Comment by matthewdgreen | Hacker News Reader

Comment by matthewdgreen | Hacker News Reader

Parent Full thread

I don’t think those signs indicate blind, unquestioning support for “scientific truths” as practiced by some mythical scientific priesthood. They are a response to a troubling political tendency that has made the evaluation of certain scientific findings into a question of political and tribal beliefs rather than one that involves proper skepticism, education and scientific investigation.

If the situation was that there was a strong evidence-based disagreement about scientific findings, those signs wouldn’t be necessary. The situation is instead that a vast number of scientists are producing findings that contradict certain political preferences, and the response has not been a deeper investigation made in good faith — but rather an outright political rejection of the findings. “I support science” is not shorthand for “I take every scientific study as fact because scientists are gods,” but rather “I am inclined to give weight to scientific arguments in either direction as opposed to political arguments that reject those findings based on non-scientific reasoning.”

the fault found in THAT line of reasoning is that the established scientific culture/institution doesn't even play ball by its own rules.

Political biases in every aspect from results interpretation to what questions we're allowed to ask, historical "just so" arguments that are sometimes falling flat (much to the detriment of the health of the planet), the lack of "negative results" studies due to funding biases and the replication crisis work together to cause a profound skepticism of the body of scientific institutional knowledge to be in some ways _more scientific_ than just reading and accepting the latest scientific news.

> The situation is instead that a vast number of scientists are producing findings that contradict certain political preferences, and the response has not been a deeper investigation made in good faith — but rather an outright political rejection of the findings.

How does a layperson make a deeper investigation on their own? The research is typically paywalled, the jargon is inaccessible, the experiments are prohibitively expensive, etc etc. In effect, there is a priestly caste, and one which arguably excludes from the debate members of the skeptic's tribe (i.e., "people the skeptic might trust to address their concerns"). This is why heterodoxy and political diversity in the academy is soooo important.

You might be inclined to say something like "conservatives select out of the academy" or "it's not my fault reality has a liberal bias" and you might even be right, but I don't think you'll find climate change very sympathetic to your position. In other words, this isn't affirmative action for conservatives, it's affirmative action (of conservatives) for the benefit of humanity.

Disclaimers: I'm picking 'conservatives' and 'liberals' here because that's clearly the bent of the article. I'm also picking climate change since it seems to be the favorite example of science denialism. Of course, contrary to the angle of this article, science denialism isn't even an exclusively conservative phenomenon (blank slatism is a favorite of anti-science leftists--including leftist academics--and flat-earth and anti-vax are bipartisan beliefs).

So I have some grad/phd student friends and a small computational lab - enough cores, ram, storage and jobbing engines that in 2000s it would have been a respectable not quite super computer but none the less very powerful system capable of crunching lots of numbers quickly.

My friends sometimes try to validate conclusions of the papers or just re-run the data crunching presented in those papers.

They never succeed. Either the data sets don't actually match the data sets in papers, or something in the formulas is off, or the code does not even compile.

I did computation physics in college. We had to write code as the paper of homework/tests. Our TA would never run our binaries. He would take our source code, compile it, feed it two of his data sets (we had one) and check the result. If the student's code did not compile, the student would get an F for that homework. If the student's code bombed on the data set that was shared with students, the student would automatically get an F. And only if the second data set got different results than he expected would he go chasing the errors and potentially still give a good grade. By the end of two semesters everyone knew one never tweaked the code to just run the known data set and get the known result.

Science would benefit tremendously if this was the approach.

>"Science would benefit tremendously if this was the approach."

This is just the original idea behind science as developed in the renaissance. The resistance to checking each others work is a recent phenomenon.

Edit to add a source:

> "the foundation of scientific societies compelled scientists to change the way in which experiments were presented so that not only members of the scientific community but also the inexperienced reader would be able to repeat them by following the same steps.

[...]

Robert Boyle also played an important role in the evolution of scientific writing (Gotti,1996, 2001; Hunter, 2007). Fulton (1932: 78) has highlighted the contribution of the scientist, describing Boyle’s undertaking “to establish science as an integral part of the intellectual life of ordinary men”. He encouraged scientists to write in a language that their contemporaries could understand, and he undoubtedly did more than anyone of his time to make scientific methodology part of the intellectual equipment of educated men.

[...]

According to Gotti, one of the features of experimental essays is brevity, but this brevity of form does not imply brevity in the exposition of the subject. On the contrary, experiments are usually reported in full, furnishing the reader with all the minutest details. The aim of such a report is, after all, to make it possible for the experiment to be repeated, making the precision of the description an important factor." https://doi.org/10.14198/raei.2007.20.06

Yeah, and sometimes the researchers are actively hostile to laypeople questioning their priestly caste.

For example, there's a fairly nasty and debilitating condition called CFS which respected British psychiatrists long ago decided was psychosomatic. So a few years ago, they ran a high-profile clinical trial called PACE which tried to prove this by testing their preferred psychiatric treatments, and naturally it concluded they were right. It was also fairly shoddy, which shouldn't surprise anyone on HN. In particular, the results in the final paper used a different and much weaker criteria for success than they'd intended to use at the start, which is a no-no. (There were a bunch of other issues with it too.)

A few trial patients and some outside scientists noticed this tried to get enough data to see what the results were if the researchers had done the scientifically valid thing and used their planned methodology, and that's when things got nasty. The researchers loudly cried to anyone who'd listen - and plenty did, because they were respected psychiatrists and the patients were officially nuts - that this was dangerous, vexatious harassment of scientists by activists, using their position to frame their pre-existing beliefs as science and questioning them as unscientific. All the proud, clever supporters of science bought into this and kept on doing so.

This was almost a decade ago now. The reason it springs to mind is that a few days ago, Reuters ran a long-form piece uncritically repeating the position that opposition to them is a dangerous attempt to silence scientists through harassment and abuse, and it made HN: https://news.ycombinator.com/item?id=19377986 Discussion ensued as to whether suing the evil people discrediting their research into oblivion was enough or whether they needed to be jailed, and why universities hadn't fired staff who disagreed with their results yet.

Oh, and the punchline to all this: the researchers were almost all male, the condition in question primarily affects women. It turns out that science, at least as experienced by poor disabled women, might be more accurately described by all those pesky postmodernist complaints from the women's studies department than it is by the scientific method.

1. You use sci-hub. Or write the researchers. Scientists don’t want to keep their work secret nor do they want a “priestly caste”, they want it out there and read. They work like dogs so people on HN can read and actually understand their work and not write about their political feelings instead.

2. Even if you’re not qualified to do the science yourself, you try to reject implausible ideas like the notion that there’s some grand scientific conspiracy against your preferred ideas. If you’re inclined to believe in some scientific result that is also the result that would be vastly preferred by industry and the richest people on earth, then you should be pretty confident that scientists who are willing to consider that outcome are getting resources to investigate it, and if those resources were producing credible results they’d be making waves in the scientific community. And if the scientific community was somehow censoring those credible results, we’d be hearing all about it.

Sci-Hub isn't exactly easy to use or recommended by the larger community a grad student is likely to encounter.

If you're smart enough to use HN's interface, you're smart enough to Google/copy/paste a DOI string into the sci-hub website. Not a criticism of you personally, but right now it's absurdly easy to bypass scientific paywalls if you're even slightly motivated.

PS Scientists almost uniformly hate the predatory scientific publishing system, want it destroyed, and aren't trying to prevent you from reading their work.

> try to reject implausible ideas like the notion that there’s some grand scientific conspiracy against your preferred ideas

You mean like academics who push blank slatism to prop up dubious political outcomes? Or people who push notions of racist math or sexist physics?

The questions are 1) how/why do our institutions allow these frauds to be so prevalent/powerful 2) how do we make the necessary changes 3) how do we help laypeople distinguish between ideologues and real academics?

Reply on news.ycombinator.com