CERN team claims measurement of neutrino speed >c

[dan_b]

I'm not arguing that GPS clocks aren't used. I'm arguing that GPS clocks in an application requiring nanosecond-level synchronization of distant points is rare. Thus far, nobody has mentioned one.

Hi Vanadium,

I kind of lost track of the good stuff but now you sheared the sheep and I read it ALL. Er, I think I did! I didn't see a straight answer to your question so here's one: You know all those radio astronomers? Many of them use Very Long Baseline Interferometry, because they can. Nanosecond timing over long baselines? You bet. Academic? Sure, but there's a lot of them and they make good resolution radio-frequency pictures which proves their timing must be good, or else the pictures would be bad. They time lock the receivers local oscillators over huge distances so the individual receiver signals can be coherently combined. Tons of resolution that way. Not so much receiver net gain, of course, since they don't have enough money to pave over the whole world with antennas. :-) It's not all that rare.

 

[PatrickPowers]

What makes you think that 160 researchers from 30 institutions and 11 countries working for 5 years would have missed something like this, if it has any value?

Quite so. My Bayesian prior that neutrinos would move faster than light is close to zero. However my Bayesian prior that an error would be overlooked by this team is also close to zero.

So I don't worry about it, and will wait and see. Let people who are paid for it do the work.

 

[lalbatros]

A simple analysis gives no chance at all for the OPERA result to be relevant.
You can find it in this arXiv paper:

http://arxiv.org/PS_cache/arxiv/pdf/1110/1110.5275v1.pdf

Have a look at fig 4 from this paper:

[PLAIN]http://img543.imageshack.us/img543/3149/operadelays.jpg

Can you guess which of the red or blue curves isthe OPERA best fit?
If you can't make your choice, then this probably indicates the OPERA result is irrelevant.

 

[DevilsAvocado]

... Can you guess which of the red or blue curves is

Is this supposed to be a tricky question?

A blind coconut could see that the x-axis is ns, and which curve gets in/out first...

I think someone on BBC promised to eat his underwear live, if the OPERA result holds, I’m almost about to do the same on that the red curve is the OPERA result...

 

[lalbatros]

Is this supposed to be a tricky question?

A blind coconut could see that the x-axis is ns, and which curve gets in/out first...

I think someone on BBC promised to eat his underwear live, if the OPERA result holds, I’m almost about to do the same on that the red curve is the OPERA result...

I changed my question, but it is still easy to answer.
A blind coconut also knows that the OPERA team claimed a FTL result.
Maybe I should remove any comment.
The picture speaks for itself: CERN stumbled on a coconut.

Seriously, how is it possible to claim a six-sigma result on this basis?
Lies, damned lies, and statistics!

 

[DevilsAvocado]

CERN stumbled on a coconut.

 

[pnmeadowcroft]

You can find it in this arXiv paper:

http://arxiv.org/PS_cache/arxiv/pdf/1110/1110.5275v1.pdf

Nice paper H. Bergeron, he's good, really good. Presents the kind of analysis of the statistical that seemed appropriate for the original paper. He had to collect his data values from poor quality grafs, and still managed to demonstrate how the original 60ns result can be arrived at, as well as some of the weaknesses of the approach.

If the main paper does go to publish, I do hope it will come with full data, and the complete statistical calculations.

I've seen something else too, but I cannot quantify the effect yet. The height of the bars in Fig. 12 of the main paper effectively represents the chances of an event being misclassified into an earlier or later 50ns segment. However, due the the steep nature of the leading edge of the pulse, the probability of misclassifications resulting a higher number in a given 50ns segment is greater than the probability of misclassifications resulting in a lower number in a given segment. This would bias the curve fit at the leading edge towards a shorter flight time.

This can all be accounted for, but it is unclear from Fig. 12, again put the key data in the paper

 

[dgwsoft]

Faster-than-light neutrino experiment to be run again

With nanosecond proton pulses. As reported by the BBC

http://www.bbc.co.uk/news/science-environment-15471118

 

[danR]

This may be 'overly speculative', or already addressed, but assuming the result to stand, is the measured c in a real vacuum, the same as what it would be in an absolute vacuum devoid of zero-point fluctuating electromagnetic fields that would minutely reduce the propagation of light, analogously to the impedance of light through the electromagnetic fields always present in matter?

Neutrinos would not see any impedance, neither through matter nor in vacuo.

Even leaving aside zero-point fluctuations, empty space is not empty of electromagnetic noise, hence field fluctuations, from all manner of sources. If not a factor, I assume both sorts of things have been long accounted for in theoretical considerations of the value of c. Perhaps someone would know that.

 

[danR]

This is a systematic effect. You can take that to the bank.

They don't see a velocity dispersion. By itself, that's a huge problem. If you want to argue that not only are neutrinos faster than light, but they all travel at the same speed regardless of energy, you have to explain why the neutrinos from SN1987A arrived on the same day as the light did, instead of (as the Opera data would indicate) four years earlier.

Did anyone look for a signal four years earlier? Did a small, separate class of neutrino-component of the supernova emission behave differently from the rest of the pack?

 

[PAllen]

Did anyone look for a signal four years earlier? Did a small, separate class of neutrino-component of the supernova emission behave differently from the rest of the pack?

See #237 for one set of arguments on this.

 

[danR]

See #237 for one set of arguments on this.

Perhaps I misunderstand those discussions, or some of them were culled out in the thread housecleaning.

#237 seemed to discuss a separate high-energy event. I'm wondering if (for some delete-worthing speculative reasoning) SN1987a spat out 2 different type/energy neutrino components, from, perhaps, two different production mechanisms in the same detonation, either simultaneously, or closely sequential (<2 minute, say).

Obviously the posited 'superluminal' neutrino packet will have no accompanying light signal. That would come substantially at the same time as the main, and recorded, light/neutrino group.

We can't ask someone to look for something when EM has not heralded it, of course. Does someone keep records of neutrino events and spikes? Is there a spike buried in the historical data somewhere around 4 years before SN1987a? Without knowing, the CERN question is not answered by the supernova.

Otherwise it's rather like the drunk looking for his keys only under the proverbial lamppost.

"Is that where you lost them?"
"No, but the light is better here."

 

[Parlyne]

The SN1987a neutrino burst wasn't the sort of event that would have spent decades hiding in the data until someone went hunting for it. (You can see reproductions of the original data in this iop pdf about the event: iopscience.iop.org/1742-6596/120/7/.../jpconf8_120_072001.pdf[/URL].) It's hard to imagine that an earlier but related burst could have been totally overlooked, even if one of the detectors involved in 1987 wasn't online yet.

 

[Parlyne]

This may be 'overly speculative', or already addressed, but assuming the result to stand, is the measured c in a real vacuum, the same as what it would be in an absolute vacuum devoid of zero-point fluctuating electromagnetic fields that would minutely reduce the propagation of light, analogously to the impedance of light through the electromagnetic fields always present in matter?

Neutrinos would not see any impedance, neither through matter nor in vacuo.

Even leaving aside zero-point fluctuations, empty space is not empty of electromagnetic noise, hence field fluctuations, from all manner of sources. If not a factor, I assume both sorts of things have been long accounted for in theoretical considerations of the value of c. Perhaps someone would know that.

Short answer: Quantum field theory says "no."

Longer answer: In QFT the effect of vacuum fluctuations on the free propagation of a particle show up either as corrections to the particle's mass or as a rescaling of the field that the particle is an excitation of. The later case won't actually affect the propagation speed (for a particular energy); so, it's only necessary to look at the mass corrections. The thing is, for a gauge boson (or, for that matter, for a fermion) the mass corrections are proportional to the "bare" mass (that is, the mass that you'd see if there were no screening due to the corrections). This means, in particular, that quantum corrections cannot give mass to an otherwise massless particle. In other words, vacuum fluctuations will not affect the propagation speed of a massless particle.

Now, we can certainly consider the possibility that photons are actually massive; but, that actually leads to a different problem. Massive particles have energy-dependent speed; and, we've measured the speed of light (directly or indirectly) over more than 15 orders of magnitude in photon energy and not seen deviation. In fact, this lack of energy (or, more directly, frequency) dependence is the best reason to think that what we've measured really is the speed of light in a vacuum, since any physical medium has a frequency-dependent index of refraction, leading to frequency dependence in the speed of light through the medium.

