The Instigator
SargonOfAkkad
Pro (for)
Winning
4 Points
The Contender
GarretKadeDupre
Con (against)
Losing
0 Points

The Hafele-Keating Experiment Supports Special Relativity

Do you like this debate?NoYes+3
Add this debate to Google Add this debate to Delicious Add this debate to FaceBook Add this debate to Digg  
Post Voting Period
The voting period for this debate has ended.
after 2 votes the winner is...
SargonOfAkkad
Voting Style: Open Point System: 7 Point
Started: 1/3/2015 Category: Science
Updated: 2 years ago Status: Post Voting Period
Viewed: 2,684 times Debate No: 67781
Debate Rounds (4)
Comments (12)
Votes (2)

 

SargonOfAkkad

Pro

Ave

The resolution of this debate states that "The Hafele-Keating Experiment Supports Special Relativity". Pro will take the position that the Hafele-Keating experiment supports the theory of special relativity. Con will take the position that the Hafele-Keating experiment does not support the theory of special relativity. The burden of proof is on Pro. This means that if both arguments are equally matched, then Con wins by default.

The first round of this debate is for acceptance. The second round will begin with my opening argument, followed by back and forth rebuttals throughout the remainder of the debate. Voters should vote for the debater who has a preponderance of the evidence (i.e. when weighted on balance, the arguments favor their side over the opposing side).

The Hafele-Keating experiment refers to the experiment conducted by Joseph Hafele and Richard Keating, where Hafele and Keating measured the times on atomic clocks from as they were stationary, traveling westward, and traveling eastward. Hafele and Keating found that the three sets of measurements were inconsistent with each other. This experiment was later published in the journal Science, and is now taken as support for special and general relativity.

Although the word "supports" is more metaphorical than analytic, I believe we may all decipher its meaning on our own, at an intuitive level. However, the operational definition I will offer is that an experiment supports a scientific theory if it is consistent with the predictions of that theory.

Special relativity refers to Einstein's theory of special relativity. It postulates that the laws of physics are invariant across all inertial reference frames, and that the speed of light in a vacuum is the same for all inertial observers. A prediction of special relativity is a phenomenon called "time dilation". In special relativity, time dilation entails that clocks moving with respect to an inertial frame of reference will be measured as running more slowly. In this debate, I will not attempt to distinguish between rivaling interpretations of special relativity, as they are all empirically equivalent.

I appreciate Con's acceptance of this debate, and look forward to a civil, thoughtful, and productive conversation.

Vale

GarretKadeDupre

Con

I accept.
Debate Round No. 1
SargonOfAkkad

Pro

Ave

Since I consider this resolution to be incredibly easy to defend, I will present a simple argument in favor of it.

P1: If the Hafele-Keating experiment affirms the predictive power of special relativity (i.e. its consistency with experiment), then the Hafele-Keating experiment supports special relativity.

P2: The Hafele-Keating experiment affirms the predictive power of special relativity.

C: The Hafele-Keating experiment supports special relativity.

In propositional logic, this is the modus ponens form of argument, where if p implies q, and p is true, then q is true. Since the argument is obviously logically valid, the only task that remains is to defend the truth of each premise.

P1


In the contemporary paradigm of the philosophy of science, the best indication of a good scientific theory is that it has predictive power. A scientific theory with predictive power is able to generate testable predictions. Predictive power differs from explanatory power in that it does not merely explain past events with the theory, but actually predicts future events. For this reason, predictive power is generally favored over all other aspects of a scientific theory. As the philosopher of science Alex Rosenberg stated in an interview with 3AM Magazine, “Science is common sense recursively reconstructing itself. The reconstruction reflects the fallibility of common sense. Insistence by science on the tentativeness of its results at its ever-shifting research frontier, is what gives us confidence that after repeated test the parts most distant from that frontier are unlikely to be called into question.”. In other words, confidence in science comes from repeated testing of the predictions that a scientific theory makes. For this reason, one cannot have a good scientific theory without predictive power. Therefore, an indication of a good scientific theory is its predictive power.


What would predictive power look like in practice? If an experiment was conducted, and the outcomes were consistent with the theory (i.e. the outcome is what you would expect if the theory is true), then the experiment affirms the predictive power of the theory, indicating that it is a good scientific theory. From this, predictive power is consistency with experiment. In round one, I defined “supports” as “an experiment supports a scientific theory if its outcomes are consistent with that theory”. Based on this definition, we can conclude that “supports” also means “affirms predictive power”. Therefore, the resolution is equivalent to stating “The Hafele-Keating experiment affirms the predictive power of special relativity”.


P2


The Hafele-Keating experiment supports special relativity if its outcomes are consistent with the predictions of special relativity. What predictions would special relativity make for the Hafele-Keating experiment? Special relativity would predict that the rate of the clock is greatest when the observer is at rest with respect to the clock. In the Hafele-Keating experiment, the inertial reference frame was represented by the United States Naval Observatory. Hafele and Keating put four atomic clocks on airplanes, and then took the time that they measured after traveling around the world in an eastward, and then westward, direction. Since special relativity predicts that the rate of the clock is greatest when the observer is at rest with respect to the clock, then it would also predict that the clocks will lose time in the eastward direction and gain time in the westward direction. Did this actually take place? Yes. Hafele and Keating found that the clocks lost time in the eastward direction and gained time in the westward direction. As Hafele and Keating state in the abstract of their 1971 Science publication, "Relative to the atomic time scale of the U.S. Naval Observatory, the flying clocks lost 59 ± 10 nanoseconds during the eastward trip and gained 273 ± 7 nanoseconds during the westward trip". This means that the Hafele-Keating experiment was a success for special relativity, as it affirms the predictive power of the theory.



