As an apostate, I view government laboratories as state-funded monasteries or Oz-like sandboxes in which the creative power of a few minds trained to worship abstraction is imposed on armies of lower-ranking individuals who lack a complete understanding of why they are doing what they do.
“The time has come,” the Walrus said,
“To talk of many things:
Of shoes–and ships–and sealing-wax–
Of cabbages–and kings–
And why the sea is boiling hot–
And whether pigs have wings.”
The Walrus and The Carpenter by Lewis Carroll
I also subscribe to the following opinion:
“The era of Big Science has provoked criticism that it undermines the basic principles of the scientific method. Results of experiments which require massive and unique machines like particle accelerators are often difficult to verify.” Big Science – Wikipedia
Gravitational wave measurements, neutrino research, colliders, complex climate simulations, etc. are all open to this sort of criticism. In particular:
The colliders have spent the past 50 years validating a reverse-engineered theory called the “Standard Model” and obvious applications of the standard model or of collider technologies are not abundant. They claim to have spun off internet technologies, but that really over-sells the significance of throwing a log on an already raging fire. Spending billions on a dead end project with no useful spin-offs could be considered evil since there are so many needy people in the world.
Like LIGO and LHC, neutrino or WIMP research like OPERA, XENON, and IceCube are also operations dedicated to finding a needle in a haystack, and there is an enormous incentive to find a piece of hay that looks like a needle, but these are not the only areas within Big Science research of questionable utility.
The big lasers developed by Big Science have been used unsuccessfully to try to reach the high temperatures necessary to ignite fusion and the technology used to build them has been applied to unsuccessful aircraft carrier mounted laser weapons. Materials scientists plan to use the big lasers to get pictures of materials and biologists plan to use the big lasers to get pictures of proteins and bacteria, but, in general, the pictures aren’t that good and the “small science” methods used in university labs are often much better and cheaper.
I think that the biggest criticism of “big science” is that once you invest a bunch of money in something really expensive, you tend to insist that it is useful, even if it isn’t. That sort of endemic inefficiency is certainly bad, even evil from a certain perspective. What might be truly evil is the group delusion that tends to take hold when maintaining and building a large machine of questionable utility. Students and workers train themselves to suspend disbelief and focus on their narrow, ant-like roles in the larger project. This narrow focus bypasses the natural skepticism which any good scientist must always have. Thus, an enterprise which purports to train the next generation of scientists ends up training groupthinkers who lack big-picture perspective. Science isn’t about team-spirit, but team-spirit infects a lot of big science work and that seems somewhat evil and monolithic to me.
What you see in CERN’s particle detector diagrams are layers and layers of material which take impossible to measure particles and convert them into something measurable. For example, a scintillator takes a high energy photon and turns it into lower energy photons which are converted into electrons which make an electric current which can be sent through some amplifiers and then into an analog digital converter which will give you some numbers which you can read out on your computer.
A question which does not get asked in physics circles is: do impossible to measure things objectively exist? I think this is a question worth asking, especially since headlines are made every time a group of physicists claims to have measured a new particle or a new property of a particle – pentaquarks, Higgs, massive neutrinos, WIMPs, oh my!
Since these unmeasureable things are also very rare, another question worth asking, is: what does it mean to measure something which is buried in noise? Out of millions of collisions per second in a CERN detector, thousands of collisions are tracked, and of the thousands, only a few are usable for analysis. One million to one is quite a challenging signal to noise ratio.
To parse all of the data which comes out of one of these monster detectors, they start at the outer layer and work their way in. The positions of the measurement devices on the outer layer have to be precisely known with respect to the devices in the inner layers as does the timing of all of the signals. They match up particles detected on the outer layer with particles detected in the inner layer by using probability statistics (because quantum particles have a lot of intrinsic uncertainty). Many particles make it through the detector unseen and any data which doesn’t match their model is thrown away (that doesn’t sound scientific). The outer layer is something like a drift chamber and it only sees charged particles. The middle layer is full of steel and scintillators and it is the only component which claims responsibility for converting unmeasurable neutral particles into measurable, charged particles. The layer closest to the collision is the vertex detector. I have experience detecting the position of nano Coulomb charge bunches with small antennae and few-micron accuracy is what you can do over a range of a couple of millimeters. Vertex detectors have to detect single charges and they claim few-micron accuracy over a large volume of space surrounding the collision point.
There are many ways in which someone can make a mistake in identifying a particle. For example, a photon is a neutral particle with no charge but also no mass. Really high energy photons might not be detected because they fly right through matter. That makes any collision for which a photon goes undetected rather worthless because you need ALL of the information to reconstruct a collision.
