"Breaking an Addiction To False Certainty Is as Hard as Breaking Any Other Addiction"

"But the first step is admitting you have a problem."

A nice line, I think, from a post at Astral Codex Ten; the rest of the post ("The Phrase 'No Evidence' Is A Red Flag For Bad Science Communication") strikes me as very good, too. An excerpt, though the article offers far more detail than that:

Science communicators are using the same term—"no evidence"—to mean:

1. This thing is super plausible, and honestly very likely true, but we haven't checked yet, so we can't be sure.
2. We have hard-and-fast evidence that this is false, stop repeating this easily debunked lie.

This is utterly corrosive to anybody trusting science journalism.

Thanks to Prof. Glenn Reynolds (InstaPundit) for the pointer.

Comments

[mydisplayname]

When does popular science become religion -- a matter of faith versus empirical evidence?

Eugenics is an example of science as an unacknowledged religion. Practiced by most academics of the 1920s through (at least) the mid-1950s, we still accept the religion of eugenics, established by Buck v. Bell, as part of American judicial thought. We do this knowing that eugenics is pseudo-science -- a mere matter of faith of a few perhaps well-meaning, but definitely misguided, adherents.

What prevents the establishment of another well-intentioned, yet potentially misguided, national religion: more specifically, do we reverse judgments when just one more booster, uniformly injected into every deserving soul, fails to vanquish the CoViD devil? More generally, what -- other than knowledge gained over many decades and many atrocities -- distinguishes the fight against degenerate offspring from the fight against CoViD?

[Dan Schiavetta]

Sorry if science doesn’t provide enough “certainty” for you.

[Sarcastr0]

Sometimes the prevailing science has been wrong.

That does not mean you should assume that the prevailing science is always wrong.

For instance, your choice of example makes me think you may have decided you don't like what science is saying and thus you have decided to turn up your skepticism.

[Jmaie]

Eugenics is simply the concept that traits are heritable, there is nothing "pseudo" about it. Modern morality has rendered the subject taboo, but hasn't disproven the idea.

[Jmaie]

This was meant as a response to mydispplayname...

[Brett Bellmore]

Indeed, eugenics is a really, really bad example of an 'unacknowledged religion'. It wasn't given up on because it was proven wrong, only because the Nazis gave it a really bad name.

[Krayt]

Eugenics is simply the concept that traits are heritable, there is nothing "pseudo" about it. Modern morality has rendered the subject taboo, but hasn't disproven the idea.

Heritability of traits is Darwinian evolution, as opposed to Lamarkism, where traits are developed through use, like a giraffe supposedly stretching his neck longer, and that feeding back into genes for inheritance.

Nobody says eugenics wouldn't "work". It is taboo because it is applied in ways where the applier is, curiously, always with the superior genes.

Slavery "works" in exactly the same way, as does a dictator killing his political enemies.

[TwelveInchPianist]

"Sometimes the prevailing science has been wrong.

That does not mean you should assume that the prevailing science is always wrong."

Now who is arguing that we should do that.

People generally use the appearance of a prevailing scientific consensus as one factor when choosing how to evaluate scientific claims.

I haven't seen any credible claims that this is the wrong approach, but boy do people get upset when others use that approach to come to conclusions that they disagree with.

[Sarcastr0]

Commenters here invoke eugenics all the time for why to disbelieve some scientific postulate or other.

Just like commenters here use thalidomide as an example of why not to trust the FDA.

Yes, it is very silly. Yes, people do it all the time. Yes, it is not exclusive to the right, but the attack on expertise sure is asymmetric between the two parties.

[Magister]

Since the US FDA did not approve thalidomide, that would be not just very silly but a completely incorrect example.

[Sarcastr0]

Yes, it was all the UK. But for those who want to attack expert institutions, any port in a storm.

Welcome to the Conspiracy comentariat.

[Rossami]

Oddly, I have never seen anyone here use the thalidomide example to attack the FDA. It appears that you're the one going for "any port in a storm". Come on, Sarcastr0, your arguments are usually better than this.

[WuzYoungOnceToo]

Come on, Sarcastr0, your arguments are usually better than this.

Given that his arguments are almost always straw men I must assume that by "better" you mean "less obvious in their dishonestly fallacious nature".

[Sarcastr0]

Do you recall it being invoked to attack the CDC? I mean, that is even less on point.

Thanks for the endorsements of my usual arguments, though!

[iowatwo]

but the attack on expertise sure is asymmetric between the two parties.

A not so subtle switch. From questioning science, (the process of constent testing(questioning). To the questioning of "expertise". Or, Trust me, I'm from the government, and I'm here to protect you.

[Sarcastr0]

Applies to both policymakers and scientists. And judges. And lawyers.

[Squirrelloid]

Consensus is not evidence. When I evaluate scientific claims, I prefer to go to the primary literature and/or data. Now, frequently a consensus is based on evidence, but sometimes that evidence is pretty tenuous. (Or a consensus is asserted where there isn't one).

Does no one remember 'there's no way Covid-19 originated from a lab' "consensus" from the start of the pandemic? Yeah, that consensus was at least premature, if not potentially wrong.

[David Nieporent]

When I evaluate scientific claims, I prefer to go to the primary literature and/or data.

I don't know you, so maybe that's true, but most people who claim that actually mean "YouTube videos," not "primarily literature and/or data."

[David Nieporent]

*primary.

Darn autocorrect.

[Brett Bellmore]

I recall reading earlier this year an article relating research into 'science skeptics'. And they were horrified to discover that the skeptics were actually quite often actually behaving rationally, and approaching science in a fairly sophisticated manner.

COVID Skeptics Don’t Just Need More Critical Thinking

[David Nieporent]

Typical Brett. Setting aside that the article did not show "skeptics" [sic] doing any such thing, it does not show anyone being "horrified" at all.

