Hacker News Reader

Comment by seibelj | Hacker News Reader

Even in hard sciences, there are papers that everyone in the niche knows is false, but fear to speak out about or publish a paper refuting it for fear of their career being damaged.

My wife is a biochemist, and during her PhD they were working on an area of research with a handful of labs publishing about it. One lab in particular was known for a decent amount of questionable publications, but the PI was a big deal and no one would officially question anything. So they would whisper amongst each other and just ignore that paper (and all the additional papers built on top of it) because they all knew it was bullshit.

View on HN

Wow, that's a clear sign science has become too political. Proving other scientists wrong used to be how careers were made, not something to fear.

All the labs working on this spoke to each other, collaborated occasionally, and saw each other at conferences. So making a big stink about this wouldn’t have made life easier. Plus peer reviewers for submitted papers are “hidden” but because the field is so niche they are likely people you know, and they know who you are when you publish (reviews aren’t double-blind), so it could hurt your career if the PI became angry at you and started bad-mouthing you to to the other labs.

I understand, but science isn't supposed to make life easier, it's supposed to progress mankind's knowledge and that necessarily entails eliminating bad information from the body of knowledge. Avoiding challenging bad papers because someone you know wrote it and you're worried they might not like you for it; well, yuk.

No, something like this can spiral: they can write bad reviews about your work, tanking your publication rate, which makes it harder to find funding etc...

You've missed the point: getting along with your peers is the most important thing. What is truth?

No I haven't...

> getting along with your peers is the most important thing

No it isn't, hence the yuk. Anyone who thinks getting along with your peers is more important than getting the science right, isn't someone I trust doing science. The most important thing is getting the science right, full stop, end of story.

The problem appears to be is that anyone who can't "get along with their peers", or otherwise "social game", is not getting close enough to attempt to do any kind of science/serious project/study in the first place.

So you would argue that Galileo Galilei was a bad scientist? He was imprisoned for disagreeing with the pope. I suppose you'd have to argue that a good scientist in the middle ages had to be in good standing with the church and monarchy?

Galileo only needed basic tools (like feathers, balls, an inclined plane, and a handheld telescope) to pursue science.

If he'd needed access to impossibly expensive equipment available only in a few palaces around Europe, politics might have stopped him before he ever got started.

You can't predict what the next great discovery will be, or where it'll come from. Generally however, they're on the theoretical side, not the experimental side, so your point doesn't really stand. Those big machines rarely do little more than confirm exiting theories and those theories can very well come from one person who is anti-social.

Science is about being right; not about getting along with your peers.

What he actually did was closer to calling the pope a moron.

Science has always been political, within itself and in relation to the rest of society.

https://en.m.wikipedia.org/wiki/Oil_drop_experiment

Obviously, hence the word "too".

Your link is not a story about politics, it's a story about psychological bias.

Just because Feynman diagnosed the gradual charge drift as internal pressure doesn't make it internal pressure. The effect would be the same between fearing one was wrong and fearing one would be condemned and punished as wrong - changing the data to be closer to the official line.

Authority is always political.

It's not a story about authority, either, if by authority you mean something narrow enough to always be political.

If you don't think Millikan's authority as a scientist was not a factor then ¯\_(ツ)_/¯

That's a broad enough definition of authority that all authority no longer originates from politics. When scientists are behaving correctly authority is awarded on scientific merit irrespective of politics.

> When scientists are behaving correctly authority is awarded on scientific merit irrespective of politics.

I suppose then one can conclude that scientists have behaved incorrectly since Aristotle's time.

Political has multiple meanings in English, and I think this use is legitimate.

This could be just anchoring bias. But I think it's much more likely that people after Millikan did not want to look bad in front of their professional peers for fear of social and professional consequences. Which I think is something we can reasonably call political.

But I think it's more likely that the bias was internal, in which case it could not be reasonably called political. Without arguing further we can agree that the parent's claim was unfounded, because the linked article did not demonstrate that the motivation was social.

(The reason I know we can agree is that unfounded is not the same as false, it just means it's a non-statement.)

We cannot agree on that. Feynman's talk is mainly about social phenomena, so it's entirely reasonable to assume that this anecdote is also meant to illustrate a social point: http://calteches.library.caltech.edu/51/2/CargoCult.htm

https://en.wikipedia.org/wiki/Brouwer%E2%80%93Hilbert_contro...

