Hacker News Reader

Comment by hinkley | Hacker News Reader

Sabine Hossenfelder basically posted a rant about particle physicists on Youtube last week, about how they are doing science wrong.

Basically she said you are supposed to change your theory based upon results that don't match the theory, and what they have been doing for decades is looking at gaps in the results, and imagining particle theories that don't alter the expected results in the currently explored energy levels, just in the unexplored ones. Then when people get funding to look into these frontiers/gaps, and find that everything looks the same, they adjust their theories to push them out into the gaps.

That's not Science, she emphatically states.

For myself it sounds like saying, "well you don't know Unicorns don't exist because we haven't looked everywhere, what if they're in <unexplored forests in the Congo>?" Then the Congo gets explored and they say, "well maybe they're in Antarctica".

View on HN

Eh, your analogy isn't apt. The problem in physics is that the space of possible theories has been searched so exhaustively that what these physicists are doing is just about the only choice.

To make your analogy close to physics -- imagine you found a unicorn horn and can tell from the biological makeup that it is indeed a horn grown from some sort of horse-like mammal. If you searched 99.9999% of the earth for animals that might have grown that horn and found nothing then the best thing to do would be to continue looking within that 0.00001%.

We can't figure out what comes after the standard model of particle physics. But we know it's wrong. e.g. we know black holes exist and something has to happen within them. But we don't have a theory that can predict what that is. e.g. we found a unicorn horn. So we have to desperately search for it.

Hossenfelder is right that it's shitty to have to keep pushing the energy boundaries in which we hope we'll see some interesting behavior. But there's literally nothing else to do. We've searched 99.99999% of earth and there's no unicorn. Do we start denying the existence of the unicorn horn that we're holding?

idk the amount of experiments / work vying for funding is far bigger than what can be funded. and if a big chunk of the funds is allocated to building bigger accelerators then a lot of experiments or other work might not get funded. im not saying its better to fund other projects over the next big accelerator or dark matter detector, but i dont think its right to say there is nothing else to do besides that...

There's other physics to do, but it's not fundamental physics. The science questions for which we need accelerators can't be answered any other way.

> The problem in physics is that the space of possible theories has been searched so exhaustively that what these physicists are doing is just about the only choice.

Can't say I agree with that one. I think we're capping out on the physics that Einstein and ilk laid out for us to explore. This is an ideal time in the next 20-40 years for a new revolutionary genius to figure out the next big thing.

I'm not going to make any sci-fi prognostications but I have a feeling like there's a new threshold to cross coming up, one of those unknown unknowns that we'll suddenly figure out thanks to a flash of insight by the right person being in the right place at the right time kind of things.

Counterpoint: there are major unresolved puzzles in symplectic geometry, and every indication is that they touch on deep physics. It could easily take research mathematicians fifty more years to build a clear picture of what's going on.

The state of physics these days is we have 100 physicists of Einstein's calibre alive right now, but not enough experimental data for them to make any headway on fundamental questions. For all we know, someone already published the theory of everything, but because there was no way to test it, it got ignored completely.

This is the exact problem with modern science.

Yes, you really should go and look back at that horn and thoroughly recheck if it really is what it seemed like to people several decades ago.

Serial contrarian jumps on table and shouts. More at 11.

A telltale sign of a crank is attacking career scientists, which is her brand. What insight is she offering? Destroying is easy, creating is hard. Lies spread faster than truth. In general she is a force of confusion rather than enlightenment. It's too bad because she seems capable, if only she'd follow through rather than chase headlines.

>> What insight is she offering?

For a while she was offering an honest look at a field going nowhere, and suggestions as to how they might better spend time and money.

Now I think she's just profiting from her new popularity by doing a not so good science/tech show.

A "Career Scientist" is very, very bad news!

A career contrarian can be even worse. A career scientist actually needs to justify their output, to their peers and the people signing the cheques.

And if not career scientists, who exactly do you think should be making a living... Doing research?

Undergrads? The admins? The timecube guy? Nobody?

There's a lot of problems with science as it's done today, but that snipe is a banal, tribalistic take that is all-too-common.

That would've been a great reply had the title of the main post not been "Cargo Cult Science".

For me, a "career scientist" is someone that publishes papers solely to hold a position at an org / institute / uni / etc (yes, yes, everyone needs food on the table... not the point here. One can still have integrity). The career scientist couldn't care less about being right, only about how many citations he's gonna get and whether anyone's going to discover any flaws in his paper.

Why? Do you believe the same is true for other occupations as well? So does a doctor just care how many prescriptions they make, or a police person about how many people they arrest? Have you considered that people aspire to other goals except money, despite they being paid for it?

