Imagine the following scenario:

Philip and Samantha are considering two hypotheses, h1 and h2, that are causal explanations of some piece of evidence, e. The first hypothesis explains e in terms of a single ultimate cause, c1, while the second hypothesis explains e in terms of two ultimate causes, c2 and c3. Which hypothesis should they favour? Samantha appeals to the theoretical virtue of simplicity in favour of h1. But then Philip appeals to a rival theoretical virtue, namely complexity, in favour of h2. Samantha then goes on to defend her theoretical virtue by pointing out that reality is not infinitely complex, but clearly has a degree of simplicity. The theoretical virtue of simplicity is thus clearly grounded in the nature of reality. Philip agrees to some extent, but claims that the same is true for his theoretical virtue, because reality also has a degree of complexity. At this point, they get stuck and become unsure if they should favour the simpler hypothesis h1 or the slightly more complex hypothesis h2.

At first glance, there is no way to resolve this issue. Both supposed theoretical virtues seem to be rooted in the nature of reality. Reality is characterized by both simplicity and complexity. Hence, we cannot really decide if we should favour the simpler or the more complex hypothesis. But perhaps a close analysis of e and the causal explanations in question is a way out of this impasse?

There is indeed an interesting relationship between simplicity and causal sufficiency, on the one hand, and complexity and causal insufficiency, on the other. If c1 were a sufficient cause of e, then, plausibly, e would have no other causes. After all, the entirety of e would be caused by c1, leaving no room for other causes. Its entire existence would be derived from c1. Consider the following analogy: If you paint your entire room blue, then no space is left for someone else to do part of the painting; you have already done it all. The causation of e by c1 would be ‘complete’ and ‘exclusive’ in a similar way. The sufficiency of c1’s causation of e would rule out the complexity of multiple causes. On the other hand, if we could locate a feature of e that could not be caused by c1, then this would be a sign that e has other causes as well, on the assumption that all of e has a cause.

But this does not get us very far. After all, what if we cannot discover whether or not e has a feature that could not be caused by c1? Then we must consider if c1 carries within it the principle of its own existence. Without it, c1 could not be the sufficient cause of e, since something else would be needed in addition to c1 for e to exist. But suppose that both hypotheses indicate that the causes they invoke are self-explained in such a way. Then we are left with (at least) two options: (1) introduce additional pieces of evidence in favor of the hypothesis, or (2) highlight how the hypothesis explains the falisity of its rival. Whether these avenues will be sufficient to resolve the debate is however unclear. Perhaps in the end the discussion will turn on who has the better candidate(s) for an ultimate reality (c1 vs. c2 & c3). As you can tell, I am still thinking through these issues.