Open discussion

This page is for open discussion, as a subpage of the analysis page supra. Comments here are open to anyone. Note that requires moderator approval of initial comments, which will normally occur within a day. Subsequent comments from the same source will be automatically approved, subject to moderator revision.

Some elements of comments here may be copied to the analysis page where relevance is established; the analysis page, however, is not to be used for advocacy of any point of view; it is for seeking consensus and, failing that, documenting positions neutrally.

Comments that have been moved here using Tako Movable Comments are not appearing in the open page (but only in the administrative page edit interface). I do not know why, and this has worked before. Pending resolution, here is the discussion copied and pasted from the admin interface:

Abd ulRahman Lomax In reply to Jed Rothwell.This interchange will be moved to a subpage, because it has obviously become a debate. The goal here is not debate, but discussion and sharing of information, evidence, and points of view.

First of all, this questioning of existence of concepts is ontologically naive. They exist as concepts. Further, there is apparently real recombination, that is not actually controversial, the issue is level, not existence. There are analyses that establish an upper bound for it, as I recall, based on certain experimental results. Are you denying this, Jed? As to CCS, there is, again, no doubt that in some experiments, shifting the location of heat generation could affect results. Again, the issue is not reality, but level and specific application. Without this recognition, debates become a juvenile “Is So!” “Is Not!” with no resolution. We can do far better than that. We can drive a stake through the heart of zombie arguments, nailing them to the wall for future reference, without insulting anyone. (gad, gruesome metaphor, sorry! shades of a Bones episode on Louisiana Voodoo.)

Some years back, a skeptic and I were discussing SRI experiments on Wikiversity. He came up with a whole series of possible artifacts. Some were totally preposterous (Radium contamination!) But, hey, he was thinking, and I wanted to encourage that. Eventually, with P13/P14 he came to miscalculation of input power due to bubble noise, and he had a Rube Goldberg explanation of why this only appeared in the deuterium cell and not hydrogen (smaller bubbles, less noise!)

Great. I didn’t think so, but … this was taken to Dieter Britz, who actually wrote a study based on experimental data he had. Because this comes up in the current discussion, that’s quite useful. This is a far deeper and more satisfactory response than “You’re Wrong!” We also got comments by experienced electrochemists on the CMNS mailing list. Pons and Fleischmann use the same method of measuring input power, controlled constant current plus averaging of voltage. It’s a powerful and effective method. It is not followed by all researchers.

This is the kind of response I want to see to questions and critiques. It recognizes the question, it does not ridicule it. Then it answers the question, and because of who Dieter Britz is, it more or less nails it.

Jed Rothwell In reply to THH.[Admin: this comment was in response to]

You wrote:

“My method here is to raise issues I’m not sure about that are not settled in the paper.”

I suggest you read the paper first, and then raise the issues. Speculating about something you have not read will only confuse the matter, and waste time.

“As you have noted we have a putative mechanism for liquid-phase water loss without salt loss. Whether it could be significant later on – we do not yet know.”

Fleischmann and I do know. You don’t, because you have not read the paper. So, you have no basis for an opinion, and your saying this here contributes nothing to the discussion.

Jed Rothwell In reply to Abd ulRahman Lomax.You wrote:

“Shanahan’s CCS and ATER certainly do exist.”

Where is the experimental evidence showing they exist?

Jed Rothwell In reply to THH.You wrote:

“Re: X does not exist, because of calibration. We have not yet looked at calibration: so though you may be correct that is for the future.”

What is the point of saying this?

I have looked. Martin Fleischmann looked, and described the reasons why calibrations rule this out. Since you have not looked, you have no basis to contradict Fleischmann or me, or to doubt what we say. I am expressing my opinion. You are saying “you might be wrong.” Yes, anyone might be wrong about anything. Pointing this out contributes nothing to the discussion.

This is like saying, “there may be an undiscovered error in this experiment.” That is true, but it applies to every experiment since Newton. It cannot be tested or falsified, so it is not scientific, and it contributes nothing to the discussion.

THH In reply to Jed Rothwell.Re: X does not exist, because of calibration. We have not yet looked at calibration: so though you may be correct that is for the future.

Re: CCS does not exist for your stated reason here, again we have not yet looked at the part of the paper that describes this: so though you may be right it is for the future.

Re these general features: yes overall I agree with them, with the caveats mentioned.

Abd ulRahman Lomax In reply to Jed Rothwell.I don’t want these detail subpages to become debates. The goal is to understand the paper, so, in the first phase, we will be summarizing it and explaining what it says — and sometimes, what it does not say. THH has raised an ordinary concern, though he has not been quantitative about it (which could be difficult). I am sure that we will be looking at all possible error sources for the calorimetry, including those addressed by F&P and those not. These matters raised by THH belong on the list.

If a problem is “invented,” and there is literature on it, it’s been considered before and isn’t spurious. Basically, a problem can easily exist as some imagined possibility but, in practice, have negligible effects. When we have a list, we can go down the list and research each item.

You have not signed up for the Review Committee. I’m considering moving comments from non-members to a special place. You are highly knowledgeable; my preference would be that you sign up, by commenting as a member on Review Committee. That will give you a vote in polls and your participation will be very, very welcome, I’m sure. But accusing THH of having a “bad habit” — in these study pages, this is not the blog — is inappropriate, member or not. “Making up problems” is a welcome task here. Attempting to prove this or that, at this point, is not.

