Occam's Razor is probably one of the most misused, maligned and over-quoted theories on the web today.
The original Latin states "Pluralitas non est ponenda sine necessitate" (Plurality should not be posited without necessity) This means “One should not increase, beyond what is necessary, the number of entities required to explain anything.” In its simplest form, Occam's Razor states that one should make no more assumptions than needed. This is referred to as the law of parsimony. Over the centuries, this little equation has endured, and is now usually stated as:
When multiple explanations are available for a phenomenon, the simplest version is preferred. For example, a charred tree on the ground could be caused by a landing alien ship or a lightning strike. According to Occam's Razor, the lightning strike is the preferred explanation as it requires the fewest assumptions.
How I See It
Why anyone in the 20th century would still use Occam’s ridiculous philosophy is beyond me. The Razor is the rule of thumb for linear equations, but it leaves a lot to chance as well, and fails to credit it. You can't use the Razor for real-world equations, because nothing in the real world is equal.
It may take you 35 minutes to drive to work doing the speed limit on a typical day with medium traffic and no percipitation. However, you cannot ALWAYS count on the 35 minute rule because right out of the gate reality conspires against you just by existing.
A car accident occurs and slows traffic down to a crawl.
A rainstorm hits, or perhaps fog, and people drive a quarter slower than the normal speed.
You get behind an elderly driver going half the speed limit and can't pass for 15 minutes.
You get in front of a Cop and slow down out of fear reflex.
And in all of the above examples, Occam's Razor accounts for NONE of them. Not that it's William of Ockham's fault, he absolved himself of any guilt at the beginning of the law itself: "All things being equal". It's like a built-in nolo contendre clause.
The actual writing of Occam says something a little different, though: “For nothing ought to be posited without a reason given, unless it is self-evident (literally, known through itself) or known by experience or proved by the authority of Sacred Scripture.” He was a monk, after all, and for all he cared the only truly necessary entity is God; everything else is contingent.
Also, it's important to note Bertrand Russell (he of the Nobel Prize and gluey leftist politics) did the most widely accepted translations.
Occam’s Razor is often referred to as the Law of Parsimony, however, this is just an example of scientists trying to over-theorize one side of the story, (which is at best a shakey theory) and hammer it into the general consensus as fact or law. It’s not a LAW at all!! It could be argued that Occam’s Razor is a utopian calculator for a world that doesn’t exist. The reason for this is that, although the need for simplistic terms and explanations is essential, it is only one half of the coin. There are plenty of everyday occurrences that are easily explained, but only through complex understanding of the modern world. For instance, until the last century, it was commonly thought that maggots spontaneously appeared from rotting meat. We all know why they appear NOW, of course, but back then, you couldn’t have proved it for anything…And they used Occam’s Razor at the time!
Occam’s Razor does nothing to calculate the X-factor, the unknown quantity, the random chromosome. The Razor states, "When you hear hoofbeats, think horses, not zebras.” This has truth, but only in certain parts of the world. For instance, in
I recently doscovered that a Scientific Congress has determined this "Law" should actually be changed to the "Principle" of Parsimony.
Close guys, but still no bananas.
16 comments:
Hmm. I think maybe you over-simplified this explanation.
Using the razor, of course.
The hoofbeats example, for instance, need only a knowledge of where one is to properly apply it, and what the local wildlife is apt to be. (And it's not really an application of the razor -- more properly, it would be, "If you hear the sound of hoofbeats, it is probably caused by an animal with hooves ..." What the animal might be depends on other factors.)
That traffic is slower in Seattle this morning than yesterday is more likely due to drivers of automobiles than it would be to a dead elephant blocking the road.
Slow drivers, whatever their reason for being slow, are a more simple explanation than that a small plane had to make an emergency landing on the freeway, though it is possible.
Not a law, just a most-likely scenario principle.
Nothing to get all het up over ...
It's not a law so much as a principle of not making things any more complicated than one needs to. It's heuristic, not absolute. Use only as directed. Your mileage may vary.
I'm not sure where Ayn Rand comes in. She certainly didn't have much to do with reality. She was a good propagandist who said things that made her followers feel special and pat themselves on the back as the unsung heroes of their own dramas. Reality never really entered into what she did. She was an advertising copywriter fer crissake.
"When you hear the clopping of hooves don't look for zebras" is a standard bit of advice for interns. Their heads are full of all sorts of exotic conditions. They do not have the clinical experience check for the most likely things first. The proverb is a reminder to look for the common before checking out the far reaches of the exotic.
My father's time as an intern provides a useful example. A thirty-something Black man presented in the ER with a mild fever, pain on urination and a pustulent discharge. My dad and the other intern on duty conferred and came up with a diagnosis of acute glomerulonehpritis, a fairly uncommon inflammation of the tubules of the kidney.
