your life in this imaginary world is surely bizarre, and clearly you cannot be held responsible for your actions. In fact, it is hard to conceive how you could even form intentions without predictable patterns of cause-and-effect to learn from.

This thought-experiment reveals that creatures like us cannot form intentions (let alone act intentionally) outside a framework of reliable cause-and-effect relationships. And if intentional actions do not exist, other goods such as free will, moral responsibility, and love do not exist either.

Why is this point important? Because it lends some credence to what theists have called the free will defense against the problem of animal suffering. Advocates of this defence argue as follows: many creaturely goods such as freedom, moral responsibility, and love would not exist if there were no “natural laws” to sponsor them. In fact, it is hard to conceive how plant and animal life on Earth could exist (let alone thrive) without such a platform to support it. So if God wants to create a biological world with these goods, he has to create natural laws as well.

However, there are risks in doing so. The same natural laws that enable creaturely goods and benefit life on the earth also generate natural catastrophes (such as fatal weather anomalies, mud slides, and mass extinctions) which cause a lot of animal pain and suffering. And given how interconnected the natural world is, it is hard to see how the benefits of natural laws can be had without the negative side-effects.

Naturalists often reply that God could have made different laws which lessen the overall amount, intensity, or duration of animal suffering.[1] Such a world would have been preferable to our own, and so God should have created it instead.

Instead of evaluating this “different natural laws” response, I want to highlight a point of irony in the naturalist’s position, especially as it pertains to a seemingly unrelated piece of natural theology called The Argument from Fundamental Order (or AFO).

According to the AFO, the basic intelligibility, consistency, and orderliness of the multiverse lend support to the hypothesis of theism, insofar as its basic laws are contingent, i.e. could have been different. Now, a minority of naturalists try to counter the AFO by arguing that those laws were necessary – that no multiverse could have been governed by different ones. Or put differently, any multiverse that could have existed had to be governed by the same laws.

I hope the reader will now detect the irony of this response. If the laws of nature had to be the way they are, then one cannot fault God for failing to create different ones that reduce animal suffering. One cannot have it both ways. Either the laws are contingent and therefore call out for explanation, or they are necessary and cannot be “tinkered with” to reduce animal suffering.

What do you, the reader, think about the “different natural laws” objection to the free will defense? If sound, does the objection implicitly support The Argument from Fundamental Order?

[1] This claim is distinct from the notion that the constants and quantities of the laws of physics could have been different. That is beyond reasonable doubt, but of little help to the naturalist because there are so few hospitable universes permitted by those laws, making it hard for us to discern if a universe with slightly different parameters would (on balance) contain less animal suffering. Therefore, the naturalist’s objection must be that an entirely different set of physical laws could have obtained.

12 thoughts on “Free Will, Animal Suffering, and the Contingency of Natural Laws”

  1. True this: “This thought-experiment reveals that creatures like us cannot form intentions (let alone act intentionally) outside a framework of reliable cause-and-effect relationships. ” (If you’re ever interested, F. Hayek makes a nice argument that parallels yours, in reference to political freedom and the importance of the state’s responsibility to establish predictable laws in The Constitution of Liberty.)

    Not sure that the “different natural laws” objection to the free will defense implicitly supports AFO, though. I’m sure people can take/interpret AFO in different ways, but I take the basic objection from the naturalist here to be that the basic intelligibility, consistency, and orderliness of the multiverse do not lend support to the hypothesis of theism because they are consistently *bad*! … Well, that’s an exaggeration for effect. But the point would be that the AFO would be less convincing precisely because the kind of fundamental order that was established was one that included animal suffering, natural disasters, etc.


    1. Hi Marilie,
      I am definitely interested in the Hayek reference if you would be willing to recommend any readings!

      Regarding the “different natural laws” (DNL) objection, at most it counts against a cosmic designer with moral attributes like perfect goodness. The intelligibility, orderliness, and consistency of the laws would still signal design, even if those laws eventually “bad” side-effects. Remember, it is the contingency of the laws of nature which the DNL presupposes, and which opens the door for asking why they are orderly, etc, rather than not. This holds even if the designer is morally defective, thought clearly this picture of the designer is at odds with classical theism 🙂

      Just as an aside, in order for the DNL objection to succeed, the naturalist needs to defend the (modal) claim that (1) a different platform of laws could have existed, and (2) would have occasioned less animal suffering, all other things being equal – i.e. without thereby jeopardizing other outweighing benefits or yielding worse suffering in other areas. That ceteris paribus claim is definitely worthy of consideration, but not easy to defend, IMHO.