The consensus in the physics community helps to establish the plausibility of this interpretation. Indeed, the notion that the Hafele-Keating experiment supports special relativity is so established in the physics community that it is taught in introductory textbooks. As Physics: Principles and Applications by Douglas Giancoli states, "Experiments have tested the time-dilation effect, and have confirmed Einstein's predictions. In 1971, for example, extremely precise atomic clocks were florm around the world in jet planes. The speed of the planes was much less than c, so the clocks had to be accurate to nanoseconds in order to detect any time dilation. They were this accurate, and they confirmed time dilation to within experimental error." As Stephen Hawking states in The Grand Design, "Imagine a reference clock at rest at the center of the earh, another clock on the earth's surface, and a third clock aboard a plane, flying with or against the direction of the earth's rotation. With reference to the clock at the earth's center, the clock aboard the plane moving eastward--in the direction of the earth's rotation--is moving faster than the clock on the earth's surface, and so it should run slower. Similarly, with reference to the clock at the earth's center, the clock aboard the plane flying westward--against the earth's rotation--is moving slower than the surface clock, which means that should run faster than the clock on the surface. And that is exactly what was observed when, in an experiment performed in October 1971, a very accurate atomic clock was flown around the world."


C




The conclusion follows logically from the premises. Since the conclusion is the same as the resolution, the resolution is affirmed.



Vale

GarretKadeDupre

Con

My opponent made the following argument:
  • P1: If the Hafele-Keating experiment affirms the predictive power of special relativity (i.e. its consistency with experiment), then the Hafele-Keating experiment supports special relativity.
  • P2: The Hafele-Keating experiment affirms the predictive power of special relativity.
  • C: The Hafele-Keating experiment supports special relativity.

I accept his first premise but deny his second.

The Hafele-Keating experiment does not affirm the predictive power of Special Relativity. 4 clocks were taken on 2 trips; an Eastward trip, and a Westward trip. For the Eastward trip, they predicted that the clocks would lose 59 nanoseconds relative to the ground clock (plus or minus 10). For the Westward trip, they predicted a gain of 275 nanoseconds (plus or minus 21).

Here is the source for their predictions:
http://ivanik3.narod.ru...


Let's compare the predictions with the actual results, and see how blatantly they contradict Special Relativity.

Clock 408

Eastward prediction: Loss of 59 nanoseconds
Actual result: Gain of 165 nanoseconds
  • Clock 408 was losing 1.78ns per hour relative to the ground clock when the trip began, so it should have lost 114.67ns by the end of the 65.43hr trip. Instead, it gained 50ns (it was behind ground clock by 20ns before trip, but ahead of ground clock by 30ns after trip). This proves it actually gained 165ns on the Eastward flight.
Westward prediction: Gain of 275 nanoseconds
Actual result: Gain of 101 nanoseconds
  • Clock 408 was gaining 4.84ns per hour relative to the ground clock when the trip began, so it should have gained 389ns by the end of the 80.33hr trip. Instead, it gained 490ns (it was ahead of ground clock by 490ns before trip, but ahead of ground clock by 980ns after trip). This proves it actually gained 101ns on the Westward flight.
  • Source: Same as previous.
Clock 361

Eastward prediction: Loss of 59ns
Actual result: Loss of 54ns

Westward prediction: Gain of 275ns
Actual result: Loss of 44ns
  • Source: Same type of analysis and source as previous.
Clock 120

Eastward prediction: Loss of 59ns
Actual result: Loss of 196ns

Westward prediction: Gain of 275ns
Actual result: Gain of 413ns


Clock 447

Eastward prediction: Loss of 59ns
Actual result: Loss of 96.6ns

Westward prediction: Gain of 275ns
Actual result: Gain of 26ns

Here is a graph of this data, derived from overlaying the prediction graph upon the results graph provided by Hafele in the 1971 report:

These results cannot be called evidence of time-dilation nor support for Special Relativity; in fact, Hafele said in 1971 that "the difference between theory and measurement[...] is disturbing". On both trips, one of the four clocks changed time in the opposite direction predicted; Clock 408 gained when predicted to lose going East, and Clock 361 lost when predicted to gain going West. That means 2 of the 8 changes (25%) were of the wrong numerical sign!

Of the remaining 6 changes which were of the correct sign, 4 were of the wrong
quantity. Some were wildly out of agreement; for instance, Clock 447 gained only 1/10 the amount of time predicted for the Westward trip, while Clock 120 lost three times the amount predicted for the Eastward trip!

Thus, only 2 out of 8 data points are in agreement with Special Relativity. Clearly, a 25% success rate cannot be considered support of the theory. As Hafele himself admitted in 1971:
  • "Most people (myself included) would be reluctant to agree that the time gained by any one of these clocks is indicative of anything"
I disagree; I think they are indicative of something, but it's not time-dilation; it's how unreliable the clocks were. Ironically, this unreliability was known to the experimenters beforehand, but they proceeded anyway. Quoting from the 1972 paper:
  • "[No] clocks keep precisely the same time, even when located together in the laboratory, but generally show[...] differences which in extreme cases may amount to time differences as large as 1 usec per day. [T]he relativistic time offsets expected in our experiments are only of the order of 0.1 usec per day"
Both trips lasted several days. The predicted effect was ten times smaller than the amount by which each clock could randomly vary under stable laboratory conditions. On the flights, the instability was even worse! Quoting Hafele:
  • "[C]locks[...] cannot be expected to perform as well under traveling conditions as they do in the laboratory."
How Hafele & Keating expected to obtain a statistically significant result from 4 such extremely erratic clocks, is baffling.

