Why "Science can be bought" isn't a valid argument

[TithonusSyndrome]

I know the title of this thread should be self-explanatory, but it deserves to be developed much more thoroughly in order to effectively rebut conspiracy theorists who reject peer reviewed publications from institutions with supposed links to undesirable organizations. I think that for the most part, it is rooted in a failure to comprehend that science isn't like government, or law, or business; this romantic notion that you can "follow the money" and uncover conflicts of interest has been seeded in pop culture ever since Watergate and All The President's Men, but scientific misconduct is fundamentally different than misconduct in disciplines governed by philosophical rather than natural principles.

I've roughly come up with a few of my objections to the notion that "science can be bought", so if you could help me expand on them for future reference in debates I'd gladly appreciate it.

1) If Scientists Wanted Money, They'd Have Majored In Something Else

I know that most conspiracy theorists are driven by an emotional need for simplicity in narratives, and "money rules the world" is as simplistic as it gets; but this dim quasi-Hobbesian view of human nature is wrongheaded from the get-go. There are a lot of other sources of fulfillment in life, and a lot of them are non-monetary; prestige, accomplishment, maybe even a need for understanding bordering on a form of "closure". Anyone who doesn't feel these or other motives, but has the academic potential to become a scientist, probably has a good shot at becoming a lawyer or business major. Right from square one, we've filtered out a lot of the mercenaries from the pack.

2) Too Many People Are Involved

Again, conspiracy theorists have a fairly pop-culture understanding of science, so I think they believe that a lot of scientific work is done by lone scientists working on projects all to themselves. Of course this couldn't be further from the truth, and the more people are involved in a project, the sooner the deception will be unveiled.

On top of this, we're assuming that all parties are striving to keep the conspiracy a secret; in reality, at least one of them will want the fame and glory of being a whistleblower, which is a standard flaw in all conspiracies.

3) The Risks Aren't Worth It

Science probably pays upper-middle class wages in general, but in order to make the risk of discovery worth the payout, you'd have to offer some pretty substantial enticement to your scientist bribes - enough to make the fact that they'd never work as a scientist again worth the risk. How much is an entire career's wages worth? A couple of million, multiplied by everyone associated with the project? These sums of money are compounded by the costs associated with concealing them, which as the Panama Papers revealed aren't as sure a bet anymore.

(That's also another good question - why haven't wikileaks or the Panama Papers turned up any bribed scientists?)

4) Individual Studies Can Be Dismissed As Outliers

So you've successfully bribed all the scientists in the study in question and got away with it. Well done.

Now you need to do this about another 5000 or so times in order to arrive at a consensus in favor of that. All of the complexities of said problems just literally exploded in time and money by several orders of magnitude.

This point, I feel, is the one that most scientifically illiterate people attempting to apply their "All The President's Men" logic fail to comprehend where they've gone wrong. Science needs replication, and law, business and government doesn't have much in the way of comparable processes. While it's true that there is something of a replication study deficit in much of the scientific community right now, at least that fact hasn't been obscured for the benefit of those who know what they're dealing with. Bribing a consensus into being would literally be the single most astonishing feat of social engineering in human history, even more extraordinary than concealing a September 11 conspiracy.

5) A Fraudulent Consensus Would Yield Visibly Deficient Science

Assuming it even got to this point (it never would) and you attempted to actually apply the science you'd confabulated, the buck would completely stop here. Nature would not and could not be bribed, and this would be evident in everyday life as people attempted to use products or equipment applying the pseudoscience. It just wouldn't fly.

That's not to say that bribery doesn't happen in science, of course, but there are checks and balances to catch it, and it certainly doesn't do much more than forestall the inevitable consensus. On the whole, though, the whole 'follow the money!" argument is one that I feel fails to comprehend what science is and what separates it from other human endeavors at the most basic level.

[Simon_Jester]

What can't realistically happen is for a major interest group to methodically bribe the scientific community to produce systematically fraudulent results.

But here's what CAN happen...

