Guest essay by Dr. Sharon Camp
While an educational manual for federal judges was improved when a biased representation of climate change was removed, a remaining chapter on the fundamentals of science would poison the judiciary with quackery.
The “How Science Works” chapter of the “Reference Manual on Scientific Evidence” allows for an overreliance on the unproven assumptions of computer models and an acceptance of “consensus” as proof even when contradicted by empirical evidence.
This violates tenets of the hundreds-year-old Scientific Method that require data obtained through experimentation or observation of the physical world as the means of supporting scientific conclusions. The undue regard for models and consensus is emblematic of climate alarmists whose claims fail scrupulous examination and whose views are shared by at least one of the chapter’s authors.
In defining the concepts of hypothesis, theory and scientific law, the writers exclude the elements of testing, observation and experimentation. Neither is it acknowledged that all three can be disproven, even though the ability to demonstrate the falseness of a conclusion has long been recognized as the path to advancing scientific knowledge.
Along these lines, the chapter even says that science cannot “disprove hypotheses,” although science has done so many times. A famous example is geocentricism that was supplanted in the 1500s by the heliocentric theory of Copernicus. Others include phrenology, eugenics, spontaneous generation and miasma theory of disease. All were embraced by much of the scientific community of their time – a consensus – until proven false. German scientist Alfred Wegener was dead for 30 years – and rejected by a consensus of his peers – before his theory of plate tectonics gained support in the last half of the 20th century.
The authors may misunderstand the Scientific Method. Or perhaps they believe that computer models relying on unproven assumptions that fail to explain or predict outcomes should be accepted over the extensive experimentation and observation required by the Scientific Method.
The chapter also misses the mark with this statement: “In the context of testing hypotheses, the term prediction refers to a logical consequence of a hypothesis, not necessarily what will happen in the future.”
On the contrary, a prediction is all about what will happen in the future. Either objects of different weights dropped from the same height in a vacuum hit the ground at the same time or don’t. (They do.)
A hypothesis also can be tested by determining whether it can explain something that has happened in the past.
Are the authors saying that a hypothesis paired with a model is sufficient confirmation, even when they fail to predict the future or explain the past? Or perhaps they believe that computer models predicting the climate 50-100 years in the future are more accurate than weather models that are reliable less than a week in advance — and sometimes failing that.
Apparently, there is no need to fret about such failures because the authors say: “The fact that there is room for improvement in the process of science does not necessitate distrust of hypotheses that have gained widespread acceptance in the scientific community and about which consensus has been achieved.” This is a justification for using consensus as a defense against evidence that challenges a popular narrative – a common ploy of climate alarmists.
In places, the chapter is self-contradictory. For example, the writers seem to embrace key elements of the Scientific Method when they say that a testable hypothesis must ultimately predict a set of observations that could or could not be made. This suggests a less cavalier relationship with traditional science than exhibited elsewhere, such as when key features of hypotheses and theories are omitted from their definitions. Unfortunately, the contradiction conveys a desire “to have it both ways” more than a commitment to rigorous scientific inquiry.
Ultimately, the chapter’s confusions and inordinate 62-page length muddy a reader’s understanding of the Scientific Method. More problematic, the writers introduce wrongheaded practices like the overuse of models and consensus in settling important questions. Such departures from the empirical tradition of Isaac Newton and Marie Curie make for superficial and specious investigations of reality known as pseudoscience.
Having received objections from state attorneys general and others, U.S. District Judge Robin Rosenberg excised the offending chapter on climate change from the reference manual. This is a remedy that needs to be applied by the National Academies of Sciences, Engineering and Medicine, which still includes the chapter on its website.
In the meantime, as head of the Federal Judicial Center, the manual’s publisher, Judge Rosenberg should do the same with “How Science Works.” The judiciary and the public it serves deserve nothing less.
Sharon Camp, senior education advisor for the CO2 Coalition, Fairfax, Virginia, has a Ph.D. in analytical chemistry, Georgia Institute of Technology, and has worked for the U.S. Environmental Protection Agency and as an advanced placement science teacher.
Excellent work!
Jessica Weinkle has also written about this:
Good Points. Too many times I read or hear “scientists say” as though 2 or 10 authors in published paper speak for everyone in their field. Its not even the last word or a complete result from multiple points of view.
The basic purpose of the publication and its prominence …or not ( many are just junk science), is around the new research techniques not truth telling like a Judge might do.
Courts are an interesting in that they have evidence suppression built in. Many types of evidence are ruled by the judge as inadmissible sometimes for obvious reasons others on whim.
But at least its hearing from both sides not a one sided submission like published research
Not unlike questioning a hypotheses, challenging assumptions, and so forth.
>> perhaps they believe that computer models relying on unproven assumptions that fail to explain or predict outcomes
It is actually worse than that, as the assumptions and tunings are carefully chosen to produce the wanted output and additionally unwanted outcomes are given a low statistical weight after the modeling, this procedure contains two mathematical fallacies!
You indirectly point out that the use of statistics is also not robust science, in and of itself.
It’s a logical fallacy:
“Proof by authority” (or argument from authority/ argumentum ad verecundiam) is a logical fallacy where a claim is deemed true simply because an expert or authority figure supports it, rather than relying on evidence. While expert opinion can support a claim, it becomes fallacious when it replaces evidence, uses an irrelevant authority, or ignores expert consensus.
“proof by authority” might well be a fallacy yet it is also the only way to do science given how vast it is. If you want to do research in field X then in the amount of background material needed would mean that people would spend most of their career reproving what was already known. Things are in textbooks for a reason — people have given up trying to disprove them and so they count as consensus science. They might ultimately be proven wrong but people are actively looking to do so.
Real scientists don’t accept claims because someone said them or because a textbook printed them. They accept them because:
the underlying evidence has been repeatedly tested,the methods are transparent and reproducible,the claims have survived attempts to break them,the community has converged on them through criticism, not deferenceThat’s not authority; that’s earned reliability.
Science relies on distributed expertise and reproducible methods.
We trust consensus because it has survived scrutiny, not because it comes from authority.
Consensus is not science, it is politics. You always have to open to the possibility your premise is not entirely true.
Your first sentence – exactly what Dr S Fred Singer pointed out in his newsletter months before Naomi Oreskes came out with her anti-science 100% consensus study.
Your second – essentially what Lord Monckton pointed out right here at WUWT when he hit back at SkS’ John Cook’s idiotic notions about consensus opinion.
The question is then how does one know that “the community has converged on them through criticism”? Nobody has the time to check everything before doing research in an area. Practically people just read a textbook or two and they start from there. Science is built on trust.
If you care, you check.
The problem is that way to many climate scientists only care whether they paycheck clears.
Climate scientists have destroyed trust in “their” science.
Take anything said by a self-style “climate scientists” with a handful of salt.
Verification establishes trust. Acceptance because it is printed is not verification.
Izaak, you are either stupid or gullible.
“Things are in textbooks for a reason” is nonsensical waffling. Authors can write books about anything they like – they might claim that adding CO2 to air makes thermometers hotter!
I suppose you would be delusional enough to believe nonsense like that?
Accept reality.
That is not how it works.
Science is a framework for reasoning with provisional information. Conjectures are entertained for as long as they produce useful results, and no longer. When a conjecture that is the basis for further conjectures is shown to be incorrect it is discarded along with all further conjectures that depended on it. And for as long as a conjecture leads to further conjectures that are consistent with observation there is no justification for re-testing it.
That is not an appeal to authority; that is simply an efficient use of time and resources. Whether someone doing science knows that is how it works before they start—or finds out the hard way later—is irrelevant. They will know or find out. No authority required.
Science is a method, a philosophy, not an object; it doesn’t have size or vastness. You either apply the scientific method correctly or you don’t. The rest is just fooling yourself.
You’ve never taken any university physical science classes have you? It is that obvious.
Much of science is experimentation and replication to determine if a casual relationship is true or partly true. That is, given “x” I can reliably predict “y” using the functional relationship posed in the hypothesis (conjecture). If I can show by experiments that “x” does not reliably predict “y” , then the hypothesis is proven incorrect and changes are necessary.
Time series graphs of temperature will never prove a casual relationship since time is not a direct fundamental value in the linear temperature relationship.
That is why models will never supplant physical experiment. They only provide what their maker intends to be true. Why have the models not been proven to accurately predict real measurements? That is the question you must answer.
Jim,
you are missing the point. Science is experimentation and replication. But nobody can replicate all past experiments done in a particular field before attempting anything new. In order to do anything new you have to trust that information you have been taught was correct and start from there.
Blindly trusting without verification is not science. It is religion.
Nobody has to do that. When I learned about electromagnetic waves and Maxwell’s equations, I did not have to independently recreate all of his experiments to learn their usefulness. I didn’t have to independently recreate all the various gas law experiments to learn about them. Look up what experimental physicists do for a living.
You keep evidencing your inexperience with science and how it is done. When I design a circuit to perform a function, I generally research what has been done before. I may modify it, use better components, etc. I don’t start recreating Faraday, Ohm, and all the other pioneers work before proceeding. That is why you learn at university. Part of that learning is the ability to recognize abnormal results.
The story of Arno Penzias is illustrative. He worked at Bell Laboratories. He and his coworkers could not understand where the background noise in his microwave circuits was originating. It was an abnormal response. His search ended up in discovering background radiation from the Big Bang. He didn’t have to create the Big Bang in order to follow the result from the very beginning.
A better scientist that\n you or me has already pontificated on the relation between theory and experiment. I’ll let you research it so you might learn something.
“When I learned about electromagnetic waves”
Are the “electromagnetic waves” in the room with us right now, Jim?
“You’ve never taken any university physical science classes have you?”
When was the last time you took one, Jim? What do you think is the magnitude of the Poynting vector inside a closed box, for example?
Authority
Let’s just ignore consensus altogether.
From the Manual:
The Manual author uses a rather desperate example to support that dodgy statement viz. the discovery of Pluto.
If Uranus were observed to be dancing around the sky randomly that would have probably falsified Newtonian physics but minor deviations in the orbit merely prompted more intensive observations and discoveries that further confirmed Newton.
Maybe some fool rewrote something while the others weren’t looking?
As Einstein said “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.”
I doubt the authors know the difference between a hypothesis and a theory anyway. They certainly are ignorant of Feynman’s statement “It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are. If it doesn’t agree with experiment, it’s wrong.”
I prefer Einstein and Feynman’s observations to some anonymous American civil servants.
It doesn’t really matter. Even if some US politicians decide that pi should be 3.0 (it’s been tried in the past), Nature just laughs in their faces, and ignores their stupidity.
Nature can’t be fooled.
Good stuff, except for the claim that science has disproved eugenics. I’m not what that is supposed to mean. Eugenics is a practice (or range of practices), not a science or a conclusion. All it requires to be workable is selective breeding but can be strengthened by genetic engineering (still very primitive).
Look at the number of food crops that have been improved by eugenics, via cross breeding and selective breeding.
Suring the hominid evolution, eugenics played a big part. Traits that augmented species survivability were retained. The ability to outcompete other life forms for the resources needed for survival starts with birth rate.
It would appear that the authors of the guide are being very careful and precise. For example when they say that science cannot “disprove hypotheses,” what
what they mean is that there is always room for additional auxiliary hypotheses that allow you to claim that the original hypothesis was correct. The best example of this would be the claim that phlogiston had negative mass. That was only introduced after it was found that objects became heavier after being burnt and “releasing” phlogiston.
Similarly when they say that “In the context of testing hypotheses, the term prediction refers to a logical consequence of a hypothesis, not necessarily what will happen in the future.” they just mean that experiments don’t have to be carried out.
So the statement “Either objects of different weights dropped from the same height in a vacuum hit the ground at the same time or don’t. ” is missing the clause about whether or not somebody actually drops them. But the prediction is exists even if the experiment is never performed.
The Manual author is more specific saying that it is not possible to disprove an hypothesis by observations.
The essay mentions a few examples falsifying that statement, the history of science is replete with more.
Einstein said something about a paper with 100 authors, that it only takes one author to prove him wrong, and that 100 cannot prove him right.