[Absaroka]

From the linked article:

"Well, I’m less of an expert on those groups, but just looking at this from a history of science or science and technology studies perspective, what stuns me are how all these groups use concepts from those academic disciplines. In one of these COVID skeptic groups, they were invoking Thomas Kuhn on scientific paradigm shifts! All this rhetoric about what expertise means, and the kinds of scientific knowledge that’s valid … they are all doing STS. Which is really interesting, and also horrifying"

(see last sentence; I put in the whole para for context. Don't know/care how much it support's Brett's thesis, but the lady being interviewed says she was horrified)

[Squirrelloid]

YouTube videos are not evidence. I mean the scientific literature and/or, when possible, the datasets published or referenced in the literature.

[jb]

"When I evaluate scientific claims, I prefer to go to the primary literature and/or data."

This is maybe fine at the margins--for a disputed claim you care a lot about where you have enough related knowledge to be able to make sense of the primary literature. But obviously it doesn't scale at all--there's a lot of science going on in the world, after all, and even a polymath isn't going to be able to make much of a dent in the stream of claims that are meaningfully relevant to their life.

The bigger problem is that when you're "doing your own research" it's extremely easy to engage in confirmation bias, even inadvertently. The normal scientific publication process with peer review is going to subject findings to a fair amount of skepticism and pressure testing, whereas armchair scientists tend to go looking for evidence to support their point of view and often misapply data in ways that a more rigorous process would prevent.

[Squirrelloid]

I'm only an 'armchair scientist' insofar as I'm not actively working as a scientist. I was educated and trained as a scientist.

And the degree of effort involved depends on the field of study and the nature of the literature. It's not that difficult to check the methods, statistics, and assumptions of key studies in many fields, and literature surveys can frequently point you to the most important studies. (You'd be surprised what makes it through peer review, sometimes, or just how limited some results actually are. There's a reason several fields have known reproducibility problems.)

But more importantly, I don't trust science journalists to actually understand the science they're writing about. I do trust scientists to explain what they did, what hypotheses were tested, and what they concluded based on that. I sometimes find reasons not to trust their conclusions, but the work is all there to evaluate.

(Obviously, I only check those things where my interest is raised sufficiently).

As far as confirmation bias, the scientists aren't exactly immune to several forms of bias. (Failure to adjust p-values for multiple comparisons, for example, is basically a type of p-hacking, and it's endemic in many literatures). A critical reader should approach all papers by trying to demolish them by finding methodological and statistical errors. That never gets you to a 'true' result, but can get you to 'i don't trust this result'.

[bernard11]

Now who is arguing that we should do that.

Lots of people, including commenters here, ISTM.

We have too many examples of arguing that "the experts can be wrong," without providing any reason why they might be wrong in the particular case.

[VinniUSMC]

Easy to say when your definition of expert repeatedly appears to be "people who say things that comport with my beliefs, because am an adherent of 'Science'!"

No evidence presented by any expert is taken as evidence presented by an expert, when you can just dismiss the expert as "obviously not an expert because what they say is wrong".

[Currentsitguy]

No, but I think that the term "settled science" is used to stifle legitimate debate.

[Sarcastr0]

The contours of legitimate debate (and thus the contours of settled science) is a difficult question.

As any scientist can tell you, there are cranks out there. I used to get e-mails from them even when I was in grad school.

But also the incentives in academia are to fight for your point of view, and a good way to do so is to argue the illegitimacy of countervailing papers.

So it's hard to get an objective read on where the contours of debate should lie.

Long way to say, you are right. But settled science is not always wrong either.

Specifically when it comes to climate science, the skeptics seem...highly motivated by external forces, whether partisanship or money.

[GKHoffman]

"Settled" science is really not settled. Every hypothesis may be falsified. Theories (i.e. hypotheses that many experiments have failed to prove false), even "laws," are subject to the great debunk. Newton's theory of gravity was settled law for 200+ years, then it got unsettled.

[Sarcastr0]

Yes, everyone knows this. But also extraordinary claims require extraordinary evidence.

[VinniUSMC]

Extraordinary claims, which upend hundreds of years of scientific theory, require irrefutable evidence.

I think this is what you mean to say:

Counter-claims, refuting the extraordinary claims of the Science! religion that go against decades of scientific understanding of virology and immunology require censorship and the most extraordinary of evidence, until the extraordinary claims of the Science! religion can no longer maintain themselves and fall like a house of cards, at which point, it is then okay to present the counter-claims as infinitesimally plausible, even though the Science! claim was thoroughly debunked.

[MatthewSlyfield]

"Settled science" is a unicorn, and any one who uses it in an argument is a anti-science moron.

[David Nieporent]

Maybe so, but I'm still not investing in your perpetual motion machine no matter how much you beg.

[TwelveInchPianist]

"But also the incentives in academia are to fight for your point of view, and a good way to do so is to argue the illegitimacy of countervailing papers."

Why has no one advanced in academia by pointing out that the academic consensus that false rape claims are rare is based on a misinterpretation of studies that find that the rate of confirmed false claims is low.

I mean, that seems like low-hanging fruit if people are interested in getting at the truth.

[Kazinski]

The problem with climate science is that it isn't really science, it's flawed computer modeling trying to simulate science.

There is some warming effect from CO2, I'll call that settled. But there is absolutely no settled science that pins it down any more than .25 - 4.5C TCS.

If there were any settled science that explained why we went from the medieval warming period, to the little ice age, and then the planet started warming over 200 years ago when it couldn't have been CO2 then I might be convinced.

But the view of the experts is we don't know what else could have caused the recent warming, so it must be anthropogenic isn't very convincing. Especially since their predicted signature that would nail it down, the tropospheric hot spot turns out not to be a thing.

In fact I'd just like to see an explanation of why we had a rapid rise in temperature 1910-40's, then cooling to the late 70's, now warming again.

[Stephen Lathrop]

If there were any settled science that explained why we went from the medieval warming period, to the little ice age, and then the planet started warming over 200 years ago when it couldn't have been CO2 then I might be convinced.

Why isn't that just a description of a warming trend interrupted by a cooling episode, and then continuing? If so, all you have to do is account for the cooling episode.

[Kazinski]

In the 30 years from 1910-1940 CO2 went up only 14 parts per million, 297-311 and we had rapid warming of .8c.

