History of Science
Controversy in Chemistry: How Do You Prove a
Negative?The Cases of Phlogiston and Cold Fusion**
Jay A. Labinger* and Stephen J. Weininger*
http://www.uaf.edu/chem/481-482-692-Sp06/pdf/labinger-1.pdf
I think this article deserves a closer look. It
relies heavily on Simon, a very good source. It's
also possible to misinterpret Simon, I've seen that on Wikipedia.
For our second study we have again
chosen two stories that appear to be
about as disparate as one could possibly
arrange. One, the overthrow of the
theory of phlogiston, dates from the
origins of modern chemistry, and is now
universally considered a central development
therein. The other, the cold
fusion episode, is only 15 years old, and
is now generally (though by no means
universally) considered as just a stumble
in the long historical march of chemistry.
Why have we paired them? Because we
feel that, as with the previous study,
closer examination reveals certain connections
that are instructive for a general
understanding of how controversies play
out and, in doing so, serve as a powerful
engine for the advancement of science.
The question of what counts as evidence
is important here as well, but we will
focus in particular on the persistence of
belief, associated with the difficulty of
demonstrating the non-existence of a
theoretical or hypothetical entity.
This was written in 2004, before the DoE review
came out, which they anticipate. Certainly it was
true, then, and possibly now -- though far less
clear -- that CF was considered a "stumble." In
fact, don't we agree on that? We just have,
perhaps, a different understanding of exactly
what the stumble was. Many sources have, in fact,
regretted what happened in 1989-1990, with even
skeptics acknowledging that the reaction to CF
was exaggerated and even "believers"
acknowledging that Pons and Fleischmann had made
mistakes. Nobody currently thinks that that first
press conference was a great idea. But,
seriously, were the attorney's wrong to insist
upon it? Yes, it turned out to be an error, but
based on the knowledge that they had at the time,
was it? Pons and Fleischmann were hot on the
trail of a tiger, and it looked like they might
have it by the tail. But even though they saw the
tiger, and were able to record some of its
characteristics, it got away. The attention,
however, caused many to start looking for tigers
where nobody had looked before.
It's easy to imagine that the attorneys, and P&F,
though that with more funding, which did arrive
as a result of the announcement, they'd soon have
the tiger in a cage. It was a gamble, in
hindsight, and even when a gamble is a good bet,
it's still a gamble and can fail.
I read this symmetrically, whether the authors
intended that or not. The persistence of belief
is a real phenomenon that affects scientists as
well as others. Everyone, really, for persistence
of belief is necessary to a degree. It is only
when we lose context and mistake belief for
evidence and proof that we go astray. Proofs are
based on assumptions, and assumptions can be
incorrect. Certainly that is what happened with
cold fusion and the theory used to reject it.
Theory was applied outside the realm where it was
well-established, on the margins.
Imagine if somehow relativistic phenomena had
been overlooked, so Newtonian mechanics reigned
supreme. Then, as would be inevitable,
eventually, some experiments moved into the realm
where relativistic effects would be measurable.
And it was found that measured acceleration no
longer was equal to force divided by mass. It
would be pointed out, it's easy to understand,
that this violated the law of conservation of
energy, a law that, by that time, would be
thoroughly well-established. But it doesn't
violate that law; the assumption of violation
would be based on assumptions proceeding from experience at lower velocities.
It was known that quantum mechanics was an
approximation, but had it been proven that the
approximation was adequate to predict fusion
cross-section under all the different possible
conditions in condensed matter, which might
involve configurations of matter not even
contemplated? Had anyone ever calculated, using
quantum field theory, what would happen under
various conditions of confinement of multiple
deuterons? I don't think that happened until the
1990s. And the results were different than
expected, and confirmed, as a possibility, what
Fleischmann had observed. This theory is still
not demonstrated to be the mechanism behind cold
fusion, but I raise it to show the reason why
theory should lead to caution, but should not
completely trump experiment, ever. The
application of theories is based on assumptions,
and we don't know what causes experimental
results until we have much more information than we may initially have.
