Why is there so much crackpot physics?

[catsmate]

As far as I can tell, most physics crackpottery comes from outside the scientific community.

I think most of it comes from those on the periphery of science, the engineers, PMs and others with just enough knowledge to make mistakes and not enough to understand how little they really know.

 

[lpetrich]

I think most of it comes from those on the periphery of science, the engineers, PMs and others with just enough knowledge to make mistakes and not enough to understand how little they really know.

I concede there. Many physics crackpots are engineers, and engineering is a sort of applied science.

 

[Sam H]

(some mainstream scientists who have indulged in crackpottery...)

How would that indicate that they are to blame?

When the attack comes from within it has a much more devastating effect than otherwise. A slight diatribe on the Democratic Party by Obama is a thousand times more harmful that thousands of attacks by Rush Limbaugh.

As far as I can tell, most physics crackpottery comes from outside the scientific community.

That is true. However, if a physicist promulgates crackpottery, people listen more credulously. Just look at the effect The Tao of Physics has had on the general public. It has been translated into dozens of language and sold millions of copies.

 

[Sam H]

It's possible to build and analyze Crookes tubes using Newton's Laws. ... That all works just fine as long as v << c.

I am not questioning Newton's Laws. They work not only in Crookes tubes, but ANYWHERE in the MACROSCOPIC universe as long as v<<c.

As it so happens, Crookes et. al. only had access to voltages in the 10kV range, so v << c is all they ever got.

And that (rather, voltages in a few Mv range) is also all Ernest Lawrence got in his cyclotron.

IF higher-voltage power supplies had been available, they would have started seeing deviations from F=ma, ...

Deviations, yes.

... and that could have been an experimental discovery of SR.

Discovery of SR, no. Not by any stretch of imagination! Leverrier saw a deviation of Newton's laws in 1845, but that did not lead to SR. It led to the discovery of Neptune. A deviation could lead to infinitely many directions!

The point is that to construct a machine that can accelerate particles close to the speed of light YOU NEED RELATIVITY. In other words, if you are NOT equipped with the theoretical knowledge that Einstein and others discovered DEDUCTIVELY in 1905, you CANNOT construct an accelerator (linear or circular) that is capable of achieving relativistic speeds. That is why cyclotrons were changed to synchrotrons (to synchronize the motion with relativistic deviations from F=ma), and that is why in the construction of the 2-mile linear accelerator at Stanford, relativistic effects - such as length contraction - had to be taken into account. So Crookes, et. al. would NEVER have reached relativistic speeds because they did not know SR ... even if they had trillion volts available to them - whose achievement is another story of dependence on SR!

 

[ben m]

The point is that to construct a machine that can accelerate particles close to the speed of light YOU NEED RELATIVITY. In other words, if you are NOT equipped with the theoretical knowledge that Einstein and others discovered DEDUCTIVELY in 1905, you CANNOT construct an accelerator (linear or circular) that is capable of achieving relativistic speeds!

Seriously, no you don't. If you built a non-relativistic accelerator, and tried to run it harder and harder, the "anomaly" that you would see is that particles travel slower than you expect given their energy or momentum. Somebody would make up a fudge factor---let's call it "gamma"--to describe the difference between the expected momentum and the observed momentum at a given velocity. Comparing experiments with nuclei, protons, electrons, it'd be clear that gamma is a function of velocity only. Lacking a good theory, people would suggest various functional forms for the mysterious "gamma", and one of the reasonable generic choices would be gamma=1/sqrt(1-a*v^2). (Of course other functions would be suggested too.) As the data get more and more precise, especially at higher v, the mysterious constant a would be measured to be very, very close to 1/c^2.

Similar experimental-deviations would establish a fudge factor distinguishing measured kinetic energies from Newton's E = 1/2mv^2. Gamma would turn up in the description of these measurements, too, and the simple KE=(gamma-1)mc^2 formula would be a good fit to this data.

That's not an actual "theory of relativity", but it gives you everything you need to build accelerators. It tells you how to set the magnets, accelerating cavities, frequencies, etc., to build linacs, or cyclotrons, or synchrotrons, at any energy you like. To build a high-energy accelerator you only need to know the laws of kinematics (i.e. the energy/velocity/momentum relations) in your own reference frame. You do not need to have the complete set of reference-frame transformation laws. It just doesn't come up.

The data themselves will point to v=c being a speed limit for particles of all masses; and c was previously known to be relevant only to studies of light and electromagnetism. It's pretty obvious that the presence of c in the mystery "gamma" factor would inspire thinking about the relationship between Maxwell's Equations and mechanics. It's pretty obvious that the new p/v/E formulas aren't invariant under Galilean transformations, too. Maybe we get straight to SR from there.

If not, once we're building accelerators, that means we're smashing nuclei together, which means we're obtaining short-lived particles (pions, nuclear excited states, etc.) that decay in-flight in our spectrometers and detectors. That gives us the means of measuring decay lifetimes, which (to everyone's surprise) will appear to be velocity-dependent, and gamma will turn up yet again. If that doesn't put the last piece into the puzzle, I don't know what would.

 

[Sam H]

Seriously, no you don't. If you built a non-relativistic accelerator, and tried to run it harder and harder, ...

Sorry, but I'll have to stop you right here! You cannot make a non-relativistic accelerator run harder and harder. You need relativity to do so.

What you CAN do is invent a non-relativistic machine that runs hard enough and try to make it run harder and note that it fails, and if you are lucky enough, you may discover WHY it fails. So you go back to the drawing board and try a new machine that MAY work for the maximum speed of the first machine, but you'll witness its failure at yet higher speeds. Construct a third machine ... I am reminded of the monkey and the typewriter. The monkey may eventually type a sentence, but the odds are very small!

 

[DeiRenDopa]

Sorry, but I'll have to stop you right here! You cannot make a non-relativistic accelerator run harder and harder. You need relativity to do so.

What you CAN do is invent a non-relativistic machine that runs hard enough and try to make it run harder and note that it fails, and if you are lucky enough, you may discover WHY it fails.

To quote a certain Sam H, sorry, but I'll have to stop you right here!

One thing almost any of the superb experimentalists of a century or so ago would have very quickly discovered is an empirical relationship - several more likely - that describes this failure. It's quite possible such would be in a form similar to what ben m describes; also possible they'd be in a quite different form.

Assuming access to sufficient resources to perform the various tests which all these kinds of empirical relationships would suggest - and I think you'd be very hard pressed to make a strong case that the unavailability of such resources would be a genuine show-stopper - my guess is it'd take the physics community a decade, maybe two tops, to derive exactly the kind of relationships ben m mentioned (or their equivalents).

So you go back to the drawing board and try a new machine that MAY work for the maximum speed of the first machine, but you'll witness its failure at yet higher speeds. Construct a third machine ... I am reminded of the monkey and the typewriter. The monkey may eventually type a sentence, but the odds are very small!

Perhaps you'd benefit from getting to know some experimental physicists? Or perhaps you could get an idea of what's possible by reviewing what Galileo - to take just one example - was able to discover, in what we today call physics, with exceedingly modest equipment?

 

[edd]

Sorry, but I'll have to stop you right here! You cannot make a non-relativistic accelerator run harder and harder. You need relativity to do so.

Early accelerators were just enormous van de Graaf generators. They were one stage devices reaching MeV potentials. They didn't need the timing of multiple acceleration stages. It really isn't clear that you need SR to get to the point you can drop enough energy into a particle for its KE to be an appreciable amount of its rest mass.

Without massive specialist knowledge I am inclined to take ben m at his word.

 

[ben m]

What you CAN do is invent a non-relativistic machine that runs hard enough and try to make it run harder and note that it fails, and if you are lucky enough, you may discover WHY it fails.

Sam H, two things. First, note that since we have particles of different masses, it's possible to build very powerful, fully-functional, and nonrelativistic proton accelerators ... which have all the ingredients to take an electron to extremely high velocities.

For example, it's reasonably easy to generate 500 kV DC using a Cockroft-Walton voltage multiplier. It's reasonably easy to generate 1MV using a Van de Graaff. Any singly-charged particle, electron/proton/ion running through a 500kV potential will pick up 500 keV of energy (this is true both in Newton and in SR); electrostatic accelerators work (in terms of producing a beam) even if you have the kinetic theory wrong. Moreover, such an accelerator can fairly easily run multiple species (even e- and H- simultaneously, if you like) of different masses. However, whereas a 500kV proton will behave within a fraction of a percent of Newton's Laws, a 500 kV electron will have a magnetic rigidity 40% higher than expected, and a velocity 56% lower than expected.

This sort of behavior would be extremely easy to spot when starting up and operating such an accelerator. In particular, a measurement of the magnetic rigidity of the beam (as a function of energy) is more or less a trivial, routine operation. You turn the high voltage up to some value, then gradually turn up the B-field in an analyzing magnet (in electrostatic machines we call it the "selector") and see what magnet-current gives you what beam-transmission. I've done this calibration myself in modern electrostatic accelerators. Maybe the first time you do it, and find the beam coming through only with 40% more magnet current than you expected, you'd be tempted to write it off as "magnet nonlinearity", but since no such nonlinearity would be visible for proton/ion beams, this red herring would not stand up long.

 

[ben m]

And let me repeat that the quantity you can measure in such an instrument----the E/v/p relationships--- is the only bit of relativity you need to build an accelerator. Accelerator design and operations do not require knowledge of length and time contractions, except in very special cases.

 

[rcfieldz]

For most it appears to be a mathmatical shell game.

 

[ben m]

For most it appears to be a mathmatical shell game.

What appears to be a mathematical shell game, to whom?

 

[rcfieldz]

What appears to be a mathematical shell game, to whom?

Looking at mathmatical theories written over the past centuries many appear to be nothing more than rearranged symbolic equations with no new conclusions. The ones that do have eureka breakthroughs become forgotten when science exceeds the limitaions of its knowledge.

 

[Roboramma]

The ones that do have eureka breakthroughs become forgotten when science exceeds the limitaions of its knowledge.

For instance?

 

[TjW]

He can't remember.

 

[Sam H]

ben m, thanks for a stimulating discussion.

... it's possible to build very powerful, fully-functional, and nonrelativistic proton accelerators ... which have all the ingredients to take an electron to extremely high velocities.

For example, it's reasonably easy to generate 500 kV DC using a Cockroft-Walton voltage multiplier.

I am no experimentalist, but doesn't Cockroft-Walton voltage multiplier use diodes, which were possible only after quantum mechanics, which depended on deBroglie's hypothesis of the wave nature of the electron, which he borrowed from Einstein's relativistic formula for the energy of the photon?

It's reasonably easy to generate 1MV using a Van de Graaff.

I have been able to find references to the use of the Van de Graaff for accelerating ions and protons, but not electrons. In all the cases that I have been able to find, the velocity is too small (max of 40 MeV for the kinetic energy of the proton). If you know of an actual experiment in which ELECTRONS were accelerated by a Van de Graaff, please share it with us. Maybe there is some kind of theoretical barrier for the electrons.

... This sort of behavior would be extremely easy to spot when starting up and operating such an accelerator.

I don't question this at all. What I find it difficult to accept is spotting a deviation and coming up with the full relativity theory, which brings me back to my intention of posting my original article.

Physics has two faces: experimental and theoretical. Both are crucial for its development. I disagree as much with the statement that all discoveries could be made deductively using mathematics alone, as with emergentists like Laughlin who think that every truth can be found only through experiments.

My favorite example is the Schroedinger equation. It describes everything from hydrogen atom to diodes, to superconductors, transistors, lasers, you name it. And it was the product of Schroedinger's brain- as he was vacationing with his girlfriend (not wife!) at a ski resort. However, the raw material that his brain used to construct the equation was two things: deBroglie's wave hypothesis for the electron, and, more crucially, the fact that the hydrogen atom had a nucleus. No amount of experiment would have given the Schroedinger equation and no amount of theoretical deductive reasoning would have given us a nucleus.

Having said these, one can imagine a science fiction scientist who COULD have imagined a nucleus for the atom and a wave property for the electron, and come up with Schroedinger equation. Similarly, one can imagine a science fiction experimentalist who COULD have thought of the electron diffraction experiments - without deBroglie telling them to look for such a diffraction - ... now I'm stuck! Because I can't think of any experimental procedure which would go from electron being a wave to the Schroedinger equation.

 

[ben m]

I am no experimentalist, but doesn't Cockroft-Walton voltage multiplier use diodes, which were possible only after quantum mechanics, which depended on deBroglie's hypothesis of the wave nature of the electron, which he borrowed from Einstein's relativistic formula for the energy of the photon?

You are again making a mistake---there is a difference between "what physics topics are fully understood" and "what technologies are available to experimentalists". The "rectifier" effect of certain crystals were discovered in the 1870s, and heavily studied after that. Cockroft and Walton built their accelerator in 1932, using diodes that had been available long before Felix Bloch explained bandgaps or conduction at all, and the p-n junction that actually makes diodes work was not explained until 1938.

I have been able to find references to the use of the Van de Graaff for accelerating ions and protons, but not electrons. In all the cases that I have been able to find, the velocity is too small (max of 40 MeV for the kinetic energy of the proton). If you know of an actual experiment in which ELECTRONS were accelerated by a Van de Graaff, please share it with us. Maybe there is some kind of theoretical barrier for the electrons.

Electrostatic accelerators work at fixed energies, not fixed velocities. If you can charge the terminal to 20 MV, you can get 20 MeV protons (which are slow) or 20 MeV xenon ions (which are very slow) or 20 MeV electrons (which are very, very fast).

There are currently-active electron Van de Graaffs at Chalk River, TU Delft, UC Santa Barbara; the Notre Dame van de Graaff used to have an electron mode, but that capability (probably a terminal-mounted filament and an alternative corona needle) was removed in the 70s. I would virtually guarantee that the Van de Graaffs of the 1930s and 40s would have been considered as all-purpose electron/ion machines.

Nowadays, if you need a 10-30 MeV electron beam, you can build or buy a linac, which is compact and inexpensive. That's why Van de Graaffs are so rare for this purpose nowadays. But there is no engineering or physics barrier.

Why do you seem so certain that this is wrong?

I don't question this at all. What I find it difficult to accept is spotting a deviation and coming up with the full relativity theory, which brings me back to my intention of posting my original article.

"Spotting a deviation, exploring it in every way you can, and coming up with a theory that makes it make sense" is what physicists do. Yes, it's difficult, but ... this sort of thinking is the whole point of being a physicist. That is the particular sort of hard problem that the Golden Age physicists were so incredibly good at.

The "deviation" is not some tiny subtlety---"if you have the spectrometer tune juuuusst right and the moon is full and the barometer is falling". The deviation is a large, reproducible effect, seen in quantities (energy, momentum, velocity) for which we've had excellent and reliable measuring instruments for a long time. The "deviation", plotted on simple graphs---a plot of v vs E or v vs. p for example---has a big obvious feature, an asymptote, pointing towards v=c, which provides a big hint in the right direction.

If it was possible for Einstein to come up with the whole SR theory---a connection between E&M (where C comes from) and kinematics (v/p/E relations) that includes a speed limit---without strong prodding from experiments; indeed, was probably one of only of a handful of people worrying about the foundations of E&M at all. In an alternate universe where experimentalists report discovery of a particle speed limit at c, and publish curves of v-vs-p and v-vs-E and such, I would bet on multiple independent theorists coming up with the full Einstein solution.

 

[Sam H]

... In an alternate universe where experimentalists report discovery of a particle speed limit at c, and publish curves of v-vs-p and v-vs-E and such, I would bet on multiple independent theorists coming up with the full Einstein solution.

I concede to the fact that in an alternate universe it is possible that experimentalists could discover the dependence of the Lorentz factor gamma on speed. But going to time dilation and length contraction requires deductive reasoning. Here is a real historical example: Planck (a theorist) came up with the exact formula describing the black body radiation (BBR) curve based on deviations from Wien's formula. But going from that curve to the quantization formula required deduction - human mental intervention. Or do you think that there is an experimental route from BBR curve to E=hf? (Or from the wave nature of the electron to the Schroedinger equation, a question I asked in my previous post.)

Robert Laughlin, though a great physicist, is damaging the reputation of science by saying that all laws of physics could be obtained from experiments alone WITHOUT any deductive process, due to his philosophical belief in emergence. He is therefore against elementary particle physics (a reductionist discipline by nature, and therefore, the philosophical antipode of emergence), whose theoretical development required one deductive step after another. His chapter on nuclear physics starts with Hiroshima and Nagasaki and blames the two bombs on reductionism! This is more damaging to physics, and science in general, than any blow a crackpot can deliver ... because it is coming from within.

 

[ben m]

I concede to the fact that in an alternate universe it is possible that experimentalists could discover the dependence of the Lorentz factor gamma on speed. But going to time dilation and length contraction requires deductive reasoning.

I don't think any working scientist would recognize actual scientific thinking as either inductive or deductive. Unless you're Plato sitting around at the agora, or Aquinas in his cell, that's simply not a clear or useful distinction.

If you try to be strict about the inductive/deductive distinction, I think you'll find that experimentalists formally can't discover anything. J. J. Thompson was able to see a variety of bends and curves in his cathode-ray beams, and he could deduce equations for this, but comparing these equations to the behavior of a hypothetical new particle---the electron---is inductive. Therefore electrons were discovered by theory, not by experiment!

In the case of relativity: if experimentalists were sitting there publishing these energy-vs-velocity curves, theorists have to dream up a new mechanics that makes those curves make sense. You're talking about "time" as though, perhaps, it's an add-on---as though they'd write a theory that explains the space/velocity business pretty well, and not realize (until an Inductive Reasoner came along) that time could be jimmied in too. I haven't done the math, but I strongly suspect that there is not a reasonable theory that would explain this "E = gamma*m" behavior without incorporating time-dilation---and that the time-dilation part of the theory pops out pretty easily. (Not as easily as "those cathode-ray curves look like a lightweight negative particle", but pretty easy nonetheless.)

I would also point out that time-dilation is accessible to simple, direct experiments with radioactive particles. As soon as you have a 300-400 MeV proton accelerator, you're going to be detecting relativistic pions, for which it's natural to try to measure the half-life---and you'll see the (large) time-dilation effect whether or not you were looking for it. There's a crude time-dilation "discovery" you can make by measuring stopped cosmic-ray muons in the lab; that doesn't even need an accelerator. (There might be a similar way to use cosmic-ray pions, but I'd have to work out the details.)

Length contraction is very hard to measure directly, though. It's hard to get either a ladder or a barn moving fast enough.

 

[Sam H]

I don't think any working scientist would recognize actual scientific thinking as either inductive or deductive.

I never said either or and my post about the crucial role of the nucleus in quantum theory is a proof of that. Both induction AND deduction are necessary.

Unless you're Plato sitting around at the agora, or Aquinas in his cell, that's simply not a clear or useful distinction.

Or Laughlin plagued with inductivism and emergentism!

I'm still waiting to hear about how experiments can take us from the BBR curve to E=hf or from wave nature of the electron to the Schroedinger equation.