  2. Hi Thomas, I’m excited to finally have something to say about what you’ve written. I’ve been creepily following your blog in silence, and finally get to have a voice.

    I enjoyed this blog entry quite a bit, but I think you’ve sidestepped the difficult bit about animal suffering and the free-will defence. You seem to be responding to the ‘logical’ problem of natural evil, but the powerful version seems to be the ‘evidential’ problem of natural evil. As I’m sure you’re aware, this problem takes issue with natural evils which appear to be superfluous (and depending on where one takes the argument, cannot be balanced out with good). So, for example, the suffering and pain which a deer that is severely burned in a forest fire (and which dies slowly from exposure and de-hydration over the next few days) appears to be an evil which no natural law would need to be changed, or re-framed, for an omni-God to avoid. That particular deer’s suffering is a contingent fact that could have been avoid in a host of natural ways, which would not have infringed on any moral agent’s ability to act freely.

    Furthermore, I would encourage some deep reflection about causation and an Omni-God. I’m happy to concede all you have written above–I think you get this right. But, if there are causal laws in this world which can play a contributing role to evils, then we have evidence against the existence of an Omni-God. Here I specifically have in mind, sociological, or psychological laws. That Stanley Milgram, through successive iterations of his experiments, was able to bring about compliance levels of up to 90% doesn’t bode well for the robustness of humanities ability to make important moral decisions. A world in which causal laws help, assist, or are bent towards evil, is a world that seemingly could have been created better.

    Undoubtedly you’ll have a response, Thomas, and I look forward to it! I’m not sure if you want to have the discussion here, but I’ll probably be more reliable by email. Hope you’re well!

  3. Hi Tim,

    Thank you for your insightful reply! You have touched on some very important points regarding the limits of the free will explanation (or FWE) of natural suffering. I would prefer to discuss this with you on my blog, so allow me first to clarify the purpose of the FWE within the context of theodicy.

    The FWE only claims that some animal suffering is inevitable in a nomically ordered world that sponsors free agents along with other biological organisms – not that all animal suffering (or even that any particular instance of it) has to be explained in this way. Indeed, because free agents arrive rather late in our planet’s biological history, this explanation alone does not account for the long pre-human history of animal pain on earth, nor does it explain why our non-human predecessors were capable of second-order mental states of pain in the first place.

    Also, for all we know, nomic regularity might also be worth having for reasons other than the good of creaturely freedom, and some animal suffering might be permissible for reasons other than nomic regularity! For a survey of these reasons, and how they can be combined to mitigate the force of the evidential (not the logical) argument from natural evil, I’d recommend Michael Murray’s helpful book, Nature Red in Tooth and Claw (2008), especially chapter 7. My hunch is that adequate responses to the counterexample you gave, namely, the death of the fawn, can be formulated when the full gamut of theistic resources for theodicy is put to work. Nevertheless, your objection does highlight the fact that no single response, like the FWE, is going to be comprehensive.

    When it came to the fawn’s suffering, you said it appeared superfluous because it “could have been avoided in a host of natural ways, which would not have infringed on any moral agent’s ability to act freely.” By “natural ways” I assume you are referring to those which accord with natural laws and the predictable powers they describe, am I right? If so, then I am curious, what ways did you have in mind?

    Finally, when it came to natural laws, you argued that “a world in which causal laws help, assist, or are bent towards evil, is a world that seemingly could have been created better.” This statement needs a lot of unpacking, but two initial thoughts come to mind:

    First, it follows only if the animal suffering occasioned by those laws could have been prevented without sacrificing other outweighing goods or bringing about other suffering equally bad or worse. This “only if” clause is both a modal claim about possibility and a ceteris paribus claim: namely, that different natural laws could have occasioned less animal pain, all other things being equal. The truth of that claim is not immediately obvious, and requires that we explore the reliability of our modal intuitions in this context, as well as the range of explanatory resources available to the theist.

