1

Does Occam's razor have any limitations?

That is to say, are there any philosophers/philosophies which find little use for this scientific and philosophical rule? If so, who/what schools of thought?

Though you need not be concerned with the following aspect in your response, I am particularly interested when Occam's razor is applied within the field of computer science and engineering where "...the simplest of competing theories are to be preferred..." does not always produce the best results.

2
  • 3
    There is no need for limitations because there is no rule. The "simplest" and "other things being equal" are too vague to prescribe anything in particular, which would then have to be "limited". Most philosophers take the razor as just that, a vague heuristic that may give some guidance in special contexts, not a rule. See SEP, Simplicity.
    – Conifold
    Dec 11, 2020 at 23:15
  • Also see philosophy.stackexchange.com/questions/2295. There are multiple variants of statements that are loosely called Occam's razor. For questions it is this most useful to clarify which variants you want to know about.
    – tkruse
    Dec 15, 2020 at 3:36

3 Answers 3

2

I am particularly interested when Occam's razor is applied within the field of computer science and engineering [...]

An software engineer's perspective:

Occam's razor is bit more subtle than 'simplest is always the most correct'.

First: 'simplest' is very subjective. Facing a server crash, I could theorize: a bad piece of code, solar flare, a dead squirrel in the rack, the CPU falling of, or just an act of God. At some level, the first reason is the least simple. However Occam's razor doesn't call for 'simplest', but for the fewest entities. From that perspective, it recommends looking at an explanation involving just the code and the hardware it runs on before introducing a squirrel (or God) in the system.

That said, Occam's razor also says that entities shouldn't be multiplied without necessity. And if after looking at the code and the hardware configuration, I can't find a good explanation for the crash, introducing the squirrel might become necessary (yes, it did happen).

Similarly, it might be necessary to introduce relativity to Newton's gravity when it proves insufficient.

There are some pitfalls: it's easy to get enamored with the most mathematically elegant explanation, or dismiss inconvenient new data that goes against an established (or favorite) theory. But those are issues unrelated to Occam's - a scientist unwilling to consider relativity or a conspiracy theorist unwilling to not consider aliens are equally guilty.

As long as Occam's razor isn't used to justify intellectual dishonesty, it's a pretty useful guiding principle.

3
  • @DavidBlomstrom you appear to be on a crusade - have fun. May I suggest that finding out that Occam's razor can be explained as simple thought guide, rather than a convoluted theory encompassing all aspects of life, and/or a fallacy by hard science to dismiss other's arguments is itself an application of Occam's razor?
    – ptyx
    Dec 21, 2020 at 20:13
  • I haven't run into someone name dropping Occam's, no. I'm not doubting it happens - propaganda is based on fallacies, and argument from authority is a classic. But I'd think the counter would be to dismiss the fallacy rather than poor Occam. (And yes, that's hard - why fallacies are effective). If you want an example of contemporary scientists using Occam, then I'd point you at the EmDrive story. In the end, conservation of momentum wasn't violated. But it made the news because proving the simplest explanation (bad experiment) got quite hard.
    – ptyx
    Dec 21, 2020 at 22:39
  • I don't think you'll find a paper saying "here are theories A and B, they're equivalent as far as I can tell, so I'll use Occam's to proclaim A is right and B should be dismissed". That'd just be trolling for someone to figure out an experimental setup to falsify one of them and prove them wrong.
    – ptyx
    Dec 21, 2020 at 22:50
1

No, even writers who reject specific theories despite Occam's Razor still use Occam's razor all the time. That's because Occam's razor is just a principle of common sense without which no coherent argument can be made. It's a necessary part of communication and rational thought.

Without Occam's razor, an argument would always have to made by explicitly discarding infinitely many theories. Instead, when making any argument, any writer implicitly assumes his argument does not need to mention the infinitely many other more complex ways any sentence could be understood.

Though if you look for an easy example of a philosopher rejecting the result of Occam's razor, then Pascal's wager would seem to be quite obvious (though Pascal makes only this one exception, and does not generally reject the usefulness of Occam's razor).

0

Ocean's razor has the severe, but by design limitation of being a simple heuristic, a rule of thumb that, when understood, can help you prioritize your investigations.

I, too, found it pretty dumb for a long time because I was introduced to it by the movie Contact, where it's used like a proof, dismissing the alien hypothesis without inquiry, which it is not meant to do. As this movie is probably the most prominent introduction of the Razor in pop culture, I suspect Jodie Foster is sadly the reason why so many people got the wrong idea about it (but I still love you, Jodie).

It kind of disagree with your assessment about engineering. Surely, the simplest system might not always be the right solution, but the simplest system that does the job is certainly preferable to an unmaintainable, impossible to concisely document mess of a solution (notwithstanding economic constraints like reuse of legacy systems, production costs and technological debt). I have done a lot of debugging in my youth, and Occam's razor clearly helped me prioritizing my investigation by looking for silly mistakes before looking into a possible but improbable compiler bug.

2
  • Your analogy using engineering is not valid. And never mentioned nor implied that I was interested in comparing the "simplest system" with an "unmaintainable, impossible to concisely document mess". My understanding is that Occam's razor gives precedence to simplicity and that is not always an option in engineering. But thank you very much for your input.
    – tale852150
    Dec 13, 2020 at 15:46
  • That's just an addendum because you brought engineering up in the first place. Let's agree we disagree about engineering. I still think the main problem in your question is you don't understand what the Occam's razor is meant for.
    – armand
    Dec 13, 2020 at 22:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .