Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Occam's razor

  1. Jan 28, 2010 #1
    Is Occam's razor becoming nothing more than a poor cliche those days ? To be more exact, Im referring here in special to the it's more naive version for which internet seems very fond, quoting it in many different flavors time and again ... "The simplest explanation is the most plausible one".

    After all nature couldn't care less about the fact we like simple things. It proved to be complex and weird. Quantum mechanics put an end to IMO to the belief that nature is simple.
    String theories introduced us to worlds with 10, 11 , whatever dimensions.

    The correct theory from a pool of theories with equal qualities may prove time and again not to be the simplest one. Yet the cliche still stands.

    (of course, Im not referring here to the fact that Occam's razor has 0 value as a scientific result. but only the philosophical aspect)
     
  2. jcsd
  3. Jan 28, 2010 #2

    DaveC426913

    User Avatar
    Gold Member

    I don't think you're really understanding Occam's Razor. You seem to think it simply means "nature is simple". And when nature turns out to not be simple, you think Occam's Razor is non-applicable.
     
  4. Jan 29, 2010 #3
    they are still theories though
     
  5. Jan 29, 2010 #4
    I think that it keeps being used so many times, it gets simplified more and more until the real purpose of it is lost.
    It's like that game where one person whispers something to another person, then that person whispers it to the next, and so on until they get to the last person and by the time it gets there, the message is sometimes completely different from what it was originally.
     
  6. Jan 29, 2010 #5
    Do you know of any working explanation of physical phenomena at the atomic level that is more simple than quantum mechanics?
     
  7. Jan 29, 2010 #6
    No. Did Einstein knew one when he stated "God does not play dice" ? |I dont know, but this never prevented him from never being really at ease with the probabilistic nature of quantum mechanics. He basically rejected Occam's Razor in this case (or embraced it, probably thinking that something simpler than a probabilistic nature must be at work) and continue to pursue something he never found, developing further his own theories.

    IMO the moment in time most applicable for this rule of thumb was back then , in the first quarter of the last century. From a long time we dont really need heuristics and rules of thumb for QM, we believe its true because the outstanding predictions it made and the phenomenal accuracy with which they where verified.

    Anyway I found a nice paper which discusses all I wanted to know. It has a mini-chapter on the myth of simplicity and simplicity criteria which are interesting to read.

    http://www.scientificexploration.org/journal/jse_21_1_gernert.pdf [Broken]
     
    Last edited by a moderator: May 4, 2017
  8. Jan 29, 2010 #7
    Occam's razor does not put any limitations on the existence of a simpler explanation. It is also does not make any assertion concerning the nature of the physical world.So if the phenomena explained by quantum mechanics could me explained by a simpler theory with the same results then by occam's razor the simpler one probably the real explanation.The problem is that two theories that would always give the same result would be indistinguishable so saying that one is the real theory is nonsense anyway. The usefulness of Occam's razor comes out in a situation like the following: If prediction of a theory is contradicted by 100 experiments then the theory is probably wrong. If 99 good experiments give the same result and 1 contradicts them then that 1 experiment was not done properly.
     
  9. Jan 29, 2010 #8
    Sure, its probably the case with Lagrangian formalism and Newtonian formalism. Good point.

    What do you think about a case like in which exist more theories, each perfectly anchored in the science of the day (i.e not crackpot stuff) , both sane mathematically and consistent, with a different degree of complexity, but none of those able to do any predictions whatsoever which can be verified by experiment. (like the situation that we dont have predictions which can be verified today from String Thories).

    Does it worth invoking Occam's razor and focus on the less complex one ?
     
  10. Jan 29, 2010 #9

    Occam's razor advises us to choose the simplest expanation which is consistent with our observations. If the simplest explanation is very complicated, that's ok as long as there isn't a simpler option. And of course if we can find further information to distinguish between explanations then we have no need to use Occam's razor. You're right that from a pool of theories the correct theory may not be the simplest, but we are advised to choose the simplest one if we have no other means to distinguish them.

    However, I would say that other factors such as the aesthetic appeal of a theory probably come into play in real life. And the simplicity of a theory seems like a subjective measure which people may disagree on.
     