Another good example is a pion. It looks almost exactly like an electron and it is difficult to distinguish the two. Whenever an electron or a pion is misidentified, this will give an incorrect result for a reconstructed collision.
A CERN scientist might argue: Just because it is possible to trick yourself, that does not mean that we have done so in our detections of things like the Higgs particle. The standards for establishing the existence of a particle are very high: five sigma.
I would answer: But what about the theta plus pentaquark? By mining noisy, old data, it was ‘detected’ with five sigma significance, yet further investigation showed that all eleven of the ‘detections’ had been fooling themselves.
Behold the story of the pentaquark:
When you collide a pair of protons or neutrons, a bunch of smaller particles spill out.
These smaller particles exist as pairs of particles which are spinning in opposite directions. They have short lifetimes and they can never exist individually, so they quickly disappear.
Since these particle pairs seem to come out of the proton and neutron collisions, many scientists decided that meant the protons and neutrons were composed of these smaller particles.
These scientists then set about figuring out how these particles might be arranged if they existed in a stable form within the protons or neutrons.
Let’s put aside the fact that one cannot directly observe what is inside of a proton or a neutron and that it is quite a logical leap to assume that the extremely unstable particles that come out of a collision exist in a stable form within the non-collided particles.
It is possible that they exist in a completely different form, but that would be boring because it isn’t a testable hypothesis.
In search of new, testable hypotheses, the bored scientists got creative.
“What if these unstable particles exist in a stable form with a more exotic configuration? In a proton, we think that there would be 3 of these unstable particles, but what if there could be 5 of these unstable particles in a stable configuration? That would be a new form of matter called a pentaquark – very exciting.”
In 2002, scientists in a Japanese lab (LEPS) looked at their old data and found evidence of a pentaquark with close to 5 sigma significance – the gold standard of particle physics research.
Over the next few years, eleven other laboratories around the world looked back through their old data and, sure enough, they found pentaquarks too! The results ranged from 5 sigma significance up to an even better 8 sigma, but surely eleven laboratories couldn’t all be deluding themselves at the same time!?
Sorry to spoil the party, but they can.
By 2008, a scathing review of these results’ poor methodology was published along with several, less noisy experiments which looked for the pentaquark and didn’t find it.
The party is over. Eleven, top physics labs have successfully deluded themselves with 5 sigma significance and crafty data slicing and dicing. This is very disappointing…. and embarrassing.. especially since so many other major results like the Higgs are held up as being unimpeachable because of their 5 sigma significance.
But I bet you didn’t see any headlines about this embarrassment because Big Science PR has done a good job of whitewashing over it. This, however, didn’t stop physicists from whispering about the debacle.
CERN to the rescue!!!!!! CERN can save the high energy physics community’s reputation! The more random data you have, the easier it is to increase the ‘significance’ of a result and CERN has lots of data. With enough random data, you can always find what you are looking for if you are generous with how you define ‘noise’ and creative with how you slice and dice with efficiency vs. purity cuts.
In 2015, CERN’s LHCb found pentaquarks in their data with 9 sigma significance!!! This either proved that CERN can find whatever it wants in its data or it proved that the experiments which showed the pentaquark did not exist were wrong. I suspect that the former is true because I wonder what the expected probability of seeing the particle was in their application of Bayes’ theorem. It is hard to tell from their article.
In practical terms, they went through their old data and traced particle paths back until they found a collection of tracks from 5 different quarks that seemed to originate from the same general location – therefore, there must’ve been a pentaquark at that location.
Hmph. Must’ve. Unicorns are full of pixie dust and I saw 5 trails of pixie dust leading to a single point, therefore there must’ve been a unicorn which exploded.
I am being a curmudgeon. Maybe pentaquarks exist and I’m just being stubbornly skeptical of human error rates in the statistical analysis of large datasets. I’ve become biased after seeing too many stupid people who thought they were smart.
Rather than digging deeper into the details of particle physics experimental design, perhaps it is helpful to back up and ask a simpler question. Are there particles with charge but no mass?
Photons don’t have a charge, but they are officially without mass within big-bang, standard-model physics, whereas, they have a little bit of mass within steady-state, tired-light physics.
Z bozons don’t have a charge, but they aren’t really particles – they are force mediators or so-called “virtual particles” that pop into and out of existence over an extremely short timescale. When they exist, they borrow energy from the vacuum and temporarily violate conservation of energy.
Neutral pions are strange things composed of undefined amounts of different quarks and no one seems to care that no objects like that exist anywhere other than in the land of fundamental particles.
In the standard model, the neutrino is massless, but in the world of neutrino research, a neutrino is a chargeless particle with a tiny bit of mass. In the world of cosmology, the neutrino is whatever you want it to be.
