The Role of Philosophy in Science: Separating Fact from Fiction

[Aquamarine]

Here is a short description of Minimum Message Length induction, a good candidate for a formal replacement of Ockham's razor.
http://hawthorn.csse.monash.edu.au/mml

Examples of how it have already been used in real-world science:
http://www.csse.monash.edu.au/~lloyd/tildeMML/Intro/

More on MML:
http://en.wikipedia.org/wiki/Minimum_message_length
http://www.csse.monash.edu.au/~lloyd/tildeMML/

But there are competing theories to MML. Malcolm R Foster have some overviews on his homepage.
http://philosophy.wisc.edu/forster/default.htm

Some examples of how Ockham's razor have been important in physics in choosing between theories:
http://www.abc.net.au/rn/science/ockham/stories/s118778.htm

But the claim that Ockham's razor is the scientific method is stronger than some anecdotes. As I showed in the Ptolemic example, any theory can be modified to fit all empirical data, at the cost of complexity. No arguments against this have been stated. And then the only way to choose between competing theories is Ockham's razor.

Regarding point 1, se above. And note that since falsification have already been proven false as an universal criteria, there is no competing theory with Ockham's razor for how science works. Regarding point 2, the modern versions of Ockham's razor may choose a more complex theory if it better fits the data.

 

[marlon]

So now you're having to resort to attacks on mistakes and typos? This is sad Marlon. There is no topic worth ruining your objectivity over to this extent. Just relax. People will disagree with you all your life. And some of them will be right.

Really ? Thanks for this new insight...

I assume you mean the people participating in the science forums? I would hope they are pointing to relevant text when the only other option is to write 600 pages of explanations.

No, they are not. They are actually helping people out with specific solutions to real science problems and questions. Ofcourse we could refer to whatever text that there is out there on the web, but then what is the use of this Forum?

Sometimes my cat wonders the same thing.

Why am i not surprised ?

Good one. How can I possibly counter such strong arguments. Maybe if I use "hahaha" a little more, that would help the strength of my arguments?

Thanks, besides you don't have to counter anything. What arguments ? Please be more specific. hmmm, i could have sworn i read that somewhere else...
Maybe you should be more silent instead of using hahaha more. Just a suggestions to you...i can be wrong ofcourse

Three concrete examples of what? Libraries full of explanations on how philosophy and science relates?

YES INDEED. Please don't tell me what i asked and use that as an answer. It makes you sound real credible and knowledgeable to all readers of this thread.
Besides, no need to bring in Zz., i am only asking this question to you and your immense library...

So please fill in : 1) 2) and 3)

Thanks in advance for the clear knowledge...

marlon

 

[ZapperZ]

Here is a short description of Minimum Message Length induction, a good candidate for a formal replacement of Ockham's razor.
http://hawthorn.csse.monash.edu.au/mml

Examples of how it have already been used in real-world science:
http://www.csse.monash.edu.au/~lloyd/tildeMML/Intro/

More on MML:
http://en.wikipedia.org/wiki/Minimum_message_length
http://www.csse.monash.edu.au/~lloyd/tildeMML/

But there are competing theories to MML. Malcolm R Foster have some overviews on his homepage.
http://philosophy.wisc.edu/forster/default.htm

Some examples of how Ockham's razor have been important in physics in choosing between theories:
http://www.abc.net.au/rn/science/ockham/stories/s118778.htm

But the claim that Ockham's razor is the scientific method is stronger than some anecdotes. As I showed in the Ptolemic example, any theory can be modified to fit all empirical data, at the cost of complexity. No arguments against this have been stated. And then the only way to choose between competing theories is Ockham's razor.

Regarding point 1, se above. And note that since falsification have already been proven false as an universal criteria, there is no competing theory with Ockham's razor for how science works. Regarding point 2, the modern versions of Ockham's razor may choose a more complex theory if it better fits the data.

Other than the fact that at least ONE of your links is dead, NONE of what you have listed here applies to the selection of a THEORY. You have listed how DATA is analyzed! Now, unless you are claiming that physics is nothing more than simply a collection of data and unrelated set of disjointed information (a common misconception for anyone who has not studied physics), then what you have tried to present here is quite irrelevant.

I am still amaze that you make all these claims, and yet you cannot clearly, on your own, produce a distinct, clear example. The Ptolemic example isn't valid, because again are doing "data fitting", and comparing a carefully-formed idea based on the knowledge at that time, with a "made-up" idea. It is why, for example, Intelligent Design is NOT a science, while Evolution is! Intelligent design claims they can explain ALL of the very same observation that Evolution already can. But you will notice that ID is NOT considered to be a scientific theory to be considered on the same level as Evolution. ID is a MADE-UP theory that allows untested intrusions. You, on the other hand, would include it and use Occam's razor as the reason why it not accepted. This is false!

Again, how about finding something in physics within the past century? Explain to me clearly how Occam's razor was used there! I can't believe you cannot even find ONE, considering that you claim it is a universal "law", even when physicists did not realize they're using it. I tried giving you one, but it is obvious that while you insist that us physicists do our "homework" by reading up on all of these sites, you don't seem to think you need to even bother reading up on the what "high-Tc superconductors" and the issues surrounding it are.

Zz.

 

[Aquamarine]

Other than the fact that at least ONE of your links is dead, NONE of what you have listed here applies to the selection of a THEORY. You have listed how DATA is analyzed! Now, unless you are claiming that physics is nothing more than simply a collection of data and unrelated set of disjointed information (a common misconception for anyone who has not studied physics), then what you have tried to present here is quite irrelevant.