In the next 30 years 1940-1970 CO2 increased much more rapidly at more than double the previous rate to to 340ppm, yet we had cooling not warming of -.15 degrees, in fact it kept cooling until the late 70’s, but I wanted to keep it at consecutive 30 year intervals.

When they can adequately explain that, then maybe we can consider the science settled.

[Kazinski]

I had a typo above, the smithed trend from 1910-1940 was .63 Not .8, but that’s still almost half the warming we’ve seen since 1880, before CO2 could have been a significant factor.

You can check the data here:
https://www.ncdc.noaa.gov/cag/global/time-series/globe/land_ocean/12/11/1910-1940?filter=true&filterType=binomial

[Stephen Lathrop]

Kazinski, it would be easy to address all your suggestions with answers which would remain plausible, speculative, and unprovable. I think it has been a mistake to suggest climate modeling as a basis for policy, and will continue to be a mistake probably forever.

I doubt that evolution of climate will prove any more tractable a problem than the task of making economic theory accurately predictive. In both cases, every step in the process opens new vistas for influence by newly-created feedback loops.

I also don't think you need the models. Systematized observations, of physical changes, but particularly also of ecological changes, will be more than sufficient to supply evidence of the advantages of energy policy change. Getting it done quickly enough to avoid horrific public costs from continued bad policy will remain a problem.

[Absaroka]

"Why isn't that just a description of a warming trend interrupted by a cooling episode, and then continuing? If so, all you have to do is account for the cooling episode."

When you have a predictive model, whether it is a big computer model or a simple equation, you are positing that your model has captured the fundamental behavior of the system, and thus the predictions it makes about the future are valid.

Suppose you write a computer program to predict eclipses. And it predicts 4 out of 5 past eclipses. Should I trust it as predictive of future eclipses? Of course not - if your model accurately captures the physics of planetary motion it will give accurate dates for prior eclipses as well as future ones.

Saying 'well, it predicted 4 out of 5 past eclipses, so it's just a matter of accounting for the missed one' is trivially true, but glosses over that 'accounting for the missed one' may mean completely changing your model. E.g., prior to Kepler there were a lot of astronomical models that kinda sorta predicted planetary motions, but were fundamentally wrong.

So 'accounting for a cooling episode' might mean your model is mostly right, or it might mean you are missing something fundamental and important, and when you find what you missed all your predictions were wrong. Or not, and you give your model a minor tweak and now it also accounts for the previous anomalies without changing its predictions. There isn't a good way to know which is true until you do the work.

[Stephen Lathrop]

Nothing you say strikes me as mistaken. But Occam is on my side, I think.

[Absaroka]

Occam's razor says models that don't fit the existing data are likely to be predictive of future data?

Who knew!

[bernard11]

It can be, and sometimes is.

But OTOH there has to come a time when you either accept it or present really strong super duper quadruple plus evidence that it's wrong. You can't debate evolution forever, at least not as a scientific matter.

[Kevin Smith]

I think the issue is when people translate "settled science" to "settled public policy" Just because something is good science, doesn't mean public policy derived from it is necessarily good

This is where the eugenics example comes in. Genetics is the science, and it's good science. We know things like heart disease, cancer risk, and a host of other mental and physical disorders are genetic. This is all good, settled science. What doesn't follow is that eugenics is good public policy, even though it makes perfect sense based on the science

[Ben_]

"do we reverse judgments when just one more booster, uniformly injected into every deserving soul, fails to vanquish the CoViD devil?"

No, they always double down. Because, to most of the authorities, it was always about authority. And to most of their followers, it very quickly became about us versus them.

Authoritarians love us versus them dynamics because it’s easy to justify more-or-less anything by pointing at them. If they were only more like us then everything would be great, you see. But they will never be as good as us and the only way to a perfect utopia is to give the authorities more power to be used against them.

If you don't get yet another booster for an illness that presents no significant risk to you, you're them.

[Sarcastr0]

Even assuming your demonization of policymakers is correct, you are mistaking a policy decision with a research outcome.

[Ben_]

No I’m not. Policy makers mostly ignore research outcomes. That’s why we still have the TSA.

[Sarcastr0]

Policy is more than just the instantiation of scientific studies.

Good policy (i.e. independent of politics) is science + values + efficacy.

[Ben_]

And that leads to an easy way to understand why policy is generally not good: facts ignored (often selectively), values alien to the population, and usually no objective, holistic attempt to measure efficacy.

[Sarcastr0]

You have hit on ways policy can be bad.

Your decision that these apply to policies generally is just you being certain without evidence again.

Remember - to make a general point you need statistics, not just anecdotes that make you angry.

[Ben_]

No, opinions and observations do not require statistics.

It’s an Internet forum. Comments invite the reader to compare what they see to the comments' assertions.

Is policy good? If it is, then congressional approval should be high. Is congressional approval high? Is public satisfaction with government high?

[Sarcastr0]

The do require statistics if you want to base them on facts. Humans are actually pretty bad at statistics, and tend to overgeneralize a lot.

Is policy good? If it is, then congressional approval should be high.
We are not a direct democracy, because the Founders knew this kind of populist argument was nonsense. Thanks to our great Constitution, we are a careful and effective mix of populism and elitism.
Any pure appeal to one or the other misunderstands our government.

[Ben_]

No they don’t.

[Sarcastr0]

Not a fan of facts.

But then, we knew that.

[Ben_]

Facts don’t require a statistical citation.

[Sarcastr0]

Generalizations do. Going with your gut is a great way to be confident of a wrong thing.

[Hyman Rosen]

There's nothing wrong with eugenics. It got a bad name because, like Communism, it was implemented incorrectly. Eugenics is not, or should not be, "I think these people are yucky so I'm not going to let them have children." Many people would be happy if we eliminated Tay-Sachs or other genetic diseases by testing and offering women the chance to abort defective fetuses. (Not all people, of course. There are those who want to force the production of children with Down Syndrome, for example. Those people are evil and deserve to be executed.)

[WuzYoungOnceToo]

It got a bad name because, like Communism, it was implemented incorrectly.

Communism has been implemented in the only ways it can be, given the realities of human nature. Or are you defining "correctly" based on implementation by humans inhabiting some hypothetical fantasyland?

