Occam's Razor ... Always

Post Reply New Topic

Actually, it just might.

I agree with you MCalavera.
I think a lot of people misinterpret Occam's Razor as "the simplest explanation is the best explanation", when it's actually more like "the simplest form of the best explanation is the best explanation".

It's also important to remember, like you said, that Occam's Razor doesn't carve a certain theory in stone, it can change with new evidence. It's also important to recognize an unproven theory as unproven, and therefore open to reassessment. Always keep an open mind, as they say. At the same time, it makes logical sense to assume certain theories (as defined through Occam's Razor) are true for the purpose of scientific advancement until evidence reveals them to be false. Example: In the past it was perfectly logical to assume the Theory of Relativity was true, because given the evidence at hand it complied with Occam's Razor. Recently the theory has been proven to be not entirely true, so Occam's Razor no longer fully applies to it.

Although Occam's Razor can also be applied to another principle, when it comes to mathematics and formulas. The theory of gravity has recently been proven not fully true in all circumstances, but it still comes up with the correct mathematical answers to a lot of equations. Even though there's more correct formulas we can use now, it sometimes still makes sense to use the old formula for things that we know it will give the right answer for, just because it's a less complex and time consuming process.

In the end, you could say that Occam's Razor refers primarily to the issue of efficiency, but without sacrificing accuracy.

Last edited by Bitoku on 09 Jul 2013, 12:15 pm, edited 3 times in total.

Bitoku - if you look at it as 'strain the variables and find the causal thread' then I agree entirely, that's the end goal really of any honest inquiry whether scientific or otherwise.

Occam's razor only adapts to new evidence when the individual applying Occam's razor accepts the new evidence. As I stated before, Occam's razor should be used as a starting point, and not as a scientific method.

For example: There was a theory that the earth's surface is expanding, which at the time the theory was proposed was the simplest solution to continental drift. Plate tectonics came around and despite the new evidence was not accepted by the majority in the field for decades. There are still a few college professors out there who will not accept plate tectonics as anything more than quack science.
And as for subatomic models of physics, there is an extremely large group of people who teach physics that think it is all bunk because it seems too improbable, despite some very in depth evidence.

Be careful with Occam's razor, it can cut your brain.

Occam's Razor is a rule of thumb. The simplest explanation-that fits all/most of the facts- is the most likely.

Its not a substitute for investigating. It is however- a good starting point of any investigation.

Now if you're dealing with deliberate deception on the part of a criminal-then yes- a criminal will most likely make up a simple lie to explain the facts- and will avoid a complicated lie ("another employee ate your sandwich from the breakroom frig, and not me!"- and not "a space alien landed behind the store in a flying saucer and took your sandwich to alpha centauri for chemical analysis of terrestrial life").

But it doesnt follow that "because lies are most likely to be simple stories, that simple stories are therefore are more likely to be lies" as some of you seem to be saying.

You dont jump to the conclusion that since your coworker didnt mention space aliens he therefore must be part of a massive conspiracy (involving NASA, the airforce, and the Bilderburgers) to cover up the fact that space aliens are stealing sandwiches from america's lunch pales across the land!

When the evidence for plate tectonics came around, then those people should've paid attention to the new evidence and gone with the simplest explanation for it instead.

But maybe the evidence back then wasn't so convincing yet, so it depends on the circumstances.

Right now, we accept based on evidence we have that the quantum world is all about pure randomness and we do so using Occam's razor.

We accept evolution rather than invoking an intelligent designer exactly because of Occam's razor. Otherwise, you wouldn't have one good argument to reject creationism or intelligent design. The whole of science depends on Occam's razor being valid.

I still don't really understand that claim. All too often the most simple hypothesis gets trumped when scientists start peeling apart variables and find out that what they thought was one process was two or three that worked in close unison. The result is a more complex answer rather than the simplest.

Seems like you're trying to say something else but still not having any luck at articulating it.

I also don't understand how Occam's razor could be considered invalid - ie. its a tool. A compass isn't made invalid by local obsolescence such as computer aided drafting, nor is it made invalid by someone who wants to draw squares with it.

Occam's razor will do in a pinch or for situations of lesser importance but it is not a one size fits all method of arriving at a conclusion.

This is exactly why Occams razor is not useful to most debates. Most debates centre around identifying which pieces of evidence are a) relevent and b) valid.

Until that is agreed upon, Occams razor is not useful.

E.g. Occams razor is not useful in the intelligent design debate as neither side can agree on the evidence. Creationists in general reject the evidence of evolution as being irrelevent or invalid. Therefore telling creationists what conclusions can be drawn from evolutionary evidence is an entirely pointless exercise that will not add to your argument at all.

I found this in the Occam's Razor wiki:

It looks like the problems with using it as a tool come from equating "fewest assumptions" with "simplest". Take crime investigation, mentioned several times in the thread. Investigators really do take great care to approach a case with as few assumptions as possible. Consider a dead body found splattered at the base of a tall building. The simplest hypothesis is that the person jumped to their death. The hypothesis that makes the fewest assumptions is that the person died as the result of striking the ground with great force. From that latter hypothesis (which is consistent with the evidence of the condition of the body) you investigate whether they jumped, fell by accident, were pushed, or even died at a different location and then were moved. So "fewest assumptions" =/= "simplest".

Examples of Occam's Razor

The Best Explanation Is Usually the Simplest

The term "Occam's Razor" refers to the philosophical idea or scientific principle that of any given set of explanations for an event occurring, it is most likely that the simplest one is the correct one. Occam's razor does not seek to offer complete and absolute proof, but to find the simplest probable answer to a question of why an event happened.

In the following examples, "a" is the simplest explanation, and is therefor the most likely to be true:

* Event: Part of the fence has fallen over. Of possible explanations:
a) The wooden fence posts had rotted and could no longer bear the weight.
b) A bunch of kids got high and tore it down.
c) A rabid moose ran through it.
d) The wind from the Black Ops stealth helicoptors chasing the rabid moose knocked it over.

* Event: The tire on my car is flat. Of possible explanations:
a) I ran over something sharp in the road.
b) Someone deliberately flattened my tire just to mess with me.
c) As "b", but it was that "minority" person living down the road.
d) As "b", but it was those Black Ops people trying to warn me against going public with my suspicions.

* Event: It is raining and I saw a bright flash through my curtains. Of possible explanations:
a) There was lightning.
b) Someone is trying to take pictures of me in my house.
c) Space aliens have abducted my neighbor.
d) As "c", but they left an exact duplicate of my neighbor so that no one would get suspicious.

* Event: A student failed the statistics test. Of possible explanations:
a) The student did not study well enough to understand the subject.
b) One of the "minority" students swapped answer sheets in plain sight without anyone else noticing.
c) The professor filled the test with trick questions designed solely to fail that particular student.
d) A clandestine mind-control experiment erased the correct answers from the student's mind.

* Event: A car rear-ended another in highway traffic during rush hour. Of possible explanations:
a) The second driver was following too close to stop in time.
b) The first driver was either a woman or was taught to drive by his mother.
c) The second driver was distracted by an alien spacecraft flying low overhead.
d) There was no accident; it was staged as a distraction from a Black Ops mission already in progress.

* Event: A loud noise is heard in an apartment that is next to a busy highway. Of possible explanations:
a) A truck backfired.
b) A bomb was dropped in the immediate area.
c) A stealth drone has fired on an innocent citizen.
d) An alien spaceship has entered the Earth's upper atmosphere.

* Event: A woman is nauseous several hours after eating at a restaurant. Of possible explanations:
a) She may have food poisoning.
b) She may have been deliberatley poisoned.
c) She is experiencing a repressed memory of childhood abuse.
d) She is suffering from stomach cancer.

* Event: A roast beef in the oven burns to a crisp after being in the oven for only one hour. Of possible explanations:
a) The cook set the wrong temperature.
b) The oven's temperature regulator is defective.
c) Someone came into the house and turned up the oven temperature.
d) The HAARP project was engaged in another successful test.

