To start with, this post will be almost pure spoiler. I’m assuming, if you got through part 1, that you’ve read Cixin Liu’s book.

I’ve gotten partway through the second book in the trilogy myself, meaning that I’ve had some additional time to think about the contents of this post, but that I don’t know the ultimate outcome of the series.

This post is addressing a central conclusion of the first book, a major piece of science fiction that I didn’t address in the previous post because it is so intrinsic to the plot. This is about the idea of the Sophon induced ‘science lock-down.’ An alien race is going to invade the planet Earth in 400 years and this race is concerned that Human technology will advance in that time to be more powerful than the alien race’s own technology, so the aliens have played a trick to prevent humans from performing fundamental scientific research in order to prevent human technology from developing.

The key of this is the idea of the “Sophon.”As mentioned in the previous post, the word ‘proton’ was chosen over the name of an actual fundamental particle in order to facilitate a wordplay in Chinese… particularly the Chinese word that got translated into English as “Sophon.” This word was chosen from a modification of the word “Sophont.” As any science fiction aficionado can tell you, this word means “intelligent creature.” A Sophon is intended to be an intelligent proton, a robot the size and mass of a subatomic particle. These Sophons are capable to some extent of changing their size and shape and can communicate back to the aliens instantaneously. Sophons can also travel, as subatomic particles, at very nearly the speed of light.

You can see right from that paragraph the first place where the Sophon (and therefore the idea of science lock-down) are broken. Sophons communicate with the aliens instantaneously by means of quantum entanglement. If you’ve read anything else I’ve written, you know how I feel about the cliche of the ‘Ansible.’ Entanglement can’t be used to pass information: the Quantum mechanics doesn’t allow for this, no matter how you misinterpret it. Entanglement means correlation, not necessarily communication. This quantum mechanical effect is an interesting and very real phenomenon, but to understand what it actually means, you need to understand more about the rest of what quantum is… the story of ‘Three Body Problem’ never goes there. I won’t go there either except to suggest learning about the Bell Inequality.

The reason that Sophons are capable of producing science lock-down is because they can falsify data coming out of particle accelerators. Sophons can fly through the sensors in particle detectors and trigger them falsely, creating intelligently designed noise. At the surface, this is a horrible prospect, making it impossible for Humans to probe the deep structure of matter and therefore attain the understandings necessary to build Sophons ourselves. Do not pass go, no ‘correct’ results means no good science!

Obviously, this looks really bad. Very interesting science fiction idea. On the other hand, it also demands a bit of discussion, both about how particle accelerators work and on how science works.

Particle accelerators are the wrecking ball of the scientific enterprise. They generate data almost entirely by accelerating charged particles up to substantial fractions of the speed of light and slamming them into each other and into stationary targets. Particle physicists are all about impact cross sections and statistical probabilities of outcomes. The gold standard of a discovery in particle physics is a 5-sigma observation. ‘Sigma’ is, of course, standard deviation, which is a statistical standard by which scientists use the Gaussian statistical distribution to judge probability of occurrence –it’s the Bell Curve. Average is the peak of this curve, while one standard deviation is either one sigma to the left or right of average. Particle physics is set up around a simple statistical weight tabulation which can be couched as a question: “How likely is it that my observation is false/true?” If an event observed in the accelerator is spurious –that is, if the event is noise– the statistical machinery of particle physics places it close to the peak of the Bell Curve, that is at the average, which is to say that the event observed is ‘not different’ from noise. A 5-sigma event is an event which has been so well observed statistically that the difference from noise is five standard deviations from the peak of the Bell curve out into the tail (99.9999% of the curve’s area is captured within this extent of the tail!) This is essentially like saying that a conclusion is better than 99% certain to be NOT false.

Do you know how big a particle accelerator data set is? They include billions of events. Particle accelerators run for months to years on end, collecting data automatically 24 hours a day. And, the whole enterprise is based on the assumption that every observation independently might be a false outcome. Statistical weight determines the correctness of an observation. Physical theory exists to model both the trends and noise of an experiment.

As I said above, the purpose of the Sophons is to produce false results within the sensors of an accelerator’s detector apparatus. The most major detection devices in the modern systems are calorimeters and photomultipliers. Calorimeters simply detect heat deposition within the sensor volume while photomultipliers give a small current when they are perturbed by a passing electric charge. Usually, detector assemblies contain layers of sensors wrapped around the collision target where photomultipliers form multiple inner layers and calorimeters reside around the outside of the whole assembly. There are usually also magnetic fields applied through the detector so that charged particles will tend to follow curving paths as they pass outward through the different layers away from the collision site. There are other detector technologies and refinements of these ideas, but this gives a basic taste.

Here is the ALTAS detector at the Large Hadron Collider:


Using this layered design, photomultipliers can resolve the path of outward flying particles, determining their charges based upon their path curvature through the magnetic fields established by the solenoids and then the calorimeters determine how much energy was in the particle when that particle heats the calorimeter upon crashing into it. Certain particles types penetrate shields differently, necessitating layers of calorimeters with different structural characteristics in order to resolve different particle types. Computers correlate detection traces between the layers and tabulate what heat depositions relate to which flight paths. Particle physicists can then do simple arithmetic  to count up all the heats and all the charges on all the particles detected for one collision event and deduce which subatomic particles appeared during a particular collision. Momentum and energy/mass get conserved relativistically while charge is directly conserved and you simply add up what went in in order to account for what comes out during a collision.

In order to falsify data within such a detector, the smart subatomic particle, the Sophon, would need to fly back and forth through the detector layers, switching its charge polarity between passes and somehow dumping heat into calorimeters without being destroyed or lost in some way. How the Sophons get their kinetic energy is somewhat opaque in the story and I spent some time abortively rereading the TBP trying to figure this out, but it can be assumed that they possess a self-contained power supply which enables them to either recharge themselves from their surroundings, or simply dip into a long term battery reserve whenever they need it. They are clearly able to accelerate to highly relativistic velocities in a self-contained manner since they flew across the void from the alien homeworld to Earth, and then slowed down without external assistance at Earth. You could presume that they are able to write completely fake collision events into the detector, pretending to travel wrong velocities and masquerading as false charges and masses.

Now, like I said, this is terrible! The experiments can’t always give reliable results. Never mind that the real experiments must always be filtered for the fact that false results exist in the data set anyway.

In the paragraph above, I said “can’t always give reliable results” because the real data set of collision events still exists behind the fake data set. The Sophon flying back and forth can’t prevent real particle collisions from occurring and also interacting with the detector. The particle physicists would actually know right away that something isn’t right with the systematic structure of the experiment because they know how many particles are in their particle beams and also know the cross-sections of interaction, meaning that they start the experiment knowing statistically how many collision events to expect in a unit of time: Sophon interference with the experiment would only increase over the expected number. What you get is two overlapping data sets, one that is false and one that’s true. If the false data is much different from the true data, you inevitably bin them as distinct results because they would create a bimodal distribution to your data set… some measurements add up to five-sigma toward one result while a distinct set will ultimately add up as five-sigma toward something distinctly different. Then, you just let the theorists work out what’s what.

In the story, the scientists just throw up their hands and declare ‘sophon barrier’ saying that science ‘can’t advance’ because it can’t discern correctness.

This prospect has really kind of sat in the back of my mind, nagging me. I’m not completely certain that the author understands the overall scientific mindset or philosophy. Science starts out assuming that all results might be false! Having a falsehood layered on top of other potential falsehoods is really not that deterring to me, particularly since the scientists know the Sophon interference is present by the end of the story. Science as a process is intrinsically concerned with error checking and finding systematic interference, even intelligent fabrication of data within the scientific community –you think the Sophons are bad: somebody simply altering the data set as they see fit, completely independent of the experiment, is worse. And, we deal with this in reality! At least with the Sophons, a real data set must sit behind the mixture of false events. If the data set is merely bimodal or multimodal with statistics backing up each conclusion, you design experiments to address each… at some point, consistency of a result must ultimately dominate. Sorting out this noise would take time, but it would be unable to stop progress overall, especially since the scientists know the noise is present!