There is one more important fact to be noted. Although all 4 clocks
demonstrated instability (they kept randomly alternating between gaining and losing time before, during, and after flights), one of the clocks was very stable compared to the others. This was Clock 447:
  • "[R]ate changes that are noticeably larger than those typical in the laboratory occurred for each clock during at least one of the trips, except for clock 447."
It is no coincidence that this clock also demonstrated the least alleged time-dilation. Here is a graph taken from the 1972 report illustrating this; I've highlighted Clock 447's data in green.


Note how straight each of the three segments for Clock 447 are compared to the other clocks; this illustrates the stability of this particular clock. All three segments are roughly parallel to each other, meaning it's drift rate was roughly constant during the entire experiment, and one can draw an imaginary line from beginning to end connecting the three segments, and see how their is no abrupt step up or down, contrary to what would be expected had time-dilation actually occured.

The other clocks drifted very erratically, and by quantities that dwarfed the expected relativistic effect.

Premise 2 of my opponent's argument is thusly falsified. The Hafele & Keating Experiment is not evidence of Special Relativity. All it demonstrates is that the clocks used were so unreliable as to render them useless for the purposes of the experiment.

Thanks to A.G. Kelly, PhD, whose paper brought this issue to my attention and saved me the trouble of having to do all of the math myself: http://www.cartesio-episteme.net...
Debate Round No. 2
SargonOfAkkad

Pro

Ave

(I switched from 12pt to 14pt font to make the text larger. I know what it's like to read a lot of complex information written in tiny text.)

I would like to note that Con is basing many of his arguments off of the link he offers at the bottom. Firstly, he fails to explicitly attribute any of this to the author, which I find concerning. Secondly, the author himself is a conspiracy theorist who denies that Hafele and Keating even published the numbers that they actually measured. This is a radically unhistorical view of Hafele and Keating that is not accepted by the consensus in either the history of science or science itself. Furthermore, despite the fact that he has a PhD, he provides no evidence of any legitimate academic credentials or publications. I myself did a search on google scholar and could not find any scientific publications from him. For these reasons, the validity of this source is seriously in doubt.


Con’s first argument is that the times measured on the clocks were different from the times predicted by special relativity. He claims that these numbers are based on Hafele and Keating’s paper, even linking us the paper for us to verify his claim on our own. Unfortunately, Con’s claims are totally inconsistent with Hafele and Keating’s paper. Here is an image of their paper stating the results of their experiment.






From "Around-the-World Atomic Clocks: Observed Relativistic Time Gains". Published in the journal Science in 1972. [1]

Con claims that clock 408 gained 165 nanoseconds when traveling in the eastward direction. However, Hafele and Keating’s results clearly state that clock 408 lost fifty-five seconds in the eastward direction, as special relativity would predict. Con claims that clock 408 gained 101 nanoseconds in the westward direction. The actual number was 266 nanoseconds, but even if it was 101 nanoseconds, special relativity predicts a gain in nanoseconds in the westward direction. For this reason, Con supports special relativity when he states that clock 408 gained milliseconds in the westward direction.


Con claims that clock 361 lost 54 nanoseconds in the eastward direction. Again, he is supporting special relativity here, as special relativity predicts a loss of time when traveling in the eastward direction. Con claims that clock 361 lost 44 nanoseconds in the westward direction. This is false, as Hafele and Keating clearly state in the above image that clock 361 gained 284 nanoseconds in the westward direction, as special relativity would predict.


Con claims that clock 120 lost 196 nanoseconds in the eastward direction. Again, he is supporting special relativity here, as special relativity predicts a loss of time when traveling in the eastward direction. Con claims that clock 120 should have gained 275 nanoseconds in the westward direction, but actually gained 413 nanoseconds. In reality, clock 120 gained 277 nanoseconds.


Con claims that clock 447 lost 96 nanoseconds in the eastward direction. In actuality, clock 447 lost 51 nanoseconds in the eastward direction. Both numbers, however, are consistent with special relativity’s prediction of a loss of time in the eastward direction. He claims that Hafele and Keating predicted clock 447 to gain 275 nanoseconds, but measured that it only gained 26 nanoseconds. However, their paper clearly shows that the measured a time gain of 266 nanoseconds.


Con claims that clock 447 was the “most stable and “experienced the least time dilation”. Since it experienced the least time dilation, Con takes this as proof that time dilation is just caused by clock instability. His argument is clearly absurd, as clock 408 experienced time dilation that was approximately equal to clock 447’s, yet was described by Con as “drifting very erratically”. If time dilation is caused by instability, then we would expect the “stable” clock 447 to experience less time dilation than the “unstable” clock 408. However, they both experienced similar amounts of time dilation. Therefore, Con’s hypothesis is demonstrably false.


Con concludes by writing that “only 2 out of 8 data points are consistent with special relativity”. In reality,
all of the clocks in the eastward direction lost time, and all of the clocks in the westward direction gained time, which is exactly what special relativity would predict.


