Ockham’s Razor and the Spooks

Anyone who knows me well knows that I hate many, many things.  In a time when atheists are fighting desperately to be seen as good, kind, well-meaning folk, I brazenly admit to being an angry, pissed-off atheist.  But of all the things that anger me—whether it be God, religious fundamentalists, chiropractors, naturopaths, conspiracy theorists, or recently-pregnant women who won’t stop talking about their damned kid—few things piss me off more than people who misuse and misunderstand the concept of Ockham’s razor.  Mary Roach, the well-known pop-science author of books like Spook, Stiff, and Bonk, is the latest to anger me in such a manner.

In her book Spook: Science Tackles the Afterlife, Roach portrays herself as a sort of wishy-washy, noncommital skeptic.  This, too, is the sort of thing that angers me, as one should always devote one’s life fully to skeptical principles with flair and gusto.  However, when she totally abuses the principle of Ockham’s razor, defiling it with her innocent, questioning prose, I couldn’t help but look at Spook with a sort of enraged pity, not unlike the way one looks upon a mangy dog that has suddenly began humping a small child; it is a disgusting, wretched display, but we almost forgive it because it is too stupid to know any better.  Indeed, I almost feel saddened to have to eviscerate Roach’s well-meaning, but horribly incorrect, description of Ockham’s razor.  But it must be done, for the sake of edifying my readers and delighting myself.

For a little context, Spook essentially details the long history of pseudoscience concerning the afterlife and ghosts.  She investigates soul-weighing experiments, crackpot physical theories about the survival of consciousness after death, and so on.  For me, it was a light, fun read until I finally read her description of Ockham’s razor, which caused me to have a raging seizure and dress my dog up as an effigy of her book while beating it relentlessly.  She explains the principle in a chapter about the capacity for infrasound (low-frequency and inaudible sound waves) to induce paranormal experiences, like sensations of a felt presence.  She ends the chapter like so:

If you ask me which is the more likely explanation—infrasonics or spirits—I will tell you to apply the wisdom of Occam’s razor, a principle which holds that the simplest, the least far-fetched, of two competing theories is the place to put your money. But depending on who’s shaving, Occam’s razor yields manifestly different views. To those who believe in an afterlife, the most straightforward explanation for hearing your dead dad is that you’re hearing your dead dad’s spirit. Infrasonics and vibrating eyeballs and fight-or-flight responses would, given this particular worldview, seem to be needless and unlikely complexities (p. 237).

You can't use Ockham's razor on just anything, especially suicide attempts.

Mary Roach’s understanding of Ockham’s razor is distressingly common.  Few people understand what the principle actually entails, largely owing to an unfortunate and misleading sound-byte interpretation frequently rendered as, “The simplest explanation is the best.”  Of course, worded like that, the principle is indeed stupid and ambiguous and should be pelted with baboon feces.  After all, if mere simplicity provides the royal road to truth, then that crazy, complicated science stuff, which is inundated with complex equations and statistics, goes right out the window in favor of the Ken Ham/religious mystic view of reality, complete with men riding atop saddled dinosaurs and the Earth resting upon an infinite series of stacked tortoises.  Truly, if the simplest explanation really were the best, then this overly simplistic paraphrasing of Ockham’s razor would suffice.  Unfortunately, Ockham’s razor doesn’t rely only a vague, amorphous conception of simplicity; instead it relies on a very precise and complex definition of exactly what makes one theory more epistemologically complex than another.

Roach goes wrong in just this way, by assuming an ambiguous definition of Ockham’s razor in terms of whether a theory is “simple” or “far-fetched”.  The principle is best explained in its more complex, fleshed-out form (as opposed to the walking-dead zombie form used by Roach) to avoid such confusions.  Basically Ockham’s razor states that the theory that makes the fewest assumptions, given the current state of evidence, is the best theory.  In essence, the principle only demands that the claims of a theory be founded in some sort of evidence or fact, as those theories that proliferate in hypothesized, assumed entities for which there is no evidence are less likely to be true.  For instance, if I explain the water cycle with stages of precipitation and evaporation, this water cycle theory is more parsimonious than a hypothesis that posits a water cycle model with the stages of precipitation, evaporation, and magical elves.  The latter model would be rejected by Ockham’s razor because it posits magical elves when there is no additional evidence in support of their existence, and because the model that posits only precipitation and evaporation adequately models the water cycle without positing any elves or magic.  In a nut shell, Ockham’s razor tends to lead to truth because it prohibits baseless assumptions, and as we all know it is much easier to be incorrect when we come to conclusions based on assumptions for which we have no evidence.  If you want to add magical elves to your theory, find evidence of magical elves.  It’s really quite simple.