Now, giving false data is actually somewhat different than prohibiting data collection. This facet is somewhat unclear to me by the story –my memory fails. You can imagine that the Aliens realize that the humans know about the tampering and rather than leaving humans with a data set that contains some good data, they would simply have their Sophons swamp the detectors. In this, the Sophons fly back and forth within the detector giving so many false events that they prohibit the detector from being able to trigger for the resolution of real events. They could simply white us out!

While this would indeed be a bad thing, it would have a sort of a perverse effect on a real scientist. Consider: you know how fast your instrument triggers and you know the latency required for it to recover… this gives you a measure for how quickly and in what frequency the Sophon must act! You can just imagine the particle beam physicist salivating at the prospect of his Nobel prize in the nascent field of Sophon physics. Imagine the flood of grant proposals around the subject of baiting a Sophon into a particle beam line by the performance of basic science only to try to turn the particle beam against the Sophon in order to smash it apart and see how it works!

Really, if you were a high energy physicist and you knew unequivocally that a smart particle was flying around inside your instrument, how could you not be trying to figure out a way to probe it? It’s like getting Maxwell’s demon handed to you on a shiny platter!

A realistic outcome here is actually not the prohibition of science. It would be an arm-wrestling match with the Aliens: at the very best, leaving us with a partial data set that we can ultimately advance with, or giving us the chance to probe the Sophons directly.

The prospect of probing the Sophons directly contains the danger that it would be hard to distinguish engineered results from real ones, but every demonstration by the Sophons of some other confusing behavior is in fact data itself. The author made a huge argument in “Three Body Problem” that Sophons are typically point-like and would probably subscribe to the notion that they can’t be probed since they would essentially have no collision cross-section; I would resist this idea because it either violates or misunderstands quantum mechanics, which I detailed a bit in the previous post. The author might even suggest that Sophons can’t be probed because they can dodge collisions with other particles in the collider, but I would doubt that simply because of the inability for the Sophon to know things about other particles due to simple quantum mechanics and the affect of relativity altering the rates of information flow: the decision would need to be made very quickly and it would have a built in imprecision from Uncertainty! Moreover, the more time the Sophons spend performing confusing behavior in order to foil their own direct examination, the less time they can spend faking data in the experiments directed at basic research. As you may be aware, machines like the LHC are actually devoted to many lines of research simultaneously and physicists are remarkably adept at piggybacking one experiment on top of another in order to conserve resources and obtain additional bang for the same buck.

One final aspect of the “science lock-down” which I take some umbrage with is the notion that only particle accelerators are responsible for fundamental research. They aren’t. There is a huge branch of physics and chemistry probing quantum mechanics based on spectroscopy. Lasers are unequivocally a quantum mechanical device and much probing into basic quantum mechanics is performed by some variation on the theme of lasing. The Nobel prize winning discovery of the Bose-Einstein condensed matter phase did not occur in a super-collider, it occurred on an optical bench. Most super precise clock mechanisms used by the human race at this point are optical devices and have absolutely nothing to do with particle accelerators –optical gratings and optical metrology are driving the expansion of precision measurement! The leaps which are in the process of producing quantum computers (one device the author specifically prohibits in book 2 under the science lockdown!) are not being made at particle accelerators at all: they are being made in optical lattice traps on lab benches and in photo-etched masks used to produce nano-scale solid state resonators. We are currently in the process of building analog quantum computers for the purposes of simulating quantum chromodynamic systems using optical and nano-resonator devices… and this development has nothing to do with particle accelerators, except as a means of reproducing results! The author made the argument that humans couldn’t build massive super-collider accelerators, Synchrotrons and Linacs, fast enough to match the production capacity that the Aliens have for making the sophons needed to foil these instruments, but the author never even touched on the rapidly expanding field of plasma wake field acceleration, which uses lasers to accelerate particles to relativitistic speeds in bench-top apparatuses for a fraction of the price of a super-collider.

The bleeding edge of physics is very multi-pronged; the Higgs boson discovery carried out in a synchrotron may someday be reproduced by a bench-top plasma wake field accelerator for a tiny fraction of the price. Can ‘locking down’ big particle accelerators like the LHC prohibit the extensive physical exploration that is occurring due to a mostly unrelated black swan technological development like lasers? I really don’t think it can. Tying one arm behind your back leaves you with the other arm. It’s true that the mothballing of the superconducting super-collider in the United States prevented humans from definitively discovering the Higgs boson for more than a decade, but that isn’t to say that there aren’t other avenues to the same discovery.

Do I think that science lockdown is possible by the means suggested by the author? Not really. And, especially not for devices like quantum computers, which is one critical development that the author suggests is prohibited by sophon interference in the second book.

Don’t get me wrong, this is a good piece of science fiction and it’s a wonderful thought experiment, but like many thought experiments, it’s arguable.

edit 2-16-17

I saw a physics colloquium yesterday delivered by a Nobel prize winner. His lab is currently working on a molecular spectroscopy experiment directed at measuring the electric dipole moment of the electron. A precision measurement of this value ties directly to the existence (or not) of supersymmetric particle theory… which is one candidate expansion of the Standard Model of particle physics. This experiment is not being done in a super collider, but on an optics bench for a fraction of the price. Experiments like this one completely invalidate the thesis of Three Body Problem: that by locking down colliders that there is no other way for particle physics to advance. There are other ways that are comparatively cheap and requiring less resources and manpower. Physics would find a way.

edit 3-8-19:

I wanted to expand on a criticism that came up in the comments:

It may not even require to alter results. Just mess with computer memory like bit flips to render results meaningless. This maybe easier than to change all and faster. Our ECC engines in ram will not be able to correct all.

I think this is a very fair comment. Our instrumentation is not restricted solely to the sensor end of the instrument. Why not interfere with the direct functioning of the computers that are recording the data? In the comments, I point out that a “bit” is not a discrete object. This response doesn’t really tackle the suggestion in terms of my overall thesis, that science itself would be hard to stop.

The simple fact here is that “bits” in a computer are perhaps one of the best understood physical objects in existence. They are not discrete in the manner of –say– an electron; they are an engineered assembly of many fundamental parts. We have been working on this engineering very very hard for the better part of something like 80 years now. A “bit” in the discrete sense is an assemblage that can assume two states where the energy difference between these two states is tuned by a huge amount of engineering and expertise. The reality is that the energy difference between these states has been reduced along with the size of the device, making smaller bits easier to flip and smaller CPUs much more energy efficient. However, not all bits are equal; they vary in patterning size, in response rate and in the energy required to set them. A Sophon has the energy to flip retinal trans to cis or about 3.5×10^-19 joules (it can write images in the human eye), which is at about the same energy is would be the bandgap in a semiconductor, but this bandgap can be tuned by doping and adjusted the bias voltages. You simply start looking for the settings that Sophons don’t have the power to effect.

If the ‘bits’ of a computer are a target for the Sophons, these can also be used to probe Sophon behavior simply by including computers with varying types of CPUs in a target that the Sophons are sure to try to take down, like an accelerator. We have decades of experience building these devices where electronic warfare is a potential use for them. Building resilient bits is done.

Now, part of the reason that particle accelerators are a target of Sophons is because the human industrial capability to build accelerators is very small and can’t overcome the trisolarian capacity to build Sophons. The same is not true for computer parts: we can build computers tremendously quickly, in frightening bulk… if the Sophons are swamping the accelerators by toying with the computers driving those instruments, you just start building computer redundancy until you swamp the Sophons, say by parallelizing your data collecting equipment and encoding data by encryption across many bits. Worse, if you can come up with a simple desktop experiment that has parallel utility to an accelerator, which can be done as I mentioned above, you just start mass producing it and watch for when the Sophons stop being able to effect all the duplicates.

Because it would require less energy and less numbers to effect, I still feel that the weakest link is probably the sensor. A Sophon is tailor-made to cosplay as a subatomic particle; my feeling is that it would be less ideal at trying to affect the higher energies and numbers necessary to disrupt digital electronics directly.

Published by foolish physicist

Low level academic enthralled with learning how things work.

Join the Conversation


  1. Great piece . I was so looking forward to starting this series but the “magic” sophons are driving me nuts. It took me 30 seconds and a quick search to find articles about why they are nonsense. Suspension of disbelief is all good and well in a Universe where magic exists, or superheroes etc but this is supposed to be “hard sci-fi” and trying to get through it is as arduos as that time I gave Dan Brown a go!!