Finally, the presence of other EM radiation is space is basically irrelevant, since light obeys the principle of superposition (well, up to highly suppressed corrections due to fermion loop diagrams).

 

[danR]

Short answer: Quantum field theory says "no."

Longer answer: In QFT the effect of vacuum fluctuations on the free propagation of a particle show up either as corrections to the particle's mass or as a rescaling of the field that the particle is an excitation of. The later case won't actually affect the propagation speed (for a particular energy); so, it's only necessary to look at the mass corrections. The thing is, for a gauge boson (or, for that matter, for a fermion) the mass corrections are proportional to the "bare" mass (that is, the mass that you'd see if there were no screening due to the corrections). This means, in particular, that quantum corrections cannot give mass to an otherwise massless particle. In other words, vacuum fluctuations will not affect the propagation speed of a massless particle.

Now, we can certainly consider the possibility that photons are actually massive; but, that actually leads to a different problem. Massive particles have energy-dependent speed; and, we've measured the speed of light (directly or indirectly) over more than 15 orders of magnitude in photon energy and not seen deviation. In fact, this lack of energy (or, more directly, frequency) dependence is the best reason to think that what we've measured really is the speed of light in a vacuum, since any physical medium has a frequency-dependent index of refraction, leading to frequency dependence in the speed of light through the medium.

Finally, the presence of other EM radiation is space is basically irrelevant, since light obeys the principle of superposition (well, up to highly suppressed corrections due to fermion loop diagrams).

Sounds good. Thanx

 

[danR]

The SN1987a neutrino burst wasn't the sort of event that would have spent decades hiding in the data until someone went hunting for it. (You can see reproductions of the original data in this iop pdf about the event: iopscience.iop.org/1742-6596/120/7/.../jpconf8_120_072001.pdf[/URL].) It's hard to imagine that an earlier but related burst could have been totally overlooked, even if one of the detectors involved in 1987 wasn't online yet.[/QUOTE]

It would be overlooked until someone came up with some weird data that there might have been an appetizer 4 years previous to the main course. I'm not looking for pizza before I order it.

And until CERN, no one would have any idea when to look for a signal. There could have been all kinds of random or significant bumps or spikes from 1980, say, to 1987. Now there's a place to look for one. Finding 'one' would not prove anything, of course. I'm just wondering if any did, in fact look; because the 'prior to 1987a' argument came up within a day or two on SA comments-section. It's not a new discussion, but I haven't seen an "OK, we looked, there's nothing there."

But I'll check that URL.

Edit: link asks the file be prefixed with '%PDF-', but that's not getting me there.

 

[Parlyne]

It would be overlooked until someone came up with some weird data that there might have been an appetizer 4 years previous to the main course. I'm not looking for pizza before I order it.

And until CERN, no one would have any idea when to look for a signal. There could have been all kinds of random or significant bumps or spikes from 1980, say, to 1987. Now there's a place to look for one. Finding 'one' would not prove anything, of course. I'm just wondering if any did, in fact look; because the 'prior to 1987a' argument came up within a day or two on SA comments-section. It's not a new discussion, but I haven't seen an "OK, we looked, there's nothing there."

But I'll check that URL.

Edit: link asks the file be prefixed with '%PDF-', but that's not getting me there.

Sorry. It looks like the link didn't copy right. Try going here and clicking on the pdf link. http://iopscience.iop.org/1742-6596/120/7/072001

 

[Parlyne]

It would be overlooked until someone came up with some weird data that there might have been an appetizer 4 years previous to the main course. I'm not looking for pizza before I order it.

And until CERN, no one would have any idea when to look for a signal. There could have been all kinds of random or significant bumps or spikes from 1980, say, to 1987. Now there's a place to look for one. Finding 'one' would not prove anything, of course. I'm just wondering if any did, in fact look; because the 'prior to 1987a' argument came up within a day or two on SA comments-section. It's not a new discussion, but I haven't seen an "OK, we looked, there's nothing there."

But I'll check that URL.

Edit: link asks the file be prefixed with '%PDF-', but that's not getting me there.