Not quite, but still valid. He queried why it would take 100 others to prove him wrong when it would only take one.
Heliocentrism and epicycles prove you wrong.
Actually epicycles work. You can keep adding more and more epicycles to produce better and better predictions of the planets motion. Heliocentrism is simpler and easier to understand (and correct) but that is a different thing from being more accurate.
Mr. Walton: Since the authors are being “very careful and precise”, why do you feel compelled to further explain it?
If I considered, for a moment, that your CO2 cult might have an argument, your need to add “missing clauses” to what you describe as a “very careful and precise” 62-page “science” paper reminds me that the cult is just a distraction staffed by fools. Thanks.
The article said.
Your interpretation is totally incorrect as is the document provided to the court.
Science cannot disprove a hypothesis that has no definite casual relationship. But, neither can it be proven correct without knowing exactly how the outcome is determined.
That is why you see so many weasel words in climate science. Maybe, could, should, might.
Not inconsistent with…
Karl Popper was critical of ad hoc modification of a theory in the attempt to rescue it from falsification. He said it could be done, but it reduced the strength of that theory. The more elaborate the rescue, the less the theory can be considered valid.
If it doesn’t have to follow the established precepts of Science then it isn’t Science …
it’s … Climate-Science™
I think Popper and Feynman would certainly agree that this is totally cringeworthy.
>>STORY TIP<< https://notthebee.com/article/watch-former-climate-activist-and-aoc-supporter-explains-how-she-found-out-the-global-warming-narrative-was-all-bs
The carbon dioxide theory of global warming has failed, and failed in the way that Professor Stephen Hawking said:
“Any physical theory is always provisional, in the sense that it is only a hypothesis: you can never prove it. No matter how many times the results of experiments agree with some theory, you can never be sure that the next time the result will not contradict the theory. On the other hand you can disprove a theory by finding even a single observation that disagrees with the predictions of the theory.”
There have been many failed predictions of climate doom and all of them have failed to occur.
I agree with you and not trying to be nitpicky, but aren’t “failed predictions” by definition predictions that “have failed to occur.”?
He was a brilliant physicist, but he was mistaken about this. Or at least, its worded rather misleadingly.. The argument depends on not being clear about what ‘sure; means. In any reasonable sense you can be sure, if you’ve repeated an experiment lots of times, by lots of different people in different settings, that you are sure what another trial under any of these circumstances will bring. You may still find it worthwhile to try and find out if you get a different result under some other circumstances – many scientific advances come from such experiments.
Can you be sure that the next result will not contradict the theory? If its just repetition, yes. If has some relevant difference of method or circumstance, no. As in the above para.
Can you disprove a theory by finding a single observation that disagrees with the predictions of the theory? No, probably not without a lot more work. You must always wonder whether there is a flaw in the experiment design or equipment. The notion of ‘observation’ isn’t as clear as it seems. Almost all observations are theory laden. Generally for a well confirmed theory which has a long track record you are going to require more than one observation and quite a lot more enquiry to be sure. Think about it: would you really be ready to regard the theory that vaccination confers immunity on the basis of one trial? Done anywhere? By anyone?
AGW, an example, there seems to be a lot less warming than the theory predicts. But its not an intrinsically unreasonable approach to say there is extra heat, but its gone into the ocean. In itself that’s an interesting hypothesis.
The problem arises when you cannot find it in the ocean either! You keep looking and conclude that maybe there isn’t any to find.
If you’re a climate scientist you keep looking and conclude that maybe it’s gone somewhere else so climate change is real and even worse than we thought.
‘the chapter even says that science cannot “disprove hypotheses,”’
Well, the only hypotheses that science cannot in principle disprove are not hypotheses that are worthy of consideration in the scientific method.
Science is about the framing of FALSIFIABLE hypotheses i.e. those from which specific testable predictions can be made. If those predictions fail, then the hypothesis has failed.
If you cannot design an experiment which is capable of falsifying an hypothesis under certain experimental outcomes, then the ‘hypothesis’ is merely an opinion or an article of faith.
There is also the issue of reproducability.