Taken altogether, you have a zoo of particles which don’t objectively exist, but which cause reverse-engineered calculations of noisy collisions to deliver somewhat repeatable results for a community which is highly motivated to see what it wants to see.
It used to be that an army of students would be given the data to look at and decide what was a pion and what was an electron, but now there are algorithms that do that work for them. Yet when you have an algorithm that filters out only the picture that it wants to see, are you not tricking yourself? This was certainly the case when a group of phd students thought they had a picture of black hole or when gravitational wave researchers decided that they were looking at the signals of neutron stars colliding.
Neutrinos are a particularly contentious objects because some people think that Pauli made a mistake about spin when he invented the neutrino to carry extra energy away from a neutron decaying into a proton and an electron.
“But they have been measured and observed!” a neutrino researcher would protest.
“That depends on your measurement standards,” a skeptic would reply.
The first claimed observations of chargeless fundamental particles like neutral pions, Z bozons, and neutrinos happened in bubble chambers in which charged particles make tracks, but uncharged particles don’t. Later experiments automated these sorts of observations by using the large, stretched-wire drift chambers and Cerenkov light detector chambers that you see in photos of the LHC detector halls. Massive data sets are broken up and interpreted by armies of students armed with assumptions about what they are looking for in the noise.
Experiments or models which attribute actions to invisible entities in large, complex data sets for which multiple, self-consistent interpretations are possible should be handled with care. These sorts of measurements end up telling a story based on shadows cast upon the wall of a cave. Someday we may end up seeing alternative explanations of particle mechanics developed more in terms of pilot waves or bubble and fluid dynamics in the future and I keep an open mind to the possibility that we have made a mistake about spin and neutrinos – but if you say this in polite company today, they will decide that you are a crackpot who thinks that all of particle physics research to date amounts to giving silly names to the waves on an ocean which has been stirred up by a chaotic regatta.
Before consigning me to the crackpot bin, take a close look at some neutrino science:
Ice Cube detects neutrino induced muons, but ” There are a million times more cosmic ray muons than neutrino-induced muons observed in IceCube and I’m not sure that I buy the story of how they distinguish the two.
OPERA makes a muon beam and then blocks the muons in order to make a neutrino beam. Then they turn a tiny fraction of the neutrinos back into muons to measure the neutrino beam. Gamma rays can also make muons. Looks like a minefield of errors to me. This was the experiment that claimed that they had measured faster than light neutrinos and then had to retract the claim after finding that it was due to a loose fiber optic cable.
Th particle zoo gets even more crowded when you consider WIMPs. All neutrinos are WIMPs, yet not all WIMPs are neutrinos. WIMPs are Weakly Interacting Massive Particles and they were invented to explain why galaxies appear to be glued together with dark matter – matter that light travels through without interacting.
While working for no pay during maternity leave, Helen Quinn realized that all it takes to glue a galaxy together within standard model physics and general relativity is to postulate that space is filled with virtually unmeasurable WIMPs called axions.
Several other physicists copied this idea and came up with their own variation on a virtually unmeasurable WIMP.
This sort of thinking is nothing new in physics. When Pauli couldn’t get his math to correspond to reality, he just invented the neutrino – a virtually unmeasurable WIMP. Then all of the numbers worked out.
The funny thing about virtually unmeasurable particles is that if people try hard enough to measure them, they start to see them everywhere. It is sort of like looking at sea monkeys through a magnifying glass and seeing evidence that they are playing soccer or playing video games in their parents’ basements. If you look hard enough and long enough, you will find the evidence you are looking for.
This is great because professors need ever more fools’ errands on which to send their students, and inventing more virtually unmeasurable particles for them to measure has been keeping students busy for decades.
Neutrinos are the most well-known WIMPs, so why do we need to invent new WIMPs to explain dark matter? It turns out that neutrinos are so lightweight and move so quickly that they would’ve formed clumps as the universe evolved, so they are a terrible candidate to explain dark matter in a big bang universe. The WIMPs that big bang physicists are looking for are large, slow, and heavy – yet invisible and unmeasurable!
[insane interlude] Maybe we just need to look for these particles in our kitchens and nurseries? OR are they hiding in their parents’ basements and playing video games. In an economic sense, WIMPs consume energy and thereby glue our societies together. They must be fed! Do they eat light? When are physicists going to invent an INCEL particle? Incomprehensible Nuclear Concept Evolving Lightworkers would be a good acronym. They need to get laid! Is there a virtual sex particle which carries force between INCELs? Are INCELs majorana? Supersymmetry!