The point I was trying to make is that the signal identified as being from SN1987a was so huge that no one would have missed it once they were looking at the data. Had anything comparable shown up several years earlier, it similarly would not have been missed.

 

[Passionflower]

The point I was trying to make is that the signal identified as being from SN1987a was so huge that no one would have missed it once they were looking at the data. Had anything comparable shown up several years earlier, it similarly would not have been missed.

On what do you base large quantities?
For all intents and purposes the hypothetical ftl neutrinos might come at very low rates.

 

[danR]

The point I was trying to make is that the signal identified as being from SN1987a was so huge that no one would have missed it once they were looking at the data. Had anything comparable shown up several years earlier, it similarly would not have been missed.

Right. The devil's in the details. I want/ed to clarify there may only be a small signal (if, say 'FTL' neutrinos were produced by some minor process, but all 'FTL' neutrinos have their own, and only one, version of 'c'--call it c_n), and it would be a complete waste of time to look at every Tom, Dick and Harry bump or blip in years of noise and say this or that is something. But if someone looked back at 4 years (or 3.7.., whatever it was) before and there was something interesting, but too noisy, then we could go 'Hmm...' It would look, well, interesting. If nothing more.

I'm becoming a hardened FTL-skeptic, but 'hmm...' data makes my day, if temporarily.

 

[danR]

On what do you base large quantities?
For all intents and purposes the hypothetical ftl neutrinos might come at very low rates.

You beat me to it.

 

[PAllen]

Perhaps I misunderstand those discussions, or some of them were culled out in the thread housecleaning.

#237 seemed to discuss a separate high-energy event. I'm wondering if (for some delete-worthing speculative reasoning) SN1987a spat out 2 different type/energy neutrino components, from, perhaps, two different production mechanisms in the same detonation, either simultaneously, or closely sequential (<2 minute, say).

Obviously the posited 'superluminal' neutrino packet will have no accompanying light signal. That would come substantially at the same time as the main, and recorded, light/neutrino group.

We can't ask someone to look for something when EM has not heralded it, of course. Does someone keep records of neutrino events and spikes? Is there a spike buried in the historical data somewhere around 4 years before SN1987a? Without knowing, the CERN question is not answered by the supernova.

Otherwise it's rather like the drunk looking for his keys only under the proverbial lamppost.

"Is that where you lost them?"
"No, but the light is better here."

The issue is you got a strong neutrino burst observed at time t0 (coincidentally with light from the SN). If these neutrinos arrive 3-4 years before light from the event producing them, the the event should be seen 3-4 years later optically. This would not have been missed because people were on the look (and always are) optically. So, to believe an alternative explanation than the normal one for SN, you must believe in both remarkable coincidence (very rare neutrino event timed to arrive same time as light from SN) + ineptitude of astronomers (to not see later light from this event).

This argument is based on a key assumption stated therein (#237). You can reject it if you want:

Any event producing intense neutrino burst will also produce intense EM radiation.

From what I've seen, the theorists playing with "what if it's true", do not reject this assumption. Instead they reconcile the SN either with an energy threshold effect (the SN neutrinos were order 1000x less energetic than the OPERA ones), or a matter based effect (neutrinos travel faster in matter), or a 'jump start' effect, where neutrinos only travel fast for a brief time after emission (which may be combined with an energy threshold as well).

 

[kikokoko]

about supernova SN1987A:

according to FTL data from CERN, SN1987A neutrinos
should arrive about 4 years before light

1987-4 = 1983

Of the three neutrino observatories that saw neutrinos from SN1987a,
IMB and Baksan detectors were active since 1982

So, if neutrinos had arrived in 1983,
they would certainly have been detected,
since the burst of the supernova was very evident.

All historical data were scrutinised, and nothing appears in publication

 

[danR]

The issue is you got a strong neutrino burst observed at time t0 (coincidentally with light from the SN). If these neutrinos arrive 3-4 years before light from the event producing them, the the event should be seen 3-4 years later optically. This would not have been missed because people were on the look (and always are) optically. So, to believe an alternative explanation than the normal one for SN, you must believe in both remarkable coincidence (very rare neutrino event timed to arrive same time as light from SN) + ineptitude of astronomers (to not see later light from this event).