I am still amaze that you make all these claims, and yet you cannot clearly, on your own, produce a distinct, clear example. The Ptolemic example isn't valid, because again are doing "data fitting", and comparing a carefully-formed idea based on the knowledge at that time, with a "made-up" idea. It is why, for example, Intelligent Design is NOT a science, while Evolution is! Intelligent design claims they can explain ALL of the very same observation that Evolution already can. But you will notice that ID is NOT considered to be a scientific theory to be considered on the same level as Evolution. ID is a MADE-UP theory that allows untested intrusions. You, on the other hand, would include it and use Occam's razor as the reason why it not accepted. This is false!

Again, how about finding something in physics within the past century? Explain to me clearly how Occam's razor was used there! I can't believe you cannot even find ONE, considering that you claim it is a universal "law", even when physicists did not realize they're using it. I tried giving you one, but it is obvious that while you insist that us physicists do our "homework" by reading up on all of these sites, you don't seem to think you need to even bother reading up on the what "high-Tc superconductors" and the issues surrounding it are.

Zz.

The links worked fine when I checked them again. And I suggest you read them, they are about how to select the best theory. One with examples of Ockham's razor this century, written by physicist. Foster have many example from physics this century in his articles. The Ptolemic example is correct, as others here have agreed. Regarding ID it fits the empirical evidence less well than a theory based on evolution, little need to invoke complexity.

 

[ZapperZ]

The links worked fine when I checked them again. And I suggest you read them, they are about how to select the best theory. One with examples of Ockham's razor this century, written by physicist. Foster have many example from physics this century in his articles. The Ptolemic example is correct, as others here have agreed. Regarding ID it fits the empirical evidence less well than a theory based on evolution, little need to invoke complexity.

Again, you don't seem to understand the difference between "selecting the appropriate model for set of data" and "selecting a theory to be correct". I asked you if you think physics is nothing more than just a collection of data. You never answered. If you think it is, then it would be consistent to your persistent of "curve fitting" analogy. But it would also clearly reveal your faulty knowledge of what physics is. Maybe this is why you refused to answer.

Secondly, Foster is EXACTLY describing what I just mentioned, especially in quantitative analysis of data! (Note: I graduated from UW-Madison also where Foster is a faculty member). Not only that, *I* have done the very same thing in my analysis of my experimental data. This means that I just don't pay lip service about these things, I have DONE it. However, these things are done simply to create a phenomenological model that can be checked, tested, and eventually evolved into a theory that is derived from First Principles. It is NEVER done to SELECT which theory is "correct"! This is what I've been trying to get across!

Look, if I were to give you tons and tons of links to read about high-Tc superconductors (and trust me, I am able to do that), I bet you'll roll your eyes and let a few things fly over your head. I'm not asking for links. I am asking you to PICK ONE physics example (I don't care where you hijack that example from) and tell me clearly how Occam's Razor was applied in picking out which theory is correct. Pick one from within this past century, if you please, since you obviously do not care for the example I picked already.

Zz.

 

[Aquamarine]

Again, you don't seem to understand the difference between "selecting the appropriate model for set of data" and "selecting a theory to be correct". I asked you if you think physics is nothing more than just a collection of data. You never answered. If you think it is, then it would be consistent to your persistent of "curve fitting" analogy. But it would also clearly reveal your faulty knowledge of what physics is. Maybe this is why you refused to answer.

Secondly, Foster is EXACTLY describing what I just mentioned, especially in quantitative analysis of data! (Note: I graduated from UW-Madison also where Foster is a faculty member). Not only that, *I* have done the very same thing in my analysis of my experimental data. This means that I just don't pay lip service about these things, I have DONE it. However, these things are done simply to create a phenomenological model that can be checked, tested, and eventually evolved into a theory that is derived from First Principles. It is NEVER done to SELECT which theory is "correct"! This is what I've been trying to get across!

Look, if I were to give you tons and tons of links to read about high-Tc superconductors (and trust me, I am able to do that), I bet you'll roll your eyes and let a few things fly over your head. I'm not asking for links. I am asking you to PICK ONE physics example (I don't care where you hijack that example from) and tell me clearly how Occam's Razor was applied in picking out which theory is correct. Pick one from within this past century, if you please, since you obviously do not care for the example I picked already.

Zz.

Of course science is more than a collection of facts.

Please give an exact quote from Foster that support your statement.

Regarding examples, exactly the same arguements that were used in the Ptolemaic example can be used in any scientific controvery this century. Take the probably most famous of them, Einstein against Newton.

Or from one of the lnks that you have not read:

Ockham’s Razor is also the motivation behind unification of physical theory. A good example of this came nearly 100 years ago. The German physicist Max Planck had invented an early version of the quantum theory that explained a baffling phenomenon: the speed of electrons that were thrown off when light is shone at a metal. His equations called for a new physical constant, a new constant of nature, whose value had to be found from the observations he made. But the same idea was then applied to explain the amount of radiation given off by a hot body, an electric fire, for example, and also to explain the wavelengths of light that are absorbed by hydrogen atoms. But of these further phenomena had been experimentally studied and each had required its own physical constant of nature to be set separately from the observations. The new idea related these two extra constants to Planck’s and accurately gave their values. Three supposedly separate phenomena had been shown to have the same underlying explanation. The quantum idea was rapidly accepted in consequence.