1) A wealthy interest group (e.g. the oil industry, pesticide manufacturers) can lavishly endow research groups that selectively pick people with 'good' records on an issue. These groups then have all the resources they need to locate whatever type of evidence best suits their case. They may not be able to dominate the overall scientific community's position on the issue, but they can muddy the waters enough to keep the general public from seeing a scientific consensus for what it is. This can easily result in scenarios where the science is well known, but utterly fails to have the logical public policy consequences. Ask Alyrium Denyrle about pesticides if you want an expert's example of this process.

2) When scientists rely for funding on for-profit corporations (e.g. the pharmaceutical industry), while the industry probably won't try to bribe the scientists to outright lie, it will tend to selectively fund work that it deems likely to benefit its bottom line. Thus, the funding available to research better pills to treat chronic heart conditions will tend to be far greater than the funding available to research better ways to get people to exercise and eat a healthy diet, even if the return-on-investment for the former approach would be superior.

3) In the social sciences, so much of one's results can hinge on one's starting assumptions, and it is so straightforward to fund one's research through the ideological commitment of supporters, that false concepts in social science can persist far longer than they would ever survive in, say, chemistry.

[TithonusSyndrome]

Yeah, these are all good observations inasmuch as they reveal what the maximum outermost violations of scientific ethics can realistically be, and how short they fall of the kind of nefarious bribery conspiracy nuts assume goes on.

[General Zod]

The timing of this thread is interesting.

http://arstechnica.com/science/2016/09/ ... r-decades/

Back in the 1960s, a sugar industry executive wrote fat checks to a group of Harvard researchers so that they’d downplay the links between sugar and heart disease in a prominent medical journal—and the researchers did it, according to historical documents reported Monday in the journal JAMA Internal Medicine.

One of those Harvard researchers went on to become the head of nutrition at the United States Department of Agriculture, where he set the stage for the federal government’s current dietary guidelines. All in all, the corrupted researchers and skewed scientific literature successfully helped draw attention away from the health risks of sweets and shift the blame solely to fats—for nearly five decades. The low-fat, high-sugar diets that health experts subsequently encouraged are now seen as a main driver of the current obesity epidemic.

The bitter revelations come from archived documents from the Sugar Research Foundation (now the Sugar Association), dug up by researchers at the University of California, San Francisco. Their dive into the old, sour affair highlights both the perils of trusting industry-sponsored research to inform policy and the importance of requiring scientists to disclose conflicts of interest—something that didn’t become the norm until years later. Perhaps most strikingly, it spotlights the concerning power of the sugar industry.

“These findings, our analysis, and current Sugar Association criticisms of evidence linking sucrose to cardiovascular disease suggest the industry may have a long history of influencing federal policy,” the authors concluded.

In a statement also issued today, the Sugar Association acknowledged that it “should have exercised greater transparency in all of its research activities.” However, the trade-group went on to question the UCSF researchers’ motives in digging up the issue and reframing the past events to “conveniently align with the currently trending anti-sugar narrative.”

The association also chastised the journal for publishing the historical analysis, which it implied was insignificant and sensationalist. “Most concerning is the growing use of headline-baiting articles to trump quality scientific research—we’re disappointed to see a journal of JAMA’s stature being drawn into this trend,” the association wrote.

But scientists disagree with that take. In an accompanying editorial, nutrition professor Marion Nestle of New York University argued that “this 50-year-old incident may seem like ancient history, but it is quite relevant, not least because it answers some questions germane to our current era.”

Both Nestle and the UCSF authors note that prior to the sugar industry’s involvement, nutrition research had pegged both sugars and fats as culprits behind coronary heart disease. Studies in the 1950s and early 1960s found evidence that low-fat diets high in sugars could boost cholesterol levels. The sugar industry had taken note of this, according to the historic documents. In 1964, the vice president and director of research for the Sugar Research Foundation (SRF), John Hickson, proposed that the group “embark on a major program” to dispute the data as well as any “negative attitudes toward sugar.”

Project 226

According the documents, the SRF enlisted Fredrick Stare, then chair of Harvard’s Nutrition Department, as a member of the trade group’s advisory board. Stare then put the SRF in touch with D. Mark Hegsted and Robert McGandy, members of Stare’s department. Hegsted would later go on to be the head of nutrition at the USDA. (All three researchers as well as Hickson are no longer alive.)

By 1965, the SRF funded “Project 226,” which would have Hegsted and McGandy—supervised by Stare—write a literature review that downplayed sugars’ role in heart disease and shifted blame solely to saturated fat. In return the researchers received a total of $6,500—the 2016 equivalent of $48,900.

During the write-up, which wasn’t published until 1967, the SRF’s Hickson was in frequent contact with the researchers, asking to review drafts and reminding them of the SRF’s interests. In one response to Hickson, Harvard’s Hegsted wrote, “We are well aware of your particular interest in carbohydrate and will cover this as well as we can.” After several delays in the writing, Hegsted reported to Hickson that they had to “rework a section in rebuttal” every time a new study came out supporting a link between sugar and elevated cholesterol levels.

In her editorial, Nestle concluded that “the documents leave little doubt that the intent of the industry-funded review was to reach a foregone conclusion.”

Hegsted and McGandy’s article, published as a two-part review in New England Journal of Medicine (NEJM), cherry-picked data, discounted studies that disputed its conclusion, and overstated the consistency of data suggesting that fat was the primary driver of heart disease. In conclusion, the review stated that there was “no doubt” that the only way to dodge heart disease was to reduce saturated fat.

The review made no mention of funding from the sugar industry. (NEJM didn’t start requiring authors to list conflicts of interest until 1984.)

After the review, the sugar industry continued to fund research into heart disease and other health issues. By the 1980s, few scientists focused on the role of sugar in heart disease. The 1980 Dietary Guidelines for Americans emphasized curbing fats and dietary cholesterol to prevent heart disease.

Today, Nestle points out, “the balance has shifted to less concern about fat and much greater concern about sugars.” But, the story should act as a cautionary tale of the potential harms from industry-sponsored studies.

[Ziggy Stardust]

I think that example fits rather nicely into 1) and 2) from Simon's post, from what details that article gives.

[Alyrium Denryle]

Ask Alyrium Denyrle about pesticides if you want an expert's example of this process.

*Energy crackles, a pentacle formed from lines comprised of double helixes cast in dark energy appears on the ground. A ball of radiant dark energy bursts into existence, and from its glare, a biologist appears*

I AM SUMMONED

Alright, where to even begin with this. Well, I suppose we can start with the regulatory process (US specific). A company wishing to register a pesticide is required to perform certain types of tests upon their submission. Mostly for efficacy of the active ingredient and formulation, human safety of said active ingredient, and environmental toxicity of the active ingredient.

You can see a hole right there. For some of those, only the active ingredient is tested--but not necessarily other "inactive" ingredients that might not be so inactive (sometimes they are tested, sometimes they are not. Moreover, safety trials are done for acute toxicity of dosage, not necessarily for long term or chronic toxicity. Or, for that matter, combined effects with other common pesticides.

Lets take two herbicides as an example. Roundup (active ingedient is Glyphosphate), and Atrazine.

Both passed all their tests and were registered for use in the United States.

Roundup is not too terrible on its own. Glyphosphate is harmless to pretty much everything but monocot plants. Cool. The problem is one of its "inactive" ingredients, a surfactant used to allow it to penetrate the cuticle of plants. That surfactant is lethal to amphibians on contact, in very low doses. But Roundup is still permitted for use in the US because once a pesticide is successfully registered and approved for use, it can be difficult to get that approval revoked. The other problem is that they are permitted to cherry pick what studies they submit to the USDA and EPA, so they can put their best food forward. Independent studies are not required, and the agencies cannot perform their own testing for logistical reasons.

Atrazine is another herbicide. Again, passed all its tests. However, later study (more on this later) indicated that with chronic exposure at very very low concentrations it disrupts the endocrine systems of amphibians, fish, birds, and mammals causing sex-reversal during development, as well as multi-generational increased cancer risk in mammals. Yes, you read that right. At ridiculously low concetrations (parts per million sort of low) the grandchildren of rats exposed have higher cancer risk. Worse, it evaporates into water droplets and can travel in weather systems to get rained out elsewhere. The EPA is finally reviewing its registration after over a decade of study.

