Can somebody explain what Kuhn is saying here?

[nonequilibrium]

From the wikipedia page on his "The Structure of Scientific Revolutions"

Kuhn (SSR, section XII) states that the probabilistic tools used by verificationists are inherently inadequate for the task of deciding between conflicting theories, since they belong to the very paradigms they seek to compare. Similarly, observations that are intended to falsify a statement will fall under one of the paradigms they are supposed to help compare, and will therefore also be inadequate for the task. According to Kuhn, the concept of falsifiability is unhelpful for understanding why and how science has developed as it has. In the practice of science, scientists will only consider the possibility that a theory has been falsified if an alternative theory is available that they judge credible. If there isn't, scientists will continue to adhere to the established conceptual framework. If a paradigm shift has occurred, the textbooks will be rewritten to state that the previous theory has been falsified.

Can somebody explain what is meant? The way I see it, it seems so obviously wrong: I don't see any reason why two opposing paradigms can't be tested by the same experiment... For example you had the Phlogiston theory, and the atomic theory; the earlier was proven wrong with a certain experiment, an experiment which did give a consistent result in the latter theory. Or Einstein and Newton: although they work with quite different concepts, you can figure out the result of a certain experiment in both theories and then find out how reality actually behaves, so the concept of falsifiability is certainly helpful...

I must be totally missing the point here.

 

[apeiron]

He's making the social point that given an established view of the world and a bit of evidence that seems to contradict it, we will put a lot of effort into ignoring the conflict. We will doubt the data rather than doubt the theory.

This seems bad science, but is in fact good psychology. Views of the world are the result of long experience and have proved their worth. So you don't just junk that because of some possibly mistaken passing event.

But science of course is a method for correcting for psychology. It says no, you really need to seek to disconfirm your prejudices/theories. If you have data that proves your existing view inadequate as a model, you are forced to invent some better model of the world (which you then seek to disprove).

The reality is that science moves forward with a mix of both. Mostly people cling on to their cherished beliefs (because there are no really bad consequences - their existing theories work pretty well in the areas they were developed for). And occasionally there is the kind of significant shift in paradigm where people can point to an obstinate bit of data and say that is where a scientist followed the proper method. The event becomes celebrated because it reinforces the view of how science should work.

There is a deeper point that Kuhn makes in that models are self-reinforcing because they spell out the kind of measurements that should be made in the first place. So there is a problem that models themselves don't tend to throw up the kinds of observations that would disconfirm them in a big way. They don't even generate the obstinate data because they can't even imagine the necessary measurement.

This is a methodological issue rather than a sociological one. But it is dealt with - at least at a certain level - by random variability. Give scientists enough freedom to dream up any theory they like. Then let the winners emerge.

The problem is that you then have many scientists with pet ideas for which they seek only confirmatory observations. So back to the sociology I guess.

 

[Willowz]

I don't see how psychology fits in what Kuhn is saying.

And it doesn't make much sense. I mean look at one of the reasons why the LHC was built. To verify/falsify some of the theories that are out there.

 

[disregardthat]

I'd say that while it is general consensus that general relativity falsified Newtonian mechanics, this is such an example where a paradigm shift is the actual reason for that it is claimed that it falsified a previous theory. For Newtons laws didn't necessarily have to be used while assuming Euclidean geometry, this was merely a choice in where to apply them (which worked at the time). In a sense the transition into space-time was a transition into a different context of application, not a falsification per se.

Newtons laws were generalized, not because they didn't work, but because of the transition into a different geometry required different calculations. The very notion of that Newton's laws were wrong and corrected is faulty, in my opinion.

The falsification consisted of that euclidean space was not a satisfactory choice of geometry, but the core mechanics of Newtons theory remained.

 

[apeiron]

I don't see how psychology fits in what Kuhn is saying.

And it doesn't make much sense. I mean look at one of the reasons why the LHC was built. To verify/falsify some of the theories that are out there.

Precisely. Some of the theories. Ie: the ones fitting some prevailing paradigm.

Of course, the LHC says something more about the social practice of science. When money and prestige are tied up in a big facility, it is suddenly in a lot of people's interests to create precisely the kind of theories that the facility might test.

That is another kind of psycho-sociological departure from a pure, dispassionate methodology.

 

[disregardthat]

I don't think this is a sociological or psychological issue at all. It's more a statement of how the word "falsified" is actually used, and Kuhn is stating that it makes sense to apply to a theory only when in conjunction with a rival theory.

 

[apeiron]

I don't think this is a sociological or psychological issue at all. It's more a statement of how the word "falsified" is actually used, and Kuhn is stating that it makes sense to apply to a theory only when in conjunction with a rival theory.

Well what about this bit where it is being said that people don't jump ship until there is another ship to clamber aboard (whereas they will cling to the sinking ship already holed by a disproof)...

In the practice of science, scientists will only consider the possibility that a theory has been falsified if an alternative theory is available that they judge credible.

Phlogiston appears to support the idea of the natural psychological resistance that can be shown even in the face of the "decisive experiment".

When it was shown matter gained weight on burning (instead of losing weight because of lost phlogiston), the reply was that phlogiston had a negative weight, or it was lighter than the air that took its place.

http://en.wikipedia.org/wiki/Phlogiston_theory

So the same experimental result could not immediately be used to separate the two theories. You say combustion creates a gain in weight as atoms are added, I say combustion creates a gain in weight as substance with a negative weight has been shed. Nothing is as yet falsified as two theories now predict the same observation.

