What IS occam's razor?

[Frostbite]

Reading your replies in this thread, I don't think you quite understand Occam's razor. It's not about assuming things are true or false. It's about not assuming anything that you don't need to.

You mentioned criminal investigation, so let's take an example from there. Suppose you find a man shot dead, with a gun in his hand. Was it suicide or murder? As an investigator, you might assume, for the sake of argument, that his wife killed him and tried to make it look like a suicide, and then look for evidence to support that.

But that's not the situation that Occam's razor applies to. In that case, you have two competing theories: the suicide theory, and the murder theory. The theories are not equal: the evidence might support one and not the other (and, in fact, probably will).

Occam's razor only talks about situations where the two competing theories are equivalent -- where there is no evidence that could show which one was true. That's why you hear "all else being equal" so much in casual conversation about it. It simply doesn't apply in situations where further investigation could reveal which of two competing theories is correct.

Jeremy

Yeah that makes sense. So.. Occam's Razor only applies to things that are irrelevant?

 

[toddjh]

Yeah that makes sense. So.. Occam's Razor only applies to things that are irrelevant?

Yes. It's a way of weeding out irrelevancies from scientific theories. You wouldn't think it would be that necessary, but the human race is sadly full of irrational people.

Jeremy

 

[UserGoogol]

It seems to me that a somwhat weak rephrasing of Occam's Razor may be found in the expression:

When you ASSUME, you make an ASS out of U and ME.

Although it seems better to say:

In general, one should make as few assumptions as possible.

 

[Guest]

I'll post it again:
(my own notation in bold, T means "Theory")

If theory A has M assumptions and explains event E
T(A,M)~E

and

theory B has N assumptions and explains the same event E equally well
T(B,N)~E

Occam's Razor is:
Theory A is "best" if M is less than N.

 

[RichardR]

Ah, I see. Using your linked example, (cat vs. milk fairy) Occam's Razor would discredit the milk fairy theory. However, by what you are saying, Occam's Razor would have no effect on the cat vs. a stray cat who habitually comes in through the cat door. The second theory is more complex, yet assuming these conditions are not invented (the stray does come in every day, there really is a cat door), Occam's Razor then has nothing to do with it.

We know a stray cat can exist, so OR doesn’t help us choose between our cat and the stray because neither is an additional unproven entity.

By the way, do you think that this corruption of the theory should be given it's own name (whether a theory or a fallacy) in order to distinguish itself from Occam's Razor?

It confuses the real use of OR and I believe we should explain the correct use.

If it's not a theory, what is it? It's not a proof. A guideline, perhaps?

A guideline, probably.

I say probability because, using the definition in the link you supplied, "...there is a plausible explanation that does not require the milk fairy - the cat." Therefore, the author implies that the cat is a plausible reason. More likely. Higher probability. True, the science of probability cannot be strictly applied to give a percentage, but in the most general terms, the cat is more likely to be the answer than the milk fairy. The problem here is that the original Latin translation is "Should not" but the modern English version given is "Do not". The change has made it from a suggestion or hypothesis into a proof.

The cat is plausible. Therefore we don’t need to invent a supernatural explanation. That’s why we use OR, not because the probability may be higher. In most cases OR will choose the option with the higher probability but that is not the reason we use it.

Also, your linked essay says, "Note: we haven't proven that the cat drank the milk." That means Occam's Razor is NOT a proof. Perhaps the wording of the modern version should be altered slightly to reflect this.

Agreed, it’s not proof. The words “don’t posit plurality without necessity” or “don’t invent unnecessary entities” doesn’t mention proof.

Occam's Razor can be redundant because we are attempting to determine something that eventually, in time, we will discover through scientific means.

If we prove the cat drank the milk (by setting a camera, for example), then it is redundant – we won’t need it. But it is necessary if we don’t have absolute proof.

