The perpetual “is LENR even real” argument thread.

Quote from JedRothwell

No, you don't "remember" that. You made that up. There are dozens of cases with large results, from Will, Bockris, BARC and many others. You would know that if you had ever read the literature, but you have not.

Why do you keep making up stuff? Who do you think you are fooling? What is the point?

Define "large result. And quote the study please.

My definition of large result is > 10 X average precipitation background for the year in which the experiment was conducted. (Background tritium has varied by factor of 1000X between 1970s and now).

This argument I know will not change your mind or mine, but without precise numbers what is the point?

Why do I ask you to quote numbers and link a paper? Well, you are making this claim, not me. And it is some work - though more for me than you since i do not doubt you are more familiar than me with the literature. I challenge your ability accurately to interpret it, not your knowledge or honesty. (And, yes, when you say I am dishonest, as you sometimes here have done, I realise that is what you honestly believe: it is untrue).

Quote from JedRothwell

A large result is one that far exceeds the margin of error for instrument in question. For example, with an ordinary calorimeter, it would be 0.5 W

Absolutely not. At least not always! When (as for tritium) the expected value is much less certain than the measurement accuracy, a large value must be significantly bigger than that.

I asked for figured because tritium values in water are so very variable across time and geography: in the 1970s they were 1000s of time higher than now. Unless water is detritiated it causes much uncertainty - as does the fact that tritium is fractionated by electrolysis, and by evaporation.

I realise that all of these sources of uncertainty can be addressed - and this is done by the best of the old tritium papers. What however you leave out is that when this is done the conclusion that the experiment is safe depends on a whole load of complex assumptions which may all be reasonable but are not necessarily all certain.

That complexity causes uncertainty because it leaves more room for unknown unknowns. Something you state cannot exist. Few experimental scientists would be so certain.

As always - your certainty is reasonable if the many different positive results of different types can be safely combined to make an overall conclusion more certain than the experiments. The peculiar nature of LENR where many different anomalies are unquantitative possible signatures of it and none are required signatures makes that (quite normal) enhancement of certainty not safe.

And before you mention He/excess heat. Yes that is a genuine example of a quantitative relationship that would help in adding certainty if redone reproducible set of experiments with a suitable protocol that did not indirectly select for the required result. As you might remember I have been strongly in favour of that approach as one that might deliver evidence - but when you look at the He detection required, and the atmospheric level of He, it is a very difficult experiment.

One thing that would also help is clarity about what is the nuclear reaction?

"Besides helium, other nuclear products are detected in much smaller quantities. Early in the
history, great effort was made to detect neutrons, an expected nuclear product from the d-d
fusion reaction. Except for occasional bursts, the emission rate was found to be near the limit of
detection or completely absent. This fact was used to reject the initial claim. It is now believed
that the few observed neutrons are caused by a secondary nuclear reaction, possibility having
nothing to do with the helium producing reaction. Tritium is another expected product of d-d
fusion, which was sought. Too little tritium was detected so that once again the original claims
were inconsistent with expectations. Nevertheless, the amount of tritium detected could not be
explained by any prosaic process after all of the possibilities had been completely explored. The
source of tritium is still unknown although it clearly results from a nuclear reaction that is
initiated within the apparatus. Various nuclear products normally associated with d-d fusion also
have been detected as energetic emissions, but at very low rates. Clearly, unusual nuclear
processes are occurring in material where none should be found."

Storms 2006

One of many more recent runarounds of how can the observed results be fitted to possible nuclear reactions as determined by conservation laws (ignoring can they happen or why do they decay that way).

My point is that detecting He or T quantitatively as a fingerprint of LENR required that we assume a specific reaction, and test it. That indeed is good science. But it is not LENR because the results are variable - I know of no single reaction that can explain them all.

And if it could be any reaction or indeed none - that weak predictivity makes it impossible to disprove LENR experimentally, and means that a very wide range of anomalies can be claimed as evidence for it.

That is the source of my skepticism. I am however quite optimistic about the possibility of some real definite lattice-enhanced LENR reaction causing a small subset of these results in a way that makes complete sense and is quantitatively provable or disprovable.

Quote from orsova

There’s a pattern that I’ve noticed in the way you sometimes respond here, in that you zoom out and speak in general terms about various hypothetical sources of error, and then use that discussion as the basis for rejection of specific experiments or claims.

I did not see the specifics i asked for. Jed knows his library very well and could, were he willing, provide this. If I have missed this I am sorry - the link above is a link to Jed's library.

Searching for a paper is not a great idea in this situation - i will probably get the wrong one even if given an author. Really, there is no substitute for a specific link which is what i give people when seriously looking at things.

But perhaps Jed is not interested in doing that because what separates us is generalisations. Jed looks at the collected evidence and finds it compelling because the whole is greater than the parts. i don't - becayuse the lack of specificity and coherence with specific theories makes the whole no greater than teh parts.

That could change as I've said many times. Change could come from either of:

1. A reproducible certain experiment now
2. A specific LENR theory that made falsifiable predictions

If LENR exists I'd expect both to be possible. The post-google stuff on lattice enhancement is getting close to 2. Although even though that is supported by definite reproducible quantitative results there are quite a few not yet substantiated pieces for it to even be useful LENR: and the ideas that go with it for aneutronic reactions are particularly unsubstantiated. I'd therefore not be surprised if most people here view it as a poor potential explanation.

Quote from orsova

To speak in abstract, hypothetical terms about what may or may not be wrong about whole swathes of literature is not an adequate response to the specific task of reviewing a specific paper - which is the task that you and Jed were negotiating.

And it is certainly not grounds for the rejection of the claims of any specific paper.

In this case, you cannot prove a particular with a generality.

I think this is a slight misunderstanding of the contention here.

Sure, there are a whole load of LENR papers (generally the modern ones, and some of the late F&P ones) which can be rejected. But there are also the post-F&P high quality early papers - a lot of them - which Jed puts his faith in. Those are serious and show genuine anomalies.

The question is whether those anomalies are specific to the experiments, and or whether they all have the same (presumably nuclear - but with no theory that predicts them all) cause.

The arguments on this thread that are general are essentially reiterating that point. You are right - neither Jed's contention, nor mine, repeated, adds anything. Both sides see the other as ignoring something that is obvious.

I guess I am more interested in that type of reasoning than most. It is a problem properly solved by Bayesian probability theory - but too complex for that ever to be helpful.

You are thinking that either I can find a mistake in every string LENR paper, or Jed is right. But that is not true. Those individual papers do not support LENR. They each show an anomaly. The different anomalies are not all predicted by a single LENR theory - yet. Until that happens the individual evidence in each case is just "appears to be an anomaly" and lack of clear reproducibility makes that inconclusive.

For example, Jed will say (as will everyone here): anomalous excess heat is well proven and reproducible. Yet the modern attempts to find this are all uncertain, and show results much lower than those original F&P results. I don't dispute anomalous heat from electrolytic H or D + Pd experiments. I dispute nuclear reactions as being the mots likely reason for it. I don't dispute apparent excess Tritium from the Tritium experiments - i dispute LENR as mots likely reason for it. In that case I can say in some cases what is a likely reason - e.g. electrolytic or evaporative fractionation - and in other cases just point out that contamination has not been ruled out.

On any single experiment Jed will say - you are supposing some unlikely source of error. I will say - yes but it does not have to be very likely, because the set of possible anomalous "LENR" results is quite large, and the number of error mechanism sis also quite large. You only need one error mechanism and one anomaly to generate LENR evidence.

Jed will then say - ah - but all these results add certainty. I will say - no they don't for the reason in the paragraph above.

Quote from orsova

Agree to disagree. There may not be a theory, but there is a commonality of materials, experimental design, methodology and observation. Moreover, there is also a commonality amongst the observations - all the claimed products seem to be nuclear in their origin.

I agree with what you say, but not that this makes a collectively stronger for a common nuclear origin because by definition positive LENR results are those that imply nuclear origin and therefore they are looked for (by these experiments). Any such anomaly counts as positive - any negative (no anomaly) does not count as negative. Because exactly what anomaly is expected in any individual case is not predicted by LENR. Anyway - I understand that weighting this is complex and can be argued both ways.

Quote from orsova

A) The super majority of papers that report tritium fail to adequately account for possible sources of error.

B) One paper is absolutely exemplary and conclusively shows a large, anomalous amount of tritium.

If you make an argument in general terms based on the observation of A, and find that argument to be dispositive, then you're never going to get to B, which is really the useful conclusion to get to.

That is a worthwhile point. It is true that a single killer experiment would be powerful. But not overwhelmingly so. Scientists rightly never believe single experiments. They get very interested in them, try to reproduce, try to work out theory that could be compatible with the new evidence. You can see that with the FTL neutrino experiment. Equally when the result is not predicted by any otherwise plausible theory they are rightly very cautious about its being real. As indeed the FTL experiment - so carefully checked before publication - turned out not to be real.

The point about reproduction is not just "reproduce exactly the anomaly". It is that as soon as an anomaly is reproducible you can test it in different ways and ask questions about it. That can quickly provide hints as to its origin which may be an unknown experimental artifact, or an unknown law of physics.

The biggest credibility problems for LENR is the large amount of "could be LENR" evidence. That sounds strange! It has two aspects:

(1) (this is unfair - and I agree with people here that it was until team google unfairly discriminated against) the surge of interest a long time ago with no conclusive results, nor plausible theory makes people think now "more of the same" and be overly dismissive.