What P&F found looked, to them, like "deuterium
fusion." But it did not have the stripes of that
tiger. Even though, later, it was shown that the
predominant ash was helium and the predominant
fuel was deuterium, and the excess heat
correlated well with that expected from the
amount of ash generated, this still did not show
that the actual mechanism was d-d fusion, and
it's quite likely that it isn't. It is, quite
possibly, something else, but by conservation of
energy and mass-energy equivalence, if the input
is deuterium and the output is helium, that heat ratio is required.
The mechanism, though, could make all the
difference. Besides the Tetrahedral Symmetric
Condensate theory of Takahashi, there are
theories that involve other possibilities that
would also take in deuterium and spit out helium.
If, for example, deuterons, perhaps multiple
deuterons, fuse with heavier nuclei that then
decay through alpha emission, with that energy
being dissipated either through electromagnetic
radiation that ends up as heat (i.e., EUV that
can't penetrate the cell confinement), or as the
energy of the alphas, which will also end up as
heat, we'd see the heat ratio, but not the radiation signature of d-d fusion.
And, of course, it remains possible that there is
some mechanism whereby d-d fusion does occur, but
special conditions, again unexpected, allow
direct transfer of energy to the lattice and a different branching ratio.
Labinger and Weininger are correct: it is
difficult or impossible to prove a negative. But
it's also an error to imagine that scientists
should be attempting to do so. Rather, the proper
skeptical position, real skepticism, would not
consider anything proven that hasn't been proven!
Including the negative. Instead, the question
asked would be different. Is there any reason to
believe that this phenomenon is real? How
reasonable, given all the evidence, is the
hypothesis that the experimental results are
artifact, perhaps amplified by publication bias,
wishful thinking, or even fraud?
Skepticism, especially for those not familiar
with the literature, is quite understandable.
That scientists would not investigate,
personally, every new claim is likewise
reasonable. But there is a collective
responsibility to remain open to revision of
theory. When many researchers come with
confirmations of a rejected claim, if nobody
investigates, we have and a pretense of negative
proof enthroned and in charge. And that is the
end of science on the topic, it has become pure
politics, pretending to be science.
Many scientists, indeed, understood this, hence
the allowance of seminars on LENR at conferences.
Even if only as some activity in a corner, the
door was left open, at least in theory, and, to
some degree, it obviously was open. The last
several years have seen a major uptick in the
number of papers being published by reputable
journals, and in the quality of the work and the evidence presented.
Labinger and Weininger begin by examining the
phlogiston theory and its advocates, with this remarkable comment:
Conants respectful treatment
of phlogiston and the phlogistonists was
entirely appropriate.
If they are making an analogy, how would this
apply to cold fusion? Cold fusion researchers
have been derided, have lost professional
credibility and position and access to resources,
and peer review was denied by editors based on
"policy," or such peer review as there was, at
many publications, was knee-jerk rejection and
ridicule, many of these reviews have been
published, and Simon covers some of this.
In describing the positions of "believers" and
"skeptics," Labinger and Weinninger present the
views as an impasse, but to do this, they must
neglect or underemphasize aspects of both
positions. They do acknowledge the problem:
skeptics bring up the criticisms of 1989 as if
they had never been answered, as if those "facts"
and conclusions were unquestionably valid and
accurate, such as the oft-repeated claim that
effects disappear with increased accuracy of
measurement. That's one of the Langmuir's
characteristics of pseudoscience, so it is
comforting to skeptics to believe that, even
though there isn't serious evidence for it, i.e.,
disappearance of the effect simply with more
accurate measurements, especially with regard to
the excess heat effects, that has stood the test
of time. The contrary hypothesis, that negative
replications were simply failure to reproduce the
necessary conditions (which were very poorly
understood in 1989), is clearly just as
reasonable, and Labinger and Weininger note this reasonableness.