[DaivdBehar]

Volokh is the poster boy for that aphorism. He is a deluded indoctrinator of lawyer profession garbage doctrines, victimizing hundreds of intelligent, ethical students.

[DaivdBehar]

This lawyer dumbass believes minds can be read, in forecasting, that standards of conduct should be set a fictitious character with the personality of Mickey Mouse, i.e. overly cautions.

[dwb68]

There is no evidence that the phrase "no evidence" has any impact on the reader's conclusions regarding science.

[Lee Moore]

Sometimes "no evidence" is evidence :

Policeman : “Is there any point to which you would wish to draw my attention?”

Holmes : “To the curious incident of the dog in the night-time.”

Policeman : “The dog did nothing in the night-time.”

Holmes : “That was the curious incident"

A modern illustration would be Covid and the question of natural origin v lab leak. Whatever your priors, the absence of an animal host was initially almost irrelevant to the question, but each day that has gone by - now nearly two years - without finding an animal host adjusts the priors against a natural origin.

[Sarcastr0]

Yes, you can be deductive like that. But making such a technique dispositive is really a bad idea since we never have a full picture of reality.

[Intelligent Mr Toad]

The dog's failure to bark in the night time was evidence!

And we have several animal hosts. Bats, and feral cats.

[Sarcastr0]

Apparently, as the DNA studies get more and more granular, those hosts have been ruled out as the origin of the initial species jump.

[Bob from Ohio]

"initial species jump"

Is a research lab a "species"?

The lack of a host identified after two years likely means there was no initial species jump. Is not that the logical conclusion?

[Sarcastr0]

It absolutely makes it more likely yes..

But null evidence alone is not something upon which to a concrete positive conclusion should lie.

[Life of Brian]

Well, that last sentence didn't do much for the universe, but it sure did fill up my Bingo card.

[Sarcastr0]

'We're all out of hypotheses so this must be it' is bad science.

[Squirrelloid]

I agree with you. Sadly, the positive evidence is unlikely to see the light of day, because the CCP would find it inconvenient.

At which point, we're deciding between the likelihood of two scenarios, neither of which have positive evidence.

[Sarcastr0]

Also not great evidence - speculation about the nefarious motives of government institutions to cover up the True Facts.

[Squirrelloid]

It's not really speculation. We know China covers up stuff it finds embarassing.

[Sarcastr0]

I don't trust a word they say.

I don't, however, find that to be the same as support for the contrary hypothesis.

[Squirrelloid]

Of course not. I'm not saying a Chinese cover-up means one thing or another. I'm just saying solid positive evidence is becoming increasingly unlikely.

So, for argument's sake, you've got two possible scenarios:
-transmission to humans from wild animals
-escape from a lab

We haven't identified an animal host from whom the jump to humans occurred.

China will not allow an independent audit of the likely lab.

The lab removed their web-hosted viral DNA sequences from the internet near the start of the pandemic.

The closest viral relative we know of (last i knew) comes from ~2000 miles away.

That lab was engaged in gain of function research.

A zoonotic pathogen generally experiences rapid evolution on functional sequences like binding site proteins very early in the pandemic, as it adapts to its new host. We have no evidence of that kind of evolution. The earliest cases we know of already seem to be well-adapted to humans.

None of that is positive evidence. But it's all at least mildly suggestive negative evidence.

The only things that would qualify as indisputably positive evidence would either be locating the related virus in the immediate pre-human host, or discovering the virus's documented presence during an audit of the lab. The former hasn't happened (and at this point it seems unlikely to happen). The latter is likely impossible now (if the CCP was covering it up, they've had plenty of time to scrub all traces of the work), and hasn't been allowed to happen regardless.

Note: I'd expect the CCP to cover up any possibility of a link to the lab, even if they don't believe there is one. So the CCP's actions aren't dispositive, they mostly mean we lack an important avenue for gaining positive evidence.

[Lee Moore]

Sarcastro : I don't, however, find that to be the same as support for the contrary hypothesis.

This does not seem to be logical , Captain. Assuming we have two contrary hypotheses, then any evidence against hypothesis A is evidence for hypothesis B. Not conclusive evidence, obviously, but evidence.

And hiding the evidence you've got, while it might be simply a symptom of pathological secrecy, is much more likely if the evidence is unfavorable to you than if it is favorable. Thus, had the Chinese immediately and fully opened their records to view by Westerners, that would have shifted the Bayesian probabilities in favor of a zoonotic origin. For it would have made "lab leak" less likely. Consequently, since the Chinese did the opposite, they have shifted our priors in the other direction.

We might compare with, say, the Katyn forest massacre. The Nazis went to great, and most unusual trouble, to invite neutral observers to inspect the site, in the middle of a war. If they, rather than the Soviets, had been responsible, it would have been a great propaganda risk to allow access to people who might discover some of the truth. Consequently what they did was evidence, not conclusive evidence, but evidence, that they were not responsible.

[Lee Moore]

Is a research lab a "species"?

Mice are, and lab mice can be "humanised" for research purposes.

[IPLawyer]

"The lack of a host identified after two years likely means there was no initial species jump. Is not that the logical conclusion?"

And the evidence for a lab leak is???

[Lee Moore]

And the evidence for a lab leak is???

That it's the other main possibility apart from a natural hop from bats or an intermediate species. If we confine the other possibilities - say, "it was planted by aliens", "it is the wrath of God" etc - to say 1%, then 99% is shared between zoonotic and lab leak. The lower the probability of zoonotic, the higher the probability of lab leak.

It's a bit like trying to guess whether a coin came down heads or tails, when you have a highly restrcted view. Of course "on the edge" remains a very low possibllity, but each bit of evidence you get that reduces the probability it came down heads, increases the probability that it came down tails.

If you want positive evidence of lab leak then there isn't any - not public anyway. But then there never was for zoonotic either. Zoonotic evidence was basically "this kind of thing has happened before." But exactly the same thing can be said of a lab leak. They've happened many times before. And subject to confirmation, also since, as there seems even to have been a lab leak of Covid in Taiwan.