Brouwer's career suffered as a consequence of his disagreements with Hilbert.

EDIT (with links to the political aspect): Letters written more about the politics of publishing than the math itself are referenced in a book called "The War of the Frogs and the Mice" (PDF) http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.224... (see bottom of pg 9 as displayed for the letter Brouwer wrote to Hilbert's wife)

>In the later 1920s, Brouwer became involved in a public and demeaning controversy with Hilbert over editorial policy at Mathematische Annalen,

It doesn't read like the political fights were about math. Although Hilbert disagreed with Brouwer, it sounds like the falling out was precipitated by arguing over something subjective.

Though worth pointing out "political" is ambiguous and that this a case of politics in the "office politics" sense not in the "use science to justify policy" sense

I believe this applies mostly to public science. Private science is anchored by a market-based mechanism that naturally expunges fraud.

Is this a joke? Profit motive is far more likely to put the cart before the horse: pushing for evidence to support outcomes.

I think they were referring to internal research that says 'if we do X we can sell more' which has naturally aligned incentives.

This as opposed to releasing research along the lines of 'cigarettes are fine for you! smoke more!'

> if we do X we can sell more' which has naturally aligned incentives.

Even then there are agency problems which mean that once a company gets over a certain size there is no natural alignment.

I know plenty of managers tweaking A/B tests until they get the results they want.

Easy fix: hire a statistician. :-)

This is a quote from a preeminent scientist: https://twitter.com/bmbaczkowski/status/1069891514275385344?...

"in academia overfitting can get you a nature paper, in industry get you fired"

Not so much. When you pay someone to get a result, that’s what they do.

You want to know where this is the worst? Climate science. Everyone has an axe to grind on one side or the other and moderated opinions are flamed into oblivion. We can't even talk about the science because it becomes a discussion about destroying the planet.

Isn’t the moderate position in climate science (regarding anthropogenic climate change) the mainstream consensus position? There are no two ”extreme views” with the truth somewhere in the middle. There’s the overwhelmingly supported by evidence scientific position and then there’s the pseudoscientific position.

Yes, I suppose I've conflated two things. The first is the strident shrieking of politicians and those who echo what politicians say,for which I would argue there are two extreme sides. And neither of these sides actually has constructive discussionin my opinion.

And then, there are the different sides of scientific discussion which produce the conclusions that the politicians warp and shriek about. Some of this is good science, and a rot of itis bad science. But whatever the politicians get to blather on about is what most people are aware of, regardless of whether it's good or bad.

And your comment points to part of the problem. It's science, so if we believe that we know everything, we're wrong. If we don't know everything, then that means we need to have the lattitude to argue about things. The argument is obviously not "is the earth warming?" but "how and why is the earth warming?". If your immediate response is "well, it's obviously humans producing CO2", then you have fallen into the trap that I'm trying to explain. Yes, humans produce CO2, and CO2 is a greenhouse gas. But the story is more complicated than that, and science needs room to wonder. The problem is the political shrieking dampens the curiosity and wonder that science is about.

But if Earth in ancient prehistory was dominated by CO2 that was eventually sequestered by plants under the ground, and that today our oil reserves are running low, then wouldn't it make simple sense that human use of fossil fuels must have a significant impact on atmospheric CO2 levels?

In that case, it's pretty settled that there is a significant human impact on atmospheric CO2 levels; the space for room to wander are, say, the exact distribution of CO2 released. I mean, physicists don't wonder and revise basic kinematics since that science is pretty much settled that Newtonian kinematics are very good approximations of reality.

Well, it does make sense yes, which is why that view has become accepted.

There are some facts that seem to give pause for thought though, like it apparently being much warmer than today in medieval times. This can be hard to see because temperature records don't go back that far, but we know that the Romans appeared to once grow wine near Scotland, something totally impossible today (in fact it only became possible to grow wine in the south of England very recently, a "new" thing blamed on global warming). It's unclear how the Romans could have done that unless grapes were very different or the weather was very different.

Of course a fair point, but I think it's worth considering that basic physics was conceived quite a bit longer ago than the idea of CO2 as a greenhouse gas began to spread. So physics has had quite a bit more time to settle and for people to prove it incorrect (which they haven't), while the reasoning for cause(s) of planetary warming is really still in its infancy, however sure many people think they are.

>today our oil reserves are running low

This argument is not quite up to snuff because, as it turns out, only a tiny fraction of the oil ever can be made to come out. The truth is that even after the wells have been abandoned most of the oil remains stuck in the rocks that it started in. I found this out while wondering whether or not it would be possible to burn all of the oxygen in the air (it's not, sadly ;) )

Also the oil reserves are not really running low, if they were that would be a lucky answer to carbon control.

Most of the impression of a mainstream consensus is formed when some kids polled a ton of scientists what they thought of global warming, but then restricted their published number to a small subset of those, climate scientists. Plenty of other scientists are like, hey, these guys are full of crap.

That's uhh, not true. At any rate, shouldn't we be listing to the climatologists on this matter, because their specialty is, you know, climate?

Yes it is true. The often cited statistic comes from https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1029/200...

Counting climatologists is a bad idea because they shut critics and skeptics out of the field. We know that climate models are generally wrong. Trying saying that while having a career as a climatologist.

You're really veering off into conspiracy theory territory here. It's clear you haven't the faintest how climate models or any geophysics works. Don't mistake a facility for computers for expertise in all science. I actually have a PhD in a geophysics field, I know what I am talking about here.

1. That study is 10 years old at this point. The evidence for climate change has become vastly more undeniable since then.

2. Majorities of scientists from all categories believed in climate change, even back then, according to this study.

I'm not disputing global warming. I am disputing the impression of a 97% consensus. And this is a great example of what I'm talking about, you're trying to call me a conspiracy theorist, and making up stuff about what you think I know.

No, you aren't just disputing the consensus. You are claiming that climatologists "shut critics and skeptics" out of the field, so we shouldn't listen to them. Yes, they do shut out pseudoscience, as do other legitimate sciences. That is different from disagreement on scientific facts. Climatologists disagree about very fundamental things all the time.

There are people on the record saying certain kinds of results simply shouldn’t be published.

Also, anybody who views climate prediction as intrinsically suspect pseudoscience isn’t going to want to enter the field either.

So now you are saying that climatology is literally pseudoscience.

That isn't what was said though. This entire conversation chain is painful to read, I would encourage you to read up on the principle of charity. It's a dialectic tactic that seems to be fast fading in the world of click bait and outrage, and I think everyone is suffering for it.

Putting on my conspiracy helmet and tightening it all the way, how can you prove that that's not the same as polling the Department of Constructive Mathematics about the law of the excluded middle? Every field self-selects for people who believe that whatever the field studies is actually possible to figure out, all the way from physics to Paranormal Studies. Someone who didn't believe that climate models could predict anything wouldn't make any.

Appealing to the authority of climatologists is not going to convince anybody because everyone who thinks they are authoritative already agrees that global warming is caused by humans...

Scientists who aren't climatologists have uninformed opinions about the climatology, aka the climate; they shouldn't be listened to just as you don't poll brain surgeons about the intricacies of heart surgery. Being a scientist doesn't make you qualified to speak on the climate, being a climatologist is what makes you qualified to speak on the climate. You don't ask climatologists about physics, and you don't ask physicists about the climate. If you don't have a PhD in the field, you don't belong in the big discussions where consequences matter.

So even if your story is true, those kids did exactly the right thing.

Regardless, the consensus that's bandied about comes from the IPC meta study of all published peer reviewed articles from climatologists; not from a few kids.

> being a climatologist is what makes you qualified to speak on the climate

> If you don't have a PhD in the field, you don't belong in the big discussions where consequences matter.

Why?

I think what we need is accountability - there should be serious negative consequences for being wrong proportional to how strongly you stand behind an idea.

Were all the climatologists who were confident we were facing global cooling driven from the industry? Stripped of their academic standing?

Then why would I trust the industry?

I believe the mainstream take on climate change - but it’s worth noting that you can’t even become a climatologist today if you disagree with the mainstream view of climate change.

So it shouldn’t exactly surprise us that most of them agree.

There is a definitive difference. Global warming has been confirmed across multiple different experiments that have been repeated... this study covers obscure results that are unlikely to be repeated.

It turns out, people are always political, because they're people.

Max Planck:

“A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.”

Or more succinctly, science advances one funeral at a time.

Except this isnt just "one funeral at a time." This is genocide.

True and funny. Looks like you got downvoted into oblivion. Someone being a bit too sensitive about a genocide metaphor I suspect.

A version of this happens in software too. See machine learning. There’s definitely valid field with real benefits. But also lots of BS. Nobody wants to call out their peers because everyone benefits from the hype.

> Nobody wants to call out their peers because everyone benefits from the hype.

As a counterpoint, it seems that minsky and papert's work was partly driven from their desire to stop the funding of larger perceptrons , in which they did not believe. The winter that followed their work may or may not have delayed progress in AI by years or decades.

https://www.youtube.com/watch?v=w3ues-NayAs&t=642s

The theory of deep learning was developed about 50 years ago, but useful applications weren't possible until recently because the hardware was too slow.

I'd be interested to learn about results in the "hard sciences" that I work in (mathematics, logic, program verification, typing systems), what papers "everyone in the niche knows is false", but "fears to speak out about".

Sure, mistakes happen, my own PhD contained a major blunder (discovered several years after submission -- at least somebody read it ...), but I'm extremely open about it.

FWIW, your examples aren’t hard sciences per se. The term hard science is referring to natural world subjects like physics, chemistry, biology and astronomy. It’s not a difficulty or value judgement, it’s just a term with a specific categorical meaning.

I wonder sometimes whether math isn't a natural science by accident. If 2+2=4 wasn't applicable to the real world, I don't believe we would have defined math the same way

That’s an open (and interesting) question, I think, whether there are absolute physical truths in math, and whether our basic math rules can be defined any other way. But certainly there are huge swaths of math that don’t correspond to the physical world, and there are parts of math that are decided rather than natural consequences.

Ultimately though, the math that is natural science is just called “physics”, right? ;) The reason math can’t be a hard science is because we can’t prove abstract math by physical experiment. That’s all “hard science” means - things you can show by physical experiment. We can prove math theorems by showing they’re consistent under all the other math theorems, but generally not by conducting physical tests.

Not all mathematics formalises physics. Nowadays, much math formalises other maths, e.g. set theory, category theory etc.

Right, exactly. Physics already is the subset of math that is a hard science, the rest of math isn’t.

Yet

"I will begin by saying that there is probably less difference between the positions of a mathematician and a physicist than is generally supposed, and that the most important seems to me to be this, that the mathematician is in much more direct contact with reality." -- G.H. Hardy (1922)

I suspect that to the ancient Greeks, math and natural science were not yet separate disciplines. "Geometry" means something like measuring the world. They had to be aware of the relationship between their math and the ability to function as a minimally technological society, surveying land, navigating, and so forth. As I understand it, they believed that math was telling them something deep about reality.

I think turning math into a purely abstract game came later, and we owe our fun to the discipline's more humble origins.

I think it would be more accurate to say that they were far more separate disciplines anciently, and far more rarely mixed. See for example, Euclid vs the the contents of this work of Aristotle: https://en.wikipedia.org/wiki/Physics_(Aristotle)

Now you're getting into philosophy. As a materialist, I do believe everything comes from the natural world. I take the view that if every time we combined two objects with two other objects we ended up with something other than four objects, we would not be saying "2+2=4".

There are of course other viewpoints that math has a Platonic metaphysical reality in and of itself, independent of the natural world.

Mathematics isn't a science because it doesn't provide us with empirical answers. That's precisely the difference between physics and mathematics: the former tells us what is, while the latter only tells us what must be, given certain specific assumptions.

Or more succinctly: mathematics doesn't give us truths, it gives us consequences. The statement, "2 + 2 = 4" is not an empirical fact intrinsic to the universe. It's a consequence of our definition of natural numbers. It's considered a useful definition because (among other things) of its application as a foundation for engineering and the sciences. But many other useful areas of mathematics have essentially no basis in reality.

> Mathematics isn't a science because it doesn't provide us with empirical answers.

Then physics isn't a science either since only parts of the field fits this criteria. I think you will realize that the distinction between physics and math isn't as clear cut as you imagine, it is not uncommon for physicists to publish in math journals and vice versa.

The distinction is pretty clear cut. The two disciplines are philosophically different by category, not degree. Just because much of theoretical physics is applied math and physicists can publish in math journals doesn't mean the two occupy the same domain. One is ostensibly guided by empiricism (even if it's not always possible), the other is not.

In my view, the hard sciences aren't really directly comparable. We doubtlessly have our own problems, but not necessarily the same problems as the health and behavioral sciences.

I think that the hard sciences aren't necessarily so dependent on papers derived from isolated results, or even replication of those results, because there is more of a "web" of knowledge. An idea is hopefully supported by an experiment, but possibly also by theory and multiple different kinds of experiments, sufficient to be usable by non-experts and laypeople. For instance the basic theory of electromagnetism is good enough for virtually any kind of engineering, and it would take some monumental discovery to prove it false, even if a portion of the supporting work fails upon replication.

Another thing is, perhaps because of the precision of physics, we get proven wrong every day, and it ceases to be an embarrassment. The road is littered with the wreckage of failed theories and experiments.

>Sure, mistakes happen, my own PhD contained a major blunder (discovered several years after submission -- at least somebody read it ...), but I'm extremely open about it.

So what was it?

That sounds grave. I believe in most cases the science is not questionable per se, but it passes just above the significance threshold to be considered technically valid. The problem is the extrapolations made from that, and the false media attention.

what your wife is describing is what is commonly called "progress one death at a time"

Sometimes I wonder if researchers are too conflict-averse temperamentally to effectively police themselves. I remember grad school being suffused with below-the-surface grudges, innuendo, and shunning. Those aren't healthy ways to resolve conflicts, but they seemed like the path of least resistance for people who fear the raising of voices.

Is that really temperament? I always thought of that as learned behavior necessary to survive in an academic context, where one's fate is for years in the hands of various authorities who can cause you a great deal of harm.

Not all "hard" sciences. I'm pretty sure there are no papers in physics that are well known to be wrong but people fear speaking out, though I'm open to being proven wrong.

Richard Feynman was laughed at when presenting new ideas in physics when he was an up-and-comer. The human ego tends to intervene when careers are on the line.

Laughing at true new ideas is actually the opposite behavior flaw of whispering about false ideas that are known to be false but are associated with too much social power.

It’s the same family: public judgement of a claim influenced by the person making it, rather than its contents.

Point being: Feynman’s ideas should have been judged on their merit, rather than Richard’s career. Same for the papers published by the institute mentioned above.

They eventually were judged on merit, which is a testament to the system at the expense of the characters that were involved at the time.

To be fair, Feynman wasn't above proof by intimidation, either.

I adore Feynman's work, but he was human and had human failings, too.

There is an interesting anecdote about how Millikan's oil drop experiment that measured the elementary electric charge was way off when he measured. Feynman talks about it in his book:

"We have learned a lot from experience about how to handle some of the ways we fool ourselves. One example: Millikan measured the charge on an electron by an experiment with falling oil drops, and got an answer which we now know not to be quite right. It's a little bit off because he had the incorrect value for the viscosity of air. It's interesting to look at the history of measurements of the charge of an electron, after Millikan. If you plot them as a function of time, you find that one is a little bit bigger than Millikan's, and the next one's a little bit bigger than that, and the next one's a little bit bigger than that, until finally they settle down to a number which is higher.

Why didn't they discover the new number was higher right away? It's a thing that scientists are ashamed of—this history—because it's apparent that people did things like this: When they got a number that was too high above Millikan's, they thought something must be wrong—and they would look for and find a reason why something might be wrong. When they got a number close to Millikan's value they didn't look so hard. And so they eliminated the numbers that were too far off, and did other things like that."

There are definitely papers in physics that are well-known to be wrong. I know of results published in high-profile journals where the authors have stated (in person, but not officially in a published update) that even they cannot reproduce their own results.

i agree that this is a problem in the biosciences as well.

the dead ringer for these groups is a PI who has a reputation for taking on extremely ambitious projects with all sorts of funding sources. typically, the PI is vindictive and controlling in my experience.

but yeah. you can see them from a mile away. they tend to run sweatshop-style labs where everyone is very hard pressed for results. so "results" are produced, whether or not they are real.

Why don't they publish a paper anonymously? Hopefully that would at least work for cases where "everyone" knows it is bullshit.

Isn't that fear something which the "Journal of Controversial Ideas" aims to solve?

Your story sounds like what I have seen, but biology is not a "hard science".

Biochemistry and creating pharmaceuticals is absolutely a hard science

Not at all. In the hard sciences you can run reproducible experiments and have mathematical models used to make precise predictions and come up with "natural laws". Biologists were moving in that direction up to 1940 or so but that is not what is going on now.

Here is the biggest funder of biomedical research in the world admitting that $27 billion out of $30 billion per year is wasted on unreproducible research: https://nihrecord.nih.gov/newsletters/2016/07_01_2016/story3...

And I would say that is optimistic because it does not consider misinterpreted results (interpreting your data correctly is even harder than figuring out how to perform and report reproducible experiments)...

If biochemistry is not a hard science then the only hard sciences are math and physics

I'd say both organic and inorganic chemistry are hard sciences since they make precise predictions about what will happen. But not biochemistry, and definitely not cell biology or molecular biology.

I also wouldn't consider math a science since it doesn't require comparing your results to data.

Biology is absolutely a hard science. The soft sciences are things like economics, sociology and psychology. Look up "hard science" on wikipedia. Look it up in a dictionary. Look it up in any reference work you like.

It may very well be that you have your own peculiar definition of "hard science", or that you personally disagree that biology or biochemistry should be classified that way. That's fine, you're allowed to have your own opinions. But you having your own crackpot opinions doesn't make it fact in the real world.

I am using this definition:

>"Precise definitions vary,[4] but features often cited as characteristic of hard science include producing testable predictions, performing controlled experiments, relying on quantifiable data and mathematical models, a high degree of accuracy and objectivity, higher levels of consensus, faster progression of the field, greater explanatory success, cumulativeness, replicability, and generally applying a purer form of the scientific method." https://en.wikipedia.org/wiki/Hard_and_soft_science

That same page does group biology in with physics and chemistry, but with no justification. From personal experience biology is much closer to a social science than it is to physics. I mean, most people who study biology-related stuff "hate math".

Are you honestly saying that medicine (a branch of biology) is not a "hard science"? That epidemiology, or genetics, or oncology isn't? Are you saying that when Crick and Watson deduced the structure of DNA based on the X-ray diffraction work done by Rosalind Franklin, they were not doing "hard science"?

Hell, go back further: consider Mendel's discovery of the rules of genetic inheritance (and thus founding the field of genetics) by painstakingly planting, replanting, combing and recombining pea plants. It's hard to imagine a more pure and rigorous application of the scientific method, in any field.

I'm sorry, but your position is beyond absurd. Even the page you link disagrees with you. It sounds very much like you once met a biologist who said that they didn't very much like math, and since then you've carried a personal prejudice against biologists. You admit as much in your comment.

Just move on. Someone made a lazy and uninformed opinion and then, instead of admiting fault or walking away, they struggled to justify their own ignorance.

Ironically in the same vein as the OP article!

Good advice.

Interesting you concluded my opinion was lazy and uninformed when I was the one sharing links and references. That behavior usually indicates the opposite. Indeed, that heuristic would be correct in this case. Not that it matters to the point, but I have a PhD in a biomedical field and spent many years working in it before I quit because I got fed up with how "soft" it was.

The sad part is it will never improve to become a "hard science" if people think it already is one. I mean really think about it. Biology is the study of how complex systems change over time, yet learning the primary tool we have to study/analyze change (calculus) is not a prerequisite. Doesn't that seem strange for a "hard science"?

This comment is insulting and rude. There are other ways to inform someone that you think he is wrong while being civil and mature.

That being said I get where the commenter is coming from. Much of biology is a qualitative science. Darwin's evolution was originally elucidated through descriptions not hard numbers and statistics. Taxonomy in zoology is also largely qualitative.

That comment was only insulting if you believe that biology is not a hard science; an idea which is itself insulting, dismissive and farcical.

But, if a person does share that opinion, I can see how that person might be inclined to ignore tone, and focus on supporting that opinion.

Sometimes reality can make people feel bad, that doesn't mean we should try to ignore it. Lets take the cancer research example. Using this definition:

1) Producing testable predictions

--- The predictions are almost all vague, of the form "x will be correlated to some degree with y".

2) Performing controlled experiments

--- Maybe, but I very often see that proper blinding procedures were not used.

3) Relying on quantifiable data and mathematical models

--- Mathematical models are rarely used to make precise testable predictions.

--- There was some progress on this front from Armitage and Doll way back in 1954 but interest has largely fizzled out.

4) A high degree of accuracy and objectivity

--- As mentioned in #1 and #3, making a precise prediction at all is rare. Accuracy doesnt matter much if you predict vague stuff.

5) Higher levels of consensus

--- Not sure this should be here. "Hard" vs "soft" is a matter of the procedures used, not the results.

6) Faster progression of the field

--- Not sure this should be here. "Hard" vs "soft" is a matter of the procedures used, not the results.

7) Greater explanatory success

--- Not sure this should be here. "Hard" vs "soft" is a matter of the procedures used, not the results.

8) Cumulativeness

--- The more research that has been done on cancer the more complex things have gotten, to the point where now they say "cancer is many diseases". This is the opposite of what happens when you gain a cumulative understanding and figure out "natural laws" that make things easier to understand.

9) Replicability

--- Seems to be well less than 50% of studies can be repeated (see my links here: https://news.ycombinator.com/item?id=18578282)

This is utterly incorrect. You can disagree with people without being rude.

You cannot call an idea itself insulting unless you yourself are a biased person. Ideas in itself are just statements with no malevolence intended. Only people can express malevolence.

I don't agree with him. But I see where he's coming from. A lot of biology is just observing behavior and documenting observations. It is very different from employing the scientific method in an experiment. It is also very different from trying to quantify everything. The line between hard and soft is blurry.

This describes some of the mathematical economics I've read.

Can you give an example of a testable prediction?

That there is a general tendency for the rate of profit across society to fall with time, subject to certain countervailing measures.

I would have preferred a reference to the paper making the prediction, but that sounds very vague so whatever.

>"Are you honestly saying that medicine (a branch of biology) is not a "hard science"? That epidemiology, or genetics, or oncology isn't?"

These fields in particular are not even close to being hard, especially oncology which appears to (somehow) be even worse off than psychology.[1:3]

I bet less than 1% of people with masters or greater in a bio field have used calculus for anything in the last 5 years. While of course the definition of "hard science" is arbitrary, here are some decent overviews of the differences between physics and bio:

https://www.cell.com/cancer-cell/fulltext/S1535-6108(02)0013...

https://meehl.dl.umn.edu/sites/g/files/pua1696/f/074theoryte...

Refs:

[1] https://www.nature.com/articles/483531a

[2] https://www.the-scientist.com/news-opinion/effort-to-reprodu...

[3]https://www.sciencemag.org/news/2017/01/rigorous-replication...

Uh chemistry is only predictable from first principles for diatomic molecules, or very simple extended structures like perfect bulk graphene. Beyond that chemistry is all about approximations

There is a spectrum of precision regarding the predictions you can make. At the high end is a point prediction, at the low end is "x is correlated with y somehow". In between are intervals of various sizes.

>In the hard sciences you can run reproducible experiments and have mathematical models used to make precise predictions and come up with "natural laws".

Spend time with condensed matter physicists. What you describe is not what they do. A lot of results are not reproducible.

I've been getting the impression that many areas of physics have been getting "softer", but have no direct experience. Even Feynman mentioned that physicists were discouraging replication way back in the "cargo cult science" talk.

>Even Feynman mentioned that physicists were discouraging replication way back in the "cargo cult science" talk.

If replication is the benchmark, almost no one does it (physicist or otherwise). In all my time in academia, I didn't find a single person attempting to replicate anything. You don't get funding, tenure or papers from it.

I was referring to everything else in your comment.

I've seen a couple direct replications in biomed, NIH even specifically funded some a few years back under the “Facilities of Research-Spinal Cord Injury” project. AFAIK, pretty much nothing replicated:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3246348/

https://www.ncbi.nlm.nih.gov/pubmed/18234196

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2643041/

http://europepmc.org/abstract/med/21195070

But yes it is extremely rare. It is very frustrating because it is hard to come up with a precise model when you only half-trust the data to begin with, and when the replications are (rarely) run it appears that half-trusting the reported results is optimistic...

I know of one major team in my field. I tried to passively discuss it, was graduating and I was so naive. Not a good idea! At the time I had a few papers in review.

All my work got rejected (everything was being accepted before, and the rejected batch was a full level higher quality). So yes, it did set back my career for two and a half years if I have to approximate. I am probably never submitting to certain publications ever again to avoid having them as reviewers. I am gradually moving away from their field entirely. I was so short on publication then that I couldnt even move away to nearby fields, having accomplished vengeful rejections in my field. These days I wish I had done a second PhD in a different field, just to get away from them! I'd have had a much easier career path.

This is by no means an exaggeration, I am actually downplaying what happened.

Reply on news.ycombinator.com