How do you deal with the 'science is a religion' crap?

[Dancing David]

Its not about whether a theory has been falsified. Its about whether a theory is, in principle, falsifiable.

Yup. So some religous theories are potentially falsifiable. (But proponents don't want to test them.)

The point is that you have a model that can be tested, and that you do so. The exceptions to the model disprove it or modify it.

(I am agreeing not disagreeing)

 

[Tubbythin]

However, even if you ignore that, I still ask what use it would be if a concept was only theoretically falsifiable, yet practically impossible.

Athon

You can still compare it to falsifiable theory B. Even if neither theory can be directly falsified we can evaluate which theory is, say, the simplest - say the number of free parameters or the number of assumptions we have to make for a theory to be possible. This isn't the case with unfalsifiable theory C since it has no free parameters and the assumptions are flexible so long as the final conclusion remains in tact.

 

[athon]

You can still compare it to falsifiable theory B. Even if neither theory can be directly falsified we can evaluate which theory is, say, the simplest

Parsimony is indeed useful as a means of comparing the potential of different theories. However falsification isn't useful because of parsimony - it's useful because it creates a means of removing useless ideas from the game, increasing the probability that a remaining theory could be useful.

- say the number of free parameters or the number of assumptions we have to make for a theory to be possible.This isn't the case with unfalsifiable theory C since it has no free parameters and the assumptions are flexible so long as the final conclusion remains in tact.

I'm afraid I'm not sure what you mean by 'free parameters'. Could you explain? I'm still not clear on why you're making such a strong correlation between falsification and parsimony.

Athon

 

[Tubbythin]

Odd, because right off the bat you then say-



So, you're not trying to argue the philosophy of science...you're just trying to argue the philosophy of science.

Huh.

Different definitions of philosophy I think.

I fully understand your example. And it is true - one is practically falsifiable, and the other isn't. However I was commenting on whether falsification defines science or not. It doesn't, no more than its absence might define religion.

I've already agreed that falsifiability alone does not make a theory a scientific theory. However, I struggle to see how a theory that is in principle unfalsifiable can be considered of any scientific use.

Ok, given you stated above that you weren't arguing the philosophy of science, this is really confusing. Either you're stating what science is - a philosophical argument - or you're not. Using falsification to define it (something that I've heard all too often, and is way too simplistic) won't do it, I'm afraid.

What I suppose I meant was I have no real knowledge of or the desire to discuss the philosophy of science as might be taught in a lecture course entitled "The Philosophy of Science". I have knowledge of what people mean by falsifiability, reproduceability etc from personal experience. I suppose these things relate directly to what the philosophy of science is. But not to what I was thinking of as The Philosophy of Science which is basically a subject that covers who said what and when. Any clearer? Quite possibly not.

Ok, I'll simplify it. Popper was the guy who first came up with the whole falsification thing, right? It was part of an attempt to define science as a method called the 'hypothetico-deductive method' (although it was actually another guy who came up with this method, and he just made it popular). In really simple terms, it was all a response to the decline of an earlier attempt to define science, namely something called logical positivism.

Seeing a pattern? Yeah, the falsification thing also had problems. Fast forward to today. We're still no closer to putting science into a nice, neat box using simple terms.

Maybe I have the wrong end of the stick. Maybe when you said 'such as', I mistook the statement as you (as with Perpetual Student) finding it difficult to believe something could be considered scientific if it cannot be falsified. Correct me if I'm wrong and I'll sincerely apologise.

Errm, see above?

 

[Tubbythin]

Parsimony is indeed useful as a means of comparing the potential of different theories. However falsification isn't useful because of parsimony - it's useful because it creates a means of removing useless ideas from the game, increasing the probability that a remaining theory could be useful.

Falsification allows us to remove a dud theory. But for it to be falsified it needs to be falsifiable (obviously). For the latter to be the case the theory needs to be well defined eg based on some set of equations plus certain constants and possibly some number of free parameters. (This is clearly the case with LCDM but not "Goddidit".) Once we have this well defined theory we can then compare it to another well defined theory (eg comparing LCDM to a steady-state model) and compare parsimony. Even if neither of the theories is falsifiable in practice at the current time we can still evaluate these theories in terms of simplicity and consider ways of making one or both of them falsifiable in practice. There's no guarantee it'll give the right answer but it can give a hint as to where to look.

I'm afraid I'm not sure what you mean by 'free parameters'. Could you explain? I'm still not clear on why you're making such a strong correlation between falsification and parsimony.