    Second, it is worth keeping in mind that causal laws always have trade-offs because their benefits cannot be had without side-effects. As Murray remarks, “insofar as nature is nomically regular, the laws governing the hardness of matter and the conservation of energy, laws that permit me to use my hammer to drive nails, will also allow the hammer to crush my thumb. The same will hold true with respect to laws that make possible all sorts of conditions which leads to natural evils. The same laws that allow me [and other animals to digest] food also allow [predators] to tear my flesh” (p.141). Examples could be multiplied here, but the point remains that many of the goods sponsored by causal are not jointly realizable without suffering as their byproduct. Btw, I am not convinced that omnipotence can overcome these trade-offs either. Perhaps this is where we disagree?

  4. Hi Thomas,

    Thanks for the reply. I’m afraid that in my eagerness to post, I didn’t explain myself particularly well and wasn’t as clear or precise as I should have been. Let me try again.

    In your original post, you indicated that natural laws are necessary for intentions, which in turn are required for goods such as free will, moral responsibility and love. On the FWE it isn’t clear that the benefits of natural laws can be had without the negative side-effects; it may be that some animal suffering is inevitable in a nomically ordered world.

    But, in terms of natural evil there are two distinctly different problems: the logical one and the evidential one. The FWE, as you point out in your response to me, is an adequate reply to the logical problem of natural evil only. A world which is nomically ordered and includes animal suffering is logically compatible with a good God because these laws are (seemingly) required for free will, and moral responsibility.

    However, given this reply to the logical problem, then it follows that a good God would actualize a world that minimizes the suffering that results from the nomic order. And, as I understand it, this commitment to minimizing evil requires ensuring that the actualized laws are not the cause of, or causal contributors to, superfluous suffering. Scenarios like the one involving the fawn, then, cannot involve gratuitous pain. But, that the fawn is badly burned and dies slowly from dehydration and exposure seems superfluous because we can easily imagine another possible world that is exactly like this one in all relevant respects, except that the fawn: (a) is not severely burned because it is able to escape from the forest-fire, or (b) is burned but dies more quickly/humanely. Since actualizing either of these parallel-but-slightly-better worlds would not require God to violate, or change, any natural law, or to restrict/infringe upon any moral agent’s ability to act freely, then we have strong evidence against the existence of an omni-God.
    To see that God would not need to violate any law, or instantiate different ones, consider that both (a) and (b) can be brought about by placing the fawn in the right place at the right time. Or, if you feel that some basic level of autonomy or free will should be attributed to the deer, then this could be accomplished by placing the deer in situational surroundings such that the options presented to the fawn resulted in it being led away from the fire, or to a quicker and more humane death.
    I can’t see a way to avoid this. And, further, do not find the idea that God might permit these natural evils in order to allow outweighing goods (say free-will and moral responsibility) remotely compelling. A good God should not be treating sentient creatures as a means to an end, and should not be engaged in the horrific kind of calculus that this reasoning suggests.

    Putting the evidential issue aside, I also want to advocate caution when it comes to how natural laws connect with the problems evils present. (I’m still figuring out exactly what it is that bothers me here, so please be charitable as you read this.) I don’t think that concern over the ‘goodness’ of the natural laws in this world is a conversation that only has implications for natural evil; I believe that it applies equally to moral evil. Typically moral evil is attributed to the free decisions of individuals. But, when an individual chooses to act in an evil way, they are situated in a particular context, and if they had been situated in a different context then they may not have committed the evil. In some cases, I think we can say this more strongly; it isn’t that they may not have committed the evil, it is that they would not have committed the evil. There are ‘situational pressures’ which can affect a moral agents decisions. Sometimes these pressures are benign, but sometimes they act as causal contributors to, or as a part of the necessary conditions for, the performance of moral evils. If this is correct, then there is a sense in which an agent who could have placed the individual moral agent in a different situation, or which could have created the individual in such a way as to be less sensitive to these particular situational pressures, ought to have done so. Undoubtedly a tangible example would be helpful here, and I don’t think that one will be that hard to come up with–though I won’t be doing that tonight.

    Hopefully this gives you a clearer idea of my thoughts, and as always, I look forward to your response.