They were all ready to present their findings when the Chief Resident walked in. Doctor Spellman, nephew of then-Cardinal Spellman, took one look at the patient and his female companion and said "His pecker, eh. How long have you had the clap?" At this point the interns were trying to shoo the lady out. The Doctor said "Don't bother. She's been screwing him around the clock."
That is Occam's Razor at its finest. Entities are not to be multiplied without necessity. The added ingredient is knowing what's necessary.
Well, to me the definition and use of Occam's Razor
Saying "The added ingredient is knowing what's necessary" (in Steve's reply, using both the hoofbeats and traffic) are valid points, but what I am saying is that Occam's Razor DOESN'T account for this. It says "all things being equal". My point is that you can never have such a utopian scenario to apply the Razor with in the first place.
>"It's not a law so much as a principle of not making things any more complicated than one needs to. It's heuristic, not absolute. Use only as directed. Your mileage may vary."<
This is spot-on, and I agree 100%. I just feel that I am more investigative than assumptive, and for me the Razor contradicts my methods of discovery. It's subtle, but does it seem to you that the use of it almost discourages deep investigation of anything?
Of course, the patient had the clap. However, suppose he had a disease that PRESENTED as the clap, same symptoms, yet the diagnosis was wrong: He has cancer that has spread to his bowels. Occam's Razor would lead you intuitively down the wrong path to begin with.
I compared it to the Fountianhead because of the plot twist concerning architectural integrity, the way Snyte allows his designers freedom to design in their specialties, but then combines their ideas into one finished product of clashing principles. Roark can design as he likes at Snyte’s, but he will never see a building erected as he creates it.
However, I agree with your assesment of Ayn Rand. Gahh, what a freak spank...I failed a creative writing class because I couldn't finish "Atlas Shrugged" without projectile vomiting.
Sorry, didn't complete my first sentance in the above response!!
Well, to me the definition and application of Occam's Razor are both in conflict with reality and the purpose they were meant for in the first place. Perhaps I should say the CONVENTIONAL use of the Razor, which is not the literal translation or meaning. However, I also disagree with the "It's all God's fault" attitude William of Ockham bandied about as the reason for the existence of any instance of anything beyond his realm of understanding.
Oh, I was a Rand-dite at eighteen. The Objectivists thought that Libertarians were commies. I read and re-read all her books. (And I didn't skip over the radio speech on economics, either.)
I wanted to be John Galt. Or at least Hank Reardon.
Unfortunately, the window of belief in such a system is exceedingly small and open but briefly; not helped when the chief proponents of such high-minded principles all fall down. Hollow as empty blimp hangar -- not even Mama Ayn could walk the line.
Read Nathaniel Brandon's book about it all, and also his ex-wife Barbara Brandon's. Sad.
I met Nathanial Brandon once, and my running buddy went to NYC and took classes at the Rand Institute, or whatever it was called. Named his first son "Ragnar," after the character in Atlas Shrugged.
The woman wore a cape when she went out, no lie. But it was all talk the talk. Nobody could walk that walk.
Your argument is a bit of a strawman, honestly. It's not how most people who understand the idea use it. It's not how the author meant it. If you're saying that people who do not understand an idea frequently misuse it we're on the same page. But it's a page with really big letters written in brightly colored crayons.
Consider it in your areas of expertise. Would you add curriculum if it didn't add anything to your practice or your students' skills? If someone said "I have this great idea. To make it work you have to add little green men, Budweiser that tastes better than Gulden Drak, and assume that Jakarta is in the middle of a three week blizzard." How likely are you to waste time on it compared to someone who says "As long as the sun rises in the East this will work" ?
In the case with the interns the resident's diagnosis fit the observed facts. In order for theirs to make sense the patient would have had to have a bunch of other symptoms such as severe wasting which were absent. Sure, there are other conditions which might present the same way. But the smart way to proceed is to rule out gonorrhea first.
Yep, and in medicine, the prevailing practice is differential diagnosis. You might know what you are dealing with when it comes through the door -- augenblick, the blink of an eye -- if you have enough experience, but a good doctor will go through an either-or, if-this-then-that, choice tree to rule out possibilities. You compare the signs -- your objective view -- with the symptoms -- what the patient tells you, and you add in lab tests or x-rays or whatever to winnow your choice to a diagnosis.
There are a shitload of things that cause flu symptoms, because that's how the human body tends to react to viral infection.
You call the nurse and say, "I have a headache, fever, I'm coughing, my body aches -- what might that me?
Stand by, because reading the list of what it *might* be will take the rest of the afternoon ...
But if you throw in a really stiff neck? That infection might be in the central nervous system, and you need to really-damn-quick make sure you aren't dealing with meningitis.
You will be coming in to see the doctor, we can't call a prescription in.
No violation of the razor to check your assumptions. It doesn't say the simpler choice is always the answer, just that there's no reason to assume zebras if you are on a ranch in south Texas.
>"It's not how most people who understand the idea use it. It's not how the author meant it."<
This has been my argument all along: How "most people" use it, as I have seen, is either
1: Out of context completely, but only with people who wouldn't know it anyway in order to look like academic geniuses.