By free parameters I mean values that appear in a theory that cannot be explicitly determined by the theory, they can only be measured. Eg the Standard Model of particle physics is very successful at explaining fundamental particles but has something like 20 free parameters - various masses, coupling strengths and mixing angles. A Grand Unified Theory would unite the three forces and remove several of these free parameters (by merging them in to one or two).

 

[athon]

I've already agreed that falsifiability alone does not make a theory a scientific theory.

I missed it. In which case, I apologise.

However, I struggle to see how a theory that is in principle unfalsifiable can be considered of any scientific use.

Evolution is in itself non-falsifiable. I'm aware of statements like those of J.B.Haldane's that claim if we found a modern fossil in an ancient era, it would falsify evolution. However, I disagree. I think if we found a fossil that resembled that of a modern rabbit in the Cambrian, it would be insufficient in reality for people to abandon an otherwise useful theory. It would in practice take a lot more for people to abandon an otherwise robust idea, even if there was now a fly in the ointment.

As I understand it (and I might be wrong), relativity was originally unfalsifiable. I dimly recall an article on that from its early days, although might be misremembering. I'd need to look it up again. In any case, it would mean that relativity wouldn't be science until somebody could find a way to falsify it.

That is where my problem lies with the view that falsification defines science, and that anything that cannot be falsified is useless. It's not how science works at all - I can guarentee that Darwin didn't give two hoots about whether his hypothesis could be falsified at all.

I refer to my earlier statement about metals and temperature. It was an inductive claim that all metals have a melting point, and it is technically unfalsifiable. Imagine coming across a metal that you couldn't melt - there could always be greater temperature we couldn't feasibly reach at which it would melt. Yet the claim is still useful, given our understanding of physics.

What I suppose I meant was I have no real knowledge of or the desire to discuss the philosophy of science as might be taught in a lecture course entitled "The Philosophy of Science". I have knowledge of what people mean by falsifiability, reproduceability etc from personal experience.

That actually surprises me more, as on a practical level science doesn't work on falsification. This was perhaps one of the biggest criticisms levelled against falsification - in the real world, science progresses in increments according to what a theory predicts. If a theory successfully predicts something, it gains the confidence of those who are testing it. Falsification is useful, but if an idea is presented without being shown to be falsifiable, scientists won't always dismiss it completely. Thankfully.

I suppose these things relate directly to what the philosophy of science is. But not to what I was thinking of as The Philosophy of Science which is basically a subject that covers who said what and when. Any clearer? Quite possibly not.

Yes, quite. Thanks.

I do find it surprising that you seem to be interested in how science works, yet aren't interested in learning more about the topic. It's like somebody saying they're interested in physics, but couldn't care to take the time to learning any of its laws or the works done by physicists.

Athon

 

[athon]

Falsification allows us to remove a dud theory. But for it to be falsified it needs to be falsifiable (obviously). For the latter to be the case the theory needs to be well defined eg based on some set of equations plus certain constants and possibly some number of free parameters. (This is clearly the case with LCDM but not "Goddidit".) Once we have this well defined theory we can then compare it to another well defined theory (eg comparing LCDM to a steady-state model) and compare parsimony. Even if neither of the theories is falsifiable in practice at the current time we can still evaluate these theories in terms of simplicity and consider ways of making one or both of them falsifiable in practice. There's no guarantee it'll give the right answer but it can give a hint as to where to look.

For the most part, there's no argument here. I think it is more of an issue of definition than parsimony or falsification, however. The problem with the god-of-gaps principle is that it is too vague to be of use. There is nothing in it to observe, let alone falsify, hence nothing to distinguish it as a theory. My 'pet' field of study is vitalism, for instance. In some guises it has been similar to your comparison of LCDM and 'goddidit'. It failed not because it couldn't be falsified, or even because it wasn't parsimonious. It failed because it offered nothing that could be observed to distinguish it from any other idea, so failed purely on an empirical level. In other words it couldn't even make it to the table, let alone be worded in a way that it might be tested and falsified.

By free parameters I mean values that appear in a theory that cannot be explicitly determined by the theory, they can only be measured. Eg the Standard Model of particle physics is very successful at explaining fundamental particles but has something like 20 free parameters - various masses, coupling strengths and mixing angles. A Grand Unified Theory would unite the three forces and remove several of these free parameters (by merging them in to one or two).

Ah, thanks. Appreciated.

Athon

 

[technoextreme]

The claim would be falsified if we encountered a metal substance that, when exposed to all available temperatures, remained solid. As such, the claim is a scientific one.
We continue to believe that claim because, when working with a new metal, we observe that it melts at a defined temperature.

No not really. It would be defying the laws of thermodynamics.

