Mavericks- and why science needs them.

It seems the evolution of science progress today is more dependent on expensive technology and expensive big team collaborations.

This can surely provide great focus and "reach";

Planetary space probes

Human genome project.

Yesterdays news about collaboration on the gravity detection results is in this vein.

But I guess this must constrain the areas of enquiry towards mainstream goals and inevitably collaboration means that more "maverick" scientists get sidelined.

Also of concern is the "reputation trap" and the politics of science.

Science needs creativity otherwise it risks becoming ossified.

It seems the evolution of science progress today is more dependent on expensive technology and expensive big team collaborations.

This approach can surely provide great focus and "reach";

Planetary space probes.

CERN.

Human genome project.

Yesterdays news about collaboration on the gravity detection results is in this vein.

But I guess this must constrain the areas of enquiry towards mainstream goals and inevitably collaboration means that more "maverick" scientists get sidelined.

Also of concern is the "reputation trap" and the politics of science.

Science needs creativity otherwise it risks becoming ossified.

There is still much creativity in certain theoretical areas, just look at sites such as Quanta Magazine.

However we need some mechanism to encourage mavericks in research, I guess this comes down to funding and the battle for money.

Even worse than 'Corporate Science' is when the government decides it is best placed to decide what specific research projects should be funded.

There needs to be a balance between oversight of tax payers money and stifling scientific creativity. So who is best placed to decide?

While I agree the current system sidelines maverick scientists, if I wanted to make a contrary wise argument;

In the UK I know that the Royal Society do bang the drum for "blue sky" research and leaving open the possibility for serendipity in science.

Also in the case of "mavericks" such as Rossi and BLP, although supporters can complain about lack of support by mainstream science and various establishment obstacles that they face, but both have managed to find parties willing to provide financial backing, so financing is not a show stopper.

Quote from Alan Smith

This is what I call 'Corporate Science'. Science by committee consensus seldom produces anything remarkably novel.

See the recent book by Bauer, "Science is not What You Think"

https://www.amazon.com/Science…hink-ebook/dp/B071JZQ9H1/

I was a little disappointed by this book. I wrote a brief review at CMNS, as follows:

. . . I think Bauer overstates his case in several instances. I think he is muddled. I will describe only two problems, to keep this message short.

Chapter 2 is titled Science Is Not Methodical. It begins:

"The scientific method is taught to school children and featured in social science textbooks as the way science works: Set up an hypothesis; test it; then accept or reject it, depending on whether the test supported it. Hypotheses that have passed such a test become the basis for established scientific theories. Several things are wrong with this. Conceiving and testing hypotheses is not what most scientists do most of the time. More likely they are doing what others have asked them to do: develop a better food coloring, or a paint, or a drug, or a pesticide; or analyze a competitor’s product . . . But even when research scientists are trying to expand the scope of scientific understanding, they don’t do as the scientific method would have it, they are more likely to follow a hunch: 'If I do this, something interesting will turn up' . . ."

I would describe the latter method, "if I do this . . ." as intuitive. Or as art, rather than science. I think it is true that many experiments are a combination of science and art, but I think Bauer overstates this. Even when scientists are "doing what others have asked them to do," and even when they themselves do not conceive of a hypothesis, they make use of hypotheses conceived by other people. They make use of general theory.

Bauer says that the scientific method is not taught. That is true to some extent, and it is shame. I think it should be taught. But most researchers use it even when they are not taught it. Otherwise they would flounder around trying one thing after another with no direction. Edison supposedly did that in what is now called the Edisonian method. Reading Edison's biography and his notebooks I get a sense that he knew more theory than he let on. He was exaggerating his aw-shucks down-home ignorance. Although Tesla and others said he wasted a great deal of time because he did not understand theory well enough.

As an example of a muddled discussion, Bauer rejects Popper's rule that a scientific theory must be falsifiable. Bauer says that Popper himself soon rejected that idea. I do not think he did.

I will grant that a researcher might make progress in a theory or an experimental method even when that researcher cannot think of a way to falsify it. However, the researcher would be skating on thin ice. It may not be essential that you can readily think of a way to falsify your claim, but you probably do not understand the issues if you cannot. Bauer muddles this discussion when he says that scientists sometimes believe claims that have been falsified. I take it he thinks this to demonstrate that falsification is not the be-all end-all test of validity. Then he says that these scientists do not agree the claims have been falsified. Yes, and that defeats his argument. If those hold-out scientists agreed the claims were falsified, they would cease to believe them.

I also got a sense that he thinks falsifying means you actually show there is something wrong, rather than showing you know what factor would invalidate your claim even if it is extremely unlikely that factor will arise. A famous example is what J.B.S. Haldane said would disprove Darwinian evolution: "rabbits in the Precambrian." Haldane was not suggesting that fossilized rabbits are likely to be found in Precambrian layers; he meant that if they were found this would disprove the theory.

Along the same lines, because cold fusion is an experimental finding rather than a theory, a person could disprove it by showing experimental errors in all of the major studies, such as the ones conducted by McKubre, Miles, Fleischmann and Storms. Since these groups used different calorimeter types, you would have to show a wide range of experimental errors. I do not think anyone could do this, but if they could they would disprove cold fusion. I cannot think of any other way to falsify cold fusion. The claims are predicated on the Laws of Thermodynamics, and it is not likely these will be disproved.

Quote from maryyugo

Most of the "unwashed public" out there doesn't even know that correlation doesn't equal causation . . .

Actually I agree with David Hume and with my late mother that correlation does equal causation. Usually, anyway, unless there is a common cause. (My mother was a statistician who knew from correlation.) In the book, Bauer makes a worse mistake. He fails to note that when there is both correlation and a theoretical basis for causation, causation is likely. For example, he cites global warming as merely correlated with the increase of CO2. He ignores the fact that climatology includes theories that predict CO2 will cause warming. The theories might be wrong, but if they are right, they bolster the likelihood of causation.

Quote from maryyugo

And that for most things, extraordinary claims require at least sufficient evidence.

No, they do not. I dislike this formulation. All claims, ordinary and extraordinary, call for the same level of rigorous evidence. Extraordinary claims are best supported by ordinary evidence, such as calorimetry that was perfected in 1840, and that any capable scientist could have done with assurance since 1780.

Quote from maryyugo

This would be much better stated something like: There is no way to falsify that certain materials under certain conditions give the appearance of anomalous heat production. This finding is not consistent and is not always reproducible. The methods for achieving it vary widely and often, replications by independent researchers do not work. Nobody knows what causes this phenomenon. No practical application for it has been found despite twenty years of trying during which thousands of scientists have been involved and hundreds of millions of dollars have been expended.

That's an extraordinary set of assertions. This is nonsense so concentrated I fear it might form a gravitational collapse and a kind of intellectual neutron star: stupidity so dense that no one can escape from it with an intact mind. Like Fox News.

I won't go into details, because the stupid might rub off on me, but we have here five statements totally at odds with the facts and with published claims by researchers. In other words, statements invented out of whole cloth by Yugo.

Followed by one true sentence: "nobody knows what causes it." An irrelevancy: no practical application. (How could there be?) And then nonsense about imaginary scientists and imaginary hundreds of millions of dollars.

I think the term “replication” is often used here in a way that is not really appropriate to scientific work. The gold standard of replication is that any suitably skilled researcher with access to comparable equipment and materials will obtain comparable results by performing the same experiment. It doesn’t have to work all the time. New stuff often only seems to work sporadically. But a totally different expeiment using different materials and methodology that obtains similar results does not constitute replication. It may well be additional evidence of the phenomenon of interest, but it is not replication.

Quote from interested observer

The gold standard of replication is that any suitably skilled researcher with access to comparable equipment and materials will obtain comparable results by performing the same experiment.

That's true. Confusion arises when the person fails to get comparable results. You cannot tell whether that person is not "a suitably skilled researcher," or whether the results are invalid. You have to wait until several other people try. How many is "several"? I would say 5 or 10, but it is a matter of opinion.

Quote from interested observer

It doesn’t have to work all the time. New stuff often only seems to work sporadically.

The failure rate is irrelevant. Cloning mammals only worked about once in 1,000 attempts at first. That only meant you had to make lots of attempts, preferably simultaneous attempts, to save time. When Bockris went to measure tritium from cold fusion cells, he ran arrays of 10 x 10 cells. As I recall, only about 1 in 3 produced tritium.

In the 1950s, some transistor types used to fail at high rates. As in 90% or 100% per batch. They were less reliable than cold fusion. That is one of the reasons the cost remained higher than vacuum tubes for a while. Despite the high failure rate, no one doubted that transistors were real.

Quote from interested observer

But a totally different expeiment using different materials and methodology that obtains similar results does not constitute replication.

That is true, and important.

It is sometimes difficult to judge whether an experiment is so different from the others it constitutes something new. I would say Mizuno's recent experiment is quite different from something like Fleischmann and Pons' original electrochemical technique. It is so different that if it turns out Mizuno is making a mistake, I would say that does not reflect badly on F&P, and it should not make anyone doubt their results.