    1. Well, I wouldn’t say it’s an awful piece of writing. If an author does too badly, I usually don’t finish the book and I finished this one. As they say, any sufficiently advanced piece of technology will appear to be magic; the problem most authors tend to face is that they feel compelled to try to explain how it works when many of them have little idea how pretty much everything actually works. IMHO, this author is better than some. To make bank, an author just needs to be believable to the average member of his/her target audience. I don’t think this book is a waste of time, just that it, like many others, stretches credulity;-)


  2. Thank you for sharing this. I love the trilogy, and am always happy to suspend disbelief for a good story, but I’m even happier when someone helps me learn something new in connection with something I love.



  3. Of course since there are no sophons there can be no technological block – that is a malicious rumour and everything you have said is correct.


  4. It may not even require to alter results. Just mess with computer memory like bit flips to render results meaningless. This maybe easier than to change all and faster. Our ECC engines in ram will not be able to correct all.


    1. Perhaps, but a bit is not an indivisible unit; a bit is itself a composite of many indivisible units. However that physical bit is defined, they are not the same as indivisible particles. Within a particle accelerator, the sensors are deliberately designed to enrich signals obtained from subatomic events into detectable phenomena… that is a large degree of sensitivity. Physical bits, on the other, are already macroscopic and are intended to be controlled by very much larger signal manipulations. It would take a great deal more effort to change a bit (particularly if you start talking about data that is written by a fourier-like encoding into a space of many bits, making the pattern of the bits less corruptable.) I grant you, there are many places the Sophons could interface with a sensor system, but there are drawbacks to each.


      1. My issue is the following:
        – if you can mess, at all with computer memory, why not alter the world’s financial data and cause societal mayhem?
        – If you can mess, at all, with sensors, why not mess with airplane navigational systems and make air travel unreliable?
        – If they can alter your eye to display images, what else can they do to our bodies? Could they give everyone a brain tumor?

        I don’t understand the science well enough, but on first reading it seemed to me that the plot device invented to facilitate this idea that the aliens are messing with our scientific development was so powerful, that it made the plot itself meaningless. If they actually had sophons, as described in the book, we would be pretty much doomed already. They wouldn’t need to stall our science, they could just wreck our society. Kinda ruined the book for me.


  5. The proposition was not that all science was at a dead end from the time the sophons arrived. The proposition was that there was no way to advance into fundamentally new and necessary avenues of science. An example is electrical energy generation. All our most energy dense methods of power generation still rely on burning things or boiling water. We’ve refined the reliability and miniaturisation of related machines and such, but we’re still burning things and boiling water. Everything else we have in power generation is either hoping that something will move and impart kinetic energy, or hoping an existing charge will be imparted into something else which can make use of some of that charge; and both of those options are very low on density. How can that situation be remedied? Well, we’d need some big breakthroughs related to the fundamental nature of reality, energy, and so on. While advancement in some areas would still occur, those necessary breakthroughs would be blocked. Having progressed further than humanity, the Trisolarans know precisely which types of experiments and observations need to be blocked, and thus don’t need to mess with *every* scientific endeavour. This is actually addressed in the story.


    1. And you, my friend, have missed a big part of my point. Science is so interconnected and multi-pronged that the “necessary advances” cannot be blocked by hitting only one avenue of scientific endeavor, like blocking data collection at synchrotrons. Keep in mind that this is what you’ve said: that a particular advance in understanding the nature of reality won’t happen if this one avenue is blocked, which is wrong. It’s possible to make the same advances by other cheaper, more numerous, less easily suppressed techniques (for instance: bench-top laser plasma wake field accelerators as an alternative to big synchrotrons or LINACs). This is not addressed in the story. Moreover, Mr. Liu does not understand physicists; telling us a tantalizing bit of knowledge is available and hanging it in front of our faces with confirmation that it is definitely possible is really not going to stop us from wanting to learn it; it’s going to make us try harder to learn it and push us to be more inventive. This would proliferate and accelerate innovation, not suppress it. What happens when somebody comes up with a parallel technology to a synchrotron which is too easily mass produced for the Sophons to suppress? Parallel experimental technologies already exist today. This is not how scientists were characterized in the story at all. He more or less assumes that smart people never talk to each other.


      1. Say you time travel to the 17th century, and you want to foil Newton, Faraday, Maxwell, Hooke and other contemporaries’ experiments. It may be hard.
        Now say you want to prevent termite colonies to form in that house in the middle of wilderness you’re moving to next year. The fiction in the book invites you to consider a civilization that thinks of us as “bugs”.

        Even when Wallfacer planned to destroy the Trisolarians’ answer was “the Lord doesn’t care”. Physicists in the book were committing suicide, but Luo Ji was a target of an assassination attempt.

        My interpretation from the book is that Trisolarians were afraid of discoveries that would render their invasion pointless because of the “Dark Forest Hypothesis”, not because of anything mankind could do to them.

        A stretch of imagination is required for the more informed reader to assume that aliens have circumvented no communication theorem’s limits, and that we live in a universe where local realism is violated quite casually.
        Such stretch of imagination, in turn, allows readers to assume those aliens have a radically different view of existence, which I personally find satisfying, and intriguing, like good fiction only can be.


      2. That is one way to consider it. I would still offer that by saying that “Physicists in the book were committing suicide” you’re more or less agreeing with the author’s caricature of what a physicist is and how they respond to the world around them, which I would still suggest that the author himself doesn’t really understand. Keep in mind that it can also be argued that the Trisolarians engaged their attempt to stunt human intellectual growth too late since Luo Ji, a human scientist (yes, sociology is a science), was still able to have an idea like the Dark Forest Hypothesis which more or less totally mooted the Trisolarian invasion –no need for an operable synchrotron. Doesn’t that ultimately just prove my point about Black Swans? Physicists are a varied wealth of people and many get out of graduate school with PhDs and then go to work as economists, or political lobbyists, or even go and join the clergy. I know a few that have gone into biosciences or even social sciences (a friend of mine is working in a psychology department.) Physicists are not a size six glove… they are another scientist among many who contribute to the tapestry of what humanity knows in collaboration with a lot of other members of the scientific establishment. From the events in “The Dark Forest,” it’s pretty damn clear that the Trisolarian “science lock-down” utterly failed since human science beat them before they even truly got a chance to invade.

        If the author truly wants to envision a civilization that thinks of us as bugs, why were they so easily beaten? (I haven’t read the third book yet, so maybe I’m wrong, but I’m betting the logic is consistent with both the second and the first books.) You have to admit, Luo Ji’s solution was a very easy one.


  6. I read only through the first portion of this article so I might missed something but as far as I remember sophons don’t interfere with the sensors but with the particles – it is clearly explained that they catch the particle before it collides with anything else – being able to affect the collision also. This is different than applying noise to the results – it’s substituting the results all together. They intercept the accelerated particle before it collides with anything – this is the core of the sophon blockade – it’s not creating noise on the sensors.


    1. I suggest you go back and read the book again. Your interpretation is totally aphysical and more or less makes the idea of the sophons patently impossible. What we see here is one example of the core problem of fictional writing; all readers believe they understood the author’s intent, and since the author is not bound to actually reflect the reality of a given subject, all interpretations end up being equally plausible. However, when you make an action as completely magical as this one would be, physics simply doesn’t support it any more. That violates a core tenant of science fictional writing, that the magic the author is spouting somehow follows from a scientific principle. The book explains quite clearly, with chapters of exposition, that the trisolarians target is the apparatus of terran _science_, an -instrumental- weakest link, not the unfolding of reality itself. The blockade is to prevent the forward progress of science, not to hoover up elemental particles so that terrans can’t find them. Please please, next commenter, exhibit more reading comprehension.

      I’ve debated adding a full discussion of the physics as to why this suggestion is a ridiculous one and I think it would take more time than is due. I will reduce it to a simple argument of quantum mechanics (which the author either mutilated or ignored in defining most of what Sophons can do). The Sophon magically finds and grabs the tiny undiscovered particle and throws it into it’s tiny dimensionless trunk in order to kidnap it; tiny particle then exhibits wave duality like all things with physical dimensions tend to do and tunnels out of the Sophon which –as defined by the author himself– is too small to hold onto anything with any measurable physical extent whatsoever. This, my friend, is why Sophons can’t kidnap fundamental particles.


Leave a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: