|
return to homepage return to updates
SOME INITIAL REMARKS
on RICHARD FEYNMAN
by Miles Mathis
Has any thought been given to the number of things that must remain active in men's souls that there may continue to be actual "men of science"? —Ortega y Gasset Abstract: Here I critique Feynman by looking at two of his most influential books, Lectures on Gravitation and QED. Specifically, you will find analyses of Dicke's experiment, Einstein's elevator car, Feynman's "amusing puzzle," renormalization, and the shrink and turn method.
I began reading Feynman's Lectures on Gravitation today (2/24/2004). It took me less than five minutes to find the first major flaw of reasoning in the book. I swept past the endless introductions, written by physicists obviously more interested in mathematics and field theories grounded in nothing than in presenting a meaningful and consistent overview of anything. On page 3 of Feynman's first lecture, he brings up the equality of inertial and gravitational mass, and offers as proof the experiments of Eötvös and Dicke, which he says prove the equality to 1 part in 108. He says that the experiment is analogous to a fancy plumb-bob experiment, but he fails to see the flaw in the set-up. Dicke's experiment is prepared to show a torque if the centrifugal force (due to the earth's rotation) on one mass is greater than another mass. The two masses weigh the same, so this would prove that inertial mass is not equivalent to gravitational mass. The experiment has failed to show any torque.
This experiment—and its null outcome—is similar in many ways to the Michelson-Morley interferometer experiment, which also had a null outcome. Both experiments have used the null outcome as proof of theory. But a null outcome can also be caused by an experiment that is set up to show something that cannot possibly occur. A null outcome can be a sign of a poorly conceptualized experiment. A null outcome is a precarious thing to rest a proof on, regardless. It is much different than predicting some positive outcome, and achieving that outcome experimentally. It can be shown logically that a null outcome cannot prove a theory at all. I will not pursue this proof here, hoping that it is apparent to the reader.
I will instead show how Dicke has failed to prepare a meaningful experiment, and how this failure has gone unnoticed. The key is in the sentence, "the two masses weigh the same." Dicke needed two masses that had equal gravitational components, in order to test the difference in the centrifugal components. But he assumes for some reason that the centrifugal component will show itself only when the masses are suspended. The truth is that the centrifugal force is always working, since the earth continues to spin, whether Dicke is suspending the masses or letting them rest on his scales—whatever these scales are and however they are made. That is, Dicke's scales are already measuring a resultant force, the result of gravity and spin. The only way that his initial weights could be measurements of gravity alone is if he weighed them in a field that was equivalent to the earth's in every way but one: it should not spin. I don't know where he would go in order to achieve this. His experiment therefore tells us nothing except that the resultant forces on the two weights are the same during weighing and during suspension. The only force that is lost during suspension is friction, and that would be lost by both masses equally, since they weigh the same.
Feynman says that the ratio of inertial mass to gravitational mass should be different for objects made of different substances, but he gives no physical or mechanical or theoretical reason why this is so. Inertia is resistance to a force and gravity is a force. Logically, one would expect that a mass would react to the force of gravity in the same way, and to the same degree, that it would react to any other force, centrifugal, circular, or otherwise. But whether or not that is true, the experiments Feynman is describing don't address the issue anyway. They don't successfully isolate the problem so that we can decide one way or the other.
One reason for this is that the experimenters and analysts can't seem to do a proper vector analysis. Feynman’s vector for the centrifugal force is terribly wrong, in his illustration (see below). He draws the centrifugal force vector at an obtuse angle to the gravitational force, which is a sub-collegiate mistake. The centrifugal force is a reaction to the gravitational or centripetal force, so that it must be equal and opposite to it. The vector that Feynman has drawn is a compound vector of the object’s motion due to the centrifugal force and the motion due to circular motion. The object has a tangential motion at every point due to the earth’s rotation. If this tangential motion and the centrifugal motion are added together, in a vector addition, then you get the vector that Feynman has drawn. But not otherwise. You can see how it is arrived at by studying my diagram. If gravity were suddenly turned off, the non-suspended object would fly off in this direction (r2). But this vector is not the centrifugal force.
The only possible difference between two suspended objects of different make up would be their tangential velocity in the rotating field, since all other motions are either caused by the gravitational field (g) or are a direct response to it (c or f). These latter motions must be assigned to “gravitational mass”, obviously. This means that the difference in inertial masses must be expected to show itself only in the tangential vector. Obviously this is what Dicke’s experiment is trying to isolate. But forces in circular motion also do not rely on substance; they rely only on the distance from the center. The fact that both objects are the same distance from the center of rotation must give them the same tangential velocity. Or, if they are suspended, it gives them the same reaction to this velocity. Meaning, the earth is spinning, but they are no longer being spun by the earth, since they are not in contact with it. They are only being dragged along by their filaments. These filaments should show an angle or a torque. If the filaments were cut, the objects would fly off in the direction and speed of r1. But the angles cannot be different for the two substances, since they are the same distance from the center of the earth. Their tangential velocities must be equal for this reason. Their centrifugal and centripetal velocities must be equal because they weigh the same. Therefore their resultant velocities must be equal, and there must be no difference in torques.
This means that the expectation of a torque is wrong for at least three separate reasons. Feynman draws the vector wrong, expects different substances that weigh the same to have different inertial masses, and he expects a suspended mass to act differently than a non-suspended mass, with various materials. All these assumptions are false. Both Eötvös’ and Dicke’s experiments fall to these various false assumptions. The null outcome was caused by trying to measure something that could not possibly happen. Those who know of my papers on gravitation know that I would predict that Dicke's experiment would fail for a fourth and central reason, but I don't even need to broach that reason here. I have just shown that by using the classical assumptions of the gravitational field and circular motion, the experiments should have been expected to show a null outcome. That is, using Eötvös', Dicke's, and Feynman's own reasoning, I have shown various inconsistencies in it. The problem in this instance is not that 20th century physicists needed a new theory of gravity (although that is also true, I believe). The problem here is that contemporary physicists have spent so much time learning the tensor calculus and Hamiltonion math and Fourier equations and Lagrangians and so on that they never had time to fully digest basic Newtonian mechanics. Their vector analysis here has been very slipshod, and this is due to a basic inability to apply simple math to physical situations. The modern physicist is a master of esoterica and a moron when it comes to basic physics.
Eötvös and Dicke didn’t prove the equivalence of gravitational mass and inertial mass, since their experiments never successfully isolated anything that we could call an inertial mass. When we weigh an object, we are already measuring a resultant force—we are measuring the gravitational force minus a force due to circular motion. If the earth stopped spinning, the object would weigh more on the scale, since the scale would then be feeling all the mass of the object rather than just a large fraction of it. All that happens when we suspend an object is that the negative part of this equation is allowed to express itself by a motion and an angle. The centrifugal component that we subtracted out to find the weight on the scale is allowed to push the object backwards, against rotation, and the object swings to a small angle. So the same vector subtraction is working whether the object is suspended or not. When the object is not suspended, friction keeps the ball from rolling backward. When the object is suspended, it swings a bit against the rotation of the earth.
In other words, we are given (by the weight equivalence) that both objects have the same total force on them—gravity minus effect due to inertia. Then we showed by experiment that both objects have the same inertial mass, since they swing to the same angle or create the same torque. Therefore they must have the same gravitational masses. All true. The experiment of Dicke does in fact prove all that. But that is not a proof that gravitational mass equals inertial mass. Why? Because if inertial mass were 49% or 5% of gravitational mass, all that would still be true. Look at this bar graph.
The inertial masses are equal, the gravitational masses are equal, the total masses (weights on a scale) are equal, but the inertial mass does not equal the gravitational mass.
Of course I can't blame this entire mess on Feynman. He is just reporting the accepted findings here. But I think it is a bad sign that someone with Feynman's towering intellect was not able to see through this pretty simple vector analysis. It is a bad sign that no one has been able to pick out the basic flaw here. Just as with the interferometer, no one in the 20th century retained enough basic physics or kinematics in order to penetrate the problem or the diagram. By the time of Lorentz physicists had already been buried under complex maths. They didn’t have time for basic physics, and therefore weren’t any good at it. This situation has remained up to the present time.
Before I move on, let me be clear: I believe in the equivalence of inertial mass and gravitational mass. Even more, I believe that they are the same thing. Meaning, they don't just yield the same number, they are caused by the same physical state. Having two names for the two masses was redundant from the start, so that finding that they are equivalent is just a tautology, like finding that blue=blue. Despite that, the experiments of Dicke and Eötvös do not succeed in proving the equivalence. By the current meaning of inertial mass, these experiments prove only that one inertial mass equals another. They do not prove that inertial mass is equivalent to gravitational mass. Experimental science cannot prove the equivalence, and does not need to, since the equivalence is definitional. That is, it already exists, buried in our definitions. We don't need to fine-tune our experiments, we need to sort out our theory: write it all out again, study it, and understand what it means. We have created a semantic separation where there is no physical or theoretical separation.
I don't have to go far for the next big blunder, for it is on page 4. Feynman says that the accuracy of Dicke's experiment (one part in 108, remember) "is already telling us many things." One thing it is telling us, he says, is that since the binding energy of the nuclei is one percent of the total energy of the nucleon, the accuracy of the mass equivalent of the binding energy is one part in 106. This accuracy is also a check on the electronic binding energy. And if the experiment can be pushed to 1010 that would allow a 5% check on chemical biding energy. I have just shown that Dicke's experiment is accurate all the way to infinity, since it is a tautology, so by that argument it is a "100% check" on the binding energy of the universe. But Feynman's argument is absurd even if Dicke's experiment had been true. The accuracy of the experiment says nothing about any binding energies whatsoever. The accuracy of the experiment was determined by factors specific to the operation, and is a measurement primarily of the scientist's ability to measure torque to a tiny fraction. Binding energies have absolutely nothing to do with this. Feynman is making mathematical connections between numbers that have no mathematical connection, and my exploding of Dicke's experiment only makes this more obvious.
On page 5 Feynman is already talking about antimatter and the behavior of kaons in experiments. This is his first day of class, his first lecture, about five minutes in, and already we are hearing about gravity's effects on antimatter. We have been told two glaring falsehoods; now we are presented immediately with esoterica.
[One thing I took from the introduction was Feynman's statement (written on the blackboard and never erased, according to Brian Hatfeld): "What I cannot create, I do not understand." I find this to be a perfect expression of his method, which was conceptually sloppy but mathematically complex. It was a method that relied on "creation" more than discovery, and belied a character that rushed ahead where fools rushed ahead. He often seems more concerned with self-promotion than with any search for the truth, and his example has been disastrous for physics. His quote might be changed to read, "I cannot understand it unless I created it myself." But this obviously turns science on its head, since science is not art. Science is the discovery of pre-existing fact, not the creation of fact. Scientists now are more concerned with fitting fact to theory than theory to fact, and Feynman's quote is typical of the new outlook.]
The next six lectures are on applying tensor calculus to a problem that Feynman has not even defined yet. It is really extraordinary to see the way this class was set up. Feynman has no idea what he is talking about, so he accelerates into his own field (QED) as fast as possible. He immediately quantizes gravity, applies tensor fields, and so totally confuses the students—who now cannot possibly object to anything he says, since they don't know what the hell he is talking about. First he tries to make the neutrino the force-bearing particle of gravity. He hits that theory with just enough math to muck it up, and then some more for good measure. Then he goes on to the graviton, making assumption after assumption, none of them backed up by anything. He assumes for instance that gravity is a field, that the field is controlled by particles, and that these particles will behave in ways precisely like quantum mechanics. So he carries into gravity all the baseless assumptions of QED. For instance, since particles with spin 1/2 repel like particles, Feynman assumes that the graviton must have a spin of 2—since even spins attract. Well, OK, but the electron has the spin it does simply because we defined it that way. No one really knows what that signifies or how it works. The photon is probably the force-bearing particle, but how it causes binding energy is not really known. In a classical sense, you can imagine how a force-bearing particle might cause a force-feeling particle to be repelled (by bombardment) but attractions have never been explained. So when Feynman is talking about spin 2, he is just coining a term. No one knows what spin 2 signifies, or if it signifies anything. Is a spin 2 spinning twice as fast as spin 1? No, that is not the theory. There is no theory. It is just names without concepts.