Of course, this creates further ideas to be tested and the evidence accumulates. And also an alternative theory is developed that looks a better choice. The psychological/social jump may be swift when it comes. The mob reacts. And points back at the evidence that seems decisive rather than the change in interpretive viewpoint which allowed it to seem decisive.

Not that I want to make a strong defense of Kuhn here. Like the old punctate vs graduationalist debate in evolutionary theory, it seems a false dilemma. Sometimes paradigms change more steadily, other times more abruptly.

But science is about belief as much as evidence. A mindset has a way of making its observations fit. Or else ignoring them. So the sociology is important.

I see a rabbit, you see a duck. The actual data falsifies neither view.

 

[physosophy]

His point here is that if there is an honest to goodness paradigmatic difference, then the very measurement tools themselves are part of the paradigm and therefore can't be used to test each other (otherwise it is http://en.wikipedia.org/wiki/Begging_the_question" ).

For example, you can't exactly use Newtonian physics to test Einsteinian physics, since the latter is an entirely different way of thinking. Now, they both fall under the general paradigms of math and empiricism, so you can test them with those, which they have been.

 

[MarcoD]

I don't see how psychology fits in what Kuhn is saying.

And it doesn't make much sense. I mean look at one of the reasons why the LHC was built. To verify/falsify some of the theories that are out there.

The problem is that he made two statements (well, at least two.) The first part is that tools are part of the paradigm:

Kuhn (SSR, section XII) states that the probabilistic tools used by verificationists are inherently inadequate for the task of deciding between conflicting theories, since they belong to the very paradigms they seek to compare. Similarly, observations that are intended to falsify a statement will fall under one of the paradigms they are supposed to help compare, and will therefore also be inadequate for the task. According to Kuhn, the concept of falsifiability is unhelpful for understanding why and how science has developed as it has.

The second part is on how science sociologically works.

In the practice of science, scientists will only consider the possibility that a theory has been falsified if an alternative theory is available that they judge credible. If there isn't, scientists will continue to adhere to the established conceptual framework. If a paradigm shift has occurred, the textbooks will be rewritten to state that the previous theory has been falsified.

I agree with aperion's analysis.

 

[mathal]

The break in the original quote that MarcoM pointed out is where the paradigm shift is described. Until there is a critical mass of consensus there is no real paradigm shift. Understand this, there is no direct reference to measurements in these comments of a paradigm shift; we are dealing with generalities here, and sometimes, a new perspective that is equally 'valid' i.e. not unconfirmable just 'appeals' to scientists more than the previous one.
To Willowz, your comment about LHC implies that you think new colliders are built at least to some degree to confirm or deny different prevailing theories. The only connection between theoretical physics and experimental physics is that theoretical physics points out what hasn't been absolutely discredited by present data. In that respect there is only one direction that experimental physicist can go in (with accelerators)-build more powerful accelerators to determine 'what's out there'. What I mean is, there is no direct correlation between theoreticians and experimentalists.
The original post referred to two differing paradigms being tested by the same experiment. The problem with resolving this 'problem' is that most present theories 1. can't be tested and 2. if one, by a particular test is 'true', that doesn't preclude that other theories, that don't specifically apply to this particular test, aren't necessarily still 'true'(i.e. unfalsifiable).
What I'm saying in essence is that the old model of "it's true or it's false" may not be achievable anymore.(if anyone wants to start a thread on this idea -go for it. It is debatable which thread it would stick on)
(EAAD my Acronym-for future reference only, if any, ignore it (just for my reference))
mathal

 

[LaurieAG]

Gallileo created a new paradigm with regards to our perceived priveleged location at the centre of the universe in absolute terms. Unfortunately our conceptual framework for experimental observations (made on galactic scales) necessitates observations that are made by a relatively stationary observer with respect to the objects being observed i.e. under the old paradigm.

Observations made from a point relatively stationary to the movement of mass in the universe i.e. on the Hubble flow, would be expected to return the same results for the De Sitter star pair experiment but would the results of the Twins experiment be expected to be the same?

http://en.wikipedia.org/wiki/Twin_paradox

The time differential explanation in absolute terms
...
For example, A.P. French writes, regarding the twin paradox: "Note, though, that we are appealing to the reality of A's acceleration, and to the observability of the inertial forces associated with it. Would such effects as the twin paradox exist if the framework of fixed stars and distant galaxies were not there? Most physicists would say no. Our ultimate definition of an inertial frame may indeed be that it is a frame having zero acceleration with respect to the matter of the universe at large."

This type of perceived discrepancy is what Kuhn is referring to.

 

[epenguin]

I think Kuhn essentially says that although the paradigm-shifters are the heroes celebrated by him also, they and their times and opportunities are exceptional - the ballast of conservatism is very necessary to Science. Most of the time in fact - he calls what most scientists do 'normal science'. For a few months and for a few people neutrinos went faster than light, and it would not be good if everyone rushed to judgement just because there was unanswerable experimental evidence.

Also that the model of The Decisive Experiment which textbooks tend to sanction does not describe the historical reality of how things do develop nor and understanding of how they could really.