Now let’s apply Ockham’s razor to competing explanations concerning sensations of ghostly presences, as in the example from Spook quoted above.  One hypothesis posits that known entities like low-frequency sound waves and flight-or-fight stress responses can produce these sensations, with a smattering of evidence to support the claims.  The other hypothesis posits that actual ghosts are producing these sensations, with no evidence of ghosts other than these disputed sensations.  According to Roach, Ockham’s razor cannot decide between these two hypotheses, because one’s worldview—depending on whether it accepts the supernatural or not—would influence what one considers the “simpler” theory.  But this is not a correct application of the principle, as Ockham’s razor is decidedly not ambiguous with a proper contextual definition of the word simple.  The first hypothesis posits entities that everyone agrees already exists; we have plenty of evidence demonstrating that low-frequency sound waves and flight-or-fight responses are present in reality.  The other hypothesis, however, invents a new existential entity, and further a new existential category (the supernatural), in an attempt to explain the evidence, even in spite of the fact that there is no other confirmation of the existence of such paranormal entities.  Now, if we can explain the phenomenon with already-existing entities like sound waves, there is no further need to posit anything more, much less an entirely separate and ineffable spiritual plane of reality.  The paranormal claims are thoroughly shaved away by Ockham’s razor, regardless of your worldview, because the principle depends only on reality and evidence, not ideology!

So just as we reject magic elves as necessary to explain the water cycle because already-known processes like evaporation and precipitation can explain the cycle, so too are we led to reject paranormal hypotheses concerning strange sensations of vague presences when already-known processes can account for the evidence.  Ockham’s razor, far from being a tool that can be wielded haphazardly by two sides of a dispute to come to startlingly different conclusions, actually slices open paranormal claims and reveals them for the empty, silly nonsense they are.  Like any good razor, Ockham’s razor only has one handle, and it can only be grasped from that one side.  Anyone who tries to grab it from the other end only winds up getting sliced.

Atheism and Absence of Evidence

Many people believe that the main difference between agnosticism and atheism is that atheism is dogmatic and closed-minded whereas agnosticism is more open and accepting. In reality, though, the difference between agnosticism and atheism isn’t a matter of degrees of certainty; the differences are primarily epistemological. The rowdy, unshakeable, dogmatic boisterousness associated with atheism is not a defining feature of the position, but is instead a persistent mischaracterization. In the olden days of yore, atheists were merely stereotyped as immoral, evil baby-eaters. Nowadays, we are characterized as arrogant and dogmatically certain immoral baby-eaters! This misguided stereotype must be put to rest, for we baby-eating atheists are actually quite moral and frequently embrace positions of fallibilism—the exact opposite of dogmatism!

Now, atheism is primarily seen as dogmatic because it assumes nonexistence given an absence of evidence. This is almost always the contentious point. An agnostic would assert that absence of evidence is not proof of absence, maintaining that asserting nonexistence without any evidence is an argument from ignorance. An atheist, on the other hand, finds it perfectly reasonable to conclude that something does not exist given an absence of any evidence. The atheist does not believe with certainty that God does not exist given the lack of evidence, only that this is more likely to imply that God does not exist. Thus, the point of contention between agnostics and atheists isn’t about certainty, but about whether nonexistence can be inferred from lacking evidence. As mentioned earlier, the dispute is mostly epistemological.

The problem is rather obvious. If something did not exist, we would expect to find no evidence of it, but even if the thing did exist we could expect to find no evidence of it, for the evidence could be hidden from us. Similarly, if I had no evidence that my dog could fly, and then argued, “But you have no evidence that my dog can’t fly!”, this would hardly be a convincing argument, as I’d be making the claim in the absence of any evidence, judging that my dog could fly from a position of complete ignorance.

These problems are not legitimate, and they pose no sort of barrier to professing atheism, though. While it is indeed an argument from ignorance to assert that my dog can fly in the absence of evidence, this is because a flying dog is the sort of thing that can be supported with evidence. If a dog could fly, we’d expect to find evidence of the dog zooming through the skies, perhaps dropping turd-shaped surprised on our heads. But the same is not true of nonexistence (particularly concerning the bit about shitting on one’s head from the sky). If something does not exist, would you expect the nonexistent thing to produce evidence of its nonexistence? Of course not. One would expect that the nonexistent thing would produce no evidence whatsoever. Thus, in the case of nonexistence, the absence of evidence is exactly what we’d expect to observe.

Nevertheless, it is true that we can never be certain that a thing does not exist, owing to the fact that we may merely be ignorant of the evidence. The agnostic is entirely correct to point this out. The problem, of course, is that atheists tend to take a pragmatic, fallibilistic epistemological approach, whereas agnostics seem tethered to the idea of an epistemology rooted in logical certainty. The atheist doesn’t care if his position is uncertain; he doesn’t necessarily demand logical certainty. All that matters is that the position fits the evidence better than other hypotheses. In the end, an agnostic who demands certainty from claims of nonexistence is demanding the unreasonable and the impossible.

