Occam's Razor For Dummies
copyright Anthony Horvath, 2005.
Link to Forum. | Link to Main Page | Link to other Essays
Most people dont even know that there is a word, parsimony, let alone what it means. You can find it sometimes within a discussion where something called Occams Razor has been invoked. Occams Razor, or Ockhams Razor after Wilhelm of Ockham who famously employed it, is a principle of explanation that simply put, favors the simplest explanation of a thing rather then the complex, if the simplest will cover the facts as well as the complex explanation.
So, lets say you have observation x. One person puts forward a hypothesis that explains x in one or two propositions. Another person puts forward a hypothesis that explains it in five or six. The most parsimonious explanation, Occams Razor says, is more likely to be the correct one. This is all well and good, but as employed by the skeptic, Occams Razor is the weakest tool in his belt.
You can see already that I started out using the words simple and complex in describing what the razor is but in my example I quantified it by using one or two propositions as opposed to five or six. Occams Razor is not an a priori logical law, which must be true or all reality collapses. Occams Razor is a general principle derived from our experiences (a posteriori). And we know by experience that you can have one proposition uttered to explain something that would actually be far more complicated, and thus harder to believe, then a series of propositions. As I said, Occams Razor is a general principle.
As a general principle, it is very reasonable. One might guess that the actual opportunities to employ it are not all that common, as in many cases we need not invoke any propositions to explain something, we can just buckle down and investigate the matter. Thus it should be evident that when one gets down to it, if you have to appeal to Occams Razor at all, immediate investigation is probably not possible or practical and the case is such that any number or type of propositions are allowed to flourish.
Furthermore, Occams Razor is not nearly sharp enough for daily use. It is more effective when examining cases of explanation where there is wide disparity between the rival hypotheses. Again, using numbers, if you had an explanation with three propositions and another with one hundred propositions, one will be quick to see that the one with three has probably got more going for it. But if one had three and another four, there is probably no rational reason to decide one way or the other; if the Razor is too be invoked at all, it would be cautiously: "The one with three is probably more true then the one with four." Not, "The one with three is true."
All this reasonable talk ends when confronted by its use by a skeptic.
For the skeptic, the Razor is a tool of convenience and is employed as though it were a scalpel when it fancies him even though it is a blunt instrument. Lets take a common case.
A theist, especially a Christian theist, might find that the best explanation for the origin of life on earth would be to invoke God. The blood thirsty blade wielding skeptic rushes in with his little knife and quickly performs math. Invoking God, the skeptic says, is invoking one proposition more then is necessary. Holding onto what they call the default position of naturalism allows them to think that since it is default (at least in their mind- it never occurs to them that not all share that view) their proposition is exactly zero. One is more then zero, and by Occams Razor, they exclaim, it is more likely that their explanation is more correct.
But if their explanation is exactly zero that is the same as saying they have no explanation at all. And that is exactly the situation with abiogenesis, for if the matter could be settled more directly, there wouldnt be any need for the razor at all. Some of them really do have an explanation one less then one. Others acknowledge that that is stupid and re-work it and allow that a purely natural explanation- even if not in hand- is also a single proposition, but that it is to be favored because we have found purely natural explanations for everything else. One would think it would be obvious that the argument has shifted, but sadly, you will probably have to throw up your pinky to deflect the razor anyway. They keep it in a little sheaf on their belt, and I suspect they sleep with it under their pillow.
But abiogenesis is a fine example of how Occams Razor may have two sets of propositions where even though the quantities are near in number, one explanation though possibly more in number involves far less straining against credulity then the other. What I mean is, let us grant for a moment maybe that invoking a God as an explanation for the origin of life really is one more then the naturalistic point of view.
But invoking the naturalistic point of view will necessitate many incomprehensible things down the road. One will have to possibly entertain the principle that perhaps life is self-organizing. You will have to invoke mysterious and untestable- certainly no longer observable- initial conditions for the earth. You will have billions of years of hidden processes multiplying into the billions upon billions of chance events as RNA finally makes its way to DNA, or whatever crazy idea they come up with next. Now, in this case, accepting the naturalistic point of view may be less in number, but it certainly is far more complicated and complex!
It should be like coming upon an automobile and given the two choices: 1. An intelligent agent made this object. And, 2. "Default natural processes and definitely no intelligent agents made this object" (Remember, option #2 is actually less in quantity then option #1 even though it contains at least three propositions. Well just pretend that option #2 contains less propositions then option #1). Now, it will be quite a convoluted process beyond all imagination to conceive of how something relatively simple as an automobile (when compared to a biological organism) came about by natural, unguided processes. But invoking the designer is far simpler. Option #1 is actually much more parsimonious then option #2.
If Occams Razor has any good application, it would be in modern physics where an infinite amount of universes has been invoked to explain (or as a consequence of) quantum phenomena. These universes are not detectable, but the skeptic is allowed to consider the multi-verse without it even occurring to most of them that an infinite amount of universes is far more then one God by any measure. Dont get me wrong, they can think anything that they like. What I object to is the obvious double standard: ANYTHING is to be considered, no matter how crazy, how numerous, how complicated, how preposterous, just as long as it doesnt involve any nasty overtones that might point to a God. That is why you will rarely hear a skeptic being skeptical of the multi-verse and invoking Occams Razor in regards to it.
And this brings me to what I might call the converse of Occams Razor. The skeptic thinks he is being very reasonable in eliminating God from his calculations because invoking God is one too many invocations. He thinks he is a free thinker because of this. But if he is free, he apparently is not as free as me- I get to consider one more possibility then he does. I am allowed to consider whether something is best explained by an intelligent agent, but this option is excluded by the skeptics belief that methodical naturalism will some day provide a better explanation, and by better they mean "Oh God, anything but God."
A final note should be raised. As I said, the Razor is typically invoked because the thing to be accounted for cannot be directly clarified. As a general rule, in those cases, it is certainly useful. A good example was back in the days when it was still not certain whether the earth went around the sun or the sun when around the earth. You could create a model that covered the facts with the sun going around the earth, but it required far too many ad hoc tweaks, none of which were required by the other view. At the time, the question couldnt be answered directly.
In our modern day, various naturalistic processes are invoked that are far more complicated then any that a theist might consider. A common rebuttal often comes up here that the reader should be ready for. It is Dawkins argument from incredulity. He thinks that merely being unable to believe that these processes could account for life as we know it only means that we just cant believe it ie, not a rational reason. But surely the converse is that to accept his proposition(s) would then be an argument from credulity. So, now the theist is maligned for not being skeptical enough! We should be more gullible, like Dawkins!
And here is where Occams Razor is invoked again, but note the shifting sands! In the scenario above, something like the automobile or abiogenesis is being considered, and it is explored whether or not mere quantities of propositions gives us good information on the simplicity or complexity of the claims. But if once it looks like invoking God is actually far less in number and simpler too in regards to the subject at hand, the skeptic will bump the razors use up into a new category. You thought you were evaluating abiogenesis, but now you are talking about the whole universe. And the skeptic thinks hes really got you, since God and the universe is clearly more in number then the universe. But you were talking about abiogenesis, not the universe. This is a common ploy. In the situation you were exploring, intelligent agency was more parsimonious then the alternative, so the skeptic changes the situation.
Despite all that Ive said, I dont want anyone to think that I discount the proper use of the GENERAL principle of Occams Razor. I find it has applications; I already named one- the multi-verse. The right tool for the right job: Thats real life advice. If you find Occams Razor employed in the abstract in a way that no man would employ it in real life, you should probably be suspicious.