Wednesday, August 18, 2021

Antiexplanation

If an explanation is a truth or hypothesis that removes or would remove mystery from the proposition to be explained, then an antiexplanation is a truth or hypothesis that adds or would add mystery to the proposition to be explained. Like in the case of explanations, we need to be sensitive to context with antiexplanations. That Alice dislikes bananas is, in typical contexts, antiexplanatory of why Alice ate the banana. But if we add to the background that it’s Lent and Alice wishes to do penance, then Alice’s dislike of bananas becomes explanatory.

It is widely held, though still moderately controversial, that:

  1. The fact that a hypothesis p is explanatory of some known truth is evidence for p.

A parallel claim about antiexplanations would:

  1. The fact that a hypothesis p is antiexplanatory of some known truth is evidence against p.

This sounds even more plausible than (1). In a typical context, the antiexplanatoriness of a dislike of bananas to actual consumption of a banana provides evidence that Alice who ate a banana does not dislike bananas. Similarly, the fact that Bob is in perfect health is antiexplanatory of Bob’s death, and hence if Bob has died, we have evidence that Bob’s health was imperfect.

There are lots of explanatory arguments in philosophy based on (1). But it would be worth exploring whether one can’t also give antiexplanatory arguments based on (2).

In fact, I think some fairly intuitive arguments can be rephrased as antiexplanatory arguments. For instance:

  1. Materialism is antiexplanatory of consciousness.

  2. Consciousness is a known fact.

  3. So, we have evidence against materialism.

The thought behind (3) is simply that there is intuitively something particularly mysterious about a purely material thing having a conscious point of view.

C. S. Lewis’s version of the moral argument for theism can be taken to be in part an antiexplanatory argument.

  1. Atheism is antiexplanatory of moral law.

  2. Moral law is a known fact.

  3. So, we have evidence against atheism.

Further evaluation of such arguments would call for a deeper philosophical analysis of antiexplanation and an examination of (2). This is a task worth doing. Someone should do it.

6 comments:

Benjamin Stowell said...

I've wondered if simplicity can be subsumed into explanatory power. The relatively simple theory introduces fewer items in need of explanation and therefore has greater net explanatory power over the relatively complex theory. Put another way, the complex theory contains (at least potential) anti-explanations that the simpler theory doesn't have to worry about. To use a real world example, the atheist (allegedly) gets to sit back and relax while the theist sweats about the many puzzles of God.

(Although, ironically, simplifying my theory of theoretical virtue might just affirm the intuition that simplicity is independent of explanatory power ^_^)

Alexander R Pruss said...

Actually, I am inclined to disagree with your suggestion, enough so that it makes me want to modify or clarify my account of antiexplanation. For it seems to me that a good explanation can create new, bigger mysteries. Suppose I see milk on the floor. I then observe surveillance video of three people carefully drawing out an irregular shape on the floor with a pencil, and then, with the aid of syringes, they fill this shape with milk. A simple mystery--the spilled milk--has been replaced with a bigger one--the conspiracy. But nonetheless this is a genuine explanation, albeit one that gives rise to new puzzling things. It's not an antiexplanation.

My suggested clarification is this: When I am looking for an explanation of a proposition p, an antiexplanation makes p more mysterious. My phrase "add mystery" was problematic, because one way to add mystery to p is to bring in additional mysterious propositions--say, that three people with care and precision made this puddle of milk--that call for explanation. But that's not what happens in antiexplanation. In antiexplanations, the explanandum itself--the proposition p--becomes more mysterious given additional background information.

Or something like that. I could be barking up the wrong tree entirely.

Benjamin Stowell said...

I see; so for this case the milk conspiracy is not an anti-explanation, at least not in regards to any known data. The conspiracy theory does add mystery, but that poses no threat to the explanatory power of the theory vis. the spilled milk.

Then I have changed the topic to theoretical virtue :)

I'm thinking of situations where the explanatory power is equal, but one theory is simpler than the other. Why prefer the simple theory? Maybe: Because the simple theory results in overall less mystery.

Without the surveillance data, a better theory would be that someone dropped a cup and spilled milk and didn't clean up after themselves, because in this case I don't add the mystery of the conspiracy.

swaggerswaggmann said...
This comment has been removed by the author.
swaggerswaggmann said...

3 is just wrong, as a simple computer show us how even simpler event can give a complex result. Argument from incredulity...

Quantum physics is antiexplinatory to reality, as it is less believable and add more mysteries that newtonian mechanics...

Alexander R Pruss said...

Quantum mechanics isn't introduced to explain "reality". It is introduced to explain specific experimental observations. And in that connection, see my response to Benjamin.