* Event: A dog owner comes home to the trash can tipped over and trash is scattered on the floor. Of possible explanations:
a) The dog tipped the trash can over.
b) A burglar broke into the house and sorted through the trash can.
c) A secret Black Ops squad broke into the house, sorted through the trash can, and left listening devices all over the house.
d) As "c", but they also planted incriminationg evidence to frame the owner because he "knows too much".

* Event: A ball rolls out into the road in a residential area in front of your car. Of possible explanations:
a) Some children are playing ball and it accidentally rolled into the road.
b) Some teenagers got high and are maliciously attempting to cause an accident.
c) As "b", but the teenagers are going to highjack your car.
d) There is no ball; it is a hologram projected as a distraction from a Black Ops mission already in progress.

So now you see how Occam's razor works. It usually makes sense to chose the most logical explanation for any given problem or in any given situation and odds are that you will be right most of the time.

And here is where I think "Among competing hypotheses, the hypothesis with the fewest assumptions should be selected" is a more useful translation. Using Occam's Razor to choose a hypothesis for further investigation will lead you to a correct explanation a lot more often than going with the simplest explanation. That way you use Occam's Razor as the start rather than the conclusion and you escape making lots of wrong assumptions (because you are trying to use the fewest assumptions"

I

Depending on where you live, a) and b) are equally plausible or likely. Going with a) because it is the simplest makes the assumption that the fence fell over by itself. The hypothesis that makes the fewest assumptions is that the fence is on the ground. You then investigate the condition of the wood and the area around the fence to see if it fell or was pushed. What this investigation uncovers will inform how you fix it. Does it merely need fresh wood or does it need reinforcement and maybe spiky bits so high kids leave it alone?



a) and b) are equally likely depending on circumstances. Were you driving when you got the flat? If so, that makes deliberate puncture unlikely. But if you are parked then you shouldn't just assume it was accidental unless you are in an isolated area. In urban areas, people are known to puncture tires if you park in "their" spot. One poster here even said he did that at his apartment complex.



yes, a) is most likely since flashes aren't as bright as lightening and stalkers wouldn't use flashes anyway.



Yes, a) is most likely. But professors must be careful to use the "fewest assumptions" translation rather than the "simplest explanation" translation or they will miss sophisticated cheating.



a) is certainly the most likely and simplest of the explanations you have given. But it still makes the assumption that this was an accident and so you have not given e)the second driver rear ended the first intentionally for crime purposes. This is a known crime tactic. And it is one of the countless examples why law enforcement uses "fewest assumptions", not "simplest".



Again, a) is dependent on the assumption that the noise was accidental. Sometimes that noise is a gunshot, even when next to a busy highway.



Here's where you flunk out of medical school. A doctor must always go with "fewest assumptions" rather than "simplest explanation". A resident who did what you just did when she comes into the ER could get the hospital sued.



a) and b) are equally likely. When you always assume user error, you miss broken or defective equipment. The hypothesis that makes the fewest assumptions is that the temperature was too high. That's the start of the investigation, which should include both looking at what the cook set the temperature at and also uses an oven thermometer.

*

I have had my house broken into. The burgler left an ungodly mess since he was interested in speed, not going undetected. He knocked things over. ("He" since he was later caught.....by a detective who made no assumptions.) I didn't have a dog but if I did, I would have been wrong to blame the dog. Plus then I never would have called the cops.

*

I'm starting to think you never got high as a teen. I did, often, and am not proud of what I did. Nobody got hurt, luckily.



That's how it works to lead you badly astray. But if you use the "fewest assumptions" translation rather than the "simplest explanation", it works a lot better.

Urgh... the problem is determining whether we have another heuristic at play, or whether it really is a violation of Occam's razor. The problem with simple minded application of rules is that real world problems can be complicated. That's why it's important to understand the logic, rather than to know each rule and fallacy by name and #.

So, like this author: http://plover.net/~bonds/bdksucks.html I think that most citations of logical fallacies cause more problems than solved. Occam's razor is easier to use better, but... this kind of strong position does leave me wary.

That being said, I do think that Occam's razor and good reductionism is a part of any good intellectual's toolkit and something that can be used frequently in certain contexts.

Thank you Fnord,I will now blame the HAARP project for any burned dinners.I knew it was them all along.