Not until lecture 7 does Feynman think it worthwhile to hit a few basic concepts. He has shown off for six lectures, filling the blackboard with math that none of the students can possibly follow (not because it is difficult but because it is false bombast). But at last, at the beginning of lecture 7 he intends to gloss Einstein, that "brilliant marvelous man." In the preceding three or four lectures, Feynman wanted to show that he, Feynman, could derive all of Einstein's theory without him, just coming from QED. The fact that the students are confused is nothing compared to that. Best to keep them intimidated anyway, or they might start to see through the math. Very soon you will see why I don't hold Feyman in high regard as a mathematician, and why I don't think any of his tensor calculus is even worth a response: In lecture 7 he makes some basic algebraic errors that are astonishing, and that exceed even his inability to analyze Dicke's and Eotvos' experiments in lecture 1.
Once again, rather than start with Einstein's simple elevator car in space, Feynman first mucks up the experiment by showing all the various ways it is not true in various real fields. He talks about quadrupole character, tidal forces, gravity minus gravity-prime, accelerating towards nebulae, and so on. If any of this was to be mentioned at all, it should have been mentioned later, after the basic concepts had been made clear. But this is not Feynman's method. He wants to be sure you know that he knows all the subtleties of the case before he even states the case. The whole series of lectures is therefore set up topsy-turvy. Each lecture begins with a list of esoterica that he is in on. And the whole series of lectures began, as I said, with the esoteric math, followed, in some cases, with a description of the problem to be solved. As you see, it took Feynman six lectures of fancy math to get to a description of what he was trying to solve in the first place. In some cases he never gets to a basic description of the problem—he supplies us only with the math to solve it.
Finally, in section 7.2 [Lecture 7, page 93 of the book] he presents a real problem. To teach the principle of equivalence he introduces Einstein's elevator car in space. He lets light be emitted from the top of the box and he calculates the blue shift as measured from the bottom. He says, "The time that light takes to travel down is to a first approximation c/h, where h is the height of the box." This is his equation then for light emitted at the top of the accelerating box and received at the bottom:
t = c/h
That look odd at all? t = v/x? v = xt? I thought v = x/t. To see how this should be done, we will have to do it ourselves, and then compare Feynman at each point. We can obviously not trust anything he says. The time it takes light to travel from the top to the bottom, if the box is in uniform acceleration up, is
v = x/t
c = x/t
t = x/c
x = h - Dh
Dh = gt2/2
t = (2h - gt2)/2c
2ct = 2h - gt2
gt2 + 2ct - 2h = 0
t = -2c + √4c2 + 8gh)
2g
t = √c2 + 2gh) - c
g
In this time, the box has gained this much more velocity
v = gt = √c2 + 2gh) - c
Feynman said that to a first approximation that would have been
v = gh/c
In his first equation above, t is c/h. In his second t is h/c. I have shown that the second is correct. Perhaps the first was a misprint.
Now if φ = gh, then
v = √c2 + 2φ) - c
= √1 + 2φ/c2) - c
In order to find the final transformation equation, Feynman says that all we have to do is substitute the last equation into this one:
f = f'(1 + v/c)
This equation defines the apparent change in frequency due only to relative motion. It has nothing to do with gravity, obviously, since there is no acceleration or gravity variable in it.
Substituting we obtain:
f = f'[1 + √1 + 2φ/c2) - c ]
c
f = f'√1 + 2φ/c2
That is the change in frequency from the top of the box to the bottom. Since a clock behaves like a wave, Feynman implies that the equation also applies to the period of a clock at the top of the box seen from the bottom.
f = frequency measured at the bottom
f' = frequency emitted at the top
Since f > f' time appears to go faster at the top of the box.
First of all, I find it very interesting that Feynman uses the equation f = f'(1 + v/c).
This equation means that because the bottom of the box is moving toward the ray of light, the frequency will increase. The relative motion by itself, with no acceleration, causes a blue shift. That is what this equation must mean. But he does not seem to see that a signal from a clock will act just like a ray of light, as it moves from top to bottom. The light has a frequency of 1016/s, say, and a clock is defined as 1/s, but they are both waves. If the light is blue-shifted, then the clock will be too, as he admits. He says, "the clock at the top looks bluer" (p.94, 7.2.1). A blue clock will be ticking faster. By the equation f = f'(1 + v/c), the clock will be ticker faster even without the gravitational field. If the bottom of the box was moving at a constant velocity, the equation would still apply, and the f would describe the period of the clock.
This is important since it contradicts Special Relativity. SR states that all relative motion causes time dilation, which is to say slowing of time. But here Feynman is assuming just the opposite in order to prove that GR speeds up time. Feynman assumes that motion toward in SR causes time to speed up in order to prove that motion toward in GR causes time to speed up. By carrying his equation for v into the equation f = f'(1 + v/c) he is calculating a sort of double blue-shift.
Let's try to work the problem out another way. Let's say the clock at the top has a period of t'. We want to calculate the period as seen from the bottom, t. We are given g and h. The bottom of the box receives the first tick at T1 = T1' + Δt1.
Δt1 is the time it takes for light to travel from top to bottom, after tick #1.
Δt1 = x1/c
x1= h - Δh1 where Δh1 is the distance the bottom travels up while light is traveling down
Δh1 = vavΔt1 = (v1 + v2) Δt1/2 vav stands for v average
We need a velocity variable in there instead of g, so that we can vary it from one time to the next. g does not change, and so it will show no increase over time, if such exists.
Δt1 = 2h - [(v1 + v2) Δt1]
2c
2cΔt1 = 2h - [(v1 + v2)Δt1]
2cΔt1 + v1Δt1 + v2Δt1 = 2h
Δt1(2c + v1 + v2 ) = 2h
Δt1 = 2h/(2c + v1 + v2)
The bottom receives the second tick at T2 = T2' + Δt2
Δt2 = x2/c
x2= h - Δh2
Δh2 = vΔt2 = (v2 + v3) Δt2/2
Δt2 = 2h - [(v2 + v3) Δt2]
2c
2cΔt2 = 2h - [(v3 + v2) Δt2]
2cΔt2 + v2Δt2 + v3Δt2 = 2h
Δt2(2c + v2 + v3 ) = 2h
Δt2 = 2h/(2c + v3 + v2 )
The observed period is then
t = T2 - T1
= (T2' + Δt2) - (T1' + Δt1)
= T2' + Δt2 - T1' - Δt1)
But T2' - T1' is defined as 1s.
So, t = 1s + (Δt2 - Δt1)
= 1s + [2h/(2c + v3 + v2)] - [2h/(2c + v1 + v2)]
I don't see immediately how to reduce this, but it doesn't really matter. Because what it means is clear. For one thing, it means that the clock is not just blue shifted—the clock at the top is seen by the bottom to be increasing its rate. The clock does not have a constant rate. I think you can see that T3 - T2 is going to be less than T2 - T1, so the period is going to be steadily shrinking. I did not need the equations of SR to find this out though. I did not use the transformation for frequency, like Feynman did. I did need to assume that the clock was ticking a normal rate in its own vicinity, and that we could know that rate, as did Feynman. He says (p.94),
"How much is the time difference at various points in space? To calculate it we compare the time rates with an absolute time separation, defined in terms of the proper times ds."
This is one of the clearest statements I have found in favor of my interpretation of Relativity. Feynman seems to recognize at times like this that time differences must be measured against a standard, which standard is local time. His assumption of absolute time separation is astonishing, really, since it contradicts the current interpretation of Relativity, which states that the local field is either an illusion or an impossibility. Regardless, I hope you can see that just as with SR, one must assume all fields are equivalent locally in order to calculate how they will be seen differently from a distance. Feynman understands that he can't solve the problem without "absolute time separation" and he makes that explicit in this problem.
Let's try one more time to get a simpler transformation equation. Let us try to obtain an x variable instead of a v variable. That will also allow us to solve for different times.
Δt1 = x1/c the t variable here designates the time it takes light to travel to the bottom of the box.
x1 = h - Δh1 Δh1 is the distance the bottom travels up while light is traveling down
Δh1 = gΔt12/2 the t variable here is the time it takes for the bottom to travel up
Now, set the two times equal to eachother, since the time it takes the light to meet the bottom is the same as the time it takes the bottom to meet the light.
x12/c2 = 2Δh1/g
Δh1 = -gx12/2c2
Δt1 = h/c + (gx12/2c3)
The first part of this equation is Feynman's first approximation from above. You see that it should have been h/c not c/h.