For example, a few months ago it was reported that scientists had been able to induce out of body experiences (OBE) in a woman by stimulating a region of her brain. The conclusion was that an OBE is a hallucination. A poster stated that it could mean that the scientists had stimulated the part of the brain that connects with the “soul”, and so this was not proof that the OBE is a hallucination. Of course, he was correct. But his solution required the additional entities of the soul, and a mechanism by which the soul could leave the body and return. Using OR we reject this hypothesis, since we have a natural explanation that does not require these two supernatural (read unproven), entities. Until such time as we can prove the soul does or does not exist (if we ever can), OR will not be redundant for this example.

Now then, that doesn't mean it's not important in predicting correctness. If used correctly, Occam's Razor is a great tool. However, the mistake is that people use it as proof, which should not replace scientific scrutiny and also is an incorrect usage.

Agreed.

Technically true. The theory itself, when used, does not render the user lazy, but rather lazy application of the theory can produce illogical effects. Let me correct myself. I shall say "Lazy people use Occam's Razor as proof, though the theory itself is not a water-tight solution. Such misinterpretation is a logical fallacy."

In the example above of the OBE, you’re not lazy if you invoke OR. We simply can’t prove the alternative false, but we are making stuff up if we insist tha soul is responsible.

RR: Why choose God over the pink giraffe
K: Which is the greater invention? I hope it is a moot point, I certainly don't want to start keeping score of this stuff now.

Neither has been proven to exist, so they are equivalent. The pink giraffe is often used because people have a lot of baggage over “God”, but not over a pink giraffe. The giraffe can therefore be used to get the message over.

 

[RichardR]

God was just an example. A bad example perhaps, but whatever. If criminal investigators used Occam's Razor in their reasoning, their investigations wouldn't take years or months...

Wouldn’t make any difference. OR tells you not to invent explanations for which you have no proof. Prosecutors are supposed to utilize the same reasoning.

 

[Peter Soderqvist]

What can the razor shave?

Scientific facts was close to nothing in the middle Ages, and for the same reason, the Franciscan Monk William of Occam would be in serious doubts today about a literally interpretation of Genesis, if he was faced with modern scientific facts, because the story of creation is proven false. Occam 's razor cannot shave off the rest of god's gap, because emptiness cannot be shaved! Occam 's razor is only used to shave off a wasteful conclusion about known facts, in order to make the theory more fitted to these facts, for instance; The Newtonian theory is a flawed theory, but that doesn't mean that the Newtonian physics is ruled out! Take his second law of motion for instance, which states that, the acceleration of an object is proportional to the force applied on it; it means in simpler terms that your car's acceleration is governed by the car's acceleration pedal! But it is not truth, because your car increases in mass when it accelerates! So the amount of increased mass disproves this law of motion, very slightly in ordinary circumstances, but astronomically when we approaches light speed! We use Newton's theory in simple or in ordinary situations, and Einstein's theory of relativity in extraordinary situations, there the correlation between the forces on the object and its acceleration is included!

Principia Cybernetica Web Occam's Razor
For a given set of observations or data, there is always an infinite number of possible models explaining those same data. This is because a model normally represents an infinite number of possible cases, of which the observed cases are only a finite subset. The non-observed cases are inferred by postulating general rules covering both actual and potential observations.

For example, through two data points in a diagram you can always draw a straight line, and induce that all further observations will lie on that line. However, you could also draw an infinite variety of the most complicated curves passing through those same two points, and these curves would fit the empirical data just as well. Only Occam's razor would in this case guide you in choosing the "straight" (i.e. linear) relation as best candidate model. A similar reasoning can be made for n data points lying in any kind of distribution.
http://pespmc1.vub.ac.be/OCCAMRAZ.html

 

[Keneke]

quote:
--------------------------------------------------------------------------------
Originally posted by Keneke
By the way, do you think that this corruption of the theory should be given it's own name (whether a theory or a fallacy) in order to distinguish itself from Occam's Razor?
--------------------------------------------------------------------------------

It confuses the real use of OR and I believe we should explain the correct use.