  5. Hi d.f.,
    Thanks for clarifying. Instead of responding with something lengthy, I will attempt to distill your comments to make sure I understand you. Two claims underlie the possibilities you described:

    1. The alternate possibilities (of a the fawn escaping fire or a moral agent getting to choose within a more favourable set of circumstances) can be actualized by God in a world that is exactly like this one in all the relevant respects.

    2. If (1) is true, then God ought to have created that world instead.

    Have I understood you correctly?
    – HBR

      1. Great. I think we are making progress. My next question is, then, this: Pertaining to proposition (1) above, what are the “relevant respects” in which the alternate world and our world need to be the same?

        1. Well, I wouldn’t want to claim these are all of the relevant respects, but the central point here seems to be that this parallel-but-slightly-better world would have to have (1) as much good (read: moral freedom, free will, love, etc.) as the real world, and; (2) that the subtle alterations (the deer being more severely burned, and dying more quickly) of the parallel world would not result in, or permit, greater evils.

          Ultimately, I think a formal articulation of “the two worlds are identical in all of the relevant respects” would be that the goods of the parallel world can be mapped, or correspond, one-to-one onto the goods of the real world, and that, with the exception of the deer’s suffering, the evils of the parallel world map one-to-one onto the evils of the real world.

          At least, that’s what I’d guess an atheist would reply with.

  6. Hi d.f.
    This is on the right track, but I would suggest eliminating talk of similarity between worlds (or mapping of goods) and start talking about conditions under which an instance of suffering is gratuitous. One condition, suggested by Will Rowe, is that in the alternate world, preventing the fawn’s suffering does not (broadly speaking) undermine other outweighing goods or occasion worse evils. I am unsure whether this condition is necessary or sufficient for the suffering to count as “gratuitous”, but it is clearly captures much of what we mean by the word. Would you agree with that?

    1. Hi Thomas,

      I would roughly agree with that.

      Also, since the idea of an outweighing good has been introduced, let me just comment on that. I would want to suggest that it isn’t enough that an evil is outweighed by some later good. This is the kind of horrific calculus I don’t think we as theists can reconcile with a good God. Suffering matters to the individual; that I might have to suffer terribly in this life, only to have someone else enjoy some outweighing good, seems at the very least unjust, and perhaps even an evil itself.

      Thanks for talking this through HBR–I’m enjoying it; It’s good to have another philosopher to talk this through with. Just one more thing. I’m about to go away for about a month, so I don’t know what internet availability will be like–so I may not be in touch for a while.

      Hope you’re well.

      1. Hi d.f.

        Good to hear back from you.

        I’m on board with your comments. It is not enough that an evil be necessary for some outweighing good or for preventing an evil equally bad or worse. There are other ethical constraints which a good God would satisfy. Theists of various stripes part ways at this point, however, depending on their leanings toward deontology or utilitarianism. You have probably already read my two posts entitled “Ethical Constraints on Theodicy” –

        so I won’t belabor those points here. But the majority of theists do agree that at least one further (minimal) constraint is needed: victims of undeserved (horrendous) suffering must be compensated by an equal or outweighing benefit to them. This constraint helps somewhat to ensure that they are treated as ends, not as mere means to the benefit of others. In my reading of the literature on the problem of evil, it is very controversial to insist on much more than this, though the trend seems to be that the greater one prioritizes the claims of individuals, the harder it is for theism to explain their suffering.

        Now here is where things get tricky for the fawn example you gave. Typically, the relevant ethical constraints apply to the category of persons, or at least to creatures capable of becoming persons, and these are only a small subset of the sentient beings on this planet.

        My view is that merely sentient creatures are still capable of suffering in morally significant ways so that persons have moral obligations to treat them in certain ways. But it is not clear that sentience alone suffices for being treated as an end, or if so, that the individual claims of merely sentient beings carry even remotely as much weight as the claims of persons, particularly with respect to their right to compensation for undeserved suffering. So while the compensation constraint definitely applies to human beings, more argumentation is needed on your part to warrant its application to the fawn.

        One way to avoid a complicated detour into animal ethics is to focus solely on the constraint laid out in the first paragraph of this comment, and to see what explanatory resources (if any) theism has for meeting it – since if theism cannot even meet that constraint, then further talk about compensation is a moot point! What are your thoughts?

Leave a Reply

Your email address will not be published. Required fields are marked *