 

[Tubbythin]

I missed it. In which case, I apologise.

No problem.

Evolution is in itself non-falsifiable. I'm aware of statements like those of J.B.Haldane's that claim if we found a modern fossil in an ancient era, it would falsify evolution. However, I disagree. I think if we found a fossil that resembled that of a modern rabbit in the Cambrian, it would be insufficient in reality for people to abandon an otherwise useful theory. It would in practice take a lot more for people to abandon an otherwise robust idea, even if there was now a fly in the ointment.

Sure. But if we found a rabbit and elephant and a human I think people would start to get concerned.

As I understand it (and I might be wrong), relativity was originally unfalsifiable. I dimly recall an article on that from its early days, although might be misremembering. I'd need to look it up again. In any case, it would mean that relativity wouldn't be science until somebody could find a way to falsify it.

Special relativity? I fail to see how this could be.

That is where my problem lies with the view that falsification defines science, and that anything that cannot be falsified is useless. It's not how science works at all - I can guarentee that Darwin didn't give two hoots about whether his hypothesis could be falsified at all.

I refer to my earlier statement about metals and temperature. It was an inductive claim that all metals have a melting point, and it is technically unfalsifiable. Imagine coming across a metal that you couldn't melt - there could always be greater temperature we couldn't feasibly reach at which it would melt. Yet the claim is still useful, given our understanding of physics.

But the claim that all metals have a melting point isn't a theory I wouldn't say, its a deduction from observation. We can then use this deduction to construct a theory that explains this result.

That actually surprises me more, as on a practical level science doesn't work on falsification. This was perhaps one of the biggest criticisms levelled against falsification - in the real world, science progresses in increments according to what a theory predicts. If a theory successfully predicts something, it gains the confidence of those who are testing it. Falsification is useful, but if an idea is presented without being shown to be falsifiable, scientists won't always dismiss it completely. Thankfully.

I agree that modern physics research isnt entirely about falsification. But it does play a big role. Just look at the number of experiments that are trying to "break" SR or GR or the Standard Model or QED.

I do find it surprising that you seem to be interested in how science works, yet aren't interested in learning more about the topic. It's like somebody saying they're interested in physics, but couldn't care to take the time to learning any of its laws or the works done by physicists.

There's only limited hours in a day and I prefer to spend my time learning the physics rather than learning about people who talk about physics .

 

[Perpetual Student]

athon:

It has already been stated here that by scientific falsification, we are referring to a basic scientific theory that can be potentially falsified. Your comment about metals and temperature shows a misunderstanding. The statement "all metals have a melting point" is a consequence of underlying fundamental (quantum) theories that are potentially falsifiable.

The statement "it cannot be falsified that all metals have a melting point" is like saying "it cannot be falsified that all humans are primates."

The theory of evolution is a consequence of the synthesis of basic discoveries in geology, embryology, taxonomy, genetics, etc, each of which has potentially falsifiable aspects.

Nevertheless, if every human being in earth tomorrow at 12:00 Greenwich time were to see an apparition of an omnipotent deity telling us how he created the universe and planted false evidence of evolution to test us, we would probably abandon the theory of evolution.

 

[Beth]

My 'pet' field of study is vitalism, for instance. In some guises it has been similar to your comparison of LCDM and 'goddidit'. It failed not because it couldn't be falsified, or even because it wasn't parsimonious. It failed because it offered nothing that could be observed to distinguish it from any other idea, so failed purely on an empirical level. In other words it couldn't even make it to the table, let alone be worded in a way that it might be tested and falsified.
Athon

Interesting. How do you define the theory of vitalism? I think of it as the idea that life arises spontaneously from non-living materials. But I could be completely off base about it. I don't even know why I have this hazy idea of what it is.

athon:

The statement "it cannot be falsified that all metals have a melting point" is like saying "it cannot be falsified that all humans are primates."

It seems to me that saying all humans are primates like saying all squares are rectangles. It's a tautology due to the definitions. It can never be falsified because it is a tautology.

But the claim all metals have a melting point is non falsifiable for practical reasons rather than definitional ones. Assuming that there existed a metal that did not have a melting point, we could never be certain that it will never melt or that we simply cannot heat it up to it's melting point. I understand that that example is perhaps not the best as we have theoretical reasons [i.e. at some temperatures, the atoms come apart and metals no longer exist.], but I think he is simply trying to illustrate that there are different types of unfalsifiable statements. Tautologies can never be false. But not every non-falsifiable claim is a tautology. Some could be false, we just can't tell. I think it's an important distinction.