Well, no, the modern day explanation "All other things considered" (found in movies, books, etc.) confuses moreso than an attempt to clarify the rule. If a skeptic like me uses the incorrect, more popular definition (because I had never heard the other version), then think what a True Believer could do! I think, at the least, an effort should be made to differentiate between invention and complexity. Perhaps a bit of a comment in OR definitions? Because if not, people will continue to misuse it, and that does not benefit anyone.

Whodini uses the word "assumptions". This isn't necessarily invention.
Rwald uses the word "simplest".
Boooeee and Cecil use KISS, which also refers to simplicity.

Seeing as how this seems to be the most common misinterpretation of OR (indeed, there's more of this line of thought than of the real meaning), then an extra effort should be made to look out for this type of usage.

 

[DrMatt]

I'm talking about the common version that you hear in pop culture, such as the movie Contact: "All other things being equal, the most succinct answer is usually the right one."

I think you mis-heard. I don't think you'd get such words past Jodi Foster's mouth.

If it's not a theory, what is it? It's not a proof. A guideline, perhaps?

A guideline, yes. The formal word is "heuristic", which, used in this sense, means "guideline".

Any explanation which introduces unnecessary suppositions begs that those suppositions be justified. Most folks have trouble even noticing that they're making suppositions, at least in my experience.

 

[RichardR]

Well, no, the modern day explanation "All other things considered" ....

I disagree that this is a modern day interpretation. It is simply an incorrect representation of OR. It’s wrong. (I’ve never heard “All other things considered” being used as OR, actually.)

By all means say it is wrong. My article also says that “choose the simplest” is wrong so you should also say that “All other things considered” is wrong. I don’t think it requires a new name to itself as it makes no sense, IMO.

 

[Peter Soderqvist]

The principle of parsimony regarding OBE, souls, versus the brain!

The neuronal pattern of consciousness has not been localized, so it is a misuse of Occam 's razor to alleged that, that we have a more parsimony explanation than soulism, because you have no explanation without this "pattern"! When we say; I have a brain, or my cortex is thick, etc, you have an entity already there, namely, the owner of the brain, this owner is not ruled out until you can give us a natural description about the structure and function of conscious pattern, or in other words; soulism is potentially falsifiable, or can be shaved off, if you can give us a natural neuronal explanation?

A: M. Turing as quoted from the Mind's I, a book by Hofstadter & Dennett 1981, chapter 4, page 61: I do not wish to give the impression that I think there is no mystery about consciousness: There is, for instance, something of a paradox connected with any attempt to localize it!

Soderqvist1: I am on the page 287 in the Mind's I.
Is there someone who has read Popper & Eccles The Self and its Brain?

The Self and its Brain (with Sir John Eccles), Routledge, 1984. Springer, 1977. A joint effort of Popper and Eccles to produce what they describe as a modest theory of interactionism, consisting of two parts. In Part I Popper surveys the philosophical background and criticises rival theories. In Part II Eccles provides a synopsis of the then current knowledge of brain function and anatomy, and propounds a neurological theory underpinning the interaction of the mind and the brain. The theory has some counterintuitive predictions that suggest interesting experiments. The final section consists of a fascinating dialogue between Popper and Eccles. I especially enjoyed Popper's demolition of materialism, drawing on the history of science, including the field theories of Boscovich, Leibnitz, and Faraday and Maxwell.
http://www.eeng.dcu.ie/~tkpw/

Eccles The Nobel Prize in Physiology or Medicine 1963
http://www.nobel.se/medicine/laureates/1963/

 

[Keneke]

I think you mis-heard. I don't think you'd get such words past Jodi Foster's mouth.

To quote the movie:

KITZ: Doctor, are you familiar with the scientific precept known as Occam’s Razor?

ELLIE: Yes, it means that all things being equal. The simplest explanation tends to be the right one.


http://www.unc.edu/courses/reli035/spring1998/occam.htm

 

[Keneke]

Originally posted by RichardR I disagree that this is a modern day interpretation. It is simply an incorrect representation of OR.

I am not saying it replaces the original, correct version. I am saying that it is a modern interpretation in common usage, though incorrect. Please do not misunderstand me.