They needed to get help from a philosopher, someone who does this for a living.
Its not that its impossible to falsify an hypothesis. Its that you can always rescue a falsified hypothesis by invoking another to supplement it. For instance, the circular orbit of the planets around the earth seems to be falsified by observations of their actual movement. But it was rescued by the invocation of epicycles, to explain apparent reverse motion.
It works in the sense that you end up with a theory which is consistent with the observations. But the problem is that at some point the rescuing hypotheses too need helping, and the theory gets too complicated for credibility when compared with a simpler one – elliptical orbits around the sun.
This applies even when you have two contradictory theories each of which gives a different prediction, and one’s prediction happens and the other#s does not. If you are committed enough to tinkering you can probably (in the short term at least) appear to rescue the falsified one. But again, what happens is that there get to be too many and too arbitrary and ad hoc epicyles
Popper was probably right in arguing that the way science actually progresses is by teasing out predictions of theories and seeking to falsify them – subjecting their predictions to observational testing. He was also right to argue that no amount of observations consistent with a theory can prove its truth – though they may increase our confidence that we are dealing with a well founded theory which has survived many crucial experiments without falsification,
But the chapter under discussion appears to have been written by people who have adopted and misused some philosophy of science arguments without really understanding them, and then tried to draw policy conclusions from them.
Quine-Duhem is irrelevant to policy making.
What you do not need when arguing about, for instance, the relative cost of wind and gas generation, is to have someone arguing that its possible that the apparent higher cost of wind is due to a mass optical illusion affecting only wind spreadsheets. Yes, it will rescue the hypothesis, just as you can rescue Creation by the argument that maybe the Creator took great pains on Creation Day to make sure there was a detailed apparent fossil record, so the fossil record does not support the theory of evolution. You could argue that the failure of repeated predictions of the Rapture is just due not to any problem with the theory, but just due to miscalculations of the date. This time its different, it really will happen this Thursday.
But why would you bother? And if it predicts the hot spot and there isn’t one, yes, you can dance around it with ad hoc explanations. And if, where did the missing heat go, same? But again, why bother when there is a simpler theory giving more useful predictions?
Dr.Camp ==> Absolutely marvelous. You have brought out the very dark heart of the matter — the CliSci nonsense, so very egregiously wrong — was the instant target and correctly removed — but the overall presentation of Science — also dangerously false in part — has been left in.
The details long-term ongoing failure of the scientific endeavor has been reported by the National Association of Scholars for years.in their Shifting Sands series of reports found here.
Keep up the great work at The CO2 Coalition — climate truth is winning the battle — albeit slowly.
There is an interesting example of how science actually works in a medical case, concerning prostate cancer.
It was established in the 1950s that there was some relationship between testosterone and PCa, and for many years after the discovery the standard treatment for PCa that had spread was hormonal. Surgical prstate excision for locally confined, after that move to androgen deprivation. At first the only available method was surgical castration, and you can still find some people alive today whose fathers were so treated. In fact, that seems to be a treatment which is having a modest revival in America due to side effects from Lupron, Casodex et al, and changes in Federal funding.
This went along with a natural view that once the cancer had spread there was no longer any point in local treatments. Why remove the cancerous gland, when the cancer had already spread to the bones? Surely that could make no difference.
It worked to some extent – it prolonged life and relieved symptoms in the short term, and either alone or when it was replaced with drug based androgen deprivation, basically chemical castration, it worked.
During this period testosterone was often spoken of as gasoline on the fire of PCa. The accepted theory was that simple. But then a couple of observations came up which make this theory seem at least incomplete and perhaps importantly wrong.
One is that men with very low testosterone tend to be more at risk of PCa than those with mid-range levels. Another is that men on androgen replacement or supplementation don’t seem more at risk. A third is that local treatment of the prostate, by for instance radiation, does seem beneficial in cases where its spread. All of these are inconsistent with the simplest versions of the testosterone theory.
Finally, there is work done by Johns Hopkins, in which very high levels of testosterone, way over normal levels, are intermittently administered to PCa patients. So called Bipolar Androgen Therapy. This doesn’t always work, but it a substantial number of cases does. The old theory cannot really explain this.
So the conclusion is that the relationship between testosterone and PCa is different from what we had thought and quite a bit more complicated.
It isn’t true that we just go around looking for a single observation, and if we find it, regard a theory as falsified. We are, thank goodness, a lot more sophisticated and intelligent than that in how we approach things. Nor is that we regard lots of trials as giving certainty. We are smarter than that too, we are careful about the range of circumstances which the confirmations we have actually cover.
Newtonian mechanics isn’t false, and was not refuted by Michelson Morley. It just holds in some but not all circumstances. Climate is the same. CO2 emissions almost certainly have an effect. The question is how what and how much. The chapter the post is complaining about appears to have been written by people who got their knowledge of science and the philosophy of science from journalism. Confronted with confirming observations the typical question is, in all or just some circs? And which? Confronted with falsifying ones, the typical question is, can it be reproduced, is it due to experiemental circs? And what does it really prove?
When an hypothesis fails the null hypothesis test, things are learned that lead to a (hopefully) more robust hypothesis. Not applying band aids, but a reassessment based on what is learned..
Science advances more building on what is learned from mistakes and falsified hypothesis than anything else.
Why can’t the courts do something useful. It would have made my school math classes much easier if they would declare pi to be equal to 3, exactly. And the value of e, that’s another mess.
It’s about time we got the courts involved to say what is true and what is not in the realm of science which is often so messy and not clear-cut.
There is an old story of a king who did not like pi and declared it should be 3.
The results were…. predictable.
Consensus = agreement of opinions.
The only valid consensus is, upon reviewing the results of multiple experiments performed in multiple locations by multiple investigators, the reviewers agree that none of the results to date disprove the hypothesis, thus establishing a level of confidence, but not proof, and the hypothesis is elevated to a theory.
In places across the USA, math has been modernized. It is claimed one does not have to get the right answer as long as one understands how it works.
I suggest we put students so trained in charge of the winds or airplanes and have the teachers take a flight.
Seriously, if you cannot get the right answer, how can you prove you know how it works?
This is your local automated government agent writing fuelled by the very latest AI engines.
This concept you have of ‘the right’ answer….
We fear you are in the grip of unconscious (at least we hope its unconscious) patriarchal gender denying white racialist class prejudice, and are probably transphobic, homophobic and Islamophobic with it, and a climate denier, which is a great source of regret and emotional disturbance to our automated sensibilities, so we will be sending round trained officials to help you stop disturbing our feelings and promoting social division in this way.
Please, until they arrive and have helped you, do not use the ‘r’ word, even to yourself as you are shaving. Do not use the ‘t’ word either. In fact, do not speak, even to yourself. You may want to pack while you are waiting for your visitors. Warm clothing is a good idea, and a strong pair of boots. You will be on an extended trip to somewhere quite a way further north than where you are now and will be striving with the workers and peasants….
Oh, sorry, that was another page from a previous era.
Anyway, you will be striving, and when you have well and truly strove (see how I almost fell down that hole?) you will have forgotten all of these unpleasant and divisive words, and the concepts they used to stand for.
Seems to be preaching scientism, not science i.e. believe what the high priesthood of card carrying party members say is the approved consensus, ignore what evidence, the scientific method, and dissenters say.
Thus the judiciary ‘must’ accept the Left’s climate science, trans science, Woke science … and reject those insisting on reality and logic.
You have to remember, very few people with ‘hard’ science backgrounds get into or go to law school, so by extension, most judges are a little overwhelmed when it comes to reviewing scientific evidence, when it included cause and effect. I saw this before I went to law school, when I assisted in court on technical matters as an expert witness on methods of manufacture.
In addition, I noticed how few engineers, or those trained in ‘hard’ sciences like physics or chemistry, actually got into law school, though there were many people who had biology or sociology backgrounds. The majority of law students may have had introductory science courses, but seem to have avoided higher level math, physics, chemistry or mechanics courses.
So why are we surprised that very few judges grasp to notion that science almost always remains theory, or understand that ‘the ability to demonstrate the falseness of a conclusion has long been recognized as the path to advancing scientific knowledge’?