This is either quackery or a Nobel prize worthy breakthrough. I'm inclined to go with the first.
Basically they say they can explain the galaxy rotation speed problem using some novel concepts of entropy and information instead of traditional explanation involving dark matter. If they are right, this will bag them a Nobel prize with 100% certainty.
But I'm afraid they aren't right.
* neither author is an astrophysicist
* one's interests are [1] :"Current interests include flexgrid & elastic optical networking technologies (subsystems & devices) & architectures, converged wireless-wireline access networks (5G), theory of information & entropy, cryptography & security."
* the other author describes his interests as follows [2]: "My own interest is accurate ion beam analysis (including RBS and PIXE) using a 2MV Tandem"
* 5G networks and ion beams don't have much in common with cosmology
* the second author is also considered a Biblical scholar [3]. Here's one little glimpse into how he understands the world: "The Kingdom is about the return of the glory of God. Which widgets you have in your pocket is hardly relevant to that."
Judging a paper by the background of the authors is a sure sign that you were not able to read the paper and understand it. There is no actual astrophysics in this paper, only Lagrangian mechanics, geometrical algebra and information theory. Anyone who works with ion beams uses Lagrangian mechanics. Anyone who works with rf networks uses information theory. So, I'm not sure why their backgrounds would be a problem here. And seriously, holding someone's mainstream religious beliefs against them? That's just plain uncivil. I apologize if this is harsh, but I think that responding here is better than just downvoting you.
All that said, I agree with a part of your assessment, but for other reasons. The crux of the paper is a transformation from Lagrangian mechanics (in other words, the motion of particles with no frictional forces) into a form of 'entropic mechanics' used for calculating the entropy of a certain configuration of particles. They defined this transformation using a coordinate system which matched the spiral/helical shape that they were focusing on (see Eq 9a). In a way, this is a bit of a cheat, because had they used a fixed cartesian coordinate system I suspect the model would produce highest entropy when particles aligned with the x,y,z axes.
However, the interesting part of the paper is where they make an independent prediction of the energy to change B-DNA to p-DNA. This is the only non-tautology prediction in the paper, and it is interesting that their model gives a prediction close to the measured value. This means that their alternate 'entropic mechanics' means of calculation might be useful because it could simplify complex calculations.
The galactic stuff is basically a tautology. First, they are only calculating shape, and not doing anything with the velocity distribution (which is where the dark matter problem comes from). They say that most of the entropy is in the SBH in the galactic center, but then when they try to calculate the temperature of the remaining galaxy, they make a stability assumption that just happens to give the galaxy a thermodynamic structure which matches their spiral form (section right before eq 27a). Amazingly, this give a maximum entropy when the galaxy has a spiral structure!
So, from reading the paper only once, my assessment is a) this method could be useful for calculating entropy for particles in fixed geometries, b) the galaxy stuff is crap.
> Judging a paper by the background of the authors is a sure sign that you were not able to read the paper and understand it.
Reading a paper requires some effort. The more familiar you are with the topic, the less effort you invest, but the effort is there.
It's valuable then to apply some heuristics to decide if a paper is worth reading or even skimming. One such heuristic is if the paper makes it to the front page of HN, or, say, if it's published in Nature. Based on this, a lot of people could waste some time they better spend doing something else.
If someone comes and points out some red flags, that's a service to the community. That service comes at a risk to one's own karma, as it's threading the boundary of what the HN community thinks is good discourse (for example the charity principle). Pointing out red flags without pointing out actual errors in the papers can be easily equated with ad-hominem attacks (believe me, I thought ten times before deciding about quoting the second author's thoughts on God).
I still decided that this is a good case to put my karma at risk: the claim that some new information-entropy force can explain the missing mass in the Universe is too important to ignore.
That said, your more informed analysis obviously adds much more value than my superficial one, so thank you for that.
I appreciate your measured response to my harsh criticism.
I agree with you that some claims are too important to ignore. I can tell that you believe that you were adding value, and I cannot fault you for that. However, the red flags which you see as a useful heuristic, I tend to see as anti-useful. Perhaps we may just disagree on that.
A: A CS researcher who specializes in 5G networks, and who does not describes his interests as including astrophysics, makes a breakthrough discovery that provides a correct alternative to dark matter
B: non-A
Which one is more probable? I would say B easily by a factor of 100.
Let's consider 2 other hypotheses:
C: someone who can read a paper and point to its conceptual flaws lists some red flags about its authors
D: that person actually can't read the paper
I guess D is about 100 times more likely than C, isn't it?
Considering that we all engage in probabilistic inference like that, why is it wrong to communicate its conclusions?
You may have 2 answers to that:
1. it's against the old adage to critique the work, not the person; I'm pretty sure that's the answer
2. the uneducated masses should refrain from criticizing the ivory tower. And considering that they can't really understand the math, that means they shouldn't even try to use lateral heuristics, even if they try to give them the mathematically sounding label of "probabilistic inference". If there was no Markov Chain Monte Carlo or variational Bayes involved, it's not probabilistic inference :)
7 comments
[ 0.22 ms ] story [ 28.5 ms ] threadBasically they say they can explain the galaxy rotation speed problem using some novel concepts of entropy and information instead of traditional explanation involving dark matter. If they are right, this will bag them a Nobel prize with 100% certainty.
But I'm afraid they aren't right.
* neither author is an astrophysicist
* one's interests are [1] :"Current interests include flexgrid & elastic optical networking technologies (subsystems & devices) & architectures, converged wireless-wireline access networks (5G), theory of information & entropy, cryptography & security."
* the other author describes his interests as follows [2]: "My own interest is accurate ion beam analysis (including RBS and PIXE) using a 2MV Tandem"
* 5G networks and ion beams don't have much in common with cosmology
* the second author is also considered a Biblical scholar [3]. Here's one little glimpse into how he understands the world: "The Kingdom is about the return of the glory of God. Which widgets you have in your pocket is hardly relevant to that."
[1] https://www.researchgate.net/profile/Michael_Parker12
[2] https://www.researchgate.net/profile/Chris_Jeynes
[3] https://www.churchtimes.co.uk/articles/2018/8-june/features/...
All that said, I agree with a part of your assessment, but for other reasons. The crux of the paper is a transformation from Lagrangian mechanics (in other words, the motion of particles with no frictional forces) into a form of 'entropic mechanics' used for calculating the entropy of a certain configuration of particles. They defined this transformation using a coordinate system which matched the spiral/helical shape that they were focusing on (see Eq 9a). In a way, this is a bit of a cheat, because had they used a fixed cartesian coordinate system I suspect the model would produce highest entropy when particles aligned with the x,y,z axes.
However, the interesting part of the paper is where they make an independent prediction of the energy to change B-DNA to p-DNA. This is the only non-tautology prediction in the paper, and it is interesting that their model gives a prediction close to the measured value. This means that their alternate 'entropic mechanics' means of calculation might be useful because it could simplify complex calculations.
The galactic stuff is basically a tautology. First, they are only calculating shape, and not doing anything with the velocity distribution (which is where the dark matter problem comes from). They say that most of the entropy is in the SBH in the galactic center, but then when they try to calculate the temperature of the remaining galaxy, they make a stability assumption that just happens to give the galaxy a thermodynamic structure which matches their spiral form (section right before eq 27a). Amazingly, this give a maximum entropy when the galaxy has a spiral structure!
So, from reading the paper only once, my assessment is a) this method could be useful for calculating entropy for particles in fixed geometries, b) the galaxy stuff is crap.
Reading a paper requires some effort. The more familiar you are with the topic, the less effort you invest, but the effort is there.
It's valuable then to apply some heuristics to decide if a paper is worth reading or even skimming. One such heuristic is if the paper makes it to the front page of HN, or, say, if it's published in Nature. Based on this, a lot of people could waste some time they better spend doing something else.
If someone comes and points out some red flags, that's a service to the community. That service comes at a risk to one's own karma, as it's threading the boundary of what the HN community thinks is good discourse (for example the charity principle). Pointing out red flags without pointing out actual errors in the papers can be easily equated with ad-hominem attacks (believe me, I thought ten times before deciding about quoting the second author's thoughts on God).
I still decided that this is a good case to put my karma at risk: the claim that some new information-entropy force can explain the missing mass in the Universe is too important to ignore.
That said, your more informed analysis obviously adds much more value than my superficial one, so thank you for that.
I agree with you that some claims are too important to ignore. I can tell that you believe that you were adding value, and I cannot fault you for that. However, the red flags which you see as a useful heuristic, I tend to see as anti-useful. Perhaps we may just disagree on that.
I hope to be able to change your view.
Consider hypotheses A and B below:
A: A CS researcher who specializes in 5G networks, and who does not describes his interests as including astrophysics, makes a breakthrough discovery that provides a correct alternative to dark matter B: non-A
Which one is more probable? I would say B easily by a factor of 100.
Let's consider 2 other hypotheses: C: someone who can read a paper and point to its conceptual flaws lists some red flags about its authors D: that person actually can't read the paper
I guess D is about 100 times more likely than C, isn't it?
Considering that we all engage in probabilistic inference like that, why is it wrong to communicate its conclusions?
You may have 2 answers to that:
1. it's against the old adage to critique the work, not the person; I'm pretty sure that's the answer
2. the uneducated masses should refrain from criticizing the ivory tower. And considering that they can't really understand the math, that means they shouldn't even try to use lateral heuristics, even if they try to give them the mathematically sounding label of "probabilistic inference". If there was no Markov Chain Monte Carlo or variational Bayes involved, it's not probabilistic inference :)