Atrazine is actually a really good example of how a corporation can muddy the waters without ever actually inducing fraud or corruption (except in one person). They sponsored a number of scientists to study their product after concerns were raised. Most of them were just asking the wrong sorts of questions or were using bad methodologies. One of them didn't. When he published his findings, they cut off his grant funding, but he went on a bit of a crusade after that (it got personal when Sygenta started going after him in the press and trying to intimidate him with threatening phone calls) and found other funding. Other independent scientists followed suit, and the debate broke into two camps. Those funded by the manufacturer, and those who were not. Now, I have read all their studies. I would never have published the manufacturer funded studies because their methods were shit (bad operational definitions etc). So how did they get published? Well, it turns out that the journal they were submitting their papers to has a chief editor, that chief editor is on Sygenta's payroll. On their board of directors off and on, no less.

They just picked people they knew were not up to snuff on experimental design in environmental toxicology work, there were not that many repeats on the author list except for that aforementioned editor who also had a lab. They only needed to subvert the conscience of one person to pull that off.

[Lagmonster]

Short version: Absolutely, you can buy science, but the people who say so don't know how to say so intelligently, or measure the actual real-world consequences or their severity.

Shorter version: See Alyrium's post above.

Longer short version: Eventually, bad science will be outed. Absolutely. But even when it IS, you can employ money to ensure it doesn't matter. I say this as a guy who's rolodex is half Monsanto employees.

[Khaat]

In reading all of these, I'm left with "science can be mis-represented by money", not "science can be distorted by money."

The science isn't even being twisted: AD pointed out the "pro" science for Atrazine was crap methodology, crap parameters, i.e. bad science. Cherry-picked to present a unified front, in a world where public perception is acceptance, and everyone knows (but doesn't know) that advertising is selling you things you lived without forever before they told you you couldn't live without it.

[Q99]

Another thing: Oil companies have actively sought to buy scientists. As in, they put out a bounty for climate scientists to take their side of things. It hasn't been very successful.

[madd0ct0r]

In reading all of these, I'm left with "science can be mis-represented by money", not "science can be distorted by money."

The science isn't even being twisted: AD pointed out the "pro" science for Atrazine was crap methodology, crap parameters, i.e. bad science. Cherry-picked to present a unified front, in a world where public perception is acceptance, and everyone knows (but doesn't know) that advertising is selling you things you lived without forever before they told you you couldn't live without it.

Semantics. The scientific method includes and relies on the dissemention of knowedge.
Yes the published pro papers might be poorly executed science, but spotting the flaws of a method or assumptions requires you to know more about the science than the paper writer. Ie, be one of the top few thousand experts in the world on that very specific topic. I cannot think of any topic where i would be likely to spot anything subtle, including topics i have written papers on.

[Khaat]

The scientific method includes and relies on the dissemention of knowledge.

At it's root, I completely agree with that. On the wider interpretation where a good PR-team can make [bad things] commonly accepted as true, if they play to the uneducated well enough, I disagree.

The scientific method isn't about courting the plebes, it's about finding a better model, testing it, and making that available. The "making it available" part should never be a part of how the science is actually done. I recognize that financial support (or lack of it) can and does certainly skew what science actually gets done, though. And that publications can further skew who gets published (and thus read, thus peer-reviewed).

I count on engineers to watch engineers, medical professionals to watch medical professionals, etc. Popular opinion among the ignorant should* never be a part of the peer review/science process. An uninformed opinion is mostly worthless (which I accept could very well exclude me from weighing-in on this topic, since I am not a scientist in my professional life, nor do I publish findings or review scientific journals.)

*yeah, I know, "perfect world" and all.

[madd0ct0r]

Oh agreed, but what I'm getting at is that if a bought journal editor is knowingly allowing poor papers through, the chance of that wrong information being picked up and used to support something else goes up. ESPECIALLY when that something else is only related to the topic. For example, when I was writing and researching on energy model prediction for Bangladesh, I was using papers on biofuel cultivation from other countries in the Tropics. In one case, i had two papers that described the wood output in slightly different units from marginal land plantation in India and Bangladesh. Converted into the same units, there was an order of magnitude difference between the two of them. This was not noted in either paper.
I think I put the differenance down to poaching of trees, or early felling. My theory at the time was that a tree does not bulk up linearly with time so taking a year off a five year life represented a much larger loss of wood, but again, this was never discussed, well, anywhere, as far as I could establish.