(I’ve never heard “All other things considered” being used as OR, actually.)

It is becoming more prevalent because of being in a movie. Anecdotally, I have heard it this way on TV as well. You must not watch too much TV.

People are misusing it. A small google for "complexity Occam's razor" reveals these websites misusing Occam's razor:

http://www.survivalscience.org/debunk/ww/arg03.shtml
http://members.tripod.com/loki814/joey/joey.html
http://rii.ricoh.com/~stork/OccamWorkshop.html
http://citeseer.nj.nec.com/webb96further.html
http://www-2.cs.cmu.edu/afs/cs/project/jair/pub/volume4/webb96a-html/node1.html
http://www.cwi.nl/~paulv/papers/occam.pdf

(Granted, I did find a few webpages that show the right definition.)

I don’t think it requires a new name to itself as it makes no sense, IMO.

Don't invalidate my argument just because you just can't understand it.

Being an armchair psychologist, I tend to think in terms of things under a person's radar. I'm not saying that we MUST make a name for it. I'm thinking of ways to permeate society's conscience in order to make this correction. Certainly only teaching the right definition isn't working right now tp prevent this misinterpretation.

Ware the mass media! Bane to all republicans and skeptics!

 

[Peter Soderqvist]

All other things being equal and the principle of parsimony!

TO RICHARD R AND KENEKE

Soderqvist1: My house has no cat doors!
Suppose I have a cat. One night, I leave out a saucer of milk, and in the morning the milk has gone. No one saw who or what drank the milk. Lets say there are two possibilities: The cat drank it or the milk fairy drank it? I have now two models, one house with a milk-fairy, and one house with a real cat! Everything being equal in the two houses, except these two entities, a cat and a milk-fairy! The cat hypothesis is simpler than the milk-fairy hypothesis, since in the house with a milk-fairy there is a real cat too, but in the house with a real cat, there is no milk-fairy there. The house with two entities are more complex in number than the house with only one entity, therefore; the cat hypothesis is more parsimony, and have thus a higher probability to be the real case!

Principia Cybernetica Web Occam's Razor
The principle states that one should not make more assumptions than the minimum needed. This principle is often called the principle of parsimony. It underlies all scientific modelling and theory building. It admonishes us to choose from a set of otherwise equivalent models of a given phenomenon the simplest one. In any given model, Occam's razor helps us to "shave off" those concepts, variables or constructs that are not really needed to explain the phenomenon. By doing that, developing the model will become much easier, and there is less chance of introducing inconsistencies, ambiguities and redundancies.
http://pespmc1.vub.ac.be/OCCAMRAZ.html

 

[Keneke]

Re: All other things being equal and the principle of parsimony!

The house with two entities are more complex in number than the house with only one entity, therefore; the cat hypothesis is more parsimony, and have thus a higher probability to be the real case!

Ah, but according to RichardR, what you see on that webpage may be a modified (and therefore incorrect?) version of the original theory!

 

[DrMatt]

Re: The principle of parsimony regarding OBE, souls, versus the brain!

The neuronal pattern of consciousness has not been localized, so it is a misuse of Occam 's razor to alleged that, that we have a more parsimony explanation than soulism, because you have no explanation without this "pattern"! When we say; I have a brain, or my cortex is thick, etc, you have an entity already there, namely, the owner of the brain, this owner is not ruled out until you can give us a natural description about the structure and function of conscious pattern, or in other words; soulism is potentially falsifiable, or can be shaved off, if you can give us a natural neuronal explanation?

Look, quoting loose and free language doesn't get you any brownie points. The "I" that has a "mind" is just my body. Spend a while in the stroke ward of a hospital seeing the kinds of things that happen when parts of people's brains are lost and parts of their "selves"/"minds" cease to function. There just isn't any "I" separate from me.

The existence of a "conciousness pattern" separate from the body and its behaviors is an unnecessary assumption. Occam's razor cuts out soulism because the original question--the nature of a "conciousness separate from the body"--has not been shown to be a realilty needing explanation at all.