(2) (this is fair). The more diverse poorly attested anomalies are quoted seriously as LENR the more it looks like a pseudoscience. The Rossi phenomenon. those free energy companies that have claimed commercial results for 20 years, yet whose demos are different over time and rely on ever more complex and uncertain measurements. People who look at the may wonderful not understood phenomena such as ball lightning and hypothesise that is cause by LENR. People who claim a very wide range of transmutations from uncertain experiments. This diverse evidence makes fitting a predictive theory incredibly difficult. For example, it has to be, almost any nuclear transformation can happen, in LENR experiments, while such has never been observed by those not looking for LENR, and the energy mismatch fractionation problem is uniformly solved meaning that expected and very easily measurable high energy porducts are not observed.

(2) is why I am more positive about electrolysis D+D LENR than other things. Even opening it to H+H as well as D+D reduces its credibility (not to zero - but it becomes less attractive) from my POV.

These ideas are I think not ones you can engage with unless you have a truly open mind. They relate to real uncertainty, and weighting the probability of different explanations. They do not prove LENR does not exist, they affect high likely it is to exist. And that judgement is so subjective that I don't see any scientific way it can be resolved, except in a positive way from new evidence. So I do not see LENR can ever be disproved. However, specific LENR theories - e.g. Widom Larsen - can be checked and proved or disproved. The more definite the theory the more it can be disproved. If the theory has steps where the rationale is hand waving it will be more difficult to disprove because you can reasonably posit many different mechanisms and fit whatever results occur to some variant.

Quote from orsova

Ok, but we should make some allowance for the fact that you can't get sponsored to do this work, and it's career suicide if you do. And then, on top of that, it's an extremely challenging experiment. In a way, it's not really a fair critique because the resources available to researchers haven't been continuous and stable.

Agreed. Although post-team-google this argument applies less. TG had decent funding, and tried hard to find a reproducible "define LENR" experiment. Their work showed some anomalies that were interesting and are still being actively pursued. It is always a judgement call where do you throw money: e.g. LENR or experiments that might test modified GR theories, or deliver info on neutrinos.

Quote from orsova

Besides, what do the results of recent experiments say about the quality of older experiments? You imply that recent results somehow invalidate or cast doubt on prior results. That doesn't make sense.

Again, you make a general argument: "All the recent results are worse than the older results" and use it to cast doubt on F&P's work.

The high quality post-F&P work showed results much lower than FP claimed, and to that extent invalidates their claims. You have a credibility problem if you say that both F&P claims and McKubre claims are correct, since the later experiments, with more accurate calorimetry, shows lower results than the original F&P experiments with less accurate calorimetry. Most people, looking at those experiments, would say that if LENR is real either F&P had green fingers never repeated after or F&P's claims were over-egging things a lot. Scientists tend not to believe in green fingers.

The credible modern experiments do not replicate the most interesting (but not replicable) "good quality" post-F&P results. E.g the McKubre series that has one (?) result with much higher excess heat - clearly beyond possible errors - than the others. Ed's very creditable attempts to find a "lab rat" electrolytic experiment with clear excess heat show very low excess heat results. At a level most external observers would not call certain. Those modern experiments replicate low uncertain levels of excess heat. It is frustrating - it means we must stay interested in the phenomena without thinking of it has strong evidence for LENR.

The one negative from this is hard to quantify. Any level of real nuclear excess heat should - you would expect from every plausible mechanism - be highly variable and often muhc higher than measurement errors so with effort you would expect it to be possible to find a lab rat experiment that shows it clearly. There are challenges to this - such a variable mechanism can destroy NAEs as soon as they become active etc. But it is unlikely that consistent excess heat over a long time can only be found at levels that are so low compared with experimental error. I remind you of the first sentence in this paragraph. This is a genuine argument - but it is hard to quantify its weight. Personally I vacillate - sometimes I think it has ahigh weight, sometimes low. It is ok to be uncertain.

Quote from orsova

But it's not useful unless you've actually read a specific paper and are making a specific argument about that paper.

"Maybe there's an error, because there's often an error" isn't a rigorous critique.

True. But then none of us are qualified to make rigorous critiques of these experiments. Those who are (Ed?) are subject inevitably to the exact group think that those who believe LENR is clearly real think afflicts the rest of the scientific establishment.

What would be needed to find non-LENR reasons for the anomalies is intense interest from many highly qualified who were trying hard to disprove LENR. A sort of anti-team-google. You can see easily that no-one serious would ever want to do that. You can however be sure that LENR would get that interest when it finds a lab rat experiment with certain results. those people would be piling in trying to identify the (obvious) anomaly and equally happy whether they were proving LENR or proving some boring non-LENR error mechanism.