My last example is from cosmology. When Einstein worked out his general theory of relativity and gravity early in the 20th century, and improved on Newton’s venerable theory, there was room for an arbitrary constant, known as a parameter, in his equations. To keep things simple he was tempted to put it to zero, but another consideration weighed even more heavily: he believed on philosophical grounds that the universe was unchanging on the large scale. He believed it was unchanging in how the great clusters of stars, called galaxies, relate to each other. This meant that his number could not be zero, for technical reasons.

But some years later, it was found that the galaxies were in fact all rushing away from one another. In Einstein’s mind, an informal version of the Ockham analysis immediately took place and he reverted to the value zero for his number, which is called the ‘cosmological constant’ today. In this spirit, a translation of the Latin Ockham’s Razor, ‘entia non sunt multiplicanda praeter necessitatem’ would be ‘Parameters should not proliferate unnecessarily’. This particular plot has thickened though: the value of Einstein’s cosmological constant is once again in question. Is it zero, or is it very small, and should be chosen so as to best fit the data?

We don’t know yet. This is why these questions are exciting.

http://www.abc.net.au/rn/science/ockham/stories/s118778.htm

 

[ZapperZ]

Of course science is more than a collection of facts.

Please give an exact quote from Foster that support your statement.

Regarding examples, exactly the same arguements that were used in the Ptolemaic example can be used in any scientific controvery this century. Take the probably most famous of them, Einstein against Newton.

Or from one of the lnks that you have not read:

http://www.abc.net.au/rn/science/ockham/stories/s118778.htm

I'm sorry, but this illustrates clearly why you do not understand the difference between applying Occam's Razor to a collection of data, and applying it to know what theory is correct. Forster's article on how one analyze the quantitative data is identical to this. This is what I stated as the phenomenological model.

Furthermore, do you really, seriously think that the reason General Relativity is considered to be more accurate than Newton's Gravitational theory simply due to Occam's Razor? Here's a hint: there are SEVERAL empirical data that show that Newton's Gravitational theory is less accurate than General Relativity. In other words, as I have replied to Tom earlier, it is pure down-and-dirty experimental verification that determined in the end which theory is more accurate over the other, and NOT Occam's Razor.

So now find me another one...

Zz.

 

[Aquamarine]

I'm sorry, but this illustrates clearly why you do not understand the difference between applying Occam's Razor to a collection of data, and applying it to know what theory is correct. Forster's article on how one analyze the quantitative data is identical to this. This is what I stated as the phenomenological model.

Furthermore, do you really, seriously think that the reason General Relativity is considered to be more accurate than Newton's Gravitational theory simply due to Occam's Razor? Here's a hint: there are SEVERAL empirical data that show that Newton's Gravitational theory is less accurate than General Relativity. In other words, as I have replied to Tom earlier, it is pure down-and-dirty experimental verification that determined in the end which theory is more accurate over the other, and NOT Occam's Razor.

So now find me another one...

Zz.

Regarding the empirical evidence in support of GR, this can be explained by fraud, mass psychosis, hallucinations, optical illusions, intervention by higher powers or madness. By adding all or some of this to Newton's theory, it can again fit all of the empirical evidence. But at the cost of complexity. Thus again, Ockham's razor is needed.

You have avoided answering the examples by Anthony Garret in my previous post. And failed to produce a quote by Foster. And ignore that MML is being used already to choose between theories. And failed to produce an alternative theory of how science works.

 

[ZapperZ]

Regarding the empirical evidence in support of GR, this can be explained by fraud, mass psychosis, hallucinations, optical illusions, intervention by higher powers or madness. By adding all or some of this to Newton's theory, it can again fit all of the empirical evidence. But at the cost of complexity. Thus again, Ockham's razor is needed.

This is EXACTLY what I meant by you not knowing when an idea is legitimate to be considered as a viable theory. You think that GR is on the same par as "fraud" and mass psychosis. I have proven in one of my articles that the general public is oblivious to the difference between "scientific evidence" and "anecdotal evidence". They think anything one can think of, or anything one "sees" is sufficient to be considered on par with what is dealt with in science. This is bogus! The fact that you think we need Occam's Razor to be able to pick out GR from "fraud", "mass psychosis", "hallluciantions", etc is incredible! I am in shock that this is the level of logic that I am dealing with, not to mention the level of ignorance of the workings of science.

You have avoided answering the examples by Anthony Garret in my previous post. And failed to produce a quote by Foster. And ignore that MML is being used already to choose between theories. And failed to produce an alternative theory of how science works.

Exactly! Because I am doing exactly what you are doing. And I don't need to produce an "alternative" of how science works. I am not the one who is silly enough to state that there IS a philosophy, or a "theory" on how science works. YOU DID. I never claim to be in possession of any such theory. All I'm doing is disproving what you want to claim: that simply by using Occam's Razor, one can PICK which theory is correct. You have failed to do this when I brought up the high-Tc case, and you have failed to do this when you brought up GR vs. Newtonian Gravity, needing to resort to comparing GR with... er.. "fraud" and "mass psychosis"? Come on now!

Given two or more LEGITIMATE theories to explain the same phenomenon and NOT just to come up with a phenomenological model to fit some data, use Occam's razor and prove to me you can pick which one is correct. You want an example? Apply this Occam's Razor to tell me which one is correct: Newton's formulation of classical mechanics versus Lagrangian/Hamiltonian formulation of clasical mechanics. BOTH formulation can describe ALL of the classical dynamics! Both of them agree with each other. Both are LEGITIMATE theories with clear underlying mathematical description, unlike your hallucinations and mass psychosis. Yet, they both have distinctly DIFFERENT approaches, and some would say, different "philosphy" to how each view the classical universe. Now apply your beloeved Occam's razor and tell me which one is correct.

Zz.

 

[Aquamarine]

You have no theory of how science works, using only gut instinct. Somehow, this tells you that some theories are automatically legitimate while others are not. That fraud or mass psychosis could be an explanation is automatically dismissed on no grounds. (at least concerning certain fields, which decided again by gut feeling. I guess you would be more open to this if ID researchers claimed to have proven their theory conclusively).

Also you seem somewhat fixated by high-Tc theories, probably because you know that the highly specialized area makes it impossible to discuss the details by those not experts. But note also that exactly the same arguments used on Einstein/Newton could be used here. If I wanted, I could make a lots of claims about highly specialized areas of medicine that you know nothing about, claiming that you know nothing about science. Instead, since this a general discussion about the philosophy of science, choose areas more commonly known.

I will again repeat this quote, the author is physicist:

Ockham’s Razor is also the motivation behind unification of physical theory. A good example of this came nearly 100 years ago. The German physicist Max Planck had invented an early version of the quantum theory that explained a baffling phenomenon: the speed of electrons that were thrown off when light is shone at a metal. His equations called for a new physical constant, a new constant of nature, whose value had to be found from the observations he made. But the same idea was then applied to explain the amount of radiation given off by a hot body, an electric fire, for example, and also to explain the wavelengths of light that are absorbed by hydrogen atoms. But of these further phenomena had been experimentally studied and each had required its own physical constant of nature to be set separately from the observations. The new idea related these two extra constants to Planck’s and accurately gave their values. Three supposedly separate phenomena had been shown to have the same underlying explanation. The quantum idea was rapidly accepted in consequence.

My last example is from cosmology. When Einstein worked out his general theory of relativity and gravity early in the 20th century, and improved on Newton’s venerable theory, there was room for an arbitrary constant, known as a parameter, in his equations. To keep things simple he was tempted to put it to zero, but another consideration weighed even more heavily: he believed on philosophical grounds that the universe was unchanging on the large scale. He believed it was unchanging in how the great clusters of stars, called galaxies, relate to each other. This meant that his number could not be zero, for technical reasons.

But some years later, it was found that the galaxies were in fact all rushing away from one another. In Einstein’s mind, an informal version of the Ockham analysis immediately took place and he reverted to the value zero for his number, which is called the ‘cosmological constant’ today. In this spirit, a translation of the Latin Ockham’s Razor, ‘entia non sunt multiplicanda praeter necessitatem’ would be ‘Parameters should not proliferate unnecessarily’. This particular plot has thickened though: the value of Einstein’s cosmological constant is once again in question. Is it zero, or is it very small, and should be chosen so as to best fit the data?

We don’t know yet. This is why these questions are exciting.

http://www.abc.net.au/rn/science/oc...ies/s118778.htm

Regarding Newton's or Lagrangian/Hamiltonian formulation of classical mechanics, is my understanding that the later is simpler in some aspects, which would speak in its favor if one where forced to choose. But I am not claiming that today there is an exact formulation of Ockham's razor that can decide this. My claim in only that scientists intuitively today choose when it is overwhelmingly clear which choice is simpler while still fitting the data.

I have attempted to answer all questions, while you repeatedly avoids answering mine. I repeat: You have avoided answering the examples by Anthony Garret in my previous post. And failed to produce a quote by Foster. And ignore that MML is being used already to choose between theories. And failed to produce an alternative theory of how science works.

 

[ZapperZ]

You have no theory of how science works, using only gut instinct. Somehow, this tells you that some theories are automatically legitimate while others are not. That fraud or mass psychosis could be an explanation is automatically dismissed on no grounds. (at least concerning certain fields, which decided again by gut feeling. I guess you would be more open to this if ID researchers claimed to have proven their theory conclusively).

Let's get these VERY clear:

1. YOU were the one who claimed that there is this "universal theory" of how science works. Now, is this correct, or not?

2. *I* never make any such claim, that there is this "universal theory" oh how science works. Now, is this correct, or not?

Somehow, just because I question the validity of what you are claiming, it automatically implies that *I* have a theory that I want to push? What kind of twisted logic did you apply to get to this conclusion? So please STOP asking me for an "alternative". I didn't propose ONE (Note: there's a HINT there).

Also you seem somewhat fixated by high-Tc theories, probably because you know that the highly specialized area makes it impossible to discuss the details by those not experts. But note also that exactly the same arguments used on Einstein/Newton could be used here. If I wanted, I could make a lots of claims about highly specialized areas of medicine that you know nothing about, claiming that you know nothing about science. Instead, since this a general discussion about the philosophy of science, choose areas more commonly known.

Again, you proclaim that ALL of science follows such "rule". Now, correct me if I'm wrong. If I make a claim that the speed of light is a constant in vacuum all the time and in all cases, then anyone, from a ditch digger, to a "mechanic", to a surgeon, to anyone, can point out to me where, in their profession, this claim is violated. The speed of light is not a constant only for physicists! Now examined what YOU did. You claim that in ALL of science, the ONLY way to pick the correct theory is by just applying Occam's razor.

