[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index

Re: Karl Popper/False positives...



Judy and Jack,

I just want to reinforce Jack's interpretation of epistemology here. What he writes is very recognizable from all my training in the subject. The other question is whether it is "right" or not, i.e., given that the statements do represent Popper fairly well, are they yet correct?

I'd say it is best to just accept the logic of Popper in its own right. It has been fairly well scrutinized. However, it is quite limited in many regards. Most modern scientists (who am I to know "most" - but this is my impression from reading) are no longer positivists in the HD-methodology sense, which has been attributed to Bacon and Popper among others. The idea of strict falsification of hypotheses is accepted as logically valid, but not general enough to cover all paths to truth. Even Popper recognized this later in life and was forced to add his Metaphysical Research Program (MRP) as part of the scientific enterprise. Thomas Kuhn also did this in his theory of scientific revolutions. So I don't think we need to reinvent that discourse - there is much more to science than positivism. These specific ideas about falsifiability, etc. are part of the positivist program, and picking at the logic of that is probably a fruitless endeavor. The logic will hold up, but its applicability is what is in question.

John K.

Judith Rosen wrote:
Thanks, Jack. Interesting to read. "False negatives": theories which seemed to be proven false, but were inappropriately applied, misunderstood, or otherwise bungled by how the experiments were conducted. False positives are theories which seem to have been proven correct, but are in fact not correct (at least not as general theories). The only reason they seem to have been proven correct was due to aspects unrelated to the system or the theory, such as limited sampling (a few types of only a couple classes of systems) and/or limited analysis of results, etc.
 
I tend to agree with a lot of what is ascribed to Popper as his way of looking at science, however I see some troubling inconsistencies in the reasoning used. My analysis below:

>  From:
http://www.users.totalise.co.uk/~idmon/zphil1.htm:
So, because of their generality we cannot verify
> scientific theories or even be sure they are probable with anything
> like mathematical certainty. Therefore how can we trust them?
 
Mathematical certainty is a myth, a fictional creation requiring "suspension of disbelief" just like the Harry Potter stories are/do. It's ironic that the argument used to point out the flaws in science has actually bought into the mindset of that aspect in science that causes the flaws. Anyone who trusts mathematical certainty in a complex world is in for a whole lot of side effects.
 
It is this search for generality of scientific
> theories that precludes our establishing their definite proof.
 
I see this as being backwards. The more general the applicability, the more definite the proof. It is the very specific theories which I see as being fundamentally flawed because they only apply to a narrow band of contextual conditions, which is not the way the universe is.
 
> It is difficult to trust what cannot be proved and here Karl Popper
> overturns the argument by stating that proof is not necessary but what
> is important is the concept of falsifiability
Was Karl Popper religious? Just wondering... because lately it seems to me that humanity, as a species, is engaged in a massive global effort to kill one another/itself over and out of a deep, unwavering trust in (and love for) "that which cannot be proved". Some religious description of God has been the motivation for an awful lot of intractable hatred among humans throughout our history! The above phrase really begs the question: What is "proof"? Proof, in my view, is one of those "in-the-eye-of-the-beholder" concepts, like "beauty".
 
"Falsifiability". Now, there's a word! I tend to think that's not the word they meant to use. I think they wanted something like "refutation". Falsifiability means "the ability to be falsified". That reminds me of the recent discussion here about simulation and mimicry... (which tend, in experiments, to generate "false positives" as far as objective quantifiable proof of some theory is concerned...)
 
Having said all that, though, I do agree that the more often some theory seems to be proven false, the more likely it becomes that the theory is either false or else it's not a general theory.
 
Judith
 
 
From: Jack Park
To: ***
Sent: Tuesday, January 04, 2005 12:32 PM
Subject: Re: [ROSEN] Karl Popper/False positives...

From:
http://www.users.totalise.co.uk/~idmon/zphil1.htm

> In epistemology we want to know why we think our beleifs about the
> world are true. In the work of Popper we are looking specifically
> about the theories upon which science is based. The case of scientific
> theories presents us with particular problems. These are all to do
> with the lack of any direct or definite proof or disproof of the truth
> of scientific theories, thus how do we choose between conflicting
> theories? With the absence of a decision process is science a rational
> process as we have been led to beleive, or are we faced with a kind of
> science fetishism which cloaks itself in reason but has no real validity.
>
> Science aims at theories which explain natural phenomena. More
> precisely, science is looking for theories which tell us about the
> causation of phenomena and they aim at generality, explaining a
> particular effect by showing it to be caused by some general
> regularity of nature. It is this search for generality of scientific
> theories that precludes our establishing their definite proof. We can
> examine only a very small proportion of all the cases covered by a
> general theory which is intended to apply over all space and time as
> part of its claim. So, because of their generality we cannot verify
> scientific theories or even be sure they are probable with anything
> like mathematical certainty. Therefore how can we trust them?
>
> It is difficult to trust what cannot be proved and here Karl Popper
> overturns the argument by stating that proof is not necesarry but what
> is important is the concept of falsifiability
> <http://www.users.totalise.co.uk/%7Eidmon/zlogcona.htm>. Precisely
> because of the universality of a general scientific theory which
> states that all phenomena of a certain class have such and such
> characteristics, only one counter-example is needed to show that the
> theory is false. Science should attempt to submit theories to the most
> stringent testing possible and make predictions which have not yet
> been tested. Poppers favourite example is Einsteins special theory of
> relativity which in 1905 predicted that light was bent by heavenly
> bodies, this was not tested until the eclipse of the sun in 1919. Thus
> theories which can easily be tested will be weeded out quickly if
> false and those that survive can be retained, but even these can only
> be kept on a provisional basis. The problem of induction
> <http://www.users.totalise.co.uk/%7Eidmon/zlogcona.htm> prevents the
> passing of the tests proving the correctness of the theory.
>
> Therfore science is rational in Poppers view because though complete
> proof is impossible it follows a critical approach testing theories
> against the natural world and making predictions for as yet
> undiscoverd effects.
> This thesis was first elucidated in The Logic of Scientific Discovery
> <http://www.users.totalise.co.uk/%7Eidmon/zlogcona.htm>
>
That says it better than I did. False negatives?

Jack

Judith Rosen wrote:

> *****Jack, *
> **
> *So, what Karl Popper was saying was that there's no such thing as a
> "false negative", right? There are false positives, but not false
> negatives? Hmmm..... I'm not sure I agree with that. I'll have to give
> it some thought and figure out why my alarms go off at that idea.*
> **
> *Judith*
> **
>
> Jack Park wrote: *Karl Popper also said that.
> I don't recall him using "sometimes". My recollection is that "negate
> and prove that" is the only way to "prove" anything, according to Popper.
>
> Jack
>