Honesty and playing by the rules is often the domain of the new, shiny-eyed initiates. Unfortunately, not having integrity is often easier for most people.

Lol not trying to be snarky but if you think academics push out papers for purely monetary reasons then I suspect you are very far from academia in your day to day life. Of course PIs have goals besides money! If their main goal was ever money they're an idiot for sticking to the university lab, and it's not like science profs didn't have other career opportunities available.

That doesn't change the fact that the most common non-monetary goal of academics is in fact pretty well aligned with behaviors such as: publishing very frequently/flag planting, farming citations on old publications, exaggerating work in presentations and media press releases, etc. Not all of these behaviors are inherently bad but they easily can be and too often are taken to an extreme.

Said goal is of course intellectual credit, or from a more cynical lense pure ego stroking. And it is indeed a rampant non-monetary incentive in academia.

Anyway, I don't see the problem with someone who has recently pursued a PhD or beyond in academic science ripping on "career scientists". This is the sort of (mostly) informed criticism that science should welcome. Though some are overly negative, they usually still have relevant insights into problems in their field, and it is silly to throw the baby out with the bathwater.

That is markedly different from someone that frequently criticizes scientists from the outside while lacking any actual insight into the relevant technical complications, especially if that person is casting a wide net in what they criticize.

The goal of a lot of modern scientific criticism is to further the conversation amongst those presently in the field, and help those adjacent to the field make more informed decisions - including prospective/very early career trainees, potential interdisciplinary collaborators, and anyone that has a role in directing possibly relevant funding opportunities. If criticisms get co-opted by other individuals for irrelevant goals that is unfortunate but it does not make the criticism invalid.

It's also somewhat difficult in the social media age to avoid commentary that gains steam in the scientific community from reaching broader public exposure, which indeed opens it to misinterpretation. But the alternative of walled off information a la the old school journal publication system is obviously problematic for other reasons.

That is a long winded way of saying that I think some amount of harsh publicly viewable criticism is sorely needed by a number of scientific fields right now, including my own (very much not physics). The fact that criticisms which from what I can gather where initially internally directed eventually gained such external exposure tells me that the message resonated with many scientists, even if just earlier career ones. Whether this person has started to profiteer on their cynicism is tangential to the validity of the original core message IMO.

>So does a doctor just care how many prescriptions they make, or a police person about how many people they arrest?

Someone hasn't heard of ticket quotas before have they? Or the pharmaceutical companies that were effectively buying off doctors for years by expensive trips based on their ability to prescribe the drugs they were selling. We've had to add regulations to these above two 'fields' to prevent exactly this from happening, and in many places it still does.

>The career scientist couldn't care less about being right

What you are describing is philosophically, definitively, not science. The career scientist is a scientist who is committed to the scientific method. The career crank is committed to selling the notion that they have all the answers. That could be a citation chaser or a youtube contrarian.

I agree. But unerringly honest people are vanishingly rare, and the second someone with integrity encounters a field of research that could undermine others' research, there is a silent outgrouping and sometimes expulsion. Some science just isn't allowed to breathe.

What's wrong with scientist as a career?

> A telltale sign of a crank is attacking career scientists

yeah, she should've known better than to attack a protected class

Any reason to frame this argument as contrarian? Apart from your decision to label the author so - presumably for the sake of legitimizing your argument?

She’s not the only or first one to criticize contemporary physics funding, and the focus on mathematical beauty to guide new theories. String Theory received a lot of blowback for how much time it sucked up without empirical results because the math was very appealing. Funding is finite, and so are careers.

She is a working scientist herself.

Things like this make me want to read Against Method [1]. Like, yeah that seems 'not science' in one serious sense, but also not really? It falls into what I understand of Kuhn's 'normal science'. Slightly adjusting their theories to explain existing data is exactly what you should expect scientists to do, until the gaps get so big no adjustments are sufficiently concise.

For a defense of the 'unicorns in unexplored forests', that was pretty much the approach that led to the discovery of most of the standard model.

Plus, given dark matter and dark energy, we know something must be out there, even if its not unicorns, there's something.

1. https://en.wikipedia.org/wiki/Against_Method

> Plus, given dark matter and dark energy, we know something must be out there, even if its not unicorns, there's something.

Because science is paradigmatic, this is the problem. We do not know something else must be out there because we have no direct evidence of either Dark Matter nor Dark Energy, which, btw, have absolutely nothing to do with one another. We might just not yet have the intuition to understand what is causing our perception of the need for these labels.

I, for one, am absolutely convinced Dark Matter is absolute bunk and will be widely accepted to not exist eventually within the next 50 years. The initial reason for Dark Matter's existence, to explain rotational velocities of spiral galaxies we assume spiral galaxies appear to have more invisible matter, has been explained by the fact that the initial observation was of a galaxy in isolation, and the assumption that it was an isolated inertial frame without considering the cluster of galaxies around it, which perfectly explain spiral galaxy rotational velocities without any need for Dark Matter. But because science is paradigmatic and even in the Space Age the correct information is taking decades to disseminate, many if not most are still are convinced Dark Matter is needed even though no one knows what it is. It's nothing, really, it's a phantom, imagination gone haywire. Also, as other unexplained developments occurred, these too got lumped into Dark Matter, which galaxy clusters can not explain. So the problem becomes the shame of not just saying, "we don't know... we don't understand this," and labelling it something to fake it like we do understand it. This is a flaw of modern science, or at least of cosmology,i.e. the unwillingness to accept that we don't (yet) know the precise reasons for what is observed.

That said sometime it’s worthwhile to try something completely different. If all we did was “adjust” we would still be using epicycles to describe orbits.

Feynman mentions a similar point in QED when he talks about the "science" of proton decay. The proton hasn't had the courtesy to actually decay in an experiment, so all they do is keep publishing increasingly high lower bounds on the proton's half-life.

It's fascinating that someone can make a career out of a phenomenon not being observed.

Folks talk about the lack of support for things like replication studies and the publication of negative results in the life sciences, as contributing to the widely proclaimed replication crisis.

Establishing a solid negative result that rules out hypotheses is worthwhile if unglamorous.

Citing an upper bound on the half-life is all that can be credibly reported. It's how you put an error bar on something that's non-negative but consistent with zero.

Sure, but when an event supposedly happens on average less frequently than once every 10^32 years maybe it’s time to consider the possibility that the sought phenomenon never occurs. At least that appeared to me to be what Feynman was getting at.

The life sciences equivalent would be a study of unicorn sightings. None so far, but they could be out there so please fund my research!

Placing an upper limit on the half life means that zero is within the reported bounds, so they are considering the possibility.

Whether to fund difficult research, when it's reasonable to give up, and when it escapes beyond reasonable definitions of science, are separate questions. Physics is wrestling with these questions in a number of areas, such as the search for an empirically testable theory that reconciles quantum mechanics and gravity.

The life sciences have no equivalent. Precision measurement is not their bag, which is OK. Physics is the first to arrive at such a roadblock. Maybe the search for proofs of some famous math conjectures is an analogy to draw guidance from.

Things are more complicated than that.

Something is there causing those phenomena. People have a theory that it may be unicorns, so they are out looking for unicorns. Those are supposed to be hard to find, and people are still searching on the hardest possible places that can hold them. It may also not be unicorns, on what case nobody has any idea what it may be. Some people are trying to make such theories, but none got a good one yet.

On this level of detail, there is nothing wrong at all with any party of the above. All of the problems are with finer details.

It matters when the limited resources are spent chasing unicorns that are vanishing unlikely to actually be there.

That is part of the argument. Those dollars could go other places than these giant particle accelerators that haven't shown anything much since the Higgs Boson, and no theories are more than guessing that we will.

the thing is that this pattern of pushing to unexplored energy levels did work before. Unicorns were found in the unexplored forests of Congo. In fact an entire zoo of unicorns of all shapes and forms was found, an incredible achievement that stands among the most important stuff our gray-tinted gel has managed to internalize about the universe. You can't blame modern day particle physicists for being born late for the Unicorn chasing party...

so the argurment now is, for how much longer is it reasonable to be roaming difficult terrains in the hope that we might have a repeat miracle? Its practically impossible to answer. "All Unicorns have been discovered" has been a scientist's cry as far back as Laplace.

The last time we found something that was predicted by the Standard Model was 1983, wasn't it? That's 40 years.

edit: her complaint is that once funding sources realize that 'particle physicists shill ridiculous ideas for research grants', all of the money dries up and won't return for at least a generation or two. That's an existential threat to the field. That sounds like a legitimate complaint to me. Unfortunately she's also shining a spotlight on the thing she's worried about other people noticing.

> That's 40 years.

Is that a lot? On what basis could we decide?(1) I am not saying it isn't, I am just saying it is a rather subjective call which sort of looks at the prior pace of discoveries and extrapolates a judgement.

(1) it took more than that from the first gravitational wave detector to an actual discovery

I'm going to start repeating myself in separate replies here.

Einstein didn't predict gravitational waves as a standalone theory. Gravitational waves are a prediction of General Relativity. General Relativity was shown to be plausible based on data already at hand and data they could collect.

Gravity waves and lensing and frame dragging all confirm what we already suspected to be true. Which is good and bad news because they increase the corroborating evidence for an existing theory, and they take out huge chunks of opportunity for some other theory to explain the same phenomenon but introduce new loopholes or consequences that let us do new things like build Foundation power supplies (nuclear reactors the size of gumballs) or FTL travel or explain why galaxies move 'wrong'.

Physics is not scifi. As the good professor might say, its stranger and always more surprising than the bizarre combinatorial pseudoscience our brains might concoct in the absence of experimental guard rails.

> The last time we found something that was predicted by the Standard Model was 1983, wasn't it?

The Higgs Boson was found in 2012 by the LHC and was predicted by the standard model.

Higgs et al didn't just predict the boson though, right? His team explained other properties of particle physics and their model also predicted a particle.

> The last time we found something that was predicted by the Standard Model was 1983, wasn't it? That's 40 years.

IANAPhysicist so correct me if I'm wrong but the last time we found something predicted [1] by the Standard Model was the Higgs Boson in 2012. Before that it was the Tau neutrino and the quark-gluon plasma in 2000, the top quark in 1995, and then (the discovery I think you're referring to) the W and Z bosons in 1983

[1] we mean experimental confirmation, right?

Also not a physicist, but I think the distinction being made here is that Higgs et al explained existing phenomenon that did not make sense within the existing theory, and oh by the way if we're correct we've also created a hole in the standard model where a particle exists with certain properties that were a bit nebulous.

For instance special relativity explains gravitational lensing and predicted that frame dragging would also be discovered, but also explained other problems. The theories that make a physicist's day seem to tie the past to the present and the present to the future:

This happened, we don't know why. Here's a theory that explains those measurements. Here's a prediction we can test now/with a billion dollars that will corroborate my theory, and here are some new consequences of the theory that require <future tech> to be sure that this is 'true' instead of an accident of math. Like atomic clocks or giant particle accelerators.

> special relativity explains gravitational lensing

You mean general relativity

> You can't blame modern day particle physicists for being born late for the Unicorn chasing party...

1950s, 1950, 1970s, particle physics was a fast-moving field, theory and experiment moved hand in hand. Guided by the intuition of what is mathematically elegant, you propose an extension of current theories, and soon someone in experimental physics with particle colliders publishes results that confirm your theory. Or the other way round: Experimental physicists publish new collider results, and theorists rush in to explain them with new theories.

But you had to be fast, and greedy (in a good way). If you didn't work out and publish the new mathematical expansion of current theories, someone else surely did.

It was also a good time to be a scientist in academia: Population was growing (15% per decade in USA, 10% per decade in Europe), economy was growing, and graduates from a previous (smaller) generation found jobs and careers in the expanding university system, to teach the next (larger) generation of students.

The standard model was competed in the 1970s. Theoretical particle physics was a well-oiled, battle tested fast spinning machine. Ready to continue the mad gold rush of past 30 years. Follow your intuition of mathematical elegance, propose new theories, wait for experiments to confirm them.

But experiments stopped finding traces of new particles. The standard model of particle physics was complete, and apparently all its particles were already discovered. Well, you had to wait a bit longer for: Top quark 1995, tau neutrino 2000, Higgs boson 2012.

The fast machine of theoretical physics kept rotating and spinning out mathematical structures and theories. But experiments were no longer confirming, or even guiding, the development of theory. The mode of work that had worked beautifully up to mid-1970s didn't work anymore. Supersymmetry made predictions, but all turned out to be wrong. String theory was such a complicated mathematical construction, that we don't even know if it makes predictions (in the Popperian sense) at all.

So yes, we can blame particle physics. What first worked for 30 years, hasn't been working anymore for 40 years. You should try something different.

A good summary of key attributes of that journey (why I like HN) but I would insist the conclusion doesn't follow directly and is rushed.

We know that we are far from having figured out all fundamental physics: All the "dark" stuff in cosmology points to the pieces of the puzzle not fitting well together, if at all.

The question is what sort of local experiments can help lift that next weil of ignorance and if we are collectively capable to imagine them and pursue them.

The issue with local physics is that it is what it is (as Feynman who triggered this thread might say) and what is accessible to us to explore does not match to perfection the socioeconomic convulsions of homo sapiens.

The high energy accelerator physics paradigm (which is but one of the possible local physics windows) may have been exhausted for our current cirumstances but I would bet not in general. The pattern of all physical systems we know is that if you keep pushing them they keep revealing new behavior. We may have hit an energy gap that requires too many resources to cross. But maybe the only resource missing is imagination. Who is to tell?

What is my conclusion? I would agree the success patterns of the past will not repeat on a cookie cutter basis. And theoretical physics is certainly gyrating uncontollably in the absence of experimental anchors. This only raises the bar for people to do a better job at connecting the dots.

This one? If so I would not call it a rant, she pulls no punches which is a positive thing.

"What's Going Wrong in Particle Physics?"

https://www.youtube.com/watch?v=lu4mH3Hmw2o

I don't see what the problem with Sabine's analysis is? I thought it was great.

Shouldn't particle physicists have more expectations of what they expect to find rather than we want to find X and we are going to search the entire energy frontier to find it. The problem is that the energy frontier is limitless so the ROI could be extremely poor unless we had a good idea of what we expect

> The problem is that the energy frontier is limitless so the ROI could be extremely poor unless we had a good idea of what we expect

AKA the full employment theorem of particle physics…

If we knew what to expect, it would be engineering not science. The fact is, we have tried having expectations, most notably supersymmetry, and those haven't turned out very well because we are literally venturing into the unknown. Part of the fun is that sometimes a wild theory works out (GR, Higgs) and sometimes it doesn't.

I think if we push the energy frontier one more time, there's no real reason to expect us to find anything -- but there's also no reason for us not to find anything.

Just unsub her. She picked up a lot of people by her bold way of presenting. Most discover after a while that she isnt worth watching.

I'm not even subbed. The things I am subbed to just don't generate new content at the rate that I hit youtube so I end up with suggestions.

You seem to imply, most people were smart. But statistically, most people are not.

Hossenfelder critzises physics. That is a good thing to do, regardless of whether she is always right about it.

Being smart, you can learn from her even if she is not. At the very least about how to present better.

If thousands of people have careers in unicorn-hunting, that's what will happen.

While watching her complain I kept thinking, "publish or die".

Publish (in a scientific journal) or publish (on Youtube). No hiding from the invisible freedom of the market.

I don’t get that complaint. It’s like hearing a software engineer complaining “write code or get fired”, yeah, well, that’s your literal job?

Have you been reading the threads about the reproducibility crisis that pop up here regularly? Essentially some people worry we have more people who need to write papers than we can sustain with our current research horizon, and people are making shit up or at least being overly optimistic.

As Feynman used to say, "The most important thing is not to fool yourself, and you're the easiest person to fool."

To all reproducibility problems I say “publish more”. Do you plan to do a study? Register it and publish your methodology before data collection. Have you collected data? Publish it. Have you finished the study? Believe it or not, publish it.

Do you want to do a 20 year long study on a particle collider built specifically for you right after finishing your PhD? Well, sorry, you have to build your reputation so that society sees that you are not one of those people who “are making shit up or at least being overly optimistic”.

Yea, but it's like all the software doesn't work or has huge security holes....

> For myself it sounds like saying, "well you don't know Unicorns don't exist because we haven't looked everywhere, what if they're in <unexplored forests in the Congo>?" Then the Congo gets explored and they say, "well maybe they're in Antarctica".

In India: https://en.wikipedia.org/wiki/File:Indian_One_Horned_Rhino.j...

As a counter point to this, by all accounts Millikan had worked out the charge on an electron but had not validated it by experiemnt. Knowing what thw result should be, he then designed an experiment to arri ve at this result. Not very scientific at all.

Hypothesis: the charge on an electron is X, as predicted by Y. Experiment: observe the charge on an electron. The experiment could have shown the charge to be other than X: in which case (assuming repeated experiments confirmed it) the hypothesis was wrong and a new one would have been needed. "Y" wasn't invented out of nothing: it was built on other observations, hypotheses, and theories. Seems pretty scientific to me, but I'm biased (since I'm a physicist).

I think at some point we (physicists) are just going to have to realize we're at the point where the models we have maybe should be taken for what they are (incomplete), and re-examine the data and construct new hypotheses to design experiments for. I'm certainly not the physicist to do that (I haven't been practicing for almost 15 years now, and even when I was I was not the Einstein caliber physicist to do it), but there will be someone or a group who do that. I hope soon, as I'd like to see it!

Sounds a bit like epicycles. When they found data that doesn’t match, they just added more epicycles until it did.

Took until Kepler and the idea that orbits were mostly elliptical for people to throw out epicycles altogether.

Reply on news.ycombinator.com