Up@dawn 2.0

Wednesday, April 3, 2013

14 - 3 Karl Popper on Falsification



Empirical verification comes from induction. We seem to be pattern-matching creatures - when we have a theory, we hold on tightly to it. We go to considerable lengths to seek confirmatory instances of it. Unless we are scientists, and properly meticulous, we typically discount, explain away, and reject falsifying examples. Yet, following Hume's much earlier critique of this methodology, Popper maintains that inductive logic is inherently flawed. Just because the sun has risen every day until now does not necessarily entail that it will rise again tomorrow. One instance of the sun's failure to rise will falsify the theory that "the sun always rises".

The scientist should reject theories when they are falsified. Psychological theories, however, in their attempt to explain all forms of human behaviour, can continually be shored up by subsidiary hypotheses. Exceptions can always be found. On a Popperian model, psychology resembles magical thinking: if an expected result does not manifest, explanations can be found which explain that failure away, and thus the core theory remains intact. This, Popper considered, is a weak point – the theory cannot be properly tested if it is inherently unfalsifiable.


Therefore, to Popper, exceptions immediately falsify a theory, whereas confirmatory instances are just references to experience and thus possess no logical merit. Falsification should replace induction as a core focus of the scientific method.


All that Popper is claiming is that the underlying theories are not sciences, although they may one day play a role within a scientific paradigm, a view which seems to have enraged those who want to claim scientific status for sociology, for example.

We should be looking at theories that are confirmed by risky predictions: those predictions which should have led to an instance that disconfirms the theory. Destructive testing is the only really viable scientific test: we should always aim at refutation, not confirmation, for confirmation is too easy and too weak.

7 comments:

  1. im still trying to grasp what Popper is saying. Is he saying that its better to falsify than it is to verify or confirm? Why wouldn't you want to confirm that is theory is right?

    ReplyDelete
  2. I agree with Jessica. One of Popper's example is that Premise one: "all fish have gills" and Premise two says " John is a fish." and the conclusion says "therefore John is a fish." in reality this is true but for Popper it isn't. He said "it would be absurd to say that premise one and two are true, but that the conclusion is false. Im kind of confused because if all fish have gills and John is a fish, therefore shouldn't John have gills???

    ReplyDelete
  3. He's basically saying that trying to falsify a theory is more accurate than trying to verify it. It's easier to verify a theory when you created it and understand it's limitations and whatnot, because you know what is required for it to be "confirmed". This is unlike falsification, because if you can't disprove a theory, then that is the most accurate way to tell whether or not it is relevant.

    ReplyDelete
  4. I like how Popper claims underlying theories are not sciences.but I still agree with the others in what he's saying isn't that clear.

    ReplyDelete
  5. Popper's message isn't completely clear to me but seans blogpost gave me a better insight on it.

    ReplyDelete
  6. Consider the following psychological theory. Look at this picture:
    http://highlycontagiousideas.files.wordpress.com/2012/07/tophat.jpg

    For most individuals A to B looks longer than C to D. This is a common visual "illusion". That's my psychological theory. This theory is *testable.* Popper did not rule out all psychological theories.

    Also, it's simply a matter of logic to note that *any* theory that has a potential falsifier can be saved by creating a subsidiary theory.

    For example, "all swans are white" is my theory. What if I say, "this swan is black."

    Note you can save the theory, by arguing the swan I refer to is not a swan *because* it is black. Popper argues that if we do this, we're not doing science, but merely arguing over definitions, which is not helpful.

    But, notice there is another form of escape. I could say, that swan is not really black it's just dirty. This also is problematic, but at least here, we might be able to *test* my new claim. Popper point here is that we shouldn't endlessly try to formulate new theories with the intent of *saving* our pet theories.

    What's most notable here though is that Popper claim is not really logically, but in a sense ethical. A scientist must specify a means whereby his theory can be *test* that we basically understand, otherwise the theory is suspect. We have no way to independently judge the theory for ourselves.

    Some times a theory's truth is all in the eye of the beholder. If that is the case, the theory is not scientific.

    ReplyDelete
  7. Anonymous7:40 AM CDT

    When is a theory acceptable as a true statement? Some theory are more confrontational than others and recieve more popularity than others. A theory that is not refuttable is non sciencefic, every test is an attempt to falsify a theory , confirming evidence is not a trustworthy way to determine a theory is absolute truth. The criteria of the science status f a theory is its testibility.

    ReplyDelete

Note: Only a member of this blog may post a comment.