Here
While I find myself generally in sympathy with the sentiments below there are a few specifics were I differ from the author.
QuoteIt might take more patience or suspension of disbelief than one can allow, but I think there are some subtleties that need to be teased out in this instance. The first subtlety is that you can have a real, complex, difficult-to-understand phenomenon that has yet to attract a large number of capable people who are grounded in a systematic, effective approach. To a certain extent one could argue that this was the case with chemistry itself for many decades, when it was alchemy. The alchemists saw some interesting phenomena, but it took the Enlightenment and empiricism to really hone the methods that were needed to explore it effectively.
As a historical analogy you could point out that the development of the scientific method, which has advanced the study of chemistry and many other fields, was not discipline-specific. And having done this there are many effective people who jump on "the new idea" whether that is quantum computers or carbon nanotubes.
QuoteIn addition, a field may already have attracted a few effective researchers, but the world was not ready for what they were saying at the time. Galileo suffered house arrest when he promoted heliocentrism. It took many years for the early pioneers in radiation to prove to everyone's satisfaction that radioactivity was real, and a lot of what they thought was initially wrong, hopelessly mixed up with what they had gotten right. Becquerel is credited with having discovered radioactivity in 1896; in fact Abel Niepce de Saint-Victor reported to the French Academy a very similar finding in 1858, and it is clear that he understood the implications, but this piece of history is largely unknown.In this regard, I think society is not as far along as is implied by the story we tell ourselves, where we imagine that we finally have an empirically grounded science after all these years, free from the difficulties described above.
I think this is the wrong way to view the differences between now and then. We have now individual scientists every bit as flawed and human as they ever were. There is a slight improvement in training where many of the tenets of empiricism are explicitly learnt, and often followed. The big change is in the much larger and more heterogeneous quantity of material that is now published and considered. With scale you get more differing views. With number of journals you get "safe places" for heterodoxy (current science being a classic example). With scientific citation and referencing you get multiple views of the same phenomena. All of this is far in advance of what happened at the turn of the Twentieth Century when there were but a few players, and a few places for publishing.
Thus I'm not claiming scientists now are better trained (though some would, and i'm not ruling this out). I'm saying we have more of them, and more room both for heterodox views.
Quote
It takes a lot of willpower to write off the accumulated findings in LENR studies, even if one has yet to see a study one considers flawless or above reproach. In short, one might be in a mode of writing off evidence rather than trying to understand it, however imperfectly it's been packaged.
I think this subtlety is subtler than you indicate here!
I've made the point earlier on this thread. All experimental evidence is interpretation. We assume physical laws stay constant across the timescale of the experiment so we can trust our instruments to behave predictably as they were calibrated. We make many other detailed experiment-specific assumptions. Based on these high confidence beliefs we interpret the evidence.
In this case the correct response to an experiment showing evidence of D+D -> He4 transmutation + excess heat at low temperatures and pressures will and should depend on prior belief. If this is extraordinary we subject the evidence of any one experiment to much greater scrutiny than if it is informative but not extraordinary. Apparent differences on this forum often relate to prior belief, since may here, when considering a specific experiment, do so from the position that other experimental evidence has made some sort of LENR probable, and not extraordinary.
The next part of this argument relates to your synthesis of "accumulated findings".
Where results are probabilistic and independent, quantity is as good as high precision.
For example, if a coin is tossed 10 times and ends up heads we may guess it is biassed. If this indicative experiment is repeated 3 times with the same result we can be sure it is biassed. (confidence order 1 in 10^9).
Now consider a bag of coins (analogous to LENR experiments). There are a number of different types of coin, and new coin types can be formulated and added to the pool as variants of old coin types.
LENR researchers, when they do an experiment, take a coin from the pool and flip it, recording the results. They note collectively a significant bias towards heads. Sometimes this is just luck, sometimes because the coin is truly biassed. Some types of coin are better at generating excess heads than others. They get chosen preferentially be the LENR reserchers as "good experiments with the right conditions to generate NAEs". Newly formulated coins, variants or mutations of those most head biassed, do this even more. LENR researchers also have many experiments which do not work, and show a distinct tails bias. These are not much reported, because of no interest. Also, the types of coin that lead to them are not often selected.
Now a historian, or an internet observer such as you or I, looks at the LENR experiment results and cannot fail to be persuaded by the preponderance of heads. We might, in analogy to the one coin case, reckon the chance of these happening without a real "overall head-bias' effect are vanishingly small. You can see however that the coins overall can be true, each with some independent random bias and averaging neutral. The selected coins however show a head bias, and the reported coins even more head bias. The historically remembered coins include these two biasses and have an even higher bias.
So: in the case that a phenomena is extraordinary quantity of weak evidence does not work, because it can be implicitly or explicitly linked to common selection and therefore the individual results are not independent.
I've not formalised this argument quite as clearly as is needed - but it will do - for here - for now.
