What Is "falsifiable"?

[hazelm]

This does not quite fit this category but I can not find any better place.  Maybe someone else can.

 

Everyone knows this sentence which is the test of a scientific hypothesis.   It has troubled me ever since the first time I read it.  Let's see if someone can clarify.

 

Scientific standards require that the hypothesis must be not only testable but also falsifiable.

 

I am asking what does "falsifiable" mean as it is applied to a scientific hypothesis.  Does it mean  that the hypothesis must be proven wrong?  That is the common definition of "falsifiable" but, in this case, it seems that would mean that almost all science "facts" are false. But that isn't what the sentence says. So, does it mean that someone who disagrees with your hypothesis must be able to set up his own hypothesis showing a reasonable possibility that the first hypothesis could be wrong?  

 

Ex:  Let's say someone of Einstein's ilk announces that he has found how life started.  He explains it in his usual masterminded way.  You disagree with his hypothesis.  What do you have to do to "falsify" his hypothesis?  Prove him wrong or show the possibility of him being wrong?

 

Thank you.

Edited April 2, 2018 by hazelm

[exchemist]

This does not quite fit this category but I can not find any better place.  Maybe someone else can.

 

Everyone knows this sentence which is the test of a scientific hypothesis.   It has troubled me ever since the first time I read it.  Let's see if someone can clarify.

 

Scientific standards require that the hypothesis must be not only testable but also falsifiable.

 

I am asking what does "falsifiable" mean as it is applied to a scientific hypothesis.  Does it mean  that the hypothesis must be proven wrong?  That is the common definition of "falsifiable" but, in this case, it seems that would mean that almost all science "facts" are false. But that isn't what the sentence says. So, does it mean that someone who disagrees with your hypothesis must be able to set up his own hypothesis showing a reasonable possibility that the first hypothesis could be wrong?  

 

Ex:  Let's say someone of Einstein's ilk announces that he has found how life started.  He explains it in his usual masterminded way.  You disagree with his hypothesis.  What do you have to do to "falsify" his hypothesis?  Prove him wrong or show the possibility of him being wrong?

 

Thank you.

Falsifiable means capable in principle of being falsified, if the right type of evidence to falsify it were to be found. Obviously it does not mean that that evidence has already been found, since it that were so the hypothesis would be already discredited.

 

In other words you need to distinguish between falsifiable (in principle, in the future) and falsified.

  

In order to considered scientific, a hypothesis must allow someone to conceive of an experiment or observation of nature that would test the hypothesis, such that if it failed the test it would be falsified. For instance, relativity and quantum theory have been subjected to a series of tests, to see if their counterintuitive predictions are borne out in reality. So far, all of them have been. But these theories are falsifiable because IF these tests had had outcomes in DISagreement with the theory, then they would have been falsified.

 

By contrast, the hypothesis that, say, the universe arose from the collapse of another parallel universe is unfalsifiable, because there is absolutely no way to put it to the test observationally. It is a speculation that cannot be tested. 

 

Generally, a scientific hypothesis or theory is falsifiable because it makes predictions about the outcome of future observations. For instance, the theory of evolution asserts ancestral relationships between living organisms which predict which creatures should have similar DNA and which ones should have dissimilar DNA. It also predicts what "missing links" might be found in the fossil record and the age band of the rocks in which they should be found. These prediction are tested every time DNA comparisons are done and every time a new fossil type is found. If we were to find rabbit bones in Cambrian rocks, or that the DNA of a whale was closely related to that of a fish rather than a hippopotamus, then it would be falsified.         

[hazelm]

Falsifiable means capable in principle of being falsified, if the right type of evidence to falsify it were to be found. Obviously it does not mean that that evidence has already been found, since it that were so the hypothesis would be already discredited.

 

In other words you need to distinguish between falsifiable (in principle, in the future) and falsified.

  

In order to considered scientific, a hypothesis must allow someone to conceive of an experiment or observation of nature that would test the hypothesis, such that if it failed the test it would be falsified. For instance, relativity and quantum theory have been subjected to a series of tests, to see if their counterintuitive predictions are borne out in reality. So far, all of them have been. But these theories are falsifiable because IF these tests had had outcomes in DISagreement with the theory, then they would have been falsified.

 

By contrast, the hypothesis that, say, the universe arose from the collapse of another parallel universe is unfalsifiable, because there is absolutely no way to put it to the test observationally. It is a speculation that cannot be tested. 

 

Generally, a scientific hypothesis or theory is falsifiable because it makes predictions about the outcome of future observations. For instance, the theory of evolution asserts ancestral relationships between living organisms which predict which creatures should have similar DNA and which ones should have dissimilar DNA. It also predicts what "missing links" might be found in the fossil record and the age band of the rocks in which they should be found. These prediction are tested every time DNA comparisons are done and every time a new fossil type is found. If we were to find rabbit bones in Cambrian rocks, or that the DNA of a whale was closely related to that of a fish rather than a hippopotamus, then it would be falsified.         

So, if it can be shown that either now or in the future with new tools or methods the hypothesis could be tested (there is no life in the universe except on Earth), then it is falsifiable?   If it is shown that never would we be able to confirm the hypothesis (there are not and never were black swans), the hypothesis is not falsifiable? 

 

My first example might also be not falsifiable if everyone agreed that, even with travel capabilities, we might never be able to cover he entire universe.  Right? 

 

I think I'll have to study this a bit.  Thank you.

[exchemist]

So, if it can be shown that either now or in the future with new tools or methods the hypothesis could be tested (there is no life in the universe except on Earth), then it is falsifiable?   If it is shown that never would we be able to confirm the hypothesis (there are not and never were black swans), the hypothesis is not falsifiable? 

 

My first example might also be not falsifiable if everyone agreed that, even with travel capabilities, we might never be able to cover he entire universe.  Right? 

 

I think I'll have to study this a bit.  Thank you.

Yes I think that's right. For me the best way to think of it is whether a hypothesis or theory makes testable predictions about what could be observed, at least in principle, i.e. even if the observation is one that it technically difficult to make.  

 

A hypothesis or theory that makes no testable predictions is not scientific - or not yet. Interestingly there are some speculations, for instance in cosmology, that make no testable predictions. It is possible that one day somebody may work out an observable consequence of some of these ideas that could be tested. So there can be grey areas, it seems to me. 

[sanctus]

"there is no life in the universe except on Earth" is not falsifiable like, I think you said, because it requires knowledge of the whole universe.
On the other hand "only on earth there is life in our solar system" is falsifiable, it is enough to find one bacteria on eg. Mars and the theory is falsified.

The best example is when asked to "prove there is no god" since it is a perfect example of non-falsifiable hypothesis since it requires knowledge of everything/universe in order to say that something is not there. On the other hand the hypothesis "there is god" is falsifiable (depending on definition of god I guess), since it is enough to find one instance of where god isn't.


But back to the initial question, the simple answer is: a hypothesis is falsifiable if one can potentially make an experiment which sees whether the predictions from the hypothesis are right or wrong. NB.: the "potentially" is important, just look at string-theory or super-symmetry as an example: we do not have the energy to test it yet, but if CERN 2.0 (or X.0) gets powerful enough we will ; so these theories are falsifiable even if not yet experimentally possible.

[hazelm]

Yes I think that's right. For me the best way to think of it is whether a hypothesis or theory makes testable predictions about what could be observed, at least in principle, i.e. even if the observation is one that it technically difficult to make.  

 

A hypothesis or theory that makes no testable predictions is not scientific - or not yet. Interestingly there are some speculations, for instance in cosmology, that make no testable predictions. It is possible that one day somebody may work out an observable consequence of some of these ideas that could be tested. So there can be grey areas, it seems to me. 

And never stop testing.  I say that because of an article I read just recently.  This scientist thought he had a good idea but no one could duplicate his results.  Time went by while he kept trying.  He was about to give it up when, after a long silence, someone wrote to say he got it.  Then came another.  Now they are all back in the loop with results.    You can almost hear "I told you so"?

 

So, if it is testable, then we see if others follow your steps and get the same results.  If yes, you are in business.  If no one else gets it, back to the drawing board.  And that is why the hypothesis is falsifiable?  Because his hypothesis was testable and got results regardless of whether the results were positive or negative?

[hazelm]

"there is no life in the universe except on Earth" is not falsifiable like, I think you said, because it requires knowledge of the whole universe.

On the other hand "only on earth there is life in our solar system" is falsifiable, it is enough to find one bacteria on eg. Mars and the theory is falsified.

 

The best example is when asked to "prove there is no god" since it is a perfect example of non-falsifiable hypothesis since it requires knowledge of everything/universe in order to say that something is not there. On the other hand the hypothesis "there is god" is falsifiable (depending on definition of god I guess), since it is enough to find one instance of where god isn't.

 

 

But back to the initial question, the simple answer is: a hypothesis is falsifiable if one can potentially make an experiment which sees whether the predictions from the hypothesis are right or wrong. NB.: the "potentially" is important, just look at string-theory or super-symmetry as an example: we do not have the energy to test it yet, but if CERN 2.0 (or X.0) gets powerful enough we will ; so these theories are falsifiable even if not yet experimentally possible.

I probably did say that, Sanctus.  I was getting my tongue (and brain) tangled.    You are right.  "Life in the whole universe" is not falsifiable because it cannot be tested - for now, at least.  What made me stumble around with it was thinking the day could come.  Even here on Earth, where we think we've explored the entire planet, we keep finding little tribes that we did not know were there.  There was such a very recent case in South America.

 

That said, it's your last paragraph that keeps me re-thinking it.  If the potential of a successful test in the future exists, is it falsifiable?   Maybe, sometimes it's a draw.

[sanctus]

In philosophy of science they say "a hypothesis is scientific if falsifiable" so following that reasoning one could say that whether present technology allows for the test or not does not matter, as long as it is potentially possible to test its predictions.

 

[hazelm]

In philosophy of science they say "a hypothesis is scientific if falsifiable" so following that reasoning one could say that whether present technology allows for the test or not does not matter, as long as it is potentially possible to test its predictions.

 

 

Makes sense to me.  :-)

[Moronium]

"there is no life in the universe except on Earth" is not falsifiable like, I think you said, because it requires knowledge of the whole universe.

 

 

I can't follow the reasoning here.  You wouldn't need to have knowledge of the whole universe to falsify this claim.  Any life found anywhere in the universe would falsify it.  You would presumably need knowledge of the entire universe to prove it, but the question here isn't about proof.  It's about falsification.

 

If the claim were "there is life *somewhere* in the universe," then I agree that you would need knowledge of the entire universe to disprove (falsify) it.  On the other hand, in that case, finding life anywhere would prove it.

Edited April 26, 2018 by Moronium

[JMJones0424]

"there is no life in the universe except on Earth" is not falsifiable like, I think you said, because it requires knowledge of the whole universe."

 

This question can easily be abstracted to something that we can more readily understand.  "There are no humans outside of North America" is not a falsifiable statement because in order to show it to be false, we must know everything that exists outside of North America does not include humans.  We know the statement to be false, but the statement is still not falsifiable.  However, the statement "Humans only exist in North America" can be falsified because once we find an instance of humans existing outside of North America, the claim can be shown to be false.

 

Falsifiable statements limit our claim of knowledge, while unfalsifiable statements make assertions on areas that we can not test.  Falsifiability is separate from veracity.

Edited April 27, 2018 by JMJones0424

[Moronium]

Jim, if I'm reading you correctly, then I disagree with you for the reasons set forth in post #10. I can't follow this logic at all, I'm afraid:

 

There are no humans outside of North America" is not a falsifiable statement because in order to show it to be false, we must know everything that exists outside of North America does not include humans.

 

 

It seems to me that if I can find even one person living in central america, south america, or anyplace else on the planet, then I have "falsified" the claim being made.

 

What am I missing?  Where has my thinking "gone wrong?"

Edited April 27, 2018 by Moronium

[JMJones0424]

As an addendum-

I can't follow the reasoning here.  You wouldn't need to have knowledge of the whole universe to falsify this claim.  Any life found anywhere in the universe would falsify it.  You would presumably need knowledge of the entire universe to prove it, but the question here isn't about proof.  It's about falsification.

 

If the claim were "there is life *somewhere* in the universe," then I agree that you would need knowledge of the entire universe to disprove (falsify) it.  On the other hand, in that case, finding life anywhere would prove it.

You're right, any life found anywhere in the universe would falsify the claim that there is no life in the universe other than on Earth.  But this doesn't change the falsifiability of the claim.  Say you found life on Mars, then the claim is then easily shown to be false.  But say you didn't find life until the last possible place to look, then the claim was true until you looked there.  This is why the claim is unfalsifiable.  The falsifiability of the claim is not linked to the veracity of the claim.  Instead, falsifiability is about whether a claim can be shown to be false.  There is no life outside of the Earth is not a falsifiable claim.  The only life that exists is on Earth is a falsifiable claim.

[JMJones0424]

Moronium, I think I was typing my last reply as you made your reply, but I think I answered your question. 

 

Falsifiability is not the same as veracity.  There are no members of hypography by the name of Moronium is a falsifiable statement, regardless of its veracity.  We can examine all members of hypography and show that at least one is named Moronium.

 

Moronium is not a member of any online community is not a falsifiable statement because in order to show the statement to be false, we must know all members of all online communities.

Edited April 27, 2018 by JMJones0424

[Moronium]

  But say you didn't find life until the last possible place to look, then the claim was true until you looked there.  

 

 

I'm still completely lost.  

 

1.  "Falsifiable" is not the same as "falsified" (past tense).

 

2.  I disagree with your claim about what is" true."  If there is (or isn't) life in "the last place" you looked, then the claim would always have been true (or false).  What you "know" does not dictate what is, or can be, "true."

 

....falsifiability is about whether a claim can be shown to be false.  There is no life outside of the Earth is not a falsifiable claim.  The only life that exists is on Earth is a falsifiable claim.

 

 

 

To me both claims are "falsifiable", in theory at least.  Truth be told, I really don't see any significant difference between the two claims, so I wouldn't be able to say that one is falsifiable and one isn't, under any circumstances.

Edited April 27, 2018 by Moronium

[Moronium]

Moronium, I think I was typing my last reply as you made your reply, but I think I answered your question. 

 

Falsifiability is not the same as veracity.  There are no members of hypography by the name of Moronium is a falsifiable statement, regardless of it veracity.  We can examine all members of hypography and show that at least one is names Moronium.

 

Moronium is not a member of any online community is not a falsifiable statement because in order to show the statement to be false, we must know all members of all online communities.

 

 

Well, you didn't answer all of my questions, I don't think, Jim, but I agree with others here who say that "falsifiable," in this context, simply means a proposition can, in theory, be falsified, even if we don't have the time or means to test it immediately.

 

Sometime who cared to take the time to check them all could review "all online communities," even if it would be tedious, inconvenient, and/or not worth the effort.  But that's not even the real issue.  What is theoretically possible is not always practically possible, but the practical problems, standing alone, don't invalidate the theory.

 

Falsifiability is not the same as veracity. 

 

 

 

I agree with this. In fact I made the same point in post #10.  Proving something to be false is a much different proposition than proving it to be true.  But I guess I don't agree with the conclusions you are drawing (inferring) from that distinction.

Edited April 27, 2018 by Moronium

[JMJones0424]

Pay someone to teach you then.