This argument is based on a key assumption stated therein (#237). You can reject it if you want:

Any event producing intense neutrino burst will also produce intense EM radiation.

From what I've seen, the theorists playing with "what if it's true", do not reject this assumption. Instead they reconcile the SN either with an energy threshold effect (the SN neutrinos were order 1000x less energetic than the OPERA ones), or a matter based effect (neutrinos travel faster in matter), or a 'jump start' effect, where neutrinos only travel fast for a brief time after emission (which may be combined with an energy threshold as well).

"Any event...radiation."

1.This is more than an assumption, but an implication predicated on the assumption 'event producing...intense burst. Someone above alluded to that assumption.

2.The canonical n-burst of SN1987a is well known and loved by all, but is beside the point. At the risk of wearing out a certain mod's patience with over-speculation, we probably all agree that the CERN signal is anomalous. If it is both real, and anomalous, then a purported <1987a signal may well be the signal from some anomalous process separate from the t0 signal.

3. An anomalous <1987 n-signal might indeed be (post-)heralded by some t0 light signal, buried in the well-known light signal, weaker than that signal, and/or having a different spectral distribution from that signal, assuming some process separate from, but essentially happening at the same time as the main light burst.

4. The CERN anomaly is a statistical phenomenon. Is it possible that the neutrinos that were actually conspiring in that statistical group are, in fact, produced by some different interaction?

 

[danR]

about supernova SN1987A:

according to FTL data from CERN, SN1987A neutrinos
should arrive about 4 years before light

1987-4 = 1983

Of the three neutrino observatories that saw neutrinos from SN1987a,
IMB and Baksan detectors were active since 1982

So, if neutrinos had arrived in 1983,
they would certainly have been detected,
since the burst of the supernova was very evident.

All historical data were scrutinised, and nothing appears in publication

If the scrutiny allowed for the possibility of a weak FTL-component that might be successfully parsed out of the background, very good. I'm not going to quibble over statistically worthless blips. End of discussion

 

[tmfs10]

A simple analysis gives no chance at all for the OPERA result to be relevant.
You can find it in this arXiv paper:

http://arxiv.org/PS_cache/arxiv/pdf/1110/1110.5275v1.pdf

Have a look at fig 4 from this paper:

[PLAIN]http://img543.imageshack.us/img543/3149/operadelays.jpg

Can you guess which of the red or blue curves isthe OPERA best fit?
If you can't make your choice, then this probably indicates the OPERA result is irrelevant.

Hmm...are those graphs for one beam or for all the beams that were collected for the final result?

I do agree though that the statistical analysis is the most likely place for an error followed by the GPS clock synchronization.

The next few weeks should happily settle the former.

 

[lalbatros]

Hmm...are those graphs for one beam or for all the beams that were collected for the final result?

I do agree though that the statistical analysis is the most likely place for an error followed by the GPS clock synchronization.

The next few weeks should happily settle the former.

This graph is for all beam.
It summarizes all the information available from the 16000 detected event.

For one beam of proton, that last about 10000 ns, there is not only a small probability to detect only one neutrino in Gran Sasso.
Those neutrinos that fall in the central part of the beam pulse do not contribute any information (except a little bit related to the irregularities in the beam amplitude).
Only those few neutrino that are correlated to the leading or trailing edge of the beam do contribute to the neutrino speed analysis.

The graph represents an average were the counting statistics determines the vertical error bar. This is Poisson statistics, absolute error is the square root of the count.
The horizontal error bar is determined by the precision of time and distance measurements.

The OPERA team claims a 6-sigma quality for their result.
The graph constructed by Henri illustrated quite clearly that this is impossible.
The uncertainty on the delay (or advance) is of the order of 100 ns.
Conclusion: there should be a mistake in the uncertainty calculation.
Probably a conceptual mistake, related to the interpratation and use of the likelihood function.

My current guess is that the likelihood function used by the OPERA team does not test with a 6-sigma precision their hypothesis on the neutrino travel time.
My intuition is that their likelihood function is a test for another hypothesis.
In addition, they did not test the likelihood of neutrino travel at the speed of light, or maybe I was not patient enough to find out in their paper.

I think I will read again the http://www.nr.com/" chapter about least squares.
It should taste very good combined with this OPERA statistical analysis.

 

[AncientCoder]

... or a 'jump start' effect, where neutrinos only travel fast for a brief time after emission (which may be combined with an energy threshold as well).

Programming/systems background(40+ years) so this is way outside of my field, although extremely interesting!

Since the OPERA experiment is designed to test/capture the rare
http://public.web.cern.ch/press/pressreleases/Releases2011/PR19.11E.html" , has any thought been given to the ideal that during the transformation that for a brief moment some kind of field/particle/effect/tachyon/tunneling event might have occurred that would have allowed the particle to "temporarily" seem(?) to go faster than expected?

If one travels a highway with an apparent speed, did they travel the entire highway at that speed, or did they exceed the speed limit for a certain portion, or maybe even take a shortcut?

Testing over a different distance would help clear that up, and parsing the distance might even allow the actual transformation to be captured!

 

[ZapperZ]

Programming/systems background(40+ years) so this is way outside of my field, although extremely interesting!

Since the OPERA experiment is designed to test/capture the rare
http://public.web.cern.ch/press/pressreleases/Releases2011/PR19.11E.html" , has any thought been given to the ideal that during the transformation that for a brief moment some kind of field/particle/effect/tachyon/tunneling event might have occurred that would have allowed the particle to "temporarily" seem(?) to go faster than expected?

If one travels a highway with an apparent speed, did they travel the entire highway at that speed, or did they exceed the speed limit for a certain portion, or maybe even take a shortcut?

Testing over a different distance would help clear that up, and parsing the distance might even allow the actual transformation to be captured!

The problem with this is that, this is actually quite a well-known process. Note that neutrinos are produced on demand at many facilities. MINOS, T2K, etc... etc... all have neutrinos produced by such a process. Furthermore, such knowledge is essential to be able to decipher all of the experiments so far on the mixing angle between the different neutrino flavors. Superluminal neutrinos would throw off those experiments completely that we would have noticed the absurdity of those results by now.

I realize that it is often very hard to contain oneself when something this "big" is reported. But really, this is the time where we should reign in our guesses and possibilities, and let them work this out first. We could easily be discussing a non-existent issue here.

Zz.

 

[lalbatros]

...
I realize that it is often very hard to contain oneself when something this "big" is reported. But really, this is the time where we should reign in our guesses and possibilities, and let them work this out first. We could easily be discussing a non-existent issue here.

Zz.

My understanding is that the Physics Forum mission is pedagogical and oriented to students.
How can it be a pedagogical aim to ask people to be patient and stop thinking, switch in the wait and see mode?

The OPERA paper contains a lot of issues that can be discussed usefully by undergraduated students, like the likelyhood analysis. I have no doubt that many young students could be helpful and make very interresting analysis of such topics. I am sure that the OPERA team could learn from them. I am sure that some people posting here here could write some enlightening paper. Such paper could solve an issue in the OPERA paper or could illustrate some concept on the basis of the OPERA experiment.

Restraining the discussion to a unique melting pot thread was an anti-pedagogical decision.
It discouraged in-depth discussions.

No doubt Feynman would have suggested the exact opposite approach.
Feynmann would have encouraged discussion and critical thinking, without restrain.
But I must concede that he would have already debunked the OPERA claim by now.

 

[OnlyMe]

PhysOrg.net 10/28/11 - http://www.physorg.com/news239009787.html

Scientists who threw down the gauntlet to physics by reporting particles that broke the Universe's speed limit said on Friday they were revisiting their contested experiment...

Kasavenas said CERN was making available a special form of proton beam until November 6.

The idea is to assess a modified measurement technique.

If this works, the technique will be used in a bigger, "highly important" experiment that will be carried out in April, he said.

"The idea with the new beam is to have protons that are generated in packets lasting one or two nanoseconds with a gap between each packet of 500 nanoseconds," he said.

"We will be able to measure the neutrinos one by one, but to do this we need a beam that is a hundred times less intense than the previous one."

It looks like the CERN group is already trying to reexamine the issue, in a more restricted experiment. The PhysOrg article was very brief. The above quote, is just over a third of the linked comment.

