Guest blogger John D of Philosophical Disquisitions summarizes contemporary articles in philosophy of religion in plain talk so that you can be up to speed on the God debate as it ensues at the highest levels of thought. Visit John’s blog for more helpful summaries of contemporary philosophical works.
So you remember the sloth and the cougar right? In part one, their unfortunate encounter (from the sloth’s perspective anyway) was used to introduce us to Nick Trakakis’s strong claim that theism is incapable of accounting for any amount, instance or distribution of natural evil. I suggested that Trakakis’s argument boiled down to the following:
(1) The existence of an orthodoxly conceived monotheistic god is incompatible with the existence of evil, unless that evil is necessary in order to achieve some higher order good.
(2) No amount, instance or distribution of natural evil is necessary in order to achieve some higher order good.
(3) Therefore, an orthodoxly conceived monotheistic god is not compatible with the existence of any amount, instance or distribution of natural evil.
Premise (1) might be disputed by some non-theists, but we’re going to concede it for the purposes of this series because premise (2) is where our interest lies. As noted the last day, Trakakis doesn’t offer any positive justification for (2) but, instead, tries to see if anything can be said for the contrary position. He does this by evaluating several leading theodicies and seeing whether they can show how natural evil is necessary for achieving some higher order good. If they fail, he will deem his argument to be successful.
Last time out, we looked at the soul-making theodicy. In this post we’ll look at Swinburne’s free will theodicy.
1. The Argument from the Need for Knowledge
Within the literature on the problem of evil, the free will theodicy (and “defence”) is of special significance. For many theists, free will is recognised as a great good. But it is a unique kind of great good because its existence necessitates at least some amount of evil (or so Plantinga famously argued – some might dispute this since it presupposes a libertarian conception of free will).
This seems plausible but it makes the distinction between natural and moral evil all the more important. Although the non-believer might concede that free will necessitates the existence of moral evil (i.e. evil arising from the actions of free human agents), they are wont to argue that it does not necessitate the existence of natural evil (i.e. evil that does not arise from the actions of free human agents).
Enter Richard Swinburne. For many years now, he has been arguing that the non-believer cannot make this move because natural evil is absolutely necessary in order for free will to be possible. The argument he presents is somewhat long-winded and I’m going to lay it out in all its glorious detail right now. Brace yourself.
(Note: the numbering continues from part 1, in case you are confused and I’ve fiddled about slightly with the wording from Trakakis’s presentation.)
(10) If a moral agent S is to freely (or responsibly) bring about good or evil states of affairs, S must know how to bring about good or evil states of affairs.
(11) S cannot know how to bring about a good or evil state of affairs without knowing what consequences would follow from his or her actions.
(12) The only way for S to know the consequences of his or her actions is through induction from past events.
(13) There are two relevant sorts of past event that can provide the basis for such induction: (a) the past actions of moral agents including those of the agent himself or herself and (b) evil occurrences that are the result of natural laws.
(14) Therefore, S can have the freedom to bring about good or evil states of affairs if and only if he or she can induce the consequences of their actions from (a) the past actions of other moral agents; or (b) evil occurrences that are the result of natural laws.
(15) For any token of moral evil M, there must have been a first time for that evil have been freely brought about by a moral agent. Call this M*
(16) The moral agent who brought about M* could not have acquired knowledge of the consequences of their action by induction from the actions of other moral agents.
(17) Therefore, the moral agent who brought about M* must have acquired knowledge of the consequences of their actions by induction from the outcomes of natural laws.
(18) Therefore, natural evil is a necessary precondition for bringing about the realisation of the higher order good of free will.
A couple of quick observations here before we move on to a more substantial critique. First, in premise (10) I suggest that “freely” and “responsibly” are interchangeable for the purposes of this argument. This seems right to me. I think the higher order good at stake here is that of being a morally responsible agent. Free will is simply being viewed as a necessary precondition for that higher order good. Second, the conclusion (18) directly defeats premise (2) of Trakakis’s basic argument. So Trakakis needs to somehow break the chain of argumentation proposed by Swinburne to save his argument.
Moving on to more substantial matters, it should be clear that many aspects of Swinburne’s argument are unobjectionable. He is essentially imposing a knowledge condition on the exercise of responsible agency. I don’t know of a single philosopher of responsibility who would reject the imposition of such a condition. Knowledge and control conditions have been part of theories of responsible agency since the time of Aristotle. What does seem objectionable is the particular content that Swinburne gives to the knowledge condition in premise (12). Let’s look at this in more detail.
2. Swinburne’s Knowledge Condition
Two things should be borne in mind when assessing the acceptability of Swinburne’s knowledge condition.