There's circumstantial evidence of course. It started in Wuhan. The Chinese took down their database of collected bat viruses. A proposal was made in 2018 to engineer a furin cleavage site into a bat coronavirus. But only the Chinese really know, and they're not telling what they know. Which of course is also a bit of "no evidence" evidence.

[Sarcastr0]

'That's the other main possibility' is the assumption required for lack of evidence to be dispositive by itself.

It's a heckuva assumption.

[Absaroka]

What causes other than lab leak and species jump do you think have more than a negligible likelihood?

[Sarcastr0]

I don't know, but it is also bad science to instantiate a failure of imagination as dispositive evidence.

[Lee Moore]

It would be if i was claiming it was dispositive. But I'm not. I'm just claiming it's evidence tending to increase the probability that the competing hypothesis is correct.

And - unless you're buying the Chinese frozen meat imported from abroad theory, or the deliberate bioweapon release theory, there are only two serious runners that have been offerred.

But if you are discomfited by the A or B framing, then feel free to conclude instead that evidence against A, on its own increases the probability that each of the other possibilities are correct. Thus even on this framing, the probability of B being correct increases.

If Tom Brady hurts his thumb, the probability of each of the other teams (who have not yet been knocked out) winning the Superbowl, goes up.

[Á àß äẞç ãþÇđ âÞ¢Đæ ǎB€Ðëf ảhf]

Exactly what he said: "Sometimes "no evidence" is evidence :".

[Sarcastr0]

It's evidence, with some limitations.

[Lee Moore]

It also depends on what exacly is meant by "no evidence".

You could say the Michelson-Morley experiment found "no evidence" of the aether. But they did make actual readings of the speed of light in different directions and found they were the same. So the measurements were evidence. After a while it gets a bit philosophical.

[Ben of Houston]

I will still defend that the aether is definitively proven, even if it unlike what we thought at the time.

General Relativity proved that space-time is SOMETHING. It can be bent. It can be compressed. It can be torn. The fabric of reality is just that, a fabric. The important part of the luminiferous aether is that light traveled through something, and our current understanding of space-time fills that role, even if the word itself is out of vogue.

I know, it's pedantic, but it's still something that bothers me when people claim that the aether was disproven.

[WuzYoungOnceToo]

This is correct (to the limits of precision afforded by conversational English). However...

The important part of the luminiferous aether is that light traveled through something, and our current understanding of space-time fills that role, even if the word itself is out of vogue.

This is not correct. The "Luminous Aether" was not just postulated to be an invisible medium through which light, behaving as a waves, was propagated. The hypothesis also required that the proposed medium be non-interactive with physical matter, which is most certainly NOT a property of spacetime, as you note yourself:

As you note...

General Relativity proved that space-time is SOMETHING. It can be bent. It can be compressed. It can be torn.

[not me]

Yes, the recent gravity wave experiments are essentially the old Michelson-Morley experiment, but with more precise equipment. Had they had modern equipment, but the same hypothesis, they would have detected gravitational waves, and used that to prove that the aether did exist.

[WuzYoungOnceToo]

Had they had modern equipment, but the same hypothesis, they would have detected gravitational waves, and used that to prove that the aether did exist.

I see you don't understand that whole "non-interactive with physical matter" thing either.

[Darth Chocolate]

One of the biggest problems is that the average person is ignorant when it comes to science. They hear someone with a fancy degree spout some dumbed down nonsense that sounds good so they believe it. It is simply an appeal to authority.

There is an issue when someone who actually questions the preferred narrative gets shouted down because they don’t “believe the science”.

That’s not how science works, but the average person is too stupid to be bothered about it.

[Sarcastr0]

Expertise is not science, but it is a thing that exists.

I tend to think that if you can't explain something, you don't properly understand it, but that doesn't mean I need to understand every journal article to believe something.

[Darth Chocolate]

Belief is not science. Scientists are skeptical.

And if you don’t bother to explore and understand ALL the data, both for and against, then you’re no better than a hack.

[TwelveInchPianist]

Well, there's no way people can personally observe all the data, make all the correct inferences from the data, etc.

You have to rely on the observations and inferences provided by other people and institutions. Other people aren't always reliable.

And choosing which other people to rely on is largely a non-scientific process.

That's why admonishments to "follow the science" are largely just bullshit.

[Sarcastr0]

Understand all data is not a practical demand unless you mean scientists within their discipline. Laypeople cannot understand all data on any issue, and scientists on any issue beyond their discipline. And scientists in their discipline are usually up on the relevant papers, at least the ones I talk to.

There is a further issue here, that's not exactly on point but interesting - these days data is often interpreted using proprietary algorithms no one else has access to (this is a real problem for how to keep publicly funded science open, but I also think universities have a right to keep their IP).

Back to your post, you also ignore how science actually marches on, and it's generally not paper-by-paper. It's the occasional seminal paper, and then surveys and longitudinal studies that lock in results.

[Jimmy the Dane]

HAHHAHAHAHAHAHHAH.....what a clown show you have become lately....

"Laypeople cannot understand all data on any issue..." Yeah everyone except the smart scientist could never understand data as related to an issue. What an elitist crock o' crap. If you just want to be a subject ruled by the cabal, just say so. The rest of us though will be happy to keep our critical thinking skills sharp and honed.

[Stephen Lathrop]

Just curious, Jimmy. If you had to come up with a one- or two-sentence description of critical thinking, what would you say?

[Jimmy the Dane]

The ability to see past liberal B$.

[Absaroka]

Seems like you ought to also include the ability to see past conservative BS.

(heh, will spare everyone the long story, but Jerry Falwell and the teletubbies came up a few days ago)

[David Nieporent]

Yeah everyone except the smart scientist could never understand data as related to an issue. What an elitist crock o' crap.

A perfect encapsulation of the Trumpkin worldview. But newsflash: being elite is a good thing. Some people are smarter than others. Some people (at all levels of intelligence) have more expertise than others (of similar intelligence). People who are smarter and have more expertise can better understand things than people who are dumber and have less expertise.

Unfortunately, Dunning-Kruger says that people like you don't even realize what you don't understand.