Shanahan’s CCS and ATER certainly do exist; the question is the magnitude. My sense of the literature at this point is that it is essentially negligible under most circumstances in skilled calorimetry. We will be looking at calibrations, possible calibration errors, and recombination, these are major issues, not to be waved away by saying they don’t exist.

Jed Rothwell In reply to THH.[Admin: This comment was in response to]

THH wrote:

“heat loss through the top of the cell (Kel-F plug and associated through connections) will depend on cell conditions and this (as in CCS/ATER or CCS -something else TBC) could make significant changes.”

No, that is incorrect. There are no significant effects from these factors. If there were, they would be seen during calibrations. They are not seen.

The whole point of doing calibrations is to ensure there are no problems of this nature.

You seem to have a bad habit of speculating and inventing problems which the literature shows do not exist.

Shanahan’s CSS does not exist. If it did, the cell constant would change when the location of the heat in the cell is changed, for example when calibrating with a joule heater.

10 thoughts on “Open discussion”

  1. If a problem is “invented,” and there is literature on it, it’s been considered before and isn’t spurious. Basically, a problem can easily exist as some imagined possibility but, in practice, have negligible effects. When we have a list, we can go down the list and research each item.

    Perhaps something that will explain my rather catholic attitude towards raising potential issues. there are maybe 10 or so (we will find out) issues to consider that might potentially give rise to anomalous results like those seen in this paper. I’d expect that all or certainly most will have been considered explicitly in the literature. That is useful, but it does not answer the question of whether that consideration applies under the detailed circumstances of this experiment.

    A common source of errors is when a particular experiment has characteristics that unexpectedly break usual assumptions, and this is not realise. Therefore we need to do a first principles analysis of all these mechanisms, with the help of the relevant literature, to check that indeed they are not an issue here as would be expected.

    It only needs 1 anomaly (or perhaps 1 anomaly for each distinct phase) out of many potential sources of error. Thus we need to consider low probability possibilities when there are many possible errors, especially because the positive results in this experiment will naturally condition the experiment design. Other experiments with negative results and different experimental design may never be published.

    1. Thanks, THH. Again, one goal here is to list issues, to resolve as many as possible, and to document the rest. There is a point of inadequate returns for the effort involved, but the world is vast, perhaps someone will decide to investigate minor possibilities. However, my ultimate concern is preponderance of the evidence, because the decisions to be made based on examination are civil issues: whether or not to allocate funding, for example. Those who demand “proof” for such questions can create damage. I.e., likely damage. If there is a list of remaining anomalies, with some being possible at an estimated probability of, say, 1% to 10%, A list of 5 of these may still leave a preponderance of the evidence that none are significant. And that is an aspect of what we will be looking at, significance, expressed quantitatively, where possible. Too often “possible errors” have been asserted without such analysis, when the maximum damage to conclusions about XP was small. Example: input power measurement error due to high-frequency components based on bubble noise, with an input power supply set for constant current with a response bandwidth of 1 MHz, and with noise frequencies far below that. Bubble noise is very slow, compared to a megaherz. But for someone who believes that “they must be making some mistake,” this can sound plausible. (With constant current, voltage may then be sampled and averaged over constant sampling periods, with power being, then, simply, average voltage times the constant current. This obviously does not work if current varies.)

      Once again, first things first. The first goal is to understand the paper, not to figure out what might be wrong with it. This is consensus process, which begins with understanding positions, not with figuring out what is wrong with other points of view.

      We agree on this, I think: we will enumerate those “potential issues,” then we will see, first of all, if they are covered in the paper. And then in Morrison’s critique. And then in the Reply. Then we may move on to discovering if these issues are covered elsewhere. Way down the road, we will be looking at whether or not these issues are sufficient cause to set aside the paper’s few conclusions.

      A possible goal here is a paper studying the debate, ideally with multiple authors, and hopefully the authors will be — at least originally — people with different points of view. There was a (feeble) attempt by Scientific American to do this, but it was not interactive and did not consider actual issues in depth, it just presented, perhaps, different conclusions. I’m hoping we can do far better than that.

      We may also end up suggesting specific tests for various remaining possible artifacts. Again ideally, I hope we can find agreement on these.

      Jed has suggested that some of these ideas have already been tested. You, Tom, have pointed out that being tested under one set of circumstances does not necessarily apply to all. I think you are both correct. The significance of those positions, however, may vary. We will be looking at details.

      As well, a very pesky possibility is some unidentified recombination anomaly that is very sensitive to material conditions, a kind of mirror image of the FPHE. I.e., Shanahan.

      It will be up to experimentalists — or those who fund them — to decide if it’s worth the effort to maintain testing for this.

      Too often, we lose sight of the purpose of these “debates.” For some, it’s an exercise in debate skills, in “proving” that one is right and others are wrong. While that can be fun, for a time, it is not what humanity needs. Decision-makers, those with the responsibility to allocate funding, need much better. And their standard will properly be “preponderance of the evidence.” With time, ordinary scientific process will separate the sheep from the goats.

Leave a Reply