Con quotes Hafele as stating that “
"the difference between theory and measurement[...] is disturbing".”. Hafele did state these words, but their meaning has been distorted. In the very same paper, Hafele writes that “The results of this analysis are in reasonable agreement with theoretical predictions.” He also states that "In any event, this experiment verifies unequivocally the existence of the predicted east-west directional asymmetry; only more precise magnitudes remain to be established.” Is Hafele contradicting himself in the same paragraph, or has the meaning been distorted? The latter option is abundantly clear. Dr. Thomas Swanson, professor of physics at Oregon State, says that the quote “is taken from a section where he has prefaced his discussion by saying it after the preliminary analysis; the westward time delay is centered at 150 nsec, with large error bars. In the final analysis, published in Science, it ended up at 273±7 nsec.
(the preliminary nature is abundantly clear if you read the entire paper; it is specifically discussed in the transcript of the Q&A at the end)“. He concludes by stating “There are other quotes, and not particularly careful examination is needed to show that these, too, are taken out of context so as to change the meaning”. [2]

Con quotes Hafele as stating that “
"Most people (myself included) would be reluctant to agree that the time gained by any one of these clocks is indicative of anything". Professor Swanson states that the quote “is actually taken from a broader statement about the reasoning behind taking four clocks so that they can be averaged, because the precision of one clock was insufficient to do the experiment.” [3] Hafele’s meaning has been distorted once again.


Con's argument relies on two things. 1) The interpretation of the numbers he offered and 2) The interpretation of the quotes he offered. In the case of 1), I have proven that every number offered by Con either supports special relativity or was not the number that Hafele and Keating actually measured. In every case where Con incorrectly stated the number measured by Hafele and Keating, the correct number was predicted by special relativity. In regards to two, Hafele's words are being twisted and distorted to take on a different meaning. As his actual words show, Hafele firmly beleived that his experiment offered support for special relativity.


Con's argument amounts to nothing more than false information or out of context quotes. For this reason, a Pro vote is the most justified


Vale


References

http://www.sciencemag.org...


http://www.scienceforums.net...


Ibid



GarretKadeDupre

Con

Pro thinks the numbers in his graph display the actual time changes of the clocks, and most of his argument relies on this premise. Since this premise is false, his argument collapses like a house of cards.

The numbers do not display the actual time changes measured on the clocks. The numbers are fudged, plain and simple. They are the end result of an extremely contrived, and ultimately deceptive, analysis of the raw data.

If Pro had provided the entire graph, instead of a misleadingly cropped portion, the fraudulent nature of the numbers would have been more apparent.

The top of the graph (which Pro cropped out) says:
  • Table 1. Observed relativistic time differences from application of the correlated rate-change method to the time intercomparison data for the flying ensemble.

The numbers Pro gives are the result of the raw data after being subjected to 'the correlated rate-change method'. Although the 1972 paper provides a general overview of what this entails, it doesn't go into too much detail; it doesn't show the math. So what exactly was done to arrive at these fudged numbers is unknown; its deceptive nature, however, is obvious, since I have the actual raw data right in front of me, and linked to it in my previous round. This raw data screams something very different constrated with the numbers in Pro's graph, as explained my previous round.

Here's the complete screenshot of Pro's graph:


What Pro said about the guy I gave credit to at the bottom of my last round is completely off-base. It's just poisoning the well; it doesn't matter even if the guy is schizofrenic or thinks he can turn lead into gold (neither are true, I'm just making a point), it doesn't mean that just because I gave credit to him where credit is due, my entire argument somehow crumbles.

Pro's objection to the credibility of my math is completely unfounded. I gave two examples of my complete mathematical analysis at the beginning of my round, and just gave the end results for the others since I figured too many numbers would be distracting. However, if Pro has a problem with my math (if he thinks I made a mistake), he can point to a single, tiny error. If he can't find any error, he has to drop this absurd insinuation and apologize. All the most important raw data is right here, in Table 3 of Hafele's 1971 report:



This graph can be found at the U.S. Military website here, which I already cited in the previous round, but which I will link to again: http://handle.dtic.mil...

If Pro demands that I provide all my math with a complete justification, I will do so in my final round. For the first few numbers in my previous round, I performed the math myself on the raw data to confirm that A.G. Kelly's math was correct. Since it turned out to be trustworthy, I didn't see a need to repeat the work myself for the rest of the data, so I just copied his. I repeat, Pro has not pointed to a single mathematical or conceptual error in my computations, even though I explained my math thoroughly in the case of Clock 408 (and it applies to every single clock) and all the raw data I used is in the graph I just displayed.

At most, Pro might be able to find a tiny error in my rounding of significant figures, but even that's unlikely, and it wouldn't alter the conclusion anyways.

Refuting the rest of Pro's arguments require at least some understanding of how the two papers relate to each other. Let me explain.

The experiment was performed by Hafele & Keating in 1971. Hafele (alone, it seems) wrote a preliminary report on the data. He noted that no single clock supported the existence of time-dilation, and their numbers were pretty wild (scattered all over the place). He also noted that if one were to average the data of all 4 clocks, then it would be qualitatively consistent with the predictions. In other words, on the trip predicted to lose time, the clocks, on average, lost time, and when predicted to gain time, the clocks, on average, gained time. However, the average quantity lost was within the range of experimental error:
  • "It is amusing to notice that the values for the eastward trip are in excellent agreement despite our expectation that we would not be able to detect a definite effect in this direction."
So it's not statistically significant, and therefore, utterly useless in serving as a confirmation of theory. Regarding the Westward trip, even though on average they did gain time, the average was far below the predicted amount; indeed, not a single clock gained the predicted quantity of time:
  • "the effect for the westward trip was predicted to be considerably larger and detectable, but the observed value is more than one standard deviation below the predicted value."
None of them gained enough on the Westward trip, and indeed, as noted earlier, one lost time!