[WuzYoungOnceToo]

But newsflash: being elite is a good thing. Some people are smarter than others. Some people (at all levels of intelligence) have more expertise than others (of similar intelligence). People who are smarter and have more expertise can better understand things than people who are dumber and have less expertise.

Unfortunately, Dunning-Kruger says that people like you don't even realize what you don't understand.

What does Dunning-Kruger have to say about those who don't understand the difference between "being elite" and "being an elitist"?

[David Nieporent]

Elitist is just the pejorative term used to dismiss elites.

[Ben of Houston]

Sarcastro. I might not be an astrophysicist, but if someone says the moon is made of cheese and weighs five pounds, I can call them an idiot. I might not be a doctor, but I can read.

For example, the low-fat diet was formed on the back of contradictory and insufficient medical studies. However, political pressure from the heads of the American Heart Association made the low-fat low-cholesterol diet unquestionable in the medical field for years. You could literally not get published if your data did not agree.

However, the data didn't agree. This resulted in decades of decreasing health for Americans as we increased our carbohydrate consumption in the goal of reducing fats, and subsequent huge sales of statins to reduce arbitrary cholesterol numbers. It wasn't until the 2010s that this became acknowledged. However, this wasn't based on new studies as much as properly interpreting the previous studies, which prior generations had enshrined as gold standards, but actually didn't prove their point.

I would suggest the book "Big Fat Surprise" if you're interested.

Now, when I see that same adherence and political dominance happening again, I question it. Many "peer-reviewed" studies are of shockingly poor quality, and jumping to popular conclusions even when the data does not support it is rampant. So, please forgive my skepticism.

[Sarcastr0]

Oftentimes laypeople can understand the underlying science up to, but not including the bleeding edge. Because that's where our understanding is sufficiently incomplete.

But diet studies are especially bad. That whole complicated system is largely a black box, so they're all phenomenological groping in the dark.
If you want the counter to your example of institutional assumptions, look at some of the Atkins folks' books. Same narrative - institutionally shut out by the consensus about cholesterol. But in service of a more dubious dietary postulate.

[Union of Concerned Socks]

This.

[Jimmy the Dane]

"No evidence" is a great "wag the dog" type line.

"Although there is little to no evidence to support the assertion, we are going to call what happened on January 6th an Insurrection because it fits nicely into our leftist world view..."

"No evidence supports many of the public health initiatives being used to combat Covid, but we are going to censor and fact check anyone who says anything to the contrary..."

"Evidence is racist and hence no evidence is needed to assert that we live in a white supremacist based society..."

Need more examples, just put "no evidence" into a news search....

[MollyGodiva]

Except that there is evidence that Jan 6 was an auto-coup attempt. There is evidence that public health measures, especially vaccines, to help combat covid. And there is evidence of widespread racism.

[Bob from Ohio]

"auto-coup"

All the cool kids use the Spanish term

Nonsense in any language though.

[MollyGodiva]

Self-coup, autocoup, or autogolpe. The terms do accurately describe what Trump and his supporters attempted. He was voted out and he tried to use illegal methods to remain in power. What I do not understand is how so many people who claim to lover our country and constitution are ok with what Trump and his supporters did.

[Jmaie]

Hopefully you realize it's possible to deplore what went on at the Capitol January 6th without thinking it a serious threat to the American way of life.

[Stephen Lathrop]

Too soon.

[Social Justice is neither]

Fuck off you marxist scum. You were one of the people championing the riots of the summer and aghast at any law enforcement against the rioters that destroyed billions in property, killed dozens of people and actually seized US territory. There is your insurrection.

[David Nieporent]

Sure, pollyannaism is a real phenomenon.

[Jmaie]

As insurrections go it was pathetic. We have survived and not by the skin of our teeth. No pitched battles between government forces and anti-government militias, as example. Biden sworn in, perpetrators being punished.

Bad, yes. But as my grandmother used to say, "No need to get your bowels in an uproar."

[WuzYoungOnceToo]

Sure, pollyannaism is a real phenomenon.

"Hands up, don't shoot!"

Right...?

[Jimmy the Dane]

It wasn't a coup. It was some tourists using their God given right to petition the government. Had the institutional left not engaged in the unprecedented censorship campaign that was the post-election attempt to cover up election irregularities, then I don't think these citizens would have had to resort to actually coming en masse to the Capital and making sure their collective voice was heard.

[bernard11]

You are deranged.

[Jimmy the Dane]

That isn't much of an argument, other then to let people know that you are just plain ignorant.

[Stephen Lathrop]

They came to the Capitol with a battering ram, Jimmy. If only the ones using violence to gain entry had been shot down, there would have been at least two hundred corpses on the Capitol steps.

[Jimmy the Dane]

There was no "battering ram"....Just like there was no "insurrection".

And, yes, I guess if you go shooting innocent UNARMED protesters when there are a lot of them at a protest you might get a high body count. Not sure what that proves other than you seem to be advocating the shooting of unarmed, mostly peaceful protesters. I certainly hope that is not what you are saying should have been done because that is even more extreme than anything AK has ever said around these parts....

[Absaroka]

1)SL: if this is the battering ram, they don't make battering rams like they used to (towards the end). Most people call that a cardboard shipping tube.

(that was the first video I found searching for 'january 6 battering ram; if there was another one more medieval siege like, please share!)

JD: That ain't peaceful, unarmed protesters. Inter alia, there's a baseball bat. And people who are going to a peaceful protest don't bring shields. It may not have be all the way up to Portland levels of violence - I didn't see any power tools for example - but it sure as heck is a violent riot.

[Brett Bellmore]

Basically what happened was that most of the people came to protest, a tiny minority came to riot, and the people who arrived at the Capitol after it had been opened up didn't even know they weren't supposed to go in.

[Stephen Lathrop]

Absaroka, I couldn't get through your link to take a look. What I saw was during news coverage. It was cylindrical, color sort of blue-gray, maybe 10 inches in diameter, looked about 8 feet long, wielded horizontally at waist height by multiple people. It did look kind of medieval. But I only saw it for a few seconds, and can't say it wasn't a really big shipping tube, if that's what you saw.