Suppose, for instance, we apply the agnostic’s demand for certainty to positive claims. Most agnostics would accept that the theory of evolution adequately explains the origin of species and biological variation. Of course, the theory of evolution is open to the same sorts of criticisms that agnostics apply to atheism. For example, one can never be certain that the theory of evolution is true, because we could simply be ignorant of the evidence that falsifies it. The theory of evolution, also like atheism, is underdetermined by the data. The absence of evidence for God could be explained by our ignorance of the evidence; similarly, even given all the evidence for evolution, we can invent ad hoc explanations for our biological observations that also fit the data, such as, “God created organisms in such a way that they would appear evolved.” But the fact that we can apply different explanations to the data doesn’t call the theory of evolution into doubt, because these ad hoc explanations needlessly multiply explanatory entities, calling into existence a god-like entity when there is no additional evidence of such a being. The God explanation is also untestable, because even if we observed that nothing in biology supported evolution—there were no homologies, no vestigial parts, no transitional fossils, no progression of species through time in a linear fashion, etc.—we could still say, “God created life.” An omnipotent being could create life that apears to have evolved or appears to have not evolved. No matter what we observe, this explanation fits, and so it is a much poorer explanation than evolution. Evolution, on the other hand, fits the data perfectly, even though it could have been falsified at any moment, which shows that its fitting the data is surprising and highly unlikely—unless the theory were true! When creationism fits the data it is no big surprise, as it would explain any observation whatsoever, and hence finding observational support consistent with creationism is not all that surprising or unlikely.

Obviously, the situation is the same with claims of nonexistence. If something did not exist, we would expect it to produce no evidence, and this would be the only expectation. Now, it is true that the thing could exist and we could be ignorant of the evidence. So if the thing did exist, we could expect evidence of it, or we could expect it to be hidden and hide its evidence from view. Notice that, in this case, any observation is compatible with the thing existing. But only the absence of evidence is consistent with nonexistence. This shows that assuming nonexistence is most prudent. Not only is the prediction more exact, but it is also falsifiable. In that sense, the position of atheism is far from dogmatic. Instead, it is pragmatic—the atheist assumes nonexistence, but knows this position could be falsified so long as the theist produces legitimate evidence of God’s existence. To criticize atheism for its uncertainty is thus as silly as criticizing the theory of evolution for its uncertainty. What matters is that evolution assumes the least and fits the evidence best. Similarly, atheism assumes the least and fits the evidence (or lack thereof) best.

Suppose, though, that if God existed, we wouldn’t expect evidence of his existence. That is, suppose God was totally transcendant and unknowable. In this case, the atheist can’t argue that nonexistence is the most plausible hypothesis simply because it exactly predicts the absence of evidence. An unknowable, mystical, transcendant God would likewise predict only the absence of evidence.

At first blush, this appears to be a winning rejoinder. The problem in this case, though, is that professing belief in such a deity is to essentially profess a belief in nothing. If one believes that God is totally unknowable and mystical, then to say, “God exists” amounts to saying, “I-don’t-know-what exists.” The only reason an atheist cannot disbelieve in such a deity is because the theist has not bothered to even define it. It is as if the theist were saying, “Djklerhi exists!” and then, when asked to clarify, saying that Sjklerhi is uknowable and hence the theist can’t say anything more about it. How the theist knows an unknowable entity exists is also never explained. So while rendering the deity transcendant and mystical may avoid the criticism concerning the lack of evidence, it just runs into the criticism of incoherence, and thus this is really no threat to the position of pragmatic, fallibilistic atheism.

Because claiming nonexistence based solely on the absence of evidence is justified through fallibilistic epistemology, atheism is similarly justified. Of course, throughout this post I have assumed, for the sake of argument, that there is no evidence for or against God. This is, to put it mildly, a rather generous assumption. Those conceptions of gods as omnipotent and morally perfect, for instance, run into all kinds of logical difficulties concerning what it means to be all-powerful, whether this is consistent with the capacity to do no evil, and various other self-contradictions. Those conceptions of all-powerful, morally perfect gods who created the universe, physically interact with the world, and care about humanity are also problematic, as the existence of evil, the miniscule scale of humanity’s presence in such a vast universe, and the hostility of the natural environment to life all seem to be at odds with these conceptions of God. As such, the case for atheism is really much stronger than merely asserting, “There is no evidence for God!”, as there is in fact much evidence that implies various gods do not exist at all. But in the end even if there were no evidence for or against God, atheism would still be the most reasonable position, because epistemic fallibilism is a reasonable position, and because absence of evidence is more than adequate evidence for absence.