I hate that it has come to this, but my opponent opened Pandora's Box by suggesting, without justification, that I made an error in my calculations, and also by needlessly attacking the reputation of Doctor Kelly. Clearly, this is the first time Pro's encountered the 1971 report and it's raw data (which was not in the public eye until relatively recently), and he has never felt the need to read either paper with a critical lens, or even try to understand them fully; he prefers to put his faith in Dr. Thomas Swanson's interpretation of the reports. I recommend my opponent actually read them both himself, in full, and also look long and hard at the raw data given in Table 3 of the 1971 paper; after that, he can then formulate his next argument with a necessary understanding of the subject matter.

Or, he can continue to poison the well and focus on questioning my math skills without basis. Whatever floats his boat.

Back to the story:

So, after all those discrepancies between theory and data in Hafele's 1971 analysis, he ends the report on an optimistic note by promising to do another analysis and see if he could make the results look more favorable to Special Relativity:
  • "Although the final analysis of our data is not yet completed, we have established, with an intermediate level of analysis, that portable cesium beam clocks are capable of showing relativistic effects with relatively inexpensive commercial jet flights. The results of this analysis are in reasonable agreement with theoretical predictions. However, those who doubt the validity of conventional relativity theory, and there are many people in this category, probably will not be converted by the results shown in Figure 4. Indeed, the difference between theory and measurement in Figure 4 [the color-coded figure from my previous round] is disturbing, and if our final analysis does not improve agreement, an improved version of this experiment should be given serious consideration."
Having quoted that, I feel compelled to point out why these claims are so disingenuous:
  • "we have established[...] that portable cesium beam clocks are capable of showing relativistic effects"
This is not true at all. Such optimism regarding 2 trips of 4 clocks, where the average time lost was not statistically significant, and the average time gained wasn't anywhere close to sufficient, and where 25% of the individual time 'dilations' were in the opposite direction predicted, doesn't 'establish' anything other than a blinding bias in favor of Relativity and a glaring lack of objectivity.
  • "The results of this analysis are in reasonable agreement with theoretical predictions"
Right, notwithstanding the fact that half of them were statistically insignificant (Eastward), and the other half weren't significant enough (Westward).
  • "However, those who doubt the validity of conventional relativity theory, and there are many people in this category, probably will not be converted by the results shown in Figure 4. Indeed, the difference between theory and measurement in Figure 4 is disturbing"
Finally! A glimmer of objectivity shines through.
  • "if our final analysis does not improve agreement, an improved version of this experiment should be given serious consideration."
Since nowhere can the complete math for their 'final analysis' (which was performed to arrive at the fudged numbers in Pro's graph from the 1972 paper) be found, Hafele and Keating have not demonstrated experimental support for time-dilation, and in fact, if anything, have only falsified it.

This is important to note: there was only one experiment, and it spawned the 1971 report of the raw data. The numbers in the 1972 report were manipulated in order to bring them into glowing agreement with Relativity.

I ask the voters to just take a look at the color-coded graph in my previous round, and honestly ask themselves if their is any possible way that a mathematical operation on this data could result in a genuine, non-contrived agreement with the predictions which were so obviously not in accord with the actual results.

Pro said that
  • "I have proven that every number offered by Con either supports special relativity or was not the number that Hafele and Keating actually measured."
I just proved this is false, and Pro only claims this because he thinks his graph depicts the actual data 'measured' from the clocks.

Pro also said:
  • "In regards to two, Hafele's words are being twisted and distorted to take on a different meaning. As his actual words show, Hafele firmly beleived that his experiment offered support for special relativity."
I did not distort Hafele's words. His language in the 1971 report was far more reflective of reality than in the 1972 publicaton that brought him and his coexperimenter to worldwide fame. And yes, it seems that Hafele ended up convincing himself that his results confirmed time-dilation, but if that's the case, he's deluded himself. In his desperation to force a favorable report from his data, he must have been decieved by his own contrived math. The fact that (apparently) he refuses to reveal his math to the public supports this theory.

But if Pro can find me this math, I'm willing to take this back.
Debate Round No. 3
SargonOfAkkad

Pro

Ave


The Hafele and Keating experiment’s results, as published in the 1972 journal Science, clearly confirm the predictions of special relativity. As the diagram from this paper shows, all of the clocks in the eastward direction lost time, and all of the clocks in the westward direction gained time. The consensus in physics accepts that the times measured for the individual clocks were as predicted by special relativity (within the margin of error). In order to get around this, Con conducts an elaborate conspiracy about Hafele and Keating, alleging that they forged their measurements in the 1972 publication. Instead, insists Con, we can find the real measurements in a different paper, one which shows that the experiment was inconsistent with special relativity. I will first explain why Con’s allegation of a conspiracy is plainly ridiculous. I will then explain why the measurements he cites as the real measurements are not inconsistent with special relativity. I believe that a refutation of this conspiracy theory amounts to a demolition of anything substantive Con has left to say about Hafele and Keating.