Further, because an entirely new possibility was
being considered, and the reality might turn out
to be quite complex, with multiple possible
reactions and secondary effects, the variation in
results is not surprising. Early in the
exploration of the western hemisphere, various
explorers returned with descriptions of the lands
found. Those descriptions did not always match
each other. Did this mean that they were all
mistakes? More likely, wasn't it, that they
visited different parts of this new territory.
Nevertheless, the paucity of exact replications
has been a problem in gaining respectability for
the field. It's quite understandable, though. In
the search for an amplified effect, the "working
power generator" that some seem to think
necessary, why bother repeating experiments that
showed measurable but practically useless levels
of reaction? Why not keep trying new approaches?
There is no money in replicating someone else's
experiment, unless one is supported by grants or
institutions, and that is exactly what was denied
to the CF researchers. Both U.S. Department of
Energy reviews recommended grants to investigate
the obviously open questions, and in 2004, there
is no doubt that this recommendation was sincere
on the part of half, or more, of the reviewers.
Yet no grants appeared. The work by the U.S.
Naval Laboratory at San Diego, SPAWAR, was not
funded by high-level appropriation; the
researchers worked on their own time, and they
were assisted by relatively small discretionary
funding, as well as access to sophisticated and
expensive equipment, as such a laboratory would have.
In addition, while the claim is not without any
basis that "nobody has been able to reproduce the
experiments" -- this still appears frequently in
media stories -- that claim completely neglects
that difficult experiments involving unknown
conditions can still show confirmation and
significant results through statistical analysis.
Labinger and Weininger do obviously reognize the
significance of the claim of heat-helium
correlation. They don't seem to realize the depth
of that claim, that it was not just one
researcher; that several independent groups came
up with the same results, within experimental
error, is a stunning confirmation, requiring some
serious contortions to attempt to explain it without LENR.
Overall, the article has a tenor of assumption
that LENR is an error; however, they also
recognize the problem, that this hasn't been
proven. They focus on the impossibility of proof
of a negative, but that is a bit of a red
herring. Rather, positives can be proven, or at
least covered by sufficient evidence to be
conclusive. N-rays were not rejected because of
proof that they did not exist, but by conclusive
exposure of the error in the experiments which
purported to show that they existed.
It's easy to understand why many think that this
happened with cold fusion, because of the
demonstration that Fleischmann's neutron results
were erroneous. But errors happen in experiments,
and that wasn't the original basis of
Fleischmann's work. Once their suspicion that the
excess heat they were measuring, in a significant
fraction of their experiments, was nuclear in
origin, of course they looked for neutrons, and
when one looks and encounters an artifact that
confirms the expectation, and publication is
rushed, as it was, the artifact gets reported.
But, while the calorimetry was criticized, those
criticisms were often far from cogent and
applicable, and, definitely, no "smoking gun" was
found showing them to be, as well, artifact.
The apparent reality of the DoE recommendation in
2004: more attention was paid to a
misunderstanding of the conclusions of that
report than to the actual report. The report
concluded that it's recommendations were "much
the same" as in 1989, and, since we all know how
negative the 1989 report was, it's easy to assume
that they were confirming that earlier report.
But the 2004 report and the 1989 report were like
day and night, in fact, the similarity was only
in the formal conclusion: no special federal
program, but targeted funding to answer open
questions. Excess heat was, apparently, was
accepted as a real effect by exactly half of the
reviewers. *One reviewer* changing his mind would
have made excess heat a majority conclusion, and
we could point out that there was likewise no
majority for the view that it was error.
Now, if you don't believe that excess heat exists
in these experiments, you certainly are not going
to believe that its origin is nuclear. So we
could assume that the half that rejected excess
heat, considered there to be no conclusive
evidence for it, would also reject a nuclear
origin! Thus the statement of the report that
one-third of the reviwers were "somewhat
convinced" that the origin was nuclear means, in
fact, that of those who recognized excess heat,
the hypothesis that it is of nuclear origin was considered relatively strong.