 

[DrMatt]

To quote the movie:

KITZ: Doctor, are you familiar with the scientific precept known as Occam’s Razor?

ELLIE: Yes, it means that all things being equal. The simplest explanation tends to be the right one.


http://www.unc.edu/courses/reli035/spring1998/occam.htm

Which begs the question of whether "Simple" means "Succinct".
I mean, I can succinctly say "The earth sucks", and that's more succinct than the inverse-square law of gravitation, but it's more complex, and begs a bunch of questions.

 

[Keneke]

Which begs the question of whether "Simple" means "Succinct". I mean, I can succinctly say "The earth sucks", and that's more succinct than the inverse-square law of gravitation, but it's more complex, and begs a bunch of questions.

It's a moot point. What I was trying to demonstrate is that the quote from the movie is different than the original version.

 

[synaesthesia]

Good question!

Occam's razor is not so easily stated as some posters are suggesting. Just what is involved in parsimony? How, for instance, do we balance many simple entities over fewer but less simply entities? How does parsimony relate to epistemic probability? These questions are very interesting ones, but I have not found anyone who claims to have a general answer to them.

One of the more obvious applications of the Razor is the elimination of theories that we are just as well without. This tactic can prevent considerable confusion.

Keneke,

The problem here is when people use Occam's Razor incorrectly. It is said (in that link) that some people use the razor to slash out the whole existence of God. However, "all other things" are not equal. Positions on God-created phenomena, such as the new Earth theory, do not have the scientific success of repeatability, such as evolution does. Therefore, all other things are NOT equal, therefore, Occam's Razor does not apply.

God is in my opinion the paradigm case for Occam's razor. Even if God did explain phenomenon not accessible to science (a premise which I do not accept) occam's razor would still apply in our judgements.

God is an idea which can generate an empirically equivalent metaphysical counterpart to any other theory. However, given a particular epistemological setting, the God theory is always more complicated (being arbitrarily complicated, as he is) than a naturalistic account. Accordingly, as a general rule, God is the inferior hypotheses.

The above formulation is schematic, giving more the logical and substantial idea than the rhetoric. It is, however, behind many of the most effective and cogent attacks on God.

Keneke, scientific repeatability is not a requirement for historical events. What scientists must do is develop repetable experiements demonstrating the putative processes, and find evidence that they have occured. At any rate, your argument seems to be a non-sequitur since you have yet to establish any grounds of theoretical preferability for God.

Also, Occam's Razor is a theory for the probability of a certain phenomenon existing, and not a proof in itself. One cannot prove anything if their proof includes an assumption like Occam's Razor. That is another way Occam's Razor can be misused.

Incorrect, the assumption of parsimony is always, I repeat once again, always tacitly embedded in all theory, all argument and all human rationality. If it were indeed true that no statement can be established to be true with OR figuring in our background assumptions, nothing could be established to be true.

We can find truths, therefore parsimony is a valid part of our theoretical background.

If so, how can we know that theories of differing length have the exact same amount of correctness in them? Would not a theory, in time, be proven more or less valuable to science by research and investigation alone?

Not necessarily, because there are cases of empirically equivalent theories of differing complexity. We should seek, in such cases, to eliminate hand-waving and other waste motion in order to find a tighter conceptual isomorphism between our theories and the structure of the world. Waste motion will simply serve to decieve us unless it can be shown to have definite advantages (such as cogntive efficiency.).

 

[Keneke]

You'll notice the debate has altered. please read the whole thread.

Incorrect, the assumption of parsimony is always, I repeat once again, always tacitly embedded in all theory, all argument and all human rationality. If it were indeed true that no statement can be established to be true with OR figuring in our background assumptions, nothing could be established to be true.

It is more likely the cat drank the milk than the milk fairy. We cannot prove the cat drank the milk, though. Therefore, it is not a proof.

We will assume it is the truth in our reaction, though (punishing the cat, for example), and that is the basis for human rationality and logic. I have referred to this countless times as "bridging the mental gap". OR is simply the process for narrowing the gap as much as possible.