Facebook

Tuesday, January 24, 2012

Thaaat's Not Science!

The Demarkation Problem:  Notes and Discussion of Karl Popper's Falsificationism 


Overview


One issue in the philosophy of science is known as the demarkation problem.  That is, where do we draw the line between genuine science and pseudo-science?   What's the difference between astronomy and astrology?  Chemistry and alchemy?  Medicine and homeopathy?  Psychology and Freudian psycho-analysis?  Evolutionary biology and creation "science"?  Why is the study of literature, history, or philosophy not considered a science?  Are the social sciences, science?  All these questions are relevant to the demarkation problem.  


Possible Answers


The obvious answer is that something is science if the person doing/discovering it is wearing a lab coat, works with test tubes, beakers, and Bunsen burners.  But as appealing as this solution is to our intuitions we cannot ignore the fact that alchemists also wear lab coats, use many beakers and test tubes (filled with many colourful liquids, no less!) AND use several Bunsen burners all at the same time--and lets not forget that they measure their contents at the meniscus!  They even use cool spiral shaped glass tubes that connect to other beakers.  Alas, we might need to look elsewhere to solve our demarkation problem.  


One other possible answer is to say that something is science if and only if it true.  This won't work for several reasons: (1)  Every true state statement will be considered scientific.  Consider the statement "All unmarried men are bachelors".  Under our definition of science, by making this statement we are "doing science" and making a scientific statement.  (2) This would make things like Newtonian physics and Darwin's account of evolution not science.  Even though Newtonian physics isn't perfect and was superseded by Einsteinian physics, it's still used in science and when scientists use it, they are still doing science.  Darwin theory of evolution pre-dated the discovery of DNA so while his specific account of evolution was incorrect, it doesn't seem correct to say he wasn't doing science.  Lets try a different criterion.  


How about "science uses the empirical method"?  But so do many pseudo-sciences like astrology.  Also, if we restrict science to areas that use the empirical method we must exclude things like theoretical physics.  Hmmm...


Lets try this: if there is a "preponderance of evidence" for a theory then it's science.  The immediate problem is that early scientific inquiry into a problem won't be considered science until a lot of evidence has accumulated.  Do we want to say that when a scientist (in a lab coat, using beakers and Bunsen burners)  makes a new discovery, they aren't doing science no matter how sound their method?  This doesn't seem right.  The other problem, which occurs very often in pseudo-science, is that theories can be interpreted to accept any data set.  


Here are a few quick examples:  (1) In para-psychology anytime a skeptic tries to observe or replicate the results, the subsequent absence of the effect is explained away by saying the presence of the non-believer annihilates the para-psychological effects.   This only further confirms their (now modified) theory, that para-psychological effects don't work around skeptics.  (2) Consider astrology: predictions are so general that they can be made to retroactively apply to at least one event that happens throughout the day, further confirming the theory; i.e., adding evidence.  (3)  Consider something like the pop-psychological belief that "the universe always conspires to help you".  It is easy to see how proponents can always retrofit events and interpretations to make it true.  Take two opposite situations:  you lose your key and you don't lose your keys.  Under the theory, proponents can always apply retrospective reasons for why one or the other happened (use your imagination).  That is, both A and ~A make the theory true.  But any theory that contains a statement and it's contradiction is inconsistent.  In short--it's nonsense.  But either way, the theory is confirmed. 


To summarize, just because we can find verifications for a theory does not allow us to say we are doing science or that the theory is scientific.  Here's a good illustration of the problem with using verification (or "preponderance of evidence") to gauge whether something is science.


http://www.youtube.com/watch?v=WB-qSXSsoUI


Popper's Falsificationism 


So, what do?  How ought we to distinguish science from pseudo-science?   Listen up peoples cuz my main man Karl Popper's about to drop some knowledge on your ass, then we is going to discuss it.  Karl Popper was one the first contemporary philosophers to rassle with this problem and he came up with a 7 point list to distinguish the science from the pseudo-science.  They can be summed up as follows:  "the criterion of the scientific status of a theory is its falsifiability, or refutability, or testability".    


Vas does this mean?  Lets break this shit down in to terms I understand.  I know that testability prolly has something to do with test tubes, but what about those other big words?  Popper is saying that to test whether a theory is scientific you have to test it in such a way that failure is possible.  In a more concrete form you will do something like this:  make a specific "risky" prediction about something your theory entails.  For example, if we want to test a theory of gravity we can say "this theory predicts that object z will fall distance y, in time x".  With this sort of specific testable statement there is a real chance that the theory could fail.   Or is there? Ah! ha!


It seems we have prollem.  Popper's falsificationism is certainly a good first step but there are some difficulties.  How many specific "risky" tests do we have to make before we can accept a theory as scientific?  Consider Newton's law of gravity which has been around since the 17th Century.  Presumably it had been rigorously tested many times with specific testable predictions since it's introduction to the scientific cannon.  When Einsteinian gravity was shown to be a better theory did we say that all pre-Einsteinian scientists had been doing pseudo-science?  Nope.  Here's the prollem.  There are infinitely many tests we can do on a theory, only our imagination is the limit.  Einstein had a greater imagination than all physicists prior to him, so he was able to falsify the Newtonian theory.  Who's to say that one day some hyper-imaginative person comes up with a test that falsifies Einstein's theory?  Or any of our other scientific theories, for that matter?  How many tests make a theory scientific?


Another question is why do we consider Newtonian physics to be science, even now that it has been falsified?  If theories can be falsified, yet still be considered science, on what grounds can we exclude known pseudo-science?  And related to this is why do we say that past scientists who worked off of later disproven models were doing science?  


It seems that the ability for a theory to be falsified is a start to uncovering the demarkation line, but is not nearly enough.



No comments:

Post a Comment