Given the weakness of the process used in 2004,
the clear errors made by one of the reviewers
about heat/helium correlation, compounded by an
even more serious error made by the summarizer in
reporting this, that, thus, probably the
strongest evidence for nuclear origin was not
well considered, and that to understand this
field seems to take months of work, reviewing
thousands of documents, that the knee-jerk
rejection of cold fusion clearly affected some of
the reviewers, who showed no sign of actually
considering the evidence, 2004 could be
considered a turning point for cold fusion, in terms of acceptance.
But the misstatement of the results, so common,
has made it seem otherwise. And that overall
impression probably influenced reactions to
funding applications. Labinger and Weininger come to this:
So there matters stand: no cold
fusion researcher has been able to dispel
the stigma of pathological science by
rigorously and reproducibly demonstrating
effects sufficiently large to exclude
the possibility of error (for example,
by constructing a working power
generator), nor does it seem possible to
conclude unequivocally that all the
apparently anomalous behavior can be
attributed to error. Under these circum-
stances, the DOE has decided to carry
out another review, which is underway
at the time of writing (summer 2004).
From the range of comments made in
response to that announcement,[51]
though, it seems most unlikely that their
report will settle matters ....
Now, they were incorrect about this. Effects
sufficiently large to exclude error as a
reasonable possibility were known well before
2004. While levels of heat and levels of helium
were low in the Miles work and in that of others,
they were large enough to be significant, and the
*correlation* was far stronger as evidence.
Labinger and Weininger do not seem to be aware
that Miles' work was confirmed. Sure, details of
the experiments differed. But the clustering of
these results around the Q factor for deuterium
-> helium was probably the major result that
convinced me, and I assume it has had that effect
on others. The neutrons from SPAWAR were frosting
on that cake, full circle. Yes, Virginia, there are neutrons.
Probably from secondary reactions, not a
signature of the primary reaction at all, but
definitive evidence of nuclear reactions. We can
argue, still about some of the SPAWAR findings,
is that chemical damage or nuclear effect on the
front of the CR-39, but the back side triple
tracks and other clearly anomalous radiation,
likely proton knock-ons, on the back, away from
the cathode, consistent, well above background,
over many experiments, that's another matter. I'm
still concerned about the level of replication of
this part of their results. Hence my own work.
Unlike the Galileo project experimenters, I'm
optimizing for neutron detection, forget the
alphas or whatever it is on the front side. I'll
confirm the patterns, I assume, i.e., I'll see
something on the cathode side of my CR-39, but it
won't be important. I'm adding detector stacks,
not merely single detectors. If it's energetic
neutrons, it should continue to produce results in subsequent layers.
I'm not trying to make a "working power
generator." I'm reproducing the Galileo protocol,
which is a SPAWAR protocol as far as the
electrochemistry is concerned. I'm modifying it
to make it smaller, to lower costs. That is, I'll
be aiming for the same ratio of PdCl2 to gold
wire, hence the same presumed thickness of
deposit, and the same current densities, and the
same LiCl concentration in the heavy water, as
described in the detailed Galileo protocol. I'm
standing on the backs of those who came before.
And then, assuming it works, I'll sell the kits,
that's how this is being funded. If it doesn't
work, maybe I can sell the kits to Park or Close
or one of them fine upstanding scientists.
But, first, I'll do as much as possible to get it
to work, consulting with everyone willing to
provide advice, fixing every possible error. I
want a simple, reliable kit to demonstrate a
simple, reliable indicator of nuclear reactions.
It has to work not only for me, but for others,
or this will be an exercise in futility.
*Then* people can start working on that power
generator, it won't be me, I expect. Too much
investment and too much risk involved. It might
never happen. But until enough people recognize
the possibility, for sure it won't happen.
Or maybe I'll prove that LENR is all a big
mistake, or at least this part of the evidence
for it. I rather doubt it, it's hard to conceive,
at this point, of how that could be, but it would
not be a bad outcome. Science is about reality,
not about what we want or even expect.