Notes on Popper's Conjectures and Refutations: Chapter 1
Justin Mallone
Chapter 1: Conjectures and Refutations
Method of distinguishing between science and pseudo-science (the problem of demarcation) (44):
Empirical method? Nah. Astrology looks to empirical observations too (44).
Lots of confirmations / apparent explanatory power? Nah. Marxist history, psycho-analysis, and individual psychology seem to explain everything, but that's actually a flaw. (45-47)
Example of psychological theory which says people are driven by feelings of inferiority -- so a person who drowns a child did so out of a feeling of inferiority, and so did a person who saved a child (acting on a need to prove oneself) (46)
Compare scientific theories, which take risks by making predictions which exclude the occurrence of certain results -- which are incompatible with certain results of observation (47)
Important conclusions (47-48):
1. Easy to look for confirmations of almost any theory if you are looking for them
2. Confirmations should count ONLY IF they are the result of RISKY PREDICTIONS -- that is to say, if, without the knowledge of the theory, we should have expected a different, refuting result
3. Every good scientific theory is a prohibition -- it forbids certain stuff from happening. The more it forbids, the better the theory
4. A theory not refutable by any conceivable event is unscientific
5. Every genuine test of a theory is an attempt to falsify it, and some theories are more testable, because they take greater risks.
6. Confirming (corroborating) evidence should not count unless they are the result of a genuine test of a theory -- a serious but unsuccessful attempt to falsify it
7. ad-hoc reinterpretations of theories destroy or at least lower scientific status
Summary: Scientific status is falsifiability / refutability / testability
Popper gives some examples of applying his principle.
Einstein’s theory of gravitation passes – even though the scientific instruments at the time did not allow potentially falsifying test results to be ascertained, falsifying the theory was possible.
Astrology fails – Lots of supposedly “confirming” evidence, but lots of unfavorable evidence as well. Also, very vague and non-risky predictions.
Marxist theory of history – started out scientific and were falsified (e.g. Marx’s analysis of the character of the “coming social revolution,”). Then ad hoc adjustments were made that made it irrefutable and thus unscientific.
Psychological theories popper discussed fail – they were irrefutable due to lack of conceivable human behavior which could contradict them. Clinical observations as worthless for confirming theory as astrological “evidence.” Ego, super-ego, id, as scientific as the Olympic Gods.
Myths can become testable, (Parmenides myth of the unchanging block universe à Einstein’s block universe). 50.
“I thus felt that if a theory is found to be non-scientific, or ‘metaphysical’ (as we might say), it is not thereby found to be unimportant, or insignificant, or ‘meaningless,’ or ‘nonsensical.’ But it cannot claim to be backed by empirical evidence in the scientific sense – although it may easily be, in some genetic sense, the ‘result of observation.’
…
Thus the problem which I tried to solve by proposing the criterion of falsifiability was neither a problem of meaningfulness or significance, nor a problem of truth or acceptability. It was the problem of drawing a line (as well as this can be done) between the statements, or systems of statements, of the empirical sciences, and all other statements – whether they are of a religious or of a metaphysical character, or simply pseudo-scientific.” (50-51)
Popper discusses his problem of demarcation in contrast to Wittgenstein's problem of meaning (52-54), where Wittgenstein tried to show that what he thought of as philosophical or metaphysical statements were meaningless. Wittgenstein thought that the only meaningful statements were those that were built from atomic propositions -- and these atomic propositions were at least in-principle ascertainable by observation.
As a solution to the problem of demarcation, Popper criticizes Wittgenstein's approach as being deductionist. As to the problem of "meaningful statements," Popper regarded this as a pseudo-problem.
But Popper’s argument was misinterpreted in an interesting way -- while he was arguing for testability and falsifiability as a means for demarcating between science and pseudo-science, his view was interpreted to be arguing for a falsificationist test of meaning -- in other words, that the meaningfulness or nonsensicalness of a statement could be ascertained by whether it was falsifiable or not.
For example, Wittgenstein used the example of a non-sensical pseudo-proposition such as "Socrates is identical." This is nonsense, and its negation, "Socrates is not identical," is also nonsense, and, so, the logic went, the negation of a meaningful statement would also be meaningful. But as Popper notes in n.6 on page 54, the negation of a testable (falsifiable) statement need not be testable (discussed more in The Logic of Scientific Discovery) For example, consider the statement "at least one barn is red," and its negation, "no barns are red." One requires us only to find one barn -- the other requires us to search the universe. This asymmetry in the falsifiability of negated statements makes nonsense of the attempt to make Popper's criteria for demarcating science from pseudo-science into a criteria for demarcating meaningfulness, since such an attempt relies on the symmetry of falsifiability in negated statements when none exists.
Popper notes Hume's solid philosophical criticism of induction and how it leads to an infinite regress, but then proceeds to criticize Hume's psychological account of induction on the same grounds (55-59). Hume tries to explain the belief in laws as a product of frequent repetition -- Popper notes that there can only be "repetition" in the sense of two situations being seen as similar from a certain interpretative point of view, since no two situations are precisely the same -- therefore, there must be a PoV BEFORE there can be any repetition -- thus, the PoV can't be the product of repetition (59)
Popper introduces his theory -- which is that, instead of having patterns impressed upon us through induction, we come up with and try and impose them on the world, and then have them refuted -- a process of conjectures and refutations. (60)
All observation is selective (61-62). there is no pure observation. Observation needs some task, interest, PoV, problem.
While inborn ideas silly, some inborn reactions or responses may exist. (62) One can refer to this as knowledge.
This knowledge is not a priori valid. Popper thinks there may be an inborn tendency to look for regularities.
Expectation of finding regularities is both psychologically and logically a priori, since all observation involves the recognition of similarities and dissimilarities.
Kant right that our intellect imposes laws upon nature, wrong that those laws are necessarily true (63).
Example of induction machine: could we build induction machine which could perform “inductions” and solve problems but it wouldn’t show that induction worked – cuz deciding what its “world” is, what laws it will discover, etc., involves embedding knowledge created by humans into the machine – problems of similarity / etc will have. (64)
Discusses dogmatic attitude vs critical attitude (64-66). Dogmatic attitude inevitable – critical attitude does not replace dogmatic attitude but is overlaid on top of it – needs dogmatic beliefs for “raw material” – like how science started with myths (66).
Popper says that deductive reasoning important NOT cuz it lets us prove theories but because it lets us tease out the implications of our theories, and crit them better. (67)
Popper connects demarcation and induction (69-72).. The belief in induction is fortified by the need for a criterion of demarcation which people think needs to be based on induction. (71).
Popper resolves what seem to be three conflicting principles:
1. (the argument from Hume and Born) the impossibility of justifying a law by observation or experiment
2. the fact that science proposes and applies universal laws
3. the principle of empiricism, which says that in science, observation and experiment must decide upon the acceptance or rejection of scientific statements
1 and 3 seem to clash, but don’t if we accept that the acceptance of theories in science is tentative only (71-72)
Only the falsity of a theory can be inferred from empirical evidence, and this inference is purely deductive
Popper argues that we jump from an observation to a good theory by conjecture and refutation (74).
Why prefer non-falsified statements? We’re truth-seeking! (74)
Popper rejects a probability calculus as solving the problem of induction. Science doesn’t seek highly probable theories – it seeks explanations, which are powerful and improbable theories.
He notes that every interesting and powerful statement must have a low probability, and vice versa – a statement with a high probability will be scientifically uninteresting. (77)
Jordan Talcot
On 2011-07-11, at 8:41 AM, Justin Mallone wrote:
> All observation is selective (61-62). there is no pure observation. Observation needs some task, interest, PoV, problem.
> While inborn ideas silly, some inborn reactions or responses may exist. (62) One can refer to this as knowledge.
>
> This knowledge is not a priori valid. Popper thinks there may be an inborn tendency to look for regularities.
>
> Expectation of finding regularities is both psychologically and logically a priori, since all observation involves the recognition of similarities and dissimilarities.
> Kant right that our intellect imposes laws upon nature, wrong that those laws are necessarily true (63).
What is the difference between inborn ideas and inborn reactions, responses, and expectations?
Jordan
Elliot Temple
On Jul 11, 2011, at 8:41 AM, Justin Mallone wrote:
> Chapter 1: Conjectures and Refutations
> Method of distinguishing between science and pseudo-science (the problem of demarcation) (44):
>
> Empirical method? Nah. Astrology looks to empirical observations too (44).
>
> Lots of confirmations / apparent explanatory power? Nah. Marxist history, psycho-analysis, and individual psychology seem to explain everything, but that's actually a flaw. (45-47)
A nice point that Popper made elsewhere is this:
"All swans are white" means "there are no swans, or all of them are white". It's the same thing. It allows the case of no swans.
Now imagine breaking up the universe into cubes 10ft on each side. Now consider how many confirming instances you can get by looking at these cubes. The vast, vast, vast majority contain no swans and are thus confirming instances. This is because most of the universe is empty!
Inductivists would no doubt try to object that a confirming instance consisting of a white swan is different than a confirming instance consisting of empty space. But what's the difference? How can they formulate their gut feeling into an actual principle of epistemology?
Both the empty space and the white swan are *consistent with* (not contradictory to) the "all swans are white" theory. In this sense they have the same relationship to it.
So this leads into the problem of "support". They want some consistent evidence to count as support while other consistent evidence does not. Or they want some consistent evidence count as *more* support than other. To do that they have to specify rules for categorizing. They have to say which evidence is which and why, and how to tell. They have to provide instructions for working it out that don't mention using gut feelings or unconscious biases.
And they have utterly failed to do that.
BTW, "no swans are white" suffers from the same flaw as "all swans are white". Empty space is consistent with it.
> Example of psychological theory which says people are driven by feelings of inferiority -- so a person who drowns a child did so out of a feeling of inferiority, and so did a person who saved a child (acting on a need to prove oneself) (46)
>
> Compare scientific theories, which take risks by making predictions which exclude the occurrence of certain results -- which are incompatible with certain results of observation (47)
>
> Important conclusions (47-48):
> 1. Easy to look for confirmations of almost any theory if you are looking for them
> 2. Confirmations should count ONLY IF they are the result of RISKY PREDICTIONS -- that is to say, if, without the knowledge of the theory, we should have expected a different, refuting result
"Count" here means (or should mean) something like "be worth paying some attention to, be interesting and relevant to the debate". Not "count as justifying".
> 3. Every good scientific theory is a prohibition -- it forbids certain stuff from happening. The more it forbids, the better the theory
> 4. A theory not refutable by any conceivable event is unscientific
> 5. Every genuine test of a theory is an attempt to falsify it, and some theories are more testable, because they take greater risks.
Vague theories avoid risks. Whenever a theory says, "X or not X" then it avoids risk. When it specifies "X for sure. Definitely not not-X" then it's at risk.
How else can risk be avoided other than vagueness (lack of specifying stuff)? In other words, is there anything that makes a theory bold other than having actual content where it asserts anything?
> 6. Confirming (corroborating) evidence should not count unless they are the result of a genuine test of a theory -- a serious but unsuccessful attempt to falsify it
Again it shouldn't count as justification.
> 7. ad-hoc reinterpretations of theories destroy or at least lower scientific status
>
> Summary: Scientific status is falsifiability / refutability / testability
>
> Popper gives some examples of applying his principle.
> Einstein’s theory of gravitation passes – even though the scientific instruments at the time did not allow potentially falsifying test results to be ascertained, falsifying the theory was possible.
> Astrology fails – Lots of supposedly “confirming” evidence, but lots of unfavorable evidence as well. Also, very vague and non-risky predictions.
> Marxist theory of history – started out scientific and were falsified (e.g. Marx’s analysis of the character of the “coming social revolution,”). Then ad hoc adjustments were made that made it irrefutable and thus unscientific.
> Psychological theories popper discussed fail – they were irrefutable due to lack of conceivable human behavior which could contradict them. Clinical observations as worthless for confirming theory as astrological “evidence.” Ego, super-ego, id, as scientific as the Olympic Gods.
>
> Myths can become testable, (Parmenides myth of the unchanging block universe à Einstein’s block universe). 50.
A bunch of ancient Greek ideas addressed *good problems* and that has helped them retain relevance today. True answers to *bad problems* are often boring, while inadequate answers to good problems are often interesting.
> “I thus felt that if a theory is found to be non-scientific, or ‘metaphysical’ (as we might say), it is not thereby found to be unimportant, or insignificant, or ‘meaningless,’ or ‘nonsensical.’ But it cannot claim to be backed by empirical evidence in the scientific sense – although it may easily be, in some genetic sense, the ‘result of observation.’
> …
> Thus the problem which I tried to solve by proposing the criterion of falsifiability was neither a problem of meaningfulness or significance, nor a problem of truth or acceptability. It was the problem of drawing a line (as well as this can be done) between the statements, or systems of statements, of the empirical sciences, and all other statements – whether they are of a religious or of a metaphysical character, or simply pseudo-scientific.” (50-51)
>
> Popper discusses his problem of demarcation in contrast to Wittgenstein's problem of meaning (52-54), where Wittgenstein tried to show that what he thought of as philosophical or metaphysical statements were meaningless. Wittgenstein thought that the only meaningful statements were those that were built from atomic propositions -- and these atomic propositions were at least in-principle ascertainable by observation.
> As a solution to the problem of demarcation, Popper criticizes Wittgenstein's approach as being deductionist. As to the problem of "meaningful statements," Popper regarded this as a pseudo-problem.
>
> But Popper’s argument was misinterpreted in an interesting way -- while he was arguing for testability and falsifiability as a means for demarcating between science and pseudo-science, his view was interpreted to be arguing for a falsificationist test of meaning -- in other words, that the meaningfulness or nonsensicalness of a statement could be ascertained by whether it was falsifiable or not.
> For example, Wittgenstein used the example of a non-sensical pseudo-proposition such as "Socrates is identical." This is nonsense, and its negation, "Socrates is not identical," is also nonsense, and, so, the logic went, the negation of a meaningful statement would also be meaningful. But as Popper notes in n.6 on page 54, the negation of a testable (falsifiable) statement need not be testable (discussed more in The Logic of Scientific Discovery) For example, consider the statement "at least one barn is red," and its negation, "no barns are red." One requires us only to find one barn -- the other requires us to search the universe. This asymmetry in the falsifiability of negated statements makes nonsense of the attempt to make Popper's criteria for demarcating science from pseudo-science into a criteria for demarcating meaningfulness, since such an attempt relies on the symmetry of falsifiability in negated statements when none exists.
Note also that the negation of a good explanation is a really bad explanation. e.g. compare
"The Earth *has* seasons because of the tilt of its axis and its orbit around the sun."
with
"The Earth *does not have* seasons because of the tilt of its axis and its orbit around the sun."
or with
"The following is false: The Earth *has* seasons because of the tilt of its axis and its orbit around the sun."
An explanation of X does not serve as an explanation of not-X. And a flat denial of an explanation is not itself an explanation (it doesn't say *why* the explanation is false, it doesn't explain why it's denying it).
> Popper notes Hume's solid philosophical criticism of induction and how it leads to an infinite regress, but then proceeds to criticize Hume's psychological account of induction on the same grounds (55-59). Hume tries to explain the belief in laws as a product of frequent repetition -- Popper notes that there can only be "repetition" in the sense of two situations being seen as similar from a certain interpretative point of view, since no two situations are precisely the same -- therefore, there must be a PoV BEFORE there can be any repetition -- thus, the PoV can't be the product of repetition (59)
yeah good argument about PoV (point of view) has to come first.
> Popper introduces his theory -- which is that, instead of having patterns impressed upon us through induction, we come up with and try and impose them on the world, and then have them refuted -- a process of conjectures and refutations. (60)
>
> All observation is selective (61-62). there is no pure observation. Observation needs some task, interest, PoV, problem.
yeah. good stuff.
> While inborn ideas silly, some inborn reactions or responses may exist. (62) One can refer to this as knowledge.
>
> This knowledge is not a priori valid. Popper thinks there may be an inborn tendency to look for regularities.
>
> Expectation of finding regularities is both psychologically and logically a priori, since all observation involves the recognition of similarities and dissimilarities.
> Kant right that our intellect imposes laws upon nature, wrong that those laws are necessarily true (63).
>
> Example of induction machine: could we build induction machine which could perform “inductions” and solve problems but it wouldn’t show that induction worked – cuz deciding what its “world” is, what laws it will discover, etc., involves embedding knowledge created by humans into the machine – problems of similarity / etc will have. (64)
>
>
> Discusses dogmatic attitude vs critical attitude (64-66). Dogmatic attitude inevitable – critical attitude does not replace dogmatic attitude but is overlaid on top of it – needs dogmatic beliefs for “raw material” – like how science started with myths (66).
that's really misleading out of context. one has to be careful saying anything good about dogmatism.
What Popper was getting at, in part, is that we can't question all our ideas at the same time. So at any time we have to use a bunch without questioning them, which could be called dogmatic (though I wouldn't prefer to call it that).
>
> Popper says that deductive reasoning important NOT cuz it lets us prove theories but because it lets us tease out the implications of our theories, and crit them better. (67)
actually we learn about the implications of theories the same way we learn everything else: by guessing them, and exposing our guesses to criticism. deduction can optionally play some partial role in this.
> Popper connects demarcation and induction (69-72).. The belief in induction is fortified by the need for a criterion of demarcation which people think needs to be based on induction. (71).
>
> Popper resolves what seem to be three conflicting principles:
> 1. (the argument from Hume and Born) the impossibility of justifying a law by observation or experiment
> 2. the fact that science proposes and applies universal laws
> 3. the principle of empiricism, which says that in science, observation and experiment must decide upon the acceptance or rejection of scientific statements
> 1 and 3 seem to clash, but don’t if we accept that the acceptance of theories in science is tentative only (71-72)
> Only the falsity of a theory can be inferred from empirical evidence, and this inference is purely deductive
>
>
> Popper argues that we jump from an observation to a good theory by conjecture and refutation (74).
>
> Why prefer non-falsified statements? We’re truth-seeking! (74)
>
> Popper rejects a probability calculus as solving the problem of induction. Science doesn’t seek highly probable theories – it seeks explanations, which are powerful and improbable theories.
> He notes that every interesting and powerful statement must have a low probability, and vice versa – a statement with a high probability will be scientifically uninteresting. (77)
it's more than that. there's no actual proposal of *specifically* how to use probability calculus to solve the problem that actually works.
-- Elliot Temple
http://elliottemple.com/
Justin Mallone
(I think) an "inborn idea" would be some big complex thing that people think is inborn, like "romance is good."
A reaction/response would be like suckling instinct, or sugariness producing pleasurable sensation, or something. So simple, low-level, not very complex stuff.
Hmm, re-reading Popper, is a bit more ambiguous. Page 62 of C&R third para:
> The theory of inborn ideas is absurd, I think; but every organism has inborn reactions or responses; and among them, responses adapted to impending events. These responses we may describe as 'expectations' without implying that these 'expectations' are conscious. The new-born baby 'expects', in this sense, to be fed (and, one could even argue, to be protected and loved). In view of the close relation between expectation and knowledge we may even speak in quite a reasonable sense of 'inborn knowledge'. This knowledge, however, is not valid a priori; an inborn expectation, no matter how strong and specific, may be mistake. (the newborn child may be abandoned, and starve)
I wonder what he means by "one could even argue" that baby could expected to be "protected and loved." Expecting to be loved seems kinda high level to be inborn. Maybe he meant something by it I'm not understanding.
-J