  11. Jan 29, 2010 #10

    Char. Limit

    User Avatar
    Gold Member

    Don't cut yourselves on Occam's Razor.

    Sometimes the simplest explanation really is the best. For example, the simplest explanation is that both DanP and Dave are annoyed at each other, but both had good intent.
     
  12. Jan 29, 2010 #11

    DaveC426913

    User Avatar
    Gold Member

    There will always be people who misuse platitudes. Why dwell on it?

    Has that happened? If not, why are you borrowing trouble?
     
  13. Jan 29, 2010 #12
    Actually, Wikipedia states on the Occam's razor entry that there where cases in which the rule delayed progress. An example was given : "...It originally rejected DNA as the carrier of genetic information in favor of proteins, since proteins provided the simpler explanation. "

    I do not know in how many instances this happened or how founded the claim is. I'll read this night the paper I linked in my previous posts, since it appears (from my cursory glance) to contain a larger analysis of this.

    But I do not think that this is very important (how many times happened, if at all) from a strictly philosophical point of view.

    As the current science theories become more and more complex and:

    - it will be often very difficult to decide which is the "simplest" one, because perceiving a thing as simple is highly subjective. two experts at the top of their field may very well see 2 different theories as "simple and beautiful". It wouldn't be unheard.

    - today several theories are mere mathematical models, and cant benefit yet from the advantage of making predictions which can be confirmed experimentally

    In those conditions Occam's razor is still useful or becoming a cliche. I was curious what ppl think about it.

    It's actually not borrowing trouble. It's discussion for the sake of the discussion, hence it was posted in a philosophy section.
     
    Last edited by a moderator: Feb 26, 2010
  14. Jan 29, 2010 #13
    Occam's Razor, or the principle that "plurality should not be posited without necessity", is an often abused concept on the internet.

    But it is not that the simplest hypothesis is the correct one, but that the hypothesis that makes the fewest assumption is more likely to be correct. This is usually the simplest, but not exclusively.
     
  15. Feb 25, 2010 #14

    JTE

    User Avatar

    Calling something a cliche is not the same as saying it is untrue. Worth separating these two. Occam's Razor is definitely cliche, so call it Parsimony and give it some fresh legs for a while. Politicians are good at this sort of trick.

    As for whether it's true, it's certainly the simplest explanation of why it has so often been observed to be true.
     
  16. Feb 25, 2010 #15

    Dale

    Staff: Mentor

    This is also my understanding of Occham's razor.
     
  17. Feb 25, 2010 #16
    The problem with Occam's razor is that it has not kept up with the times. In the latest TV commercials I see that razors now have 5 blades. Simple things become complicated over time. Perhaps a future version of Occcam's razor will say the physics laws tend to fit into one of the first 5 simplest explanations.

    Seriously though, what does this principle really do for us? It's human nature to keep things as simple as possible. It's human nature to seek out a simple pattern buried in a confusing situation. Sometimes we even think we see a simple pattern when it is not really there. The principle is basically useless. We do science with the scientific method. Anything important automatically falls out of that process.
     
    Last edited: Feb 25, 2010
  18. Feb 26, 2010 #17

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Why was this "delayed progress"??

    At that time, all you could find out about DNA wa that it was, largely,...INERT.

    The idea that the nucleic acid had, for example, som structural, stabilizing role was at least as consistent with available data than that it was the carrier of genetic information.

    So, I must ask you, in what way did that opinion "delay" progress?

    Was it lack of progress to decode many proteonic interactions?

    Did the idea prevent development of technology that could be used to determine the true role of DNA?
     
  19. Feb 26, 2010 #18

    Gokul43201

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    :biggrin:
     
  20. Feb 26, 2010 #19

    russ_watters

    User Avatar

    Staff: Mentor

    I'll take it a step further. If we're looking for a method which overall provides the most benefit for the least effort, then it doesn't matter if there are individual examples where it led us down the wrong path.
     
  21. Feb 26, 2010 #20

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Agreed.

    A theory that needs to get 10 assumptions proven right is more research expensive than one needing only to prove 5 assumptions, UNLESS those 10 happen to be vastly easier to verify...
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook