Ockham’s Razor is the rule that we shouldn’t assume more kinds of things than we need to.  Roughly, if a simpler explanation can do the job, then simpler is better.  It’s a pretty good rule of thumb if it is used correctly, but it usually isn’t.

I came across the following bad example in a review of a book written to celebrate the Razor:

If a friend tells you “I’ve seen a UFO!” what would you think?  It might have been an alien spacecraft -- or perhaps the friend was mistaken.  The first possibility requires numerous unproven assumptions about extraterrestrial life; the second is consistent with what we know about human fallibility.  The 14th-century Franciscan friar William of Ockham was never troubled by flying saucers, but he did see the importance of eliminating unnecessary assumptions -- the principle known as Occam’s Razor.

Now it’s true that a friend who says he saw a UFO – which means an unidentified flying object, mind you – may think that what he saw was an alien spacecraft, and yes, that would assume the possibility of extraterrestrial intelligence.  But which hypothesis assumes less – the one that assumes that there might be extraterrestrial intelligence and that this experience counts as evidence, or the one that assumes that there couldn’t be extraterrestrial intelligence so that nothing at all could count as evidence?

Assuming a negative isn’t the same as not assuming.  And it isn’t assuming less to assume a universal negative.  Besides, the idea isn't to posit the fewest suppositions, but the fewest suppositions to account for the facts; a simpler hypothesis is better only if it explains just as much or more.

Before speculating about a friend’s sighting, I would want to know exactly how what he thought he saw looked, why he thought it odd, whether anyone else saw it, whether there is any corroborating evidence, whether he is truthful, whether his observations are generally reliable, the state of his sobriety when he made the sighting, and a lot of other things – including what other hypotheses might be offered to explain it.  Note well, the reviewer has no other hypothesis.  He only assumes, Well, it couldn’t be that.

Turning from UFOs to metaphysics, the reviewer thinks Ockham’s Razor proves that things don’t have essences.  For it is much simpler not to assume essences than to assume them; the only reason we call cats “cats” is that we have placed them in the set of things we do call cats.  We don’t need to assume some mysterious thing called cat nature, or catness.  But wait a moment.  Why do we place certain things in the set of cats, and not place others there?  We don’t call them cats because we have placed them in the set; we place them in the set because they are cats.  They all do have catness.  Dogs don’t.

Pushing essences out the door turns out to be harder than it looks.  In fact, it’s something like pushing – well, cats.

Leaving the book reviewer to his devices, let’s consider another common abuse of Ockham’s Razor.  A certain attempt to refute one of the current arguments for the existence of God runs something like this:

a.  On a certain hypothesis – call it fine-tuning -- the physical constants of the universe have the precise values they would need to have to permit the existence of life like us, so there must be a First Cause to give them those values.  Call it a Creator.

b.  On another hypothesis – call it the multiverse -- the appearance of fine-tuning is an illusion, because there is actually an infinite number of universes, each one with different physical constants.  We just happen to be in one in which the values of the constants permit us to exist.  Lucky us.

c.  But the multiverse hypothesis is simpler, because it doesn’t need to assume a Creator.

d.  Therefore, we should prefer the multiverse hypothesis.

The problem with this reasoning is that the multiverse hypothesis is not simpler than the hypothesis that the universe was created.  In fact, it is both more complex, and explains less.  For in the first place, it assumes an infinity of universes which are all, in principle, unobservable, just to avoid assuming a single Creator.

And in the second place, it doesn’t avoid the Creator after all, for it still doesn’t answer the question, “Why is there something and not rather nothing?”  It merely replaces the question of why (instead of nothing) there is one universe, ordered in a particular way, with the question of why (instead of nothing) there is an infinite array of universes, each ordered to all the others in a particular way.

All other things being equal, simpler explanations are better.  But be careful about what you call simpler and what you call equal.