[Khaat]

Wouldn't (doesn't) that crooked journal editor get outed by the process, though? Wasn't there a kerfuffle a few years ago where a journal wasn't reading some of the (nonsensical) papers they had published? Someone reading their journal brought this back up them?

It sounds as though the energy model prediction was based on several fields of study. It certainly would not simplify it for non-standard units to be used in comparison, though I don't know if I would ascribe this to intentional dishonesty, or just poor form/loose discipline (laziness). I recall reading about a space mission (satellite?) error where imperial measurements were used by one part of the team, metric on the other, no one caught it -> critical mission failure. But those directly involved will NEVER make that mistake again.

[madd0ct0r]

Don't get hung up on units. One paper was measuring the wood by volume, the other by weight. Both were reasonable choices for the specific focus of that paper.

It's where one subject brushes up against another. So in Alyrium's second example, the nasty Atrazine herbicide which at very very low concentration induces sex changes. The poor paper might use a bad definition of sex change - perhaps defining it as hermaphroditisim, when that particular species of frog mostly reacts to the herbicide by exhibiting female behaviours such as calls and building nests* and only produces hermaphrodites when the eggs are exposed to a much higher concentration. Ergo. that paper would report they only find effects at this much higher concentration.

Now, perhaps I'm a second researcher, and I'm studying the impact of a proposed dam on the environment downstream. The reduction of water in the watercourse would result in an increased concentration of various chemicals between flushes. I'm trying to establish the minimum amount of water the dam needs to release and how regularly to show no impact on the downstream riparian environment.
I search for a paper or two for each of the twenty or so agricultural chemicals I may be looking at. I need their established safe concentration, break down rate, and possibly sensitivity to breakdown - if it happens faster under sunlight, heat, cold ect. I need to establish out their sources, their concentration on fields or in subsidary watercourses, the mechanism by which they enter the river - is it surface wash off off the field, is it blown sprays, is it brought in by muddy cattle fording a larger stream? Does it bind to clay or sand or is it water soluble? If I have a good idea to the source, is it seasonal? Is it only during the harvest season, or is early spring after the fields are ploughed? Is there a certain intensity rainfall required to disturb the stream's sediment or generate surface runoff on a full field? How frequent is such rainfall?

Plus there's the actual hydrodynamics of stream bed wet surface friction, wave front propagation from flushing, long term armouring effects, local vector acceleration (things can diffuse faster than the stream speed), and critical stream velocities where behaviour changes non linearly.

Already, it's a horribly complex problem. I'd be forgiven for just pulling a paper for one factoid about one herbicide out of the twenty indicators I'm looking at. Would I get lucky and find an accurate one, or the other?




* i know very little about frogs. This might be apparent

[Khaat]

Units is just the first thing came to mind. I guess I think in terms of dimensions for my physical sciences.

Yes, multi-disciplinary information integration relies on what came before ("standing on the shoulders of giants"), and accurately making use of it. I can certainly see how using others' flawed preliminary conclusions as a starting point could undermine everything.

[Alyrium Denryle]

It's where one subject brushes up against another. So in Alyrium's second example, the nasty Atrazine herbicide which at very very low concentration induces sex changes. The poor paper might use a bad definition of sex change - perhaps defining it as hermaphroditisim, when that particular species of frog mostly reacts to the herbicide by exhibiting female behaviours such as calls and building nests* and only produces hermaphrodites when the eggs are exposed to a much higher concentration. Ergo. that paper would report they only find effects at this much higher concentration.

You are almost exactly dead on.

At lower concentrations, you get cellular or behavioral sterility. The testes dont produce viable sperm, or the frog acts like a female. At slightly higher concentrations you get... ovotestes. Testes that are half-ovary with non-viable sperm, and sometimes-functional ovulation. Higher concentrations you get full sex reversal. If parameters are poorly defined, you get bad results. If the experiments are not done through the whole of development, you get bad results. If someone is just not a particularly good anatomist, they wont notice the ovotestes. If they dont run genetic tests they wont see sex reversal either (it is not even easy to do the genetic tests with frogs, they dont have a separate sex chromosome, you have to genotype the loci used for sex determination)