Con first tries to deprecate Hafele and Keating’s work by arguing that the correlated rate-change method is some kind of conspiracy used by Hafele and Keating to make their data consistent with special relativity. I myself do not exactly understand what the correlated rate-change method is, so instead of speaking out of ignorance, I consulted Tim Maudlin, professor of philosophy at New York University. He has published books on special and general relativity and is involved in debates on the foundational issues of quantum mechanics. Professor Maudlin told me that the correlated rate-change method in the paper was used to accomplish more precise numbers than standard averaging, even though standard averaging already supported special relativity. He explained to me why the basic claim of the 1972 paper is correct, and then concluded by stating that “Anyone [Con] trying to cast doubt on the basic claims of the 1972 paper by raising question about the method of statistical analysis does not know what they are talking about.” [1] Therefore, Con’s claim that the correlated rate-change was a way for Hafele and Keating to create fraudulent data is nothing more than an insult to two great and scientifically accomplished men. In reality, the correlated rate-change method was 1) used to accomplish more precise numbers than standard averaging, 2 ) consistent with standard averaging anyway, and 3) obviously not a way to cast doubt on the 1972 paper.


Con then tries to deprecate Hafele and Keating’s work by arguing that they did not precisely define correlated rate-change, but they did so quite clearly in their paper. I refer him to page 169 [2], where Hafele and Keating spend at least three paragraphs explaining what the correlated rate-change method is. On page 170 [3], Hafele and Keating explain the math behind this method as well. I should also note that Con is employing a double standard. He wants us merely trust the mathematics done by his shady and unverifiable source, yet he emphatically demands explicit mathematical proof from Hafele and Keating.


As the third and final horn of his conspiracy theory, Con posts a table (Table III) from a different paper by Hafele, and then argues that this is the raw data taken by Hafele and Keating. He then concludes that this raw data is inconsistent with special relativity. However, the diagram that he posted is obviously consistent with special relativity if one takes a look at its numbers. The final column measures the relativistic time change, and as one can see, three of the four clocks in the eastward direction lost time, for an average time loss in the eastward direction, and four of the four clocks in the westward direction gained time, for an average time gain in the westward direction. One may think that these results are not consistent with special relativity because one clock gained three nanoseconds in the eastward direction, but as Con quotes Hafele as stating, one single clock cannot prove time dilation. Why is this? It’s because one clock on its own is too imprecise to prove time dilation. Using the same reasoning, however, one clock on its own is too imprecise to disprove time dilation. Therefore, the fact that one clock out of four gained time instead of losing it is not a danger to special relativity. Con also makes a big deal out of the fact that the measured time dilation in the westward direction was one standard deviation beneath the predicted time dilation in the westward direction. This is irrelevant, as Professor Maudlin informed me that Hafele and Keating’s data has been re-verified at a higher precision since their original experiment.


I will offer another positive argument for trusting the results of the Hafele-Keating experiment. In science, if the results of an experiment can be replicated in different settings, then it is taken as proof that the initial experiment was accurate. Thankfully for special relativity, the results of the Hafele-Keating experiment can replicated in different settings. Wikipedia cites the published papers for at least five experiments which replicate the results of the Hafele-Keating experiment [4]. Therefore, according to conventional scientific thinking, we can have confidence in the results of the Hafele-Keating experiment.


In short, there is no conspiracy. There are merely two papers which are totally consistent with special relativity that Con has misunderstood as proof that Hafele and Keating are actually liars. Evidence and proper interpretation show that Con has slandered two great scientific minds without any basis.


Con admits that (on average) in the eastward direction, time is lost, while in the westward direction, time is gained. He states that the average time loss in the eastward direction is “within the range of experimental error” and not “statistically significant”. I believe that this throws around statistics vocabulary without understanding their meaning. If a result is not statistically significant, then it is indistinguishable from natural skewing. However, if a result is different from zero, then the result is due to more than natural skewing. Since a result that is different from zero is due to more than natural skewing, then it cannot be statistically insignificant, as statistically insignificant results are indistinguishable from natural skewing. Therefore, if the results of the experiment are different from zero, they cannot be statistically insignificant. Unfortunately for Con’s case, the results of the experiment are different from zero [5] , so they cannot be statistically insignificant. He states that there was an average time gain in the westward direction, but that it was below the predicted amount. I have already explained the problem with this logic earlier in the rond. Con has therefore failed to refute the average time gains and losses of the clocks, so I will begin the next paragraph with the average time gains and losses as a given.


Premise 2 of the argument asserts that the Hafele-Keating experiment affirms the predictive power of special relativity. In other words, it is consistent with the predictions of special relativity. Special relativity would predict an average time gain in the westward direction, and an average time loss in the eastward direction. The Hafele-Keating experiment, as Con himself stated, affirmed an average time gain in the westward direction and an average time loss in the eastward direction. Therefore, the Hafele-Keating experiment affirms the predictive power of special relativity. Since Con and I agreed that the experiment supports special relativity if it affirms its predictive power, the resolution is established on this basis as well as many others. Con states that it is not enough to have time dilation, and the experiment must prove time dilation as well as special relativity’s predictions for the degree of time dilation. I accept this standard, as the consensus in the physics community is that the numbers measured by Hafele and Keating are consistent with the predictions of special relativity within the margin of error.


Con states that the numbers he offered in his first round were based on A..G Kelly’s mathematics. I have already casted legitimate doubt on the accuracy of A.G. Kelly as a source, given that his academic credentials have not been verified to any extent. Con claims that this is “poisoning the well”, but this fallacy actually refers to the act of saying bad things about an opponent before they speak in order to make their arguments less convincing, not the practice of asking perfectly legitimate questions about the accuracy of a source [6]. Even if A.G. Kelly was a reliable source, he does not represent the consensus in physics, which has been in favor of special relativity for decade upon decade. For these reasons, we have no reason to trust the numbers that Con posted in R1, and plenty of reasons to trust the numbers offered by Hafele and Keating in their 1972 experiment.