 

[athon]

Sure. But if we found a rabbit and elephant and a human I think people would start to get concerned.

About the theory of evolution? Honestly, I doubt it.

Let's imagine the scenario in realistic terms - the scientific community learns through a conference of a journal report of human-like skeletons being found in a layer of rock that is dated as the Cambrian. I can all but completely guarantee that this would be weighed against the data that would state such a thing is impossible and it would be dismissed as an anomaly at best, or ignored as a hoax at worst. It might happen another few times, and unless a new theory taking this anomaly into account was proposed, it would remain an anomaly. Different people would have their opinions on it, and over a generation or two science might slide onto a new theory that is an adjustment on evolution.

In other words, the reality is that the theory would not spontaneously be abandoned because of a single observation, because theories take into account so much information. It's nice to think that this might be the case in principle, but science just doesn't work that way.

Special relativity? I fail to see how this could be.

As I said, I can't argue that one from an informed position. I recall being presented the argument once that it was similar to evolution in that there was no single observation that could be made which would (or would have on its acceptance) rendered it an unsuitable explanation.

But the claim that all metals have a melting point isn't a theory I wouldn't say, its a deduction from observation. We can then use this deduction to construct a theory that explains this result.

Ah, I see your point. I guess I see more to science than just theories. To say it isn't science if it isn't a theory, again, seems like a useless categorisation. Even if it was an agreed definition, there is no particular point at which a hypothesis magically turns into a theory. It's a gradual slide over time (some ideas progress faster than others).

I agree that modern physics research isnt entirely about falsification. But it does play a big role. Just look at the number of experiments that are trying to "break" SR or GR or the Standard Model or QED.

Absolutely no argument. Falsification is a wonderful tool for pragmatically finding ideas that can be dealt with empirically. My point isn't that it is useless, but rather that people have tried before to define the borders of science using this tool, and failed.

There's only limited hours in a day and I prefer to spend my time learning the physics rather than learning about people who talk about physics .

Heh.

I admit, I have a taste for learning about the history of scientists as much as I do science itself. But let me put it this way - it would be as if I had pushed on discussing special relativity from a theoretical point of view (which I am admittedly weaker on than I assume you would be) and stated I had no wish to actually learn why it was developed and what has been said on it before. I'm sure you might find it odd.

Nonetheless, I'm glad we resolved this and both learned some new things from it.

Athon

 

[athon]

athon:

It has already been stated here that by scientific falsification, we are referring to a basic scientific theory that can be potentially falsified. Your comment about metals and temperature shows a misunderstanding. The statement "all metals have a melting point" is a consequence of underlying fundamental (quantum) theories that are potentially falsifiable.

The statement "it cannot be falsified that all metals have a melting point" is like saying "it cannot be falsified that all humans are primates."

See my response to Tubbythin.

The theory of evolution is a consequence of the synthesis of basic discoveries in geology, embryology, taxonomy, genetics, etc, each of which has potentially falsifiable aspects.

I'm confused - you just said above -

It has already been stated here that by scientific falsification, we are referring to a basic scientific theory that can be potentially falsified.

Now you're saying it's not the theory that need be falsified, but that the 'discoveries' themselves (I'm going to assume you mean the deductions made from the discoveries, as falsifying a discovery doesn't make sense). Correct me if I've made a mistake here, but which is it?

Nevertheless, if every human being in earth tomorrow at 12:00 Greenwich time were to see an apparition of an omnipotent deity telling us how he created the universe and planted false evidence of evolution to test us, we would probably abandon the theory of evolution.

Again, as I said in my above post, I doubt that very much.

A range of explanations would be proposed that attempted to reconcile that observation with past ones. People would claim it was a mass hallucination, or a hoax. If it repeated, there would be attempts to study it. There'd be nothing in an observation of a figure appearing out of nothing (regardless of what it said) that would contradict all of the data we have supporting evolution.

Athon

 

[athon]

Interesting. How do you define the theory of vitalism? I think of it as the idea that life arises spontaneously from non-living materials. But I could be completely off base about it. I don't even know why I have this hazy idea of what it is.

It depends on which time period we're talking about. In some contexts, it's used to refer to things such as Aristotle's three spirits (animal, vegetable and rational). In others it's used to refer to vague 'fluids' such as the elans vitale, responsible for movement or sensations.

Yet its most recent form was to describe why cells can reproduce and differentiate. Basically it was a god-of-gaps answer that stated since it was impossible that mechanisms involving chemistry and physics could be responsible for making a cell split in half, and then again, and then again...on and on...there must be a non-physical phenomenon at work. In essence it wasn't a theory in the scientific sense as it wasn't clear enough to be tested. It was an argument from ignorance more than anything. But the history of it is sure interesting.