But seriously, what do WIMPs have to do with the Higgs? Since they both fill all of space, shouldn’t they be related to one another in some way? As in many physics problems, WIMPs and Higgs turn out to be two incompatible ways to say the exact same thing. The Higgs is like the viscosity that slows down or starts up rotational motion and mass is defined by rotational+linear motion. WIMPs’ mass comes from weak rotational+linear motion, so if a Higgs exists, WIMPs can’t exist because coupling to the Higgs field would quickly extinguish them. Likewise, if WIMPs exist, then the Higgs can’t exist because it is stupid to have two incompatible descriptions of the properties of space – one relating to nuclear spin and galactic spin and the other relating to why all fundamental particles except for WIMPs have mass. If one describes the motion of the hole and the other describes the motion of the stuff around the hole, it is impossible to describe both things in one language. The solution to this problem that physicists have come up with is that the Higgs and WIMPs can coexist as long as they never see one another.
Can we all agree that this story is ridiculous? Instead of talking about WIMPs maybe we could instead talk about how to properly apply relativity within astronomy or about light having a tiny bit of mass.
If photons have a tiny bit of mass, will they interact with other photons instead of just traveling past one another in a superposition of states? Does this interaction qualify as a particle? Gamma ray photons interact with one another by producing an electron and a positron, but we’ve never seen evidence of lower energy photons interacting. Why wouldn’t their collisions make smaller, unmeasurable sorts of particles like WIMPs? This would sure solve some problems for tired light cosmology.
This story has elements of self-consistency and inconsistency; WIMPs + tired light and the Higgs + big bang are two ways to say the same thing, but why do physicists insist on speaking in five, different languages at the same time? It is as though the tower of Babel fell down when the first nuclear bomb went off.
Now, about XENON – although the experiment is trapped deep beneath a mountain, it is not to be confused with Xenu of Scientology. While Xenon is a positively scintillating gas with a very cool name and an amazingly stable reputation, the idea of trying to identify whether evidence of a Xenon decay is caused by a virtually unmeasurable WIMP or by some more mundane, measurable phenomenon should cause you to cringe. If it does not cause you to cringe, you are probably a physicist who makes his living by virtually measuring unmeasurable things. Maybe you work for LIGO or you analyze images of black holes or you measure Higgs particles or pentaquarks. If you don’t cringe when you see this picture of an invisible neutrino, there isn’t really anything I can write that will deter you.
Something impossibly cringeworthy is precisely what the XENON collaboration has claimed to have measured with 3.5 tons of Xenon gas and a detector array. However, measuring the impossible did not deter the brave Italians in their underground lair at Gran Sasso. Did the air conditioning flip on unexpectedly and cause the detection? Who cares? When funding is threatened, anything is bound to be reported as a measurement. Publish or perish is the name of the game and nobody in science is going to quit the field of battle with honor. That would be unsportsmanlike. Fight to the death! Laus et Gloria!
If you made it this far without getting exhausted, you deserve a medal. Instead of resting on your laurels, buy one of my works of fiction and relax a bit while I go back over this article and attempt to make it more scholarly and cohesive. Or, keep reading. The conclusion is rather pithy.
Sabine Hossenfelder recently published a book called Lost in Math. It laments the “overproduction of worthless predictions” made by particle theorists. Whereas I focused here on the shortcomings of experimental work, she focused on the shortcomings of theoretical work and while I agree with her conclusions, I don’t think she goes far enough in her criticism of academic physics. She still acts as a cheerleader for many ‘big science’ experiments that appear illigitimate to my eye and this is no surprise because theorists are not good at identifying bad experimental design, even if they can identify bad theory design.
There are so many degrees of freedom within particle physics models that it is possible to make any number of predictions which could potentially be tested in some very distant, ideal future and when you have thousands of young people who are making large numbers of predictions which are either non-testable or testable-in principle-only, you end up with a community of people shooting in the dark and calling it theoretical physics.
That is one big way to be a bad scientist, but there are more. If you fine-tune your theory to match the data, that is also illegitimate. At CERN, they call some of the particle physicists “ambulance chasers” because whenever there is a rumor of some weird data, a hundred young people jump all over it and try to write a theory which matches it.
Looking at some of Dr. Hossenelder’s other blog posts, I am sympathetic with her criticism of the catch-all, fudge factor which is dark matter and of the cloak of legitimacy which the physics journals’ demanding style gives to even the worst nonsense.
I, however, take a more pessimistic view that the whole particle physics enterprise maintains its legitimacy by the sunk cost fallacy and the stylistically high bar to entry into the competition for physics glory. It is quite sad how many years of life have been poured into this stuff. Exacting style requirements and high bars should be the stuff of the Olympic games, not science.
The image in the header comes from Franco Banfi.