I am not a particle physicist, but it seems that if they are able to detect neutrinos from a two nanosecond burst, they may be able to experimentally prove or disprove, systematic errors in the original data.

 

[ZapperZ]

My understanding is that the Physics Forum mission is pedagogical and oriented to students.
How can it be a pedagogical aim to ask people to be patient and stop thinking, switch in the wait and see mode?

The OPERA paper contains a lot of issues that can be discussed usefully by undergraduated students, like the likelyhood analysis. I have no doubt that many young students could be helpful and make very interresting analysis of such topics. I am sure that the OPERA team could learn from them. I am sure that some people posting here here could write some enlightening paper. Such paper could solve an issue in the OPERA paper or could illustrate some concept on the basis of the OPERA experiment.

Restraining the discussion to a unique melting pot thread was an anti-pedagogical decision.
It discouraged in-depth discussions.

No doubt Feynman would have suggested the exact opposite approach.
Feynmann would have encouraged discussion and critical thinking, without restrain.
But I must concede that he would have already debunked the OPERA claim by now.

There is a difference between making an analysis of something versus making outright guesswork.

There's a lot about the OPERA paper that is not very clear. I had stated earlier about the fact that in cases like this, typically a longer paper will come out with all the gory details that isn't contained in the first paper. So even if you want to make an analysis, what you are given is sufficiently vague that you end up making way too many guesses to make any kind of rational, accurate analysis.

As part of learning, a student, and even the general public, should also examine the nature of the source! This is also something that we try to impart within this forum. You simply should not read something and be blind to not only the quality of your source, but also to what extent you can safely extrapolate what you read. I'm not saying one should discuss this result. I'm just saying that, at some point, one HAS to face the fact that not a lot can be gathered out of what has already been given. We do not have the OPERA people giving further details, and those who know aren't talking... yet!

I've sat in a seminar here done by the MINOS people. Considering that they would be the ones who would be very familiar with what was done and the nature of the result without being part of the OPERA collaboration, even THEY are very hesitant to make any kind of judgement on the result because they just had way too many questions about the paper that was very unclear. Even when pressed, they can only make qualified guesses on where they think the uncertainty would have crept in. So if these experts are that hesitant, and would rather take a wait-and-see attitude to see how this all will work out (and the fact that they will make their own tests), who are we to sit here and think we can make any more meaningful discussion based on something that has yet to be deem to exist?

The OPERA paper has a lot of details that are missing. Period! This is even before one considers that the results must be verified independently. To me, the most important lesson that a student or the public can learn out of this is that just because someone or some group says something, it doesn't make it so, no matter how "prestigious" that person or group is. That is the process that all of science has to go through, unlike what is done in politics, the media, etc... In some cases, it is perfectly fine to NOT draw up any kind of conclusion or make a decision one way or the other until one gets more information. Making a decision based on incomplete, or even faulty information can be as bad, if not worse, than not making a decision at all.

Zz.

 

[tmfs10]

The OPERA paper has a lot of details that are missing.
Zz.

Giulia Brunetti's PhD thesis http://operaweb.lngs.infn.it:2080/Opera/ptb/theses/theses/Brunetti-Giulia_phdthesis.pdf" has more details.

 

[lalbatros]

I understand you well ZapperZ.

However, this also amounts to saying that the OPERA paper was premature.
Personally, I prefer to see the available information, which is sufficient to start some analysis, and certainly for students to start some investigation or some study.
This is specially appropriate in a forum which is not a peer-reviewed journal and where reputation has less importance.

In addition, it may well be that in this situation, the specialists are inhibited by their fear to be wrong, while students and young people do not fear mistake so much.
Well, maybe I am wrong on this point.
After all, the OPERA team did present this result.
A huge majority of people think is not more than a mistake.
The new experiment launched by OPERA has the taste of a confession.

 

[Buckleymanor]

according to FTL data from CERN, SN1987A neutrinos
should arrive about 4 years before light

One of the main points seems to be that neutrinos are produced from higher or lower energy values.
SN1987A neutrinos have a lower energy than those used at the OPERA experiment.
The input of energy having an end result in there speed.
Could the OPERA team produce neutrinos with a lower energy input than those allready used record the TOF to see if they are slower and rule out some timeing anomalies.