Rough draft: First pass.
Consider the following two lists of evils:
List A1. The suffering and death of a fawn caused by a forest fire due to a relatively rare natural event.2. The death of an explorer by a volcano in a remote and unoccupied region.3. The suffering caused by an extremely rare birth defect.4. A death from being hit by a relatively small meteor fragment.List B5. The suffering caused by the mechanisms of pleasure and pain to condition the behavior of sentient creatures.6. Suffering caused by predation.7. The suffering caused by innate mechanisms in the cognitive architecture of humans that naturally and reliably cause out-group hostility and genocide.8. The suffering caused by sickness and death due to microbes in many natural bodies of water.
The traditional distinction between moral and natural evil treats all instances of evil on both lists as roughly the same, viz., as just a bunch of instances of natural evil. This is bad. For intuitively, the evils on List B are relevantly different from those on List A, and in a way that is significant. In particular, natural evils on List A seem like one-offs in the normal course of things, while those on List B are a constitutive part of the normal course of things. To put it in terms of a popular idiom: List-A evils are bugs in the system of nature, while List-B evils are features. I therefore propose that we mark the distinction between the two types of evil with some labels. Call evils of the sort on List B structural evils, and call evils List A non-structural evils.
As a first approximation, structural evils are characterized by at least the following three features:
1. They are a species of natural evil.2. They are caused by structural features of the universe or a specific portion thereof.3. If left to run their course, such features either (a) reliably produce suffering/harm in human or non-human creatures or (b) significantly raise the probability of suffering/harm.
The structural/non-structural evil distinction holds out the promise of an advance in the problem of evil debate. For discussion of the problem of evil not infrequently focuses on non-structural evils. But these can seem like one-off evils, in which case one might naturally infer that they are foreseen but unintended evils in a universe that generally runs in a way that supports the well-being of its creatures. By contrast, it's intuitive that structural evils are such that, if God exists, then they are foreseen and intended, thereby eliciting a natural presumption of depraved indifference or actual malice. As such, they seem to be a much more formidable category of evil to account for on the hypothesis of theism.