2: Illogically applying it to an event or scenario that is also suceptible to chance - The one thing the Razor doesn't calculate, and cannot be ignored.
As to how the author MEANT it, he was trying to streamline explainations for events in life to a minimum, because he was a minimalist (Franciscan monk, life of poverty, etc.) William frequently did this in many of his other writings, it's nothing new. It's difficult to pinpoint what the exact cause is, living in the 13th century or his devotion to God, but what he is trying to accomplish with this little equation is...Impossible. Life is NOT streamlined and the Devil isn't rsponsible for headaches and erections.
William was also petty enough to try to push his theories over the accepted science of the day (not saying much, it's still medieval Europe in the grip of the 13th century after all) When he was excommunicated (typical) he wrote a treatise on why the Pope was a heretic and tried to have HIM excommunicated.
>"Consider it in your areas of expertise. Would you add curriculum if it didn't add anything to your practice or your students' skills?"<
No, but I wouldn't want/need Occam's Razor to identify that. What you present is exactly what I would use for streamlining WITHOUT the help of Occam. I don't whittle down every little thing that I don't use, there are variations of techniques that go slightly south of what the original said to do. There are drills I keep on standby that work the same thing my usual drill does. I do this because not every example or explaination I give will fit everybody, and I need variety to teach and train well.
The Razor doesn't come right out a SAY this, of course, but I believe it suggests that retaining anything not directly related to what you are minimalizing is wrong, and should be avoided. I could be wrong on that one, but it sure looks like The Razor doesn't advocate "standby" reasoning.
>"Budweiser that tastes better than Gulden Drak"<
Now you're just being nasty, Todd. There's no need to sink into vulgarities.
>"You compare the signs -- your objective view -- with the symptoms -- what the patient tells you, and you add in lab tests or x-rays or whatever to winnow your choice to a diagnosis."<
But this is INVESTIGATIVE, not ASSUMPTIVE. The Razor advocates streamlined minimalist reality, this process has used too many elements to be considered.
>"No violation of the razor to check your assumptions. It doesn't say the simpler choice is always the answer, just that there's no reason to assume zebras if you are on a ranch in south Texas."<
Absolutely, but using Occam's Razor as the rule isn't science, in fact I would say it's something to be applied AFTER a deduction and conclusion has been reached. Not before, you can't use it to arrive and certainty, you have to investigate the cause of a thing to find the solution to a thing. A process cannot be understood by stopping it, nor can it be improved by assumption. There must be investigation.
Something I forgot to post in the original article: Going off the idea that "The simplest answer is the best"...Who determines what the official answer is the simplest? Get three people in the same room: Who knocked over the lamp? The fat guy? The twitchy guy? The girl who hates you?
You and two other people are the judges to deduce guilt. All you use is Occam's Razor.
Using the Razor only, you still won't come to an agreeable conclusion, not unless you cheat and compile evidence and investigate. Too many conflicting elements work against you to streamline without investigation, but aside from that, who YOU think is guilty will be different from your peers, and all of you used the Razor.
My point here is that simplicity, like beauty, is in the eye of the beholder. We all have different ideas as to what is the easiest method for us. Caren and I constantly butt heads over which way is a shorter drive to downtown, yet we arrive at relatively the same time. It's still individual choice, based on what we already know without investigation. You can make a logical assumption (Occam) but that is still no basis for hypothesis.
I think you're trying to stretch the concept in order to make it look bad. Of course, all broad, sweeping generalizations are wrong -- including the one I just made.
But the idea that when faced with a choice, going with the less complicated one will serve you more often than not, is a a rule of thumb, and demonstrably right more often than not.
No, not a hundred percent, nobody can predict X-factors, and sure, people misuse this and other rules of thumb all the time. But the notion goes to generalities.
If you are in South Texas and you hear hoofbeats, it is possible there is a herd of zebras out there. But that's not how the smart money bets, is it?
In our art, simple doesn't equate to easy, but simple is better almost all the time. The more complex a process, the more likely it is to leave room for failure.
So, while a lot of folks misquote or misunderstand Sir Willie's razor, it is still sharp enough to do the job more often than not, based on the way things *generally* work.
All that Nazi beer ... tsk, tsk ...
>"All that Nazi beer ... tsk, tsk ..."<
I'll bet you used the Razor to deduce that!!
So, What your saying is simplify isn't the best answer?
Unless it is. ;-)
Exactly!
No razor necessary to deduce it. I saw you drool as you poured the stuff at Todd's. I thought you were going to come as you watched it foam.
Not that it's bad beer -- aside from being stale in the bottle and all -- just that paying import prices for it is way more than it is worth.
But of course, to a beer snob, the more it costs, the better it must be ...
Steve, you are wrong on a GALACTIC scale.
...Just wrong. Because I say so.
Also disgusting.
Post a Comment