First, the knowledge supplied by induction must be sufficiently strong for us to think the agent is responsible for acting upon it. No induction from past experience will ever supply complete certainty about the consequences of our future actions. But this is not a problem: the level of knowledge required for responsibility can be relatively low on some occasions, it all depends on which outcomes the agent is being held responsible for. Despite this not being a problem, the fact that complete certainty is not required is important when considering whether there are other knowledge conditions that do not necessitate the existence of natural evil.
Second, for his argument to work, Swinburne must be proposing that induction from past events is the only logically possible source of knowledge for free agents. The claim has to be this strong because we are concerned here with an alleged logical incompatibility between theism and natural evil. Many philosophers, both theist and non-theist, reckon that as philosophically limber as he is, Swinburne has much too high a hurdle to clear on this occasion (the same could be said of Trakakis in this article).
These critics propose other logically possible sources of knowledge that are as good as, if not better than, induction from past events. What’s more, these proposals would not require the existence of natural evil. The proposals can be divided into two classes: (i) those suggesting other supernatural sources of knowledge and (ii) those suggesting non-supernatural sources of knowledge. We’ll look at both in turn.
3. Other Supernatural Sources of Knowledge
Trakakis lists six possible supernatural sources of knowledge. Each has been put forward by different critics of Swinburne over the years. Set out briefly, they are:
(a) Universal Divine Revelation
(b) Special Divine Revelation
(c) Angelic Revelation
(d) A “knowledge machine” or oracle
(e) Message-laden dreams or visions
(f) Innate Knowledge
Some of these are more important than others. Only the first and sixth possibilities will be discussed here.
Swinburne objects to universal revelation on the grounds that it would overwhelm the autonomy of the agent. The reason for this is that knowledge of God’s existence when coupled with the desire to please him and avoid divine punishment, would remove the temptation to do evil. This would undermine true freedom. I can sort of the point (it harkens back to the epistemic distance requirement in the soul-making theodicy), but as someone who was brought up to pray for God to deliver me from temptation I struggle to see how this is necessarily a bad thing. Is free will really worth it if our ultimate goal is to be rid of it? This gets into the problem of heaven, which will have to wait for another day.
Even if it is successful in relation to universal revelation, it’s not clear that Swinburne’s objection works for the other five sources of knowledge. We’ll take innate knowledge as a special case. The idea here is that we could just be pre-programmed with knowledge of the likely consequences of our actions.
Swinburne objects to this on the grounds that the certainty provided by innate knowledge would not be enough for responsible agency. The obvious reply is that innate knowledge may be no worse than induction on this score and perhaps a good deal better. Indeed, there are many cases where innate knowledge is thought more secure than inferential or inductive knowledge. A classic example is to be found in the anti-evidentialist epistemology of Alvin Plantinga.
There is another problem with innate knowledge. This one affecting its application to the present context of an argument against natural evil. This is discussed in the next section because it crops up in relation to the non-supernatural sources of knowledge as well.
4. Non-Supernatural Sources of Knowledge
Non-supernatural sources of knowledge are any sources that do not originate in supernatural agents. This means they must be attributable, at some level, to other free actions on the part of human beings. A bootstrapping process is envisioned: through one set of free acts, humans could supply themselves with the knowledge necessary for another set of free, morally evil acts.
Two possible non-supernatural sources of knowledge are discussed by Trakakis. One is that of formal scientific testing and has been suggested by Eleonore Stump; the other is that of informal trial-and-error and is suggested by Trakakis himself. In both instances, initial actions (experiments or trials) will produce consequences that provide some knowledge, which can then be used as a source of knowledge for future actions.
Both possibilities seem plausible at first glance but when you think about them in more detail they are suspicious. What could possibly provide the motivation for such experiments or trials? The answer would seem to be some innate drive for curiosity or desire satisfaction. But the desires could not stem from our own experience of pain or suffering since, for the purposes of this argument, we are assuming that the world is devoid of natural evils capable of bringing about such experiences. But then the question arises: would not the presence of such innate drives itself amount to the existence of a type of natural evil? The same question arises in relation to innate knowledge.
Trakakis suggests the following answer:
“It seems more plausible, however, to regard our natural or inbuilt dispositions as morally neutral, as being neither good nor evil. It is how we cultivate these dispositions, rather than the dispositions themselves that is morally relevant.”
I’m not sure if I’m persuaded by that response. It seems to me like the process of dispositional cultivation would implicate yet more innate dispositions and abilities, each of which would be of circumspect provenance. In other words, these cultivators might themselves be morally non-neutral and so might be instances of natural evil.
This problem goes straight to the heart of Trakakis’s overall argument. To succeed, he needs to show that a world without natural evil is actually conceivable, but even this brief consideration of what kinds of innate dispositions are allowed in such a world suggests that it would be far more counter-intuitive than we might initially suppose. To be fair, Trakakis tries to address this issue again at the end of his article. I’ll talk about this the next day.