Just like before, we find that
Δt2 = h/c + (gx22/2c3)
t1 = T2 - T1
= (T2' + Δt2) - (T1' + Δt1)
= 1s + h/c + (gx22/2c3) - h/c - (gx12/2c3)
t1 - 1s = g(x22 - x12/2c3)
x1 is bigger than x2, so t1 will be smaller than 1 second. We have proven a blue shift. But
t2 = g(x32 - x22) + 1s
2c3
x2 > x3, so we still have a blue shift, but x2 < x1, which means the blue shift is increasing with time. Notice that as the box accelerates, Δh is going to get larger and therefore x is going to get smaller. Each successive light signal from the top is going to have a shorter distance to go before meeting the bottom. With each passing moment, the clock will get bluer and bluer. Feynman and Einstein don't ever recognize this, and so they don't see it as problematical. Feynman thinks that GR implies simple (and constant) time contraction, as shown by his equations, and that SR always implies time dilation—as in the standard interpretation of Einstein. This gives him an opportunity to pursue an "amusing puzzle" in the next section (see below). But the truth is that it is quite problematical for an acceleration to imply an increasing blueshift, since we do not see such things in reality. According to Einstein's theory of equivalence, we cannot tell if we are accelerating or are in a gravitational field. But the fact is we are in a gravitational field at all times, and yet do not see things the way this experiment tells us. I feel a force pulling me down right now that we call gravity. But according to this thought experiment, I should be able to say that I am in a rocket accelerating upwards. In which case I should see a clock on the ceiling gaining time with each passing second. Even at a relatively small acceleration, I should, after a few years, be going really fast, in which case even a small initial blue shift would become obvious. And after a few decades at this acceleration, people in my rocket (people on this earth) should be approaching c. After millions of years, we should be so close to c as to not matter at all. That clock on the ceiling should be going so fast I can't read it. Why isn't it?
For instance, if the gravitational field of the earth is 9.8m/s2 near the surface, and I am 40 years old, then even if the my rocket was going zero at my birth, it would now be going,
9.8m/s/s x 40 x 365 x 24 x 60 x 60 = 1.24 x 1010m/s
That is, I would be going faster than the speed of light, if not for mass considerations. According to Feynman, I should have been gaining mass now for at least a decade, just to slow me down. All incoming light should be so blue-shifted it would have become invisible years ago, unless I had learned to see cosmic rays. If I were Feynman, I would make a mathematical game of it, and figure out how many years and months old I was when I began packing on the pounds and seeing x-rays, but I find such things juvenile, especially when they involve calculating from false theories. I have real work to do—a large part of which is becoming correcting all the mistakes of Feynman.
Current physicists might answer that we do not see blue-shifts like this because we are not really accelerating. We are in fact in a field. But if this is the case then Einstein's equivalence fails, and we cannot use math derived from blueshifts to define our field. Feynman has just used acceleration equivalence in order to derive his transformation equation. But the situation is either equivalent or it is not. If we derive equations from blueshifts, then we must explain the lack of blueshifts. If there are no blueshifts, then there is no equivalence, and we must derive equations some other way.
The whole reason Einstein postulated equivalence was so that he could calculate the potentials in just this way, and that is why Feynman quotes his method half a century later, calling it marvelous.
Feynman's amusing puzzle is to calculate the best movement of a clock in order to make it gain the most time. He assumes that greater potentials give faster clocks (GR), but that motion slows down the clock (SR). Since I have shown in another paper that SR is wrong about motion toward an observer, as with Feynman's frequency increase equation, this whole exercise is moot. Besides, I just showed how Feynman assumes that SR causes a blueshift. He uses the equation f = f'(1 + v/c) above, which is nothing more than a blueshift caused by relative motion. He assumes SR causes blueshifts on one page and that it causes redshifts on another page. This is what is called a contradiction. Not a paradox. A mistake. The Hafele-Keating experiment obviously took its cue from Feynman, though. That experiment is precisely the working out of Feynman's amusing puzzle.
[Remarks on the Lectures on Gravitation to be continued]
For the completed equations on Feynman's box, see my newer paper on the Pound-Rebka experiment. Similar equations are solved in my paper on the muon and in my paper on the equation v = v0 + at.
Last night I read Feynman's book QED. Feynman's one laudable feature is that he is straightforward. This book lays QED on the table in all its ugliness, and I can picture many physicists groaning when it came out in 1985, seeing all their wicked souls laid bare. The great lesson of this book is that QED is not a theory at all. It is a heuristic formula created ex post facto. That is, QED does not predict anything, not even probabilities. To calculate complex probabilities, the scientist must be given the probabilities of each individual event in advance, from experiment. QED then supplies the math to calculate combinations. For instance, I was struck in his example of partial reflection by glass that the number .2 was a given of the experiment. It is the outcome as well, so nothing is really achieved except an equation. This would be something indeed if the equation came with an explanation, but it does not. Feynman assigns a probability amplitude to each individual photon, then claims that many of the amplitudes "cancel eachother out." But this cancellation is given no physical meaning. For probabilities to cancel eachother out, the photons themselves have to cancel eachother out by some method. Feynman's stopwatch example implies that the method is interference of waves, since a hand going around on a watch is a wave, but Feynman refuses to assign his math to a physical situation. And he takes great joy in not doing so. He says that he prefers a Nature that does not make sense, and states that we must take her like she is. But it is not Nature that is refusing to make sense, it is Feynman. There is no reason to believe that some physical meaning could not be assigned to QED.
[For a complete solution of the problem of partial reflection by glass, go to my newer paper, where I solve that as well as two related problems.]
The problem seems to me to be that if a physical meaning is assigned, then a large part of the mystery of QED evaporates. That leaves physicists like Feynman looking like ordinary men, instead of magicians. Feynman prefers to mystify his audience, since that leaves him the only one in the room who seems to have penetrated the veil. He is obviously very proud of himself. Can you imagine Newton or Galileo mentioning in a lecture that "we got an award for that." It is all so transparently pathetic.
Feynman says on page 124, "physics is an experimental science and the framework agrees with experiment, so it is good enough for us so far." This puts the whole problem of contemporary physics in a nutshell. Of course the framework of QED agrees with experiment: it was all created after the fact to fit the experiment, and corrected when new experiments showed a need. It is a framework with near-zero conceptual content, being all heuristic math. But Feyman does not see this as a problem. For him, this is what physics is. It is worth noting, however, that this is not what physics was historically. Before the twentieth century physics was considered to be a description of the physical world—an explanation of physical phenomena. The foundation of physics was therefore theory. Theory as a body of concepts, not just a math that worked.
The only concept that QED seems to me to add to light is the concept that light has an equal probability of going all places that it can go. That is, in most situations it behaves in a nearly random manner. It reflects from surfaces at all angles, is refracted at all angles, etc. It was already known long ago that light had wave characteristics. If you add the random nature of light and its tendency to act like a gas to wave theory, you get the math of QED. QED did not even come up with this characteristic of light either. It simply combined wave theory with probability theory to arrive at a successful math. This is no small feat; don't misunderstand me. In some ways QED is as beautiful as it claims to be. But I don't think it was necessary to package it like it has been packaged. The math could have been delivered to the world without all the mysticisms and self-congratulatory pseudo-philosophy. Heisenberg and Bohr started the ball rolling with this sort of groundless anti-materialism and anti-rationalism, and it has only snowballed since then. Refusing to make sense, or to even try to make sense, has become a sign of distinction. There are close analogies in artists' refusal in the 20th century to deal with old-fashioned concepts like beauty and truth. Across all fields deconstruction has remained the fashion, and the last thing any contemporary scientist wants to be called is a classicist. Tight logical consistency and intellectual rigor have become passe, in the same way that long sentences and rhyming have become passe in literature. They are the hallmark of the fuddy-duddy. Hence you see Feynman saying, in the introduction, "what one fool can understand, another can." This is science as politics, an anti-elitism making the lecturer dear to his audience. Feynman winks at the reader, making him feel equal, while at the same time he schemes behind the scenes to make sure that no reader will possibly come away any true insight. The math, no matter how stripped down, is a bulwark against understanding—a guarantee that Feynman and his colleagues will retain all status.
Over and over again Feynman jokes about physics majors taking x-number of years just to learn the math for these calculations. But it is no joke. The audience knows that they could never weather those years, and so they are out of the loop in the end, no matter how many winks they have received. And the graduate students, too: they are kept busy with the math lest they stop for a moment to ask what it is all for. What are they proving? Only a few dozen at the top like Feynman know. That is no accident. The graduate students are much more a threat than anyone in any possible audience. A new Newton might be lurking in those graduate classes: the sooner he is demoralized the sooner the Feynmans can get backs to their Nobel Prizes in Heuristics.
And finally, Feynman states that it is the job of the physicist to ask "how" not "why". Nobody knows why, or can know, he says. Perhaps, but so many of the things he refuses to consider are questions of "how". How do photons cancel. How do particles act like waves? How do photons propagate? What geometry does the amplitude apply to? The spin? The charge? How does an exchange of photons bind an electron to a nucleus? These are not metaphysical questions. They are not outside the purview of physics. Disallowing certain questions only looks like protectionism.
Feynman admits that the one of the things he is most famous for, renormalization, is actually no more than a shell game, and probably "bad math." It is the worst kind of math, but it does not seem to matter to anyone. It has not led to any sort of contrition. QED is still considered the greatest thing ever. Feynman ranks it above GR, since it explains everything but gravity. If QED lost its mystical pretensions and found some way to tie its math to a consistent physical conceptualization, it would indeed be a grand theory. For now it is simply another brick in the wall. As a quick gloss of the bad math of renormalization, consider this explanation from Ron Schmitt:
To cancel the infinite term of the electromagnetic mass [the infinite mass of the electron, according to Maxwell's equations], the bare mass is defined to include a term that is equal but opposite in value to the electromagnetic mass. In other words, the bare mass contains a term that is infinite and negative. The positive infinity from the electromagnetic mass and the negative infinity from the bare mass cancel and the result is the finite mass that is observed in experiments. The technique, called renormalization, is not necessarily elegant, but it works to incredible precision. QED is the most accurate physical theory that mankind has produced.1
Adding an infinite term willy-nilly is "not necessarily elegant"? How about "necessarily very inelegant"? How about strictly illegal? It is also strictly illegal to assign two masses to the electron, a bare mass and an electromagnetic mass. Besides the opposite infinities, what is the difference? Nothing, since they were just manufactured. One particle cannot have two masses. And of course this mathematical fudging is precise. How are you going to make an after-the-fact push and get it wrong? You are going to correct the equation in the wrong amount? Then Schmitt has to finish with the required flattery, about the accuracy of QED, as if we should be impressed that they pushed the equations in the right amount, instead of the wrong amount.
For further proof of Feynman's sloppy and dishonest attitude toward science, look at the subtitle to the book QED: "The Strange Theory of Light and Matter." That adjective "strange" is a dead giveaway. Before the 20th century, physics was the explaining of phenomena, making them "un-strange." Physics was not the opposite of metaphysics, it was the opposite of magic. It was the opposite of the paranormal. But Feynman takes great pleasure in being paranormal. The physicists of QED (and QCD) love the adjective "strange" so much they gave it to one of their particles. Mystification has become a distinction. Most in the field of physics find this "refreshing" or "intriguing." I find it depressing and maddening. These people will join the phony artists and politicians of the 20th century in the annals of shame. In the future, no one will smile at such transparent pretension and glaring ineptitude. They only will wonder at the levels of blindness and hypocrisy that made it all possible.
It was refreshing to see Feynman conclude, as late as 1985, that a photon "went through either one hole or the other" in the two-slit experiment. Not both. This surprised me somewhat, especially after he had talked again and again about electrons going backward in time. If he would have stopped sniping at philosophers long enough to have developed a tight definition of time (and realized that such a definition was a necessary ingredient of any equation that included a time variable) he might have recognized that going backward in time is a logical impossibility. But he was more interested in having room to theorize than in being logically consistent.
[For the full solution to the two-slit problem, see my newer paper on that.]
1Ron Schmitt, Electromagnetics Explained. p. 132. http://books.google.com/books?id=fUBPN8T9bwUC&pg=PA132&lpg=PA132&dq=bare+charge&source=web&ots=_WQzOynpvX&sig=0flzHCtKRTYVupbgQX190twGNSY&hl=en&sa=X&oi=book_result&resnum=5&ct=result
If this paper was useful to you in any way, please consider donating a dollar (or more) to the SAVE THE ARTISTS FOUNDATION. This will allow me to continue writing these "unpublishable" things. Don't be confused by paying Melisa Smith--that is just one of my many noms de plume. If you are a Paypal user, there is no fee; so it might be worth your while to become one. Otherwise they will rob us 33 cents for each transaction.
|