Falsificationism J Blackmon Outline Biographical Highlights The Problem of Demarcation Inductivism Falsificationism Problems for Falsificationism Brief Bio Karl Popper 19021994 AustrianBritish ID: 249820
Download Presentation The PPT/PDF document "Karl Popper’s" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Karl Popper’s Falsificationism
J. BlackmonSlide2
Outline
Biographical Highlights
The Problem of Demarcation
Inductivism
Falsificationism
Problems for
FalsificationismSlide3
Brief Bio
Karl Popper
1902-1994
Austrian-British
Known for
falsificationism, critical rationalism.The Logic of Scientific DiscoveryRejected the popular interpretation of QMSlide4
The Problem of Demarcation
What distinguishes science from pseudo-science and, in general, from non-science?
One answer came from logical positivism and empiricism:
The inductive method,
hypothetico-deductivism
. Hypotheses are confirmed (not proven) when their logical consequences turn out to be true. Confirmation confers more reason to believe something.Slide5
The Problem of Demarcation
What distinguishes science from pseudo-science and, in general, from non-science?
Popper: “But this did not satisfy me…”
This is an understatement.
Popper holds that
confirmation is a myth.
That is to say that hypotheses are not given more justification when their logical consequences turn out to be true.Slide6
The Problem of Demarcation
The following form
H
P
P
Thus, H
is a logical fallacy. Everyone knows this. The H-D method was never supposed to offer a guarantee.
But it was supposed to show how, as more observations came in, a hypothesis might become better confirmed.
This is what Popper rejects. In addition to being logically fallacious, it’s not good science.Slide7
The Problem of Demarcation
H
1
P
P
Thus, we have more confirmation of H
1
According to H-D, each time a prediction, P, is observed, H gets more confirmation.
But as Popper insists, and as everyone else knew, H
1
could be false, and yet P occurs for some other reason.
There’s always some H
2
or H
3
(and so on) that also predicts P. And there are infinite alternatives to H
1
.Slide8
The Problem of Demarcation
Also, we are subject to confirmation bias.
Confirmation Bias: The tendency to fixate on and emphasize data that confirms one’s prior beliefs while ignoring or dismissing counter-evidence.
While the term, confirmation bias, is not Popper’s, he is distinctly aware of its power.Slide9
The Problem of Demarcation
Also, we are subject to confirmation bias.
Popper: “Confirmation is everywhere.”
But while people are inclined to see this as a good thing, this is only an illusion.
The more easily a theory can be confirmed, the more trivial and less scientific it is.Slide10
The Problem of Demarcation
Example: Freudianism
The parent who strikes the child in anger confirms the Freudian’s belief in an id, while the parent who refrains from striking in anger instead confirms the Freudian’s belief in a counter-acting super-ego.Slide11
The Problem of Demarcation
Example: Freudianism
This is
not
good news for Freud’s theory!
The Freudian sees confirmation everywhere.But this is not reason to think observation supports the theory…Slide12
The Problem of Demarcation
Example: Freudianism
Instead, it’s reason to think the theory rules our hardly any observations.Slide13
The Problem of Demarcation
Other Examples
Astrology
Adler’s Psychology
Marxism
Such theories were not good scientific theories because they rule out nothing or very little.Hardly any observation could falsify them.Slide14
The Problem of Demarcation
Imagine the weather prediction: Tomorrow we will have rain, or not.
Or the astronomer saying: Last night’s comet will reappear tonight somewhere in the northern hemisphere…Slide15
The Problem of Demarcation
Thus Popper is an inductive skeptic.
As he sees it, David Hume proved there could be no logic of inductive reasoning.
No number of confirming instances can justifiably increase our confidence in a theory.Slide16
The Problem of Demarcation
Thus Popper is an inductive skeptic.
Popper is
not simply
a
fallibilist.Fallibilism: We can never be completely certain about matters of fact.