Thanks to Con for a testy and interesting debate. Thanks to the audience for reading this debate. I'm sure it's not easy to read through this kind of material and then vote on it. I believe that the evidence in favor of the accuracy of Hafele and Keating's experiment is abundant, and that Con's reasons for doubting it rely on misunderstandings of the data and method. Whatever one may vote, I hope that they make this vote without reference to bias or preconceived notions, but with reference to evidence and logic.


Vale

References

1: Personal e-mail communication with Tim Maudlin.
2: http://www.sciencemag.org...
3: Ibid
4: http://en.wikipedia.org...
5: http://www.k1man.com...
6: http://changingminds.org...






GarretKadeDupre

Con

I appreciate Con consulting an expert source for the purposes of our debate, but the fact is his expert source provided nothing but an expert opinion. He provided Con no math, not even a single digit. So Con's assertion that the "correlated rate-change method" is honest is just an appeal to authority. This entire debate requires me to challenge the authorities named Hafele & Keating; if appeals to authority were acceptable, I would have lost this debate before we even started, which is absurd.

Con said I can find out what the correlated rate-change method is by referring to the 1972 paper. This is not true. It's not only vague, it lacks data. For instance:
  • "A piecewise extrapolation of the time trace for each clock relative to MEAN(USNO), with proper accounting for those identified rate changes, then produced the relativistic time difference listed in Table 1."
What in the world is "proper accounting" in this context? It's unclear. It could mean anything; the 1972 paper only provides a qualitative, not quantitative, description of this mysterious method which has the capacity to bring frustrating results into agreement with theory.

I ask the voters to ask themselves honestly, is it at all possible to derive a statistically significant result when looking for a relativistic effect that is 10 times smaller than the margin of error for a single clock? Of course not. Now, would increasing the sample size to 4 such clocks enable one to derive a statistically significant result? No; 4 is just not going to cut it. You would need perhaps 100 such clocks, or, even better, clocks that are at least 10 times more precise. It doesn't take a PhD in Statistics to realize this.

Look at it another way; the most reliable clock was #447, as Hafele himself admitted. It also demonstrated the least time-dilation! Clearly, this correlation between clock reliability and lack of time-dilation should undermine the credibility of using this experiment to prove time-dilation. It's funny how the authors ended their 1972 report on this note:
  • "[T]here seems to be little basis for further arguments about whether clocks will indicate the same time after a round trip, for we find that they do not."
Lol! Of course you're going to find a running clock tells a different time after a flight of several days; what else can you expect?!

Con says,
  • "One may think that these results are not consistent with special relativity because one clock gained three nanoseconds in the eastward direction, but as Con quotes Hafele as stating, one single clock cannot prove time dilation. Why is this? It’s because one clock on its own is too imprecise to prove time dilation. Using the same reasoning, however, one clock on its own is too imprecise to disprove time dilation."
On both trips, 1 out of 4 clocks changed time in the wrong direction. I'm not using this to disprove time-dilation; I'm using it to prove that this experiment is useless for demonstrating time-dilation.

Both Con and the experimenters make a big deal out of the fact that the average of the results were negative when predicted to be negative, and positive when predicted to be positive. But this is not a big deal at all. There were only 4 possibilities, and they just got lucky. It could have been positive/positive, negative/negative, or positive/negative. Instead, it was negative/positive. That's just luck (or, as Hafele put it in 1971, "amusing"). And it wasn't even quantitatively correct!

Con says,
  • "Professor Maudlin informed me that Hafele and Keating’s data has been re-verified at a higher precision"
But Maudlin is simply referring to the correlated rate-change method, which we are unable to verify! Does Con want us to take Maudlin's word for it that the method is trustworthy? If it was trustworthy, why was it not published in full in the 1972 report?

Con says other experiments confirmed this one but that's not fair; this debate is only about this particular experiment. I disagree with his assertion but it's not as if I have space to falsify all those other experiments in this last round of the debate.
  • "Con has slandered two great scientific minds without any basis."
I didn't slander anyone. I was very careful with my wording. I said the data appears to be fraudulent; I didn't say it actually is.
  • "[A] result that is different from zero is due to more than natural skewing, then it cannot be statistically insignificant"
Lol! Con doesn't understand what statistically insigificant is! Quantities greater than zero can be insignificant, as long as they are within the range of experimental error. The fact that each clock had a margin of error 10 times larger than the predicted relativistic effect, proves that any results which appear to be in favor of theory are not significant; they are meaningless.
  • "Con has therefore failed to refute the average time gains and losses of the clocks"
Are you kidding me!? You can't parade this as proof of theory whenever the quantities of gains and losses were either of the wrong amount, or within the range of error!
  • "Con states that it is not enough to have time dilation, and the experiment must prove time dilation as well as special relativity’s predictions for the degree of time dilation. I accept this standard, as the consensus in the physics community is that the numbers measured by Hafele and Keating are consistent with the predictions of special relativity within the margin of error."
In other words, even though Con has absolutely no explanation for why the amount of alleged time-dilation appears to be the wrong quantity, he will assume, based on the consensus of the physics community, that it was of the right quantitity anyway. The fact that Con repeatedly must defer to a third party's opinion instead of actually providing hard evidence, like actual numbers, reveals the feebleness of his case.

By the way, here is the graph of my numbers. The colored graph I already posted was actually of Hafele's numbers. The difference between our two graphs is that Hafele took the unsteadiness (drift-rate) of the clocks before the flights, and after, then corrected the raw time changes using the average of the two drift-rates; in my graph (and numbers) I simply assumed the instability before the flight applied throughout the flight. In neither case do the results confirm the quantitative predictions of Special Relativity.