Athon

 

[not daSkeptic]

But the claim all metals have a melting point is non falsifiable for practical reasons rather than definitional ones.

I agree it's not due to definitional reasons, but I'm not sure I agree with it being due to practical reasons. I suppose it all depends on whether one views the problem of induction as a matter of logic or practicality. I would think the former. It is illogical, in my view, to claim knowledge about the entirety of a potentially infinite set. It is impractical, yes, but only as a result of it being illogical. The very definition of infinity ensures it will never be practical.

 

[Perpetual Student]

It seems to me that saying all humans are primates like saying all squares are rectangles. It's a tautology due to the definitions. It can never be falsified because it is a tautology.

OK, fair enough. However, it would be quite similar to the statement, "all atoms can be annihilated." Both melting and annihilation are derived characteristics due to underlying fundamental physical laws. There exist countless such statements that are not fundamental theories or laws.

 

[BenBurch]

We don't need faith to believe in the power of the atom bomb.

"If the radiance of a thousand suns were to burst into the sky that would be like the splendor of the Mighty One and I am become Death, the shatterer of worlds." - Bhagavad Gita

 

[lomiller]

Ok, pretending that you're right (never read much Popper, hey?), what good would that do? If you claim all objects in the universe of a certain mass undergo fusion, it's technically falsifiable according to you. But practically it's not something that can be tested yet. Perhaps in the future, but not currently. Hence at the moment it is beyond testability, which was the main thrust of Popper's argument (and a focus of criticism against his work).

IMO Popper is largely overblown in his impact on science but what he is saying is essentially as follows. The scientific method is roughly:

Make observations
Use those observations to form a hypothesis
Develop predictions based on that hypothesis
Test those predictions against further observations

What Popper is saying is that this only works if you test can achieve a negative result. I.E. if the all of the possible outcomes of the test support the hypothesis then the test isn’t adding any new information and therefore doesn’t add any support to the hypothesis. Furthermore, if every possible test is as the one I described above, then you can't use this method to conclude anything about the hypothesis.

 

[Perpetual Student]

The theory of evolution is a consequence of the synthesis of basic discoveries in geology, embryology, taxonomy, genetics, etc, each of which has potentially falsifiable aspects.

I'm confused - you just said above -

It has already been stated here that by scientific falsification, we are referring to a basic scientific theory that can be potentially falsified.
Now you're saying it's not the theory that need be falsified, but that the 'discoveries' themselves (I'm going to assume you mean the deductions made from the discoveries, as falsifying a discovery doesn't make sense).

Correct me if I've made a mistake here, but which is it?

Nevertheless, if every human being in earth tomorrow at 12:00 Greenwich time were to see an apparition of an omnipotent deity telling us how he created the universe and planted false evidence of evolution to test us, we would probably abandon the theory of evolution.

Again, as I said in my above post, I doubt that very much.

A range of explanations would be proposed that attempted to reconcile that observation with past ones. People would claim it was a mass hallucination, or a hoax. If it repeated, there would be attempts to study it. There'd be nothing in an observation of a figure appearing out of nothing (regardless of what it said) that would contradict all of the data we have supporting evolution.

The point is that all scientific theories are testable and potentially falsifiable.
A simultaneous hallucination seen by seven billion people would certainly get our attention, although I'm sure there would be many holdouts. A task for science would then be to test the theory of a global hallucination vs. the theory of a deity. It would still be science, not religion, which does not embrace testing and potential falsification.

 

[Limbo]

Has the word scientism come up in this thread?

"The term scientism is used to describe the view that natural science has authority over all other interpretations of life, such as philosophical, religious, mythical, spiritual, or humanistic explanations, and over other fields of inquiry, such as the social sciences. The term is used by social scientists like Hayek or Karl Popper to describe what they see as the underlying attitudes and beliefs common to many scientists. They tend to use the term in either of two equally pejorative directions:

1. To indicate the improper usage of science or scientific claims as a counter-argument to appeals to scientific authority in contexts where science might not apply, such as when the topic is perceived to be beyond the scope of scientific inquiry.

2. To refer to "the belief that the methods of natural science, or the categories and things recognized in natural science, form the only proper elements in any philosophical or other inquiry," with a concomitant "elimination of the psychological dimensions of experience". It thus expresses a position critical of (at least the more extreme expressions of) positivism.

[...]

But naturally, the word scientism doesn't apply to any of you...

http://video.google.com/videoplay?docid=-6575509359036720225#