[WuzYoungOnceToo]

They came to the Capitol with a battering ram

No, they' didn't...you moron.

[David Nieporent]

Since you're just trolling, there's no need for much of an argument to refute you.

[Sarcastr0]

A lot of these are mixed question of fact and opinion. It is a common tactic of a zealot to declare it all fact, so everything can become a right/wrong binary.

[Harvey Mosley]

At least you acknowledge you're either "right" or you're wrong. 🙂

[David Nieporent]

"Although there is little to no evidence to support the assertion, we are going to call what happened on January 6th an Insurrection because it fits nicely into our leftist world view..."

I mean, no evidence other than actual video of the events, and the actual words of the people involved before, during, and after those events.

[VinniUSMC]

... other than actual video of the events, and the actual words of the people involved before, during, and after those events.

Evidence of an insurrection? Okay, sure thing, Stretch.

[Ben_]

Meanwhile people here say "no evidence" as a talking point meaning they disagree and haven't been conclusively and overwhelmingly proven wrong yet, so therefore there's "no evidence" of the thing they disagree with, regardless of the actual definition of the word evidence and how much or little evidence there is.

[Ben_]

Modern usage of "science" should be understood to mean that someone has been designated an authority (by people with a political agenda) and you have a duty to listen to them and obey them, regardless of facts or reality. And regardless of whether the figure holding the "science" flag is sane or homicidal. And regardless of who may be harmed. Authority is what matters. Nothing else.

If you aren't enthusiastic to intentionally hurt people for "science", you will be labeled "anti-science" by authoritarians.

[Jmaie]

Science be praised...

[Jimmy the Dane]

Yeah when you bring up global warming (err....the coming ice age....err.....climate change....err whatever they are calling it now) the standard leftist response is "no peer reviewed scientist agrees with that..."

Translation: No person who I personally agree is legitimate in the field of science, regardless of their actual academic credentials, who agrees with you has a legitimate opinion because it runs contrary to my deeply held political belief.

[Ben_]

If something can make life worse for Americans in the name of The Earth, it has a good chance of being called "science".

[Sarcastr0]

Congrats on your false certainty even without any science to back it up.

[Ben_]

Explain the relevance of that comment. Who is certain? About what? What are the indications of certainty? What are the indications of falsehood?

Please answer or else everyone will understand you just talk out of your ass.

[Jimmy the Dane]

It is only "false" because the clown show has drank the kool-aid and is now in double denial that everything he believes in really bunk.

[Ben_]

I just describe what I see and offer a way to understand it. If someone could communicate a different understanding that better fit the events, that would be good. Then my understanding might improve.

But the usage of "science" as a flag to wave to designate authority is clearly a thing that happens.

[VinniUSMC]

Congrats on your false certainty even without any science to back it up.

As Jimmy wrote:

"Translation: No person who I personally agree is legitimate in the field of science, regardless of their actual academic credentials, who agrees with you has a legitimate opinion because it runs contrary to my deeply held political belief."

[MollyGodiva]

The author challenges the reader to define “no evidence”.

“No evidence” means that at this time there is no scientific data to show that the assertion is true, so the best we can do is use related data to evaluate the truth of the assets.

[TangoDelta]

The problem there is that journalists, who have no related data, simply spout blind assertions. The bigger problem is when there is "no evidence" of A they typically fail to point out there is also "no evidence" of !A because it hasn't been studied. Of course that doesn't make for splashy headlines.

It's also important to point out that medical scientists aren't necessarily well informed about other sciences. For instance the early statements saying there was no evidence for airborne transmission of Covid are looking only at the medicine side. Had they take a more physics based perspective and looked at simple Brownian motion they probably would have come to a different conclusion from just looking at ballistic sneeze and cough transportation of droplets.

[MollyGodiva]

The confusion about airborne transmission has a far more complicated and interesting explanation.

https://www.wired.com/story/the-teeny-tiny-scientific-screwup-that-helped-covid-kill/

[Absaroka]

I remember reading that, it's a great article.

This part gets missed a lot: " Li convinced the administrators at the University of Hong Kong to spend most of its Covid-19 budget on upgrading the ventilation in buildings and buses rather than on things such as mass Covid testing of students. Marr reviewed blueprints and HVAC schematics with the owner of her gym, calculating the ventilation rates and consulting on a redesign that moved workout stations outside and near doors that were kept permanently open. To date, no one has caught Covid at the gym."

We have endless heated debates about masks, but not a lot of attention to HVAC. Why aren't parents at school board meetings to talk about HVAC?

[MollyGodiva]

HVAC and ventilation issues are at topic the requires a level of understanding and nuance that is completely lost on the hysterical parents at board meetings. Also the cost would be very high.

[Absaroka]

I dunno. 'How many air exchanges per hour' isn't exactly rocket science. High MERV filters aren't that expensive.

[MollyGodiva]

It is likely that for many schools the air exchange per hour is not sufficient and the cost to fix that would be high.

[Ben of Houston]

Molly is right. HVAC is neither cheap nor easy to retrofit, and schools are notorious for cheaping out on it to begin with. It's definitely something to think about in future construction (if nothing else, it might reduce flu and cold spread, and COVID is endemic now), but it's not realistic in the near term.

[Absaroka]

High MERV filters must be a lot more expensive in Houston than around here 🙂

(also, have you tracked what your local schools spent on distance learning?)

[Toranth]

In statistics, which is what most scientists claim to be using to perform their analyses, the meaning of the phrase is pretty consistent: You failed to reject the null hypothesis in this specific test (performed under these specific assumptions).
That's it.

If you go on to use different data, you haven't changed anything about the first analysis - it's still "no evidence".

Does it on test provide irrefutable proof (either way)? Nope! xkcd's green jelly beans comic is a great example of why - and also another example of why science journalism sucks.

However, repeated experiments with similar results do provide a great big flashing sign.

[Stephen Lathrop]

Science journalism and other journalism ought to be judged on similar bases. It starts by evaluating the quality and reputation of the journal. Then, as with all consequential stories, readers should look for journalist authors who show a body of work suggesting long-term engagement with the subject matter, and a successful history writing about it.

For specialized subject matter, it can be useful to seek authors with educational credentials that match them to the subject matter—lawyer journalists for legal stories, for instance; technically educated journalists for science stories. A problem in the latter instance is that education in technical habits of thought may prove journalistically helpful, but subject matter encompassed under the term "science," is so broad that the journalist himself is unlikely to match closely the exact credentials of the practitioners on whom a story may focus. In such cases, readers should look for journalists who do show technical habits of thought (however acquired), who then seek out specialists among scientists to communicate particulars during interviews.

The interviews should be reported in sufficient detail to enable readers of above-average intelligence to evaluate how the scientific processes were explored, the kinds of evidence gathered, and how both processes and evidence were communicated. For those familiar with it, the Richard Rhodes' book, The Making of the Atomic Bomb, examples that journalistic method used successfully across a broad scope. Another such book, by John McPhee, is, Annals of the Former World.

Such stories can profitably be presented as accounts relating how the information featured came to the attention of the sources reporting it—from two points of view. First, from the point of view of the scientists doing the work, and second from the point of view of the journalist collating the scientists' accounts.

Those are field marks of science journalism reported as reliably as it is practical to do it. Note that none of that can guarantee the kind of reliability which motivated science skeptics demand—science reported to a practical certainty. Science is in the business of building and critiquing systems to strive after practical certainty, not in the business of delivering it, and then closing up shop.

The certainty demand itself betrays ignorance of science so benighted that it tends to put constructive engagement beyond reach. A demand for certainty delivers willy-nilly a paradoxical demand to suspend process. That problem is not solvable by scientists, nor can it be much ameliorated by journalists, however conscientious.

[m4019597]

Evaluation of journal articles must include follow up after a year. So much gets retracted or edited. So much turns out to be fraud.

Much of the studies on which Covid policies were formed last year have turned out to be false or fraud. Including vaccine and mask policies.

Entrenched power interests are already invested in wrong conclusions. That’s why with omicron they’re doubling down on what doesn’t work. Their power is at stake.

They ignore what’s been shown to be false and retracted or disproven.

[DStraws]

Cites as to all the retractions, false and disproven--talk is cheap.

[m4019597]

Retraction Watch maintains a page of Covid related retractions and corrections. It’s hundreds of studies long.

Google it, moron.

[Sarcastr0]

Out of how many total papers?

Raw number of studies is a bunk metric.

[benji]

They weren't retracted (because they were valid at the time) but the replication crisis is a growing thing in a lot of fields.

[Michael P]

Those are just two reasonably good-faith uses of the term. Another use, arguably more common than the first but closely related to it, is to mean "I don't like the evidence that the other person presented". This is particularly common when discussing claims made by Donald Trump.

[tkamenick]

You forgot "3. There's plenty of evidence but we don't like it"

[Sarcastr0]

The science discussion is fun and all, but the big point is about humility in the face of those who disagree with you.

Your certainty is not the same as your correctness.

Lots of disagreement comes from speculation out of a set of facts. Like whether leaving off a period counts as octoring evidence.

That doesn't mean lay flat, it means make arguments. And do some critical thinking about why you think what you think; what assumptions are involved?

And good lord before you slam that link into the comment field, Google a bit to see if anyone has anything to say about it. Like how it's wrong, for instance.

[VinniUSMC]

I can picture that conversation:

Sarc: "Hey Google, does CNN say *insert thing I disagree with* is wrong?"

Google showing CNN results: "Yes, Sarc, your confirmation biases are totes correct."

Sarc: "I knew it!"

[Sarcastr0]

Great speculation.

But in a real conversation you can just iterate on the countervailing evidence.

Declaring victory and disengaging is bad form. It is also not something I do very often, sometimes to a fault.

[Artifex]

The science discussion is fun and all, but the big point is about humility in the face of those who disagree with you.

Nope, the big point is that the sarcastrian rhetoric around "No Evidence" is corrosive to honest conversation and is a red flag to those who are really trying to understand the situation.

[Sarcastr0]

Demanding evidence of factual claims is not 'corrosive to honest conversation.'

Seems to me asking for evidence shows I am trying to understand the situation.

What do you use to understand stuff, if not evidence?

[Artifex]

The article is not about demanding evidence of factual claims. The article is about using the phrase "No Evidence" as a form of motte and bailey with different meaning depending on which narrative is being pushed and noting that those dishonest word games are corrosive to honest conversation.

[Sarcastr0]

I took it as about false certainty, myself.

[Artifex]

Yes, you read the headline. It helps to actually read the linked article.

[Cal Cetín]

I hate how the other side is so rigid. Why can't they be principled like me?

[Moderation4ever]

I would suggest that another tricky word in science is "theory". A scientist will use lower case "theory" for an unknown idea that will be tested and uppercase "Theory" for an established idea, like "Theory of Evolution" or "Theory of Gravity". Both are well established scientific principles that still bear "Theory" in title.

[WuzYoungOnceToo]

A scientist will use lower case "theory" for an unknown idea that will be tested

First off, a scientist wouldn't refer to an "unknown idea" at all, because it's unknown. Secondly, a scientific idea that he/she does know about that is as of yet untested will be referred to as a "hypothsis".

"Theory" for an established idea, like "Theory of Evolution" or "Theory of Gravity". Both are well established scientific principles that still bear "Theory" in title.

Uh, an untested scientific idea (hypothesis) is most certainly NOT an "well established scientific principle".

You're one of those clowns who routinely spouts how much you love science and presumes to "educate" others on it while simultaneously demonstrating that you're profoundly ignorant about even the most basic scientific concepts.

[Cal Cetín]

RIP

https://www.findagrave.com/memorial/23903480/max-karl_ernst_ludwig-planck

[Sarcastr0]

Latest CDC data by vaccine status:

Unvaccinated: 451 cases per 100k
Vaccinated: 134 cases per 100k
Boosted: 48 cases per 100k

Unvaccinated: 6.1 deaths per 100k
Vaccinated: 0.5 deaths per 100k
Boosted: 0.1 deaths per 100k