Debate Round No. 4
12 comments have been posted on this debate. Showing 1 through 10 records.
Posted by GarretKadeDupre 2 years ago
GarretKadeDupre
Thanks dylancatlow!
Posted by dylancatlow 2 years ago
dylancatlow
RFD: Overall, I think Con argued his case better considering his inherent disadvantage (going against the consensus). Pro was often too general in his replies and didn't spend enough time addressing Con's main points. For instance, Pro says "Con states that it is not enough to have time dilation, and the experiment must prove time dilation as well as special relativity"s predictions for the degree of time dilation. I accept this standard, as the consensus in the physics community is that the numbers measured by Hafele and Keating are consistent with the predictions of special relativity within the margin of error." This was the single most important point of the debate, and yet Pro passed up the chance to explain why Con's position was wrong in detail. He also accused Con of proposing an "elaborate conspiracy" which I think was unwarranted. If anything, he was accusing them of standard scientific dishonesty. I don't think the raw data provided by Con was accurate enough to support any specific prediction on its own. The correlated rate-change method, which supposedly made up for this, was never explained.
Posted by dylancatlow 2 years ago
dylancatlow
This is probably the least sciency science debate I've ever read lol
Posted by GarretKadeDupre 2 years ago
GarretKadeDupre
dylancatlow, it's not sufficient to demonstrate time dilation. The time-dilation must be in agreement, within the range of experimental error, with the specific, quantitative predictions of SPECIAL RELATIVITY.

This debate is about Relativity's time-dilation, specifically; not just time-dilation in general.

But thanks for your analysis.
Posted by dylancatlow 2 years ago
dylancatlow
"we have established[...] that portable cesium beam clocks are capable of showing relativistic effects"

This is not true at all. Such optimism regarding 2 trips of 4 clocks, where the average time lost was not statistically significant, and the average time gained wasn't anywhere close to sufficient, and where 25% of the individual time 'dilations' were in the opposite direction predicted, doesn't 'establish' anything other than a blinding bias in favor of Relativity and a glaring lack of objectivity.

It's far easier to demonstrate the existence of time dilation in general than to confirm specific predictions. I don't think he was claiming to have verified specific predictions with this experiment. Also, I'm not sure if the fact that some clocks lost time instead of gained time (as predicted) is actually significant in and of itself. In other words, the only significance might be the absolute difference between prediction and experiment.

I think the winner of this debate will be decided by two things:

1. Whether Con's data is accurate/relevant.
2. What exactly is meant by "supports special relativity".

I think Con's data only supports the hypothesis that "time dilation occurs". It can't really get more specific than that.
Posted by GarretKadeDupre 2 years ago
GarretKadeDupre
Nevermind.
Posted by SargonOfAkkad 2 years ago
SargonOfAkkad
Why?
Posted by GarretKadeDupre 2 years ago
GarretKadeDupre
Delete the sentence "Hafele and Keating found that the three sets of measurements were inconsistent with each other" and I will accept.
Posted by SargonOfAkkad 2 years ago
SargonOfAkkad
Firstly, you are correct in stating that the invariance of the speed of light only applies to inertial reference frames, so I will gladly make this correction.

Secondly, when I state that it is "taken as support for special and general relativity", I just mean that the conventional view in physics is that this experiment supports special and general relativity. It does not mean that this conventional view is correct, leaving the resolution an open question. I hope this clarifies any ambiguity over my meaning.

As for how the Hafele-Keating experiment supports general relativity, I will point you to the fact that the Hafele-Keating experiment measured an increase in gravitational potential that sped the clocks back up, as predicted by GR.

In any case, if you are correct, you have only provided warrants for removing two sentences or so, not entire paragraphs of text.
Posted by GarretKadeDupre 2 years ago
GarretKadeDupre
There are at least a couple problems.

First, I do not see how the experiment can be taken as support for General Relativity, per your claim that it's "taken as support for special and general relativity."

Second, Special Relativity does not postulate that "the speed of light in a vacuum is the same for all observers." It only says so for inertial observers.

I prefer your make the revisions I proposed in the previous comment, and we can give our own story about the experiment in our arguments. This is because what exactly they did is up for debate.
2 votes have been placed for this debate. Showing 1 through 2 records.
Vote Placed by Paleophyte 2 years ago
Paleophyte
SargonOfAkkadGarretKadeDupreTied
Agreed with before the debate:--Vote Checkmark0 points
Agreed with after the debate:--Vote Checkmark0 points
Who had better conduct:Vote Checkmark--1 point
Had better spelling and grammar:--Vote Checkmark1 point
Made more convincing arguments:Vote Checkmark--3 points
Used the most reliable sources:--Vote Checkmark2 points
Total points awarded:40 
Reasons for voting decision: Pro successfully demonstrates that the H-K expt demonstrates SR. Con rebuts with misunderstood and misinterpreted data analysis, thus failing to defeat the resolution. Con further insults his opponent and respected scientists with needless accusations of dishonesty.
Vote Placed by dylancatlow 2 years ago
dylancatlow
SargonOfAkkadGarretKadeDupreTied
Agreed with before the debate:--Vote Checkmark0 points
Agreed with after the debate:--Vote Checkmark0 points
Who had better conduct:--Vote Checkmark1 point
Had better spelling and grammar:--Vote Checkmark1 point
Made more convincing arguments:--Vote Checkmark3 points
Used the most reliable sources:--Vote Checkmark2 points
Total points awarded:00 
Reasons for voting decision: RFD in comments