Now step back, and look at your claim from MY perspective. I look at what I do, and what I practice, and would you ever need to wonder why I would ask what I did? Let's get REAL here for a second. I see no such "rule" being applied, so I ASKED you to show how your "rule" is being applied, or have been applied, or WILL be applied to this case. It is current, still evolving, and still hot and important, with all the main characters still ALIVE! I would think you would jump at this opportunity to show that, yes, I can prove that my "rule" can make an accurate prediction of what's going to happen even BEFORE it happens (after all, we do that in physics all the time!). What could be MORE convincing than that?!

I NEVER denied the usefulness of Occam's razor, even when many physicists don't even know there's a name to what they practice. However, I claim that it is only useful in knowing how to do phenomenology. I claim and have shown examples where it is NOT used in selecting which theory is correct, and which isn't. This isn't a proposal of a theory. It is a contradiction via examples to what you want me to believe. The examples you have cited ALL contains the "rule" on how to look at a set of data, and extract the relevant information. NOWHERE, even in the one you cited here, was there a direct application of Occam's razor to select a correct theory. If you look VERY carefullly, there is ALWAYS an experimental observation that became the linchpin that distinguish two different theories. Somehow, this fact is completely overlooked!

Regarding Newton's or Lagrangian/Hamiltonian formulation of classical mechanics, is my understanding that the later is simpler in some aspects, which would speak in its favor if one where forced to choose. But I am not claiming that today there is an exact formulation of Ockham's razor that can decide this. My claim in only that scientists intuitively today choose when it is overwhelmingly clear which choice is simpler while still fitting the data.

There we go again, fitting the data. It is clear to me that you do not know the difference between phenomenology and theory. I cannot answer the rest of your questions to me if you are not able to comprehend the difference between those two.

Zz.

 

[Chronos]

Quite a spirited exchange. But, passion does not embody the principles of empiricism, rationalism or skepticism [the three pillars of the scientific method]. Interestingly enough, the body of literature on the scientific method does not mention Ockhamism as essential to the scientific method. At least not by sources from people engaged in the actual practice of science. If you accept the premise that people who actually practice science have more expertise in the matter than those who philosophize about how it should be practiced, the choice is pretty clear. Attempting to apply Ockhamism to chose between competing theories is clearly not an empirical approach. Without empiricism there is no science, just opinions. This is not to say philosophy plays no role in science. Philosophers, by definition, are commited to rationalism — which means they must scrupulously avoid all forms of unproven beliefs, illogic, and especially self-contradiction. In other words, whatever is to be believed, must be believed for specific, fundamentally sound reasons. I fail to see how Ockhamism qualifies as a specific, fundamentally sound reason to believe or disbelieve anything. It is merely a lantern.

 

[Les Sleeth]

. Philosophers, by definition, are commited to rationalism — which means they must scrupulously avoid all forms of unproven beliefs, illogic, and especially self-contradiction. In other words, whatever is to be believed, must be believed for specific, fundamentally sound reasons.

A philosopher is not required to subscribe to rationalism, at least if you mean one must in general prioritize it above being empirical. If you mean a philosopher by definition focuses on the rationality of ideas being proposed, then I suppose you are right. I myself believe a person is in general on the firmest ground when prioritizing experience first, reason second. You know, observe to find out what is true over using logic to say what is true.

 

[Chronos]

rationalism

A philosopher is not required to subscribe to rationalism, at least if you mean one must in general prioritize it above being empirical. If you mean a philosopher by definition focuses on the rationality of ideas being proposed, then I suppose you are right. I myself believe a person is in general on the firmest ground when prioritizing experience first, reason second. You know, observe to find out what is true over using logic to say what is true.

I agree, Les. Despite all the formalisms and disputes concerning what leads to discovery, empiricism is the foundation of anything remotely related to science or philosophy. I only intended to object to Ockham's razor as necessary or relevant to the scientific method. While both sound and logical, it is inadequate. Philosophy without science is opinion, and science without philosophy is merely fact without meaning. In that sense, I would say philosophy is the lantern of science.

 

[Aquamarine]

It seems that this discussion about the foundations of science stirs quite agitated feelings. In a way this is a strange since Ockham's razor as the scientific method is all that stand in the way of the postmodernistic explanations.

Let's take another approach. Ockham's razor is simple the statement that when curve-fitting, one should find the solution that best fits the data while avoiding overftting. Understand that a cuve is a theory, it is more than the data. Now, a objection is that this general goal is very vague and is not a description of how science is practiced today. This is true. The scientific method today follows one particular solution to this problem, namely cross-validation.

Cross-validation is essentially that the data should be divided into subsets and that the value of a theory is how well it predicts the data on "virginal" subsets. That is, subsets not used in making the theory.

Now this is a very powerful and robust solution of how to apply Ockham's razor. But it is in no way the only solution. And it has some great disadvantages.

One is the requirement for "virginal" data. Let's assume that an alien civilization sends a long message to Earth. The cryptographers solve this message, using all the data available. But according the cross-validation, "virginal" data never seen by the cryptographers is required to decide if this is the correct solution. So if one demands that only cross-validation should be used in science, then even if a solution if found that gives perfect meaning to a very long message, this is pseudoscience.

One other problem with cross-validation is the assumption that the subset will contain all the information of the whole set. But it may well be that the only way to find a solution is by looking at all the data available. Saving some data for "virginal" testing may make it impossible to find a solution.

Furthermore, even if cross-validation can find a solution, it is in not certain that it is more efficient at doing this compared to alternatives like MML, MMD, AIT or AIC.

So in the end, the critiques have mistaken one good solution of how to apply Ockham's razor with the whole scientific method. There may well be many different ways to apply Ockham's razor, with different ways having different efficiency depending on the particular situation.

 

[ZapperZ]

It seems that this discussion about the foundations of science stirs quite agitated feelings. In a way this is a strange since Ockham's razor as the scientific method is all that stand in the way of the postmodernistic explanations.

Let's take another approach. Ockham's razor is simple the statement that when curve-fitting, one should find the solution that best fits the data while avoiding overftting. Understand that a cuve is a theory, it is more than the data. Now, a objection is that this general goal is very vague and is not a description of how science is practiced today. This is true. The scientific method today follows one particular solution to this problem, namely cross-validation.

But this is where you KEEP missing the point. Curve fitting is NOT a theory! What you fit, isn't a theory! I fit the tunneling density of states of a superconductor as part of my graduate work. The "curve fit" is NOT A THEORY. It is a phenomenological model designed to extract useful information when there isn't an acceptable theory already in place! It is why I keep repeating that you do not understand the difference between phenomenolgy and theory.

You seem to somehow have a very jaundice view of what scientists face everyday. You think that we are often faced with a situation where we have to choose between "Newton's gravitational Law" and "fraud" or "mass psychosis". I WISH we have that kind of a choice - it would make my job a whole lot easier. The REALITY is that we don't! I can't remember reading about, or being faced with, that TYPE of choices! So your examples or analogies are completely divorced from any reality of what is being practiced in science! What we are often faced with is the very type of example I gave in choosing between spin fluctuation or phonons in high-Tc superconductors, or between Newtonian versus Hamiltonian. These ARE the types of choices we have to deal with, so deal with it! It is meaningless to come up with ridiculous scenario the very same way it is ridiculous to fit a 96 order polynomial and a 2nd order will do!

What it boils down to is the complete lack of understanding of the practice of physics. Therefore, you end up making unrealistic scenario that none of us typically are faced with. The fact that in all the examples you gave, the deciding factor is a series of empirical evidence where two competing theories deviate, and NOT an application of Occam's razor, somehow gets overlooked. There is a difference in claiming that Occam's razor is used, and claiming that occam's razor is the "universal rule" that one uses in picking out which theory is correct. I assert that the latter is false. I have shown two separate REAL (and not ridiculously made up) instances where you have failed to apply Occam's razor and pick the "correct" theory. I can bring up even MORE examples.

But I won't. There is no point in explaining this, when you are still stuck with the idea that "fitting a curve" is all that we do and that this curve fit IS a theory. It explains why there are quacks in the Physics section who have the outrageous gall to claim that QM is nothing more than "curve fitting". One tends to make such claims when one is ignorant of what it is.

Zz.

 

[Aquamarine]

I recommend that you read Foster's "Philosophy of the Quantitative Sciences: Curve Fitting and Cross Validation".
http://philosophy.wisc.edu/forster/

It is a great introduction to the philosophy of science. Read especially part two, about curve-fitting. And it should contain enough examples from physics to satisfy all. Although I don't accept his conclussions about cross-validation, due to problems stated above.

You have not given two examples where Ockham's razor have failed. Since most scientists think that the principles of cross-validation is the scientific method, and they are using it to evaluate theories, and cross-validation is in fact a valid application of Ockham's razor, then those scientists are using Ockham's razor. Including yourself.

 

[ZapperZ]

You have not given two examples where Ockham's razor have failed. Since most scientists think that the principles of cross-validation is the scientific method, and they are using it to evaluate theories, and cross-validation is in fact a valid application of Ockham's razor, then those scientists are using Ockham's razor. Including yourself.

Read these carefully:

1. I have NEVER stated that this principle isn't used, even unconsciously.

2. I have shown that "curve fitting" is NOT the same as "theory".

3. I have shown where YOU can't use OR principle in picking out which of the two mechanism is valid for High-Tc superconductor. This proves that there has to be MORE than just applying one simple principle.

4. Your selection of which of the two to pick - Newtonian laws versus Lagrangian/Hamiltonian - shows clearly that there are no such thing as a clear-cut scenario where you can simply apply a simplistic principle and choose which theory is correct.

5. I am not out here to prove that OR is wrong! <shock!> I have never made such claim even if you are forcing me to. What I objected to was your insistence that, given a number of competing theories on EQUAL FOOTING (and not some ridiculously made up theories for comparison), you can simply use OR to pick out which theory is correct. YOU have proven this to be false yourself by your inability to apply this principle in picking out the "correct" theory of superconductivity.

Zz.

 

[Fliption]

earlier you compared a philospher to a CEO and a physicist to a production line worker, how were we meant to take this analogy then?

I see what you mean now. I didn't mean that a philospher directs and controls the way a CEO does a company. I was merely drawing an analogy about the relationship between a CEO and a line worker. A CEO has no direct involvement in the line process, yet he/she does impact the operation on some level. I was just trying to get some people out of the linear thinking that can't seem to understand how anyone not working on the line could have an impact on the line.

 

[Fliption]

To be fair, I think your business analogy with philosophy presumably as the CEO may have been misleading. I think I know what you were getting at, but I can see how others would interpret the analogy in a sense stronger than what you intended.

edit: oops, I see jcsd has already pointed that out. Apparently this thread sprouted yet another page while I wasn't looking.

Yes, I caught it. I forgot that most people participating here don't seem to be very competent at interpreting analogies.

 

[Les Sleeth]

I agree, Les. Despite all the formalisms and disputes concerning what leads to discovery, empiricism is the foundation of anything remotely related to science or philosophy. I only intended to object to Ockham's razor as necessary or relevant to the scientific method. While both sound and logical, it is inadequate. Philosophy without science is opinion, and science without philosophy is merely fact without meaning. In that sense, I would say philosophy is the lantern of science.

Nicely put Chronos.

 

[Les Sleeth]

All Real Scientists Unite . . . with Philosophers!

After careful consideration, the utterly objective judges have unanimously determined that the philosophy side of this debate has won. Congratulations to the Socratic wanna-bees team! The deciding factor? Well, what tipped the scales in favor of the wisdom fellows (otherwise known as wise asses) was that the leader of the science team spent a better part of the weekend participating in a debate concerning the epistemological value of Occam's Razor. Due to his deeply passionate involvement in a philosophical debate, and quite a lively one at that, the judges had no choice but to conclude he’s finally seen the light (which the philosophy side saw along), and realizes that philosophy and science are kindred spirits, like peas in a pod, forever and ever linked in the pursuit of Truth. Let’s raise our glasses to a more collaborative future!

 

[metacristi]

Regarding falsification, here are many concrete examples that it is not how science works:
http://www.stephenjaygould.org/ctrl/gardner_popper.html

Hm...Not a good link.The mere popperian falsificationism has problems but is still the best approach we have at disposal,it only needs some 'improvments'.I think Lakatos' has already done this with enough success,though of course there is,still,no methodology without problems.

The main problems of popperianism is the situation of nonobservables and of non testable in isolation physical laws.It has more or less the same problems that 'plague' Ayer's 'weak' principle of verification though under falsificationism the problems of meaning and 'reliable verification' are dropped.Basically the same logical arguments can be used against falsificationism too.

For example,a 'classical' one,under the mere popperian account Newton's first law (the law of inertia) is metaphysical for it is not testable in isolation and therefore should not be part of science!Thus we need 'improvments' (I've proposed some) without becoming extremists and claiming that strong realism is justified by the 'no miracle' argument (unfortunately not sound,not yet at least).

Returning at the main point well,Martin Gardner put the problem as if the inductivist approach (it is however the oldest methodology since Bacon at least),in 'new clothes' this time,has proved to be superior to falsificationism.This is not true,I do not think that inductionists are in a better position now than in the time of Hume and Kant (for they induction is apriori and does not need justification).Or the so called Cambridge school (Russell,Keynes before turning to economy and so on) and Vienna Circle (logical positivists such as Neurath,Carnap,Reichenbach).Or Ayer.Indeed there is no sound argument in favor of a general inductionist method,probabilistic approaches included.

At most we can combine falsificationism with inductionism in some cases where a bayesian approach seems to support induction.This means that in some cases,when we have a wide range of corroborations (and no exceptions) the bayesian approach does enable scientists to believe that a certain law or theory is approximatively true instead of merely preferring it to all other existing,alternative,proposals as in the usual falsificationist approach.

For example if we could prove somehow that there is possible only a single main approach,all other proposals being only variations of the main approach (coming from powerful and unifying principles).Unfortunately this happens very very seldom in practice and almost never in physics,which is 'plagued' by the underdetermination of theories.

[Anyway] accepting the bayesian approach (this is a correct stance in some cases),at least punctually,in no ways imply certitudes,the bayesian approach is subjective in nature,so scientists are still fully open even to nontrivial changes in the future (at least that some attributes assigned to scientific constructs dissappear altoghether and are replaced by other).So falsificationism is retained even in such cases.There are no certitudes in science (not to mention the problems with ontological and epistemological idealism).

 

[ZapperZ]

For example,a 'classical' one,under the mere popperian account Newton's first law (the law of inertia) is metaphysical for it is not testable in isolation and therefore should not be part of science!Thus we need 'improvments' (I've proposed some) without becoming extremists and claiming that strong realism is justified by the 'no miracle' argument (unfortunately not sound,not yet at least).

Well, FINALLY you decided to use a concrete example, and an odd one too. If you think that Newton's First Law is... er... "metaphysical" and "... should not be part of science", do you feel the same way for Newton's 2nd and 3rd Laws also?

Zz.

 

[metacristi]

Well, FINALLY you decided to use a concrete example, and an odd one too. If you think that Newton's First Law is... er... "metaphysical" and "... should not be part of science", do you feel the same way for Newton's 2nd and 3rd Laws also?

Zz.

Indeed ignorance is bliss...Honestly the feeling is that I speak with the winds.Anyway those who [are] prepared to understand will understand (puting also the hand on a book on philosophy of science,it is not my goal to teach here).The rest should maturize by themselves.

 

[RetiredMD]

Indeed ignorance is bliss...Honestly the feeling is that I speak with the winds.Anyway those who [are] prepared to understand will understand (puting also the hand on a book on philosophy of science,it is not my goal to teach here).The rest should maturize by themselves.

I have enjoyed your insights Metacristi. I am not sure why ZapperZ is hostile to your input. It seems like you are more in agreement with his views than opposed to them.

Have a nice day,
Morris

 

[RetiredMD]

After careful consideration, the utterly objective judges have unanimously determined that the philosophy side of this debate has won. Congratulations to the Socratic wanna-bees team! The deciding factor? Well, what tipped the scales in favor of the wisdom fellows (otherwise known as wise asses) was that the leader of the science team spent a better part of the weekend participating in a debate concerning the epistemological value of Occam's Razor. Due to his deeply passionate involvement in a philosophical debate, and quite a lively one at that, the judges had no choice but to conclude he’s finally seen the light (which the philosophy side saw along), and realizes that philosophy and science are kindred spirits, like peas in a pod, forever and ever linked in the pursuit of Truth. Let’s raise our glasses to a more collaborative future!

Thanks for my daily laugh. You may be overly optimistic about future collaborations between science and philosophy.

Have a nice day,
Morris

 

[ZapperZ]

Indeed ignorance is bliss...Honestly the feeling is that I speak with the winds.Anyway those who [are] prepared to understand will understand (puting also the hand on a book on philosophy of science,it is not my goal to teach here).The rest should maturize by themselves.

Yikes!

I asked a question to try to understand WHY you stated that Newton's First Law is considered to be "metaphysical" and "should not be a part of science", especially when implicitly, it appears that you have excluded the 2nd and 3rd law! This appears puzzling and inconsistent to me. And all I got in return were attacks for asking!

I'm sorry I asked!

Zz.

 

[metacristi]

I have enjoyed your insights Metacristi. I am not sure why ZapperZ is hostile to your input. It seems like you are more in agreement with his views than opposed to them.

Have a nice day,
Morris

Maybe.But from what I've read from him I really doubt that he has really understood even the basics of the popperian approach.Thus I find futile to discuss with him the clear limits of [falsificationism],it's a waste of time and effort.

 

[arildno]

Maybe.But from what I've read from him I really doubt that he has understand even the basics of the popperian approach.Thus I find futile to discuss with him the clear limits of [falsificationism],it's a waste of time and effort.

Well you have proven yourself to be completely ignorant of physics; your condescending tone is that of an arrogant fool.

 

[ZapperZ]

I have enjoyed your insights Metacristi. I am not sure why ZapperZ is hostile to your input. It seems like you are more in agreement with his views than opposed to them.

Have a nice day,
Morris

Honestly, I failed to see where I am being "hostile" towards the whole posting. I was infact happy that metacristi finally used some concrete example to illustrate a point. And if you think I am not in agreement with the claim that something is "metaphysical" and thus "should not be a part of science", then you are wrong.

However, why would this stop me from questioning a section of what is being said, especially when I find it rather strange. I didn't realize that I should be forced to accept it whole or none at all. The explanation given on why JUST the 1st law is "metaphysical" and not the 2nd and 3rd is vague. Is this not a legitimate question to ask?

Oy vey!

Zz.

 

[metacristi]

Yikes!

I asked a question to try to understand WHY you stated that Newton's First Law is considered to be "metaphysical" and "should not be a part of science", especially when implicitly, it appears that you have excluded the 2nd and 3rd law! This appears puzzling and inconsistent to me. And all I got in return were attacks for asking!

I'm sorry I asked!

Zz.

I'm tired,this is the last time when I adress you.If you had been really interested you'd have tried at least to read something more on philosophy of science and tried a constructive approach (I'm not even sure if you knows the difference between the inductionist approach and falsificationism).You act as if you knew so well but in fact without really understanding...honestly the 'fun' you try to express with your words is a clear sign of your manifest ignorance on the subject,as someone more experienced in philosophy of science could easily indicate.Why should I waste my time with you?Anyway as I've already said you lack the basics in logic and philosophy of science.Obviously when someone,totally ignorant,challenge me to explain physics because otherwise he does not believe a scientific statement the wisest approach is not to do so.The best approach is that he should read by himself.Anyway there is too much to be said.

Try eblaforum suggested resources:

http://www.eblaforum.org/main/viewtopic.php?t=22

 

[ZapperZ]

You act as if you knew so well but in fact without really understanding...honestly the 'fun' you try to express with your words is a clear sign of your manifest ignorance on the subject...

Now that's funny, because I've been saying that when you make claims about science and physics in particular. It appears that the burden of "learning about things" only falls on me, but not on you.

Obviously when someone,totally ignorant,challenge me to explain physics because otherwise he does not believe a scientific statement

Pardon me, but what "scientific" statement did you just make? You said that Newton's First Law is a metaphysics and should not be considered as science. You call this a "scientific statement"? And you accuse me of being ignorant of things I'm talking about? Hello?

No, what is ignorant is the fact that you realize you are STUCK with making a statement that you cannot support. You do not want to answer my question because you cannot claim that Newton's 2nd and 3rd Laws are also "metaphysical". If it is, then you'd better get out of the building you're in, and not cross any bridges. By doing that, you automatically show (at least to anyone with a knowledge first year intro physics lesson) of the inconsistency of claiming that the 1st law is metaphysical.

Maybe you can get away with making unchecked statements like this elsewhere, who knows. But to turn this around and attack me for questioning what you just said... now that's a diversion tactic if I've ever seen one. You want me to read a philosophy of science book? Sure... only if you spend time learning Halliday and Resnick Intro Physics text.

Zz.