He is an unabashed inductive skeptic.Slide17
The Problem of Demarcation
What distinguishes science from pseudo-science and, in general, from non-science?
Popper’s Answer:
Falsificationism
demarcates the scientific theories from the rest.
Falsificationism: A hypothesis is scientific if and only if it has the potential to be refuted by some possible observation.
A hypothesis is bold to the extent that it risks falsification.Slide18
The Problem of Demarcation
The following form
H
P
~P
Thus, ~H
is logically valid.
(~ is the symbol for
not
.)
We do make scientific progress, but not by confirming hypotheses.
We aggressively attempt to
refute
our hypotheses.
We learn more and more, as time goes on, what’s false. Slide19
The Problem of Demarcation
The Holy Grail Analogy
Imagine that there are infinite grails: all of them glow for a while, but only the Holy Grail glows forever.
If you have a glowing grail, you know you might have the Holy Grail.
If it stops glowing you know you don’t have the Holy Grail, and you can drop it and look for a new glowing grail.
But you never know if you have the Holy Grail.Slide20
The Problem of Demarcation
The Holy Grail Analogy
There are infinite scientific hypotheses, but only the true one will never be falsified.
If you have and unfalsified hypothesis, you know you might have the truth.
If it is falsified, you know you don’t have the truth, and you can drop it and look for a new hypothesis.
But you never know if you have the truth.Slide21
The Problem of Demarcation
The Holy Grail Analogy
Hypothesis :: grail
Glowing :: being unfalsified
Holy Grail :: True Hypothesis
Ceasing to glow :: falsification
Blackmon:
This analogy could be better.
Godfrey-Smith writes that we carry the grail around.
But the knights should be vigorously testing the grails.Slide22
The Problem of Demarcation
Popper on Scientific Change
Stage 1: Conjecture: Offer a hypothesis H that might describe and explain something about the world. A good conjecture is a bold one, one that takes risks.
Stage 2: Attempted Refutation: Subject H to critical testing in an attempt to show that H is false. Once H is refuted, go back to Stage 1.
One prohibition…Slide23
The Problem of Demarcation
Popper on Scientific Change
One prohibition: If H predicts P and H is refuted by the observation that ~P, then you cannot offer an H* in Stage 1 that differs from H only in not predicting P.
In other words, you cannot make isolated excuses for your hypothesis.
No
ad hoc revisions to your hypothesis.Slide24
The Problem of Demarcation
Popper on Scientific Change
One prohibition: No
ad hoc
revisions!
Otherwise, you may have once had a scientific hypothesis, but you are no longer following a good scientific methodology.
Anthony
Flew’s
Invisible Gardener
H: A gardener tends this area.
P: Within a few days, we’ll see a gardener.
But instead, we never see a gardener.
H has been falsified.Slide25
The Problem of Demarcation
Popper on Scientific Change
One prohibition: No
ad hoc
revisions!
Otherwise, you may have once had a scientific hypothesis, but you are no longer following a good scientific methodology.
Anthony
Flew’s
Invisible Gardener
Ad
hoc
revision, H*: An invisible gardener tends this area.
Popper will cry foul here.
The new hypothesis should be bold, not a retreat.Slide26
The Problem of Demarcation
Popper on Marxism
Marxist Conjecture: Industrialized capitalist societies will experience a proletarian revolution and become socialist states.Slide27
The Problem of Demarcation
Popper on Marxism
Originally, Marx’s conjecture counted as a scientific hypothesis—at least if we put a time limit on it.
It is falsifiable.Slide28
The Problem of Demarcation
Popper on Marxism
Falsifiable: Just wait and see if (within a generation or whatever) the revolutions occur.
If not, the hypothesis is falsified.Slide29
The Problem of Demarcation
Popper on Marxism
So far, so good.
However, what did Marxists do when no such revolutions occurred after a period of time?
They did not treat Marxism as refuted. Slide30
The Problem of Demarcation
Popper on Marxism
Instead, according to Popper, they made an
ad hoc
revision. Slide31
The Problem of Demarcation
Popper on Marxism
H
*: Industrialized capitalist societies will experience a proletarian revolution and become socialist
states, unless welfare appeases the people.Slide32
The Problem of Demarcation
How does Albert Einstein’s General Theory of Relativity fare?Slide33
Objections and Problems
How do we recognize falsification?
Logically, observing an A that is a non-B falsifies the conjecture
All As are
Bs
.ExampleConjecture: All pieces of iron expand when heated.
Falsifier: any observation of a piece of iron which does not expand when heated.Slide34
Objections and Problems
How do we recognize falsification?
But suppose you observe what you took to be iron contracting when heated.
Should you treat the conjecture as refuted?
How can you be sure that instead this is not iron after all?
How can you be sure that your measurements of the contraction and the temperature are accurate?Slide35
Objections and Problems
How do we recognize falsification?
Generally
Maybe the conjecture that All As are
Bs
is true, but our belief that we are observing a falsifying instance (an A that is a non-B) is false.Slide36
Objections and Problems
How do we recognize falsification?
Recall
Quine’s
point about holism about testing and the
Duhem-Quine thesis: No statement is testable in isolation.When we observe the unexpected, instead of rejecting the hypothesis, we can reject one of the background assumptions.
We can even reject the description of the observation itself.Slide37
Objections and Problems
How do we recognize falsification?
Example
Conjecture: The universe is 13.8 billion years old.
Falsifier: The “Methuselah Star”, HD 140283, is 14.5 billion years old.Slide38
Objections and Problems
How do we recognize falsification?
Example
Astronomers did not reject this conjecture.
They re-examined their measurement of HD 140283.Slide39
Objections and Problems
How do we recognize falsification?
Example
Rejecting the conjecture in this case would seem rash.
What can Popper say?Slide40
Objections and Problems
How do we recognize falsification?
Here’s what Popper can’t say:
We should try to maintain our most confirmed hypotheses. If we have more reason to believe the universe is 13.8 billion years old, then we suspect the hypothesis that
HD
140283 is 14.5 billion years old. If instead, the age of HD 140283 is more confirmed, then we suspect the hypothesis about the age of the universe.Slide41
Objections and Problems
How do we recognize falsification?
Popper can’t say this because he rejects confirmation.
An
inductivist
can handle the problem of holism about testing because the inductivist is committed (right or wrong) to the idea that some hypotheses can be more confirmed than others. Whenever there is a question about which one to reject, we can choose accordingly.
Popper appears to have no such option.Slide42
Objections and Problems
How do we recognize falsification?
Popper admitted that the other assumptions were also conjectures which might be false.
He also admitted that logic could never force one to reject a conjecture instead of rejecting some background assumption.
Popper held that a good scientist would not reject background assumptions just to protect a conjecture.Slide43
Objections and Problems
How do we recognize falsification?
Popper also admitted that we cannot be certain about a reported falsifier.
His answer is that the scientist must make a decision about whether to accept the report or not.
Godfrey-Smith finds that Popper
has retreated from using logic to demarcate science from pseudo-science.Slide44
Objections and Problems
How do we recognize falsification?
Godfrey-Smith finds that Popper
has retreated from using logic to demarcate science from pseudo-science
.
Popper appears to rely more and more on method than on pure logic.Slide45
Objections and Problems
Falsificationism
cannot recognize nontrivial probability attributions as scientific.
Conjecture: Event E has a probability of 0.0001.
This has no falsifier.Slide46
Objections and Problems
Falsificationism
cannot recognize nontrivial probability attributions as scientific.
Example: If a coin is fair, then the probability of Heads 100 times in a row is extremely low.
But it is still possible.
Observing 100 Heads in a row doesn’t falsify the conjecture that the coin is fair.Are such conjectures really unscientific?Slide47
Objections and Problems
Falsificationism
cannot recognize nontrivial probability attributions as scientific.
Popper agreed that logically speaking all such hypotheses are unscientific.
However, in practice, the scientist can decide on thresholds at which nontrivial probability attributions are treated as 0 and 1.
In a given experiment, you might treat a probability less than 0.0001 as just 0.0.This is falsifiable.Slide48
Objections and Problems
Falsificationism
cannot recognize nontrivial probability attributions as scientific.
Godfrey-Smith holds that, as before, Popper has retreated from using logic to demarcate science from pseudo-science.
Popper is now relying on “good scientists” making judgment calls.Slide49
Objections and Problems
The Bridges Problem
Suppose we are going to build a bridge, and we are to choose between two theories, H1 and H2, which determine different ways of building a bridge.Slide50
Objections and Problems
The Bridges Problem
H1 has been tested many times and had passed every test.
H2 is a brand new conjecture which has never been tested.
Neither H1 nor H2 has been falsified.
Would it be irrational to use H2 to build the bridge?Slide51
Objections and Problems
The Bridges Problem
Would it be irrational to use H2 to build the bridge?
Popper never says we have
more
reason to use H1.But he does say that H1 is more corroborated.Slide52
Objections and Problems
The Bridges Problem
But he does say that H1 is more
corroborated
.
In such cases, it is rational to choose the more corroborated theory.Slide53
Objections and Problems
The Bridges Problem
So, what is
corroboration
?
Is it just confirmation or not?If it’s confirmation, then Popper has relinquished his heroic stance against inductivism.But if it’s not confirmation, then how does it help us determine whether to use H1 or H2?Slide54
Objections and Problems
Popper’s Dilemma
Either corroboration is like confirmation in that it gives us more reason to prefer corroborated theories over uncorroborated theories, or corroboration is not like confirmation and cannot be regarded as any guide to truth. [68]
If the former, then Popper can answer the Bridge Problem, but he has relinquished his inductive skepticism.
If the latter, then who cares if a theory is corroborated? Slide55
Objections and Problems
Godfrey-Smith’s Analogy
The transcript is merely a record of what you have done.
The transcript is like corroboration in this sense: it is backward-looking.
The
letter of rec usually covers both what you have done and what you are likely to do
.
The letter is more like confirmation.Slide56
Objections and Problems
Popper’s Dilemma
It appears that Popper caught between saying that we have no good reason to expect one hypothesis about bridge-building to work no matter how corroborated that theory is, or admitted that his view is in some sense “
inductivist
”.
Popper appears to be in a problematic position.Slide57
Objections and Problems
Alternatives
Hypothetico
-Deductive Method
Strong InferenceSlide58
Objections and Problems
Alternatives:
Hypothetico
-Deductive Method
Generate a hypothesis, H.
Deduce observational predictions, P, from H.
Test to see if these predictions come out true.
If they do, the H gains support.
If they do not, H should be rejected.
Return to Step 1.Slide59
Objections and Problems
Alternatives:
Hypothetico
-Deductive Method
Generate a hypothesis, H.
Deduce observational predictions, P, from H.
Test to see if these predictions come out true.
If they do, the H gains support.
If they do not, H should be rejected.
Return to Step 1.Slide60
Objections and Problems
Alternatives: Strong Inference
Godfrey-Smith calls this the “Sherlock Holmes” method.
Holmes called his reasoning deductive.Slide61
Objections and Problems
Alternatives: Strong Inference
While philosophers point out that most of his reasoning was inductive, his overall method often took the form of the process of elimination.
In that sense, it was deductive.Slide62
Objections and Problems
Alternatives: Strong Inference
Identify all possible hypotheses and their individual predictions.
Critically test them, rejecting each one until the last one is standing.Slide63
Objections and Problems
Alternatives: Strong Inference
Popper’s reply?
Popper argues that there will always be infinite other theories to choose from.
We can eliminate, but we can’t narrow it down.