Talk:2011 OPERA faster-than-light neutrino anomaly/Archive 1

Archive 1

Archive 2

Archive 3

Archive 4

The Van Elburg objection

Ronald Van Elburg is an AI researcher, not a physicist. His paper will never be peer-reviewed because he has not submitted it for peer review. I think the reference to his work is not credible. It is hard to believe 160 physicists working for several years, and the entire global community scrutinizing the results for a month, would have missed anything basic.

The core of the objection is that because the GPS satellite is moving with velocity 'v', the detector moves toward the source at that speed, and this shortens the elapsed travel time for the neutrino in the satellite's reference frame.

If the GPS satellite is moving at speed 'v', the detector is moving toward it at speed 'v', the source is moving away from it at speed 'v', and the neutrino itself is moving apparently slower by 'v'. The net flight time measured for the neutrino in this frame will be the same as in the LNGS/CERN frame (except for relativistic corrections). The neutrino leaving CERN (e1) and being detected at LNGS (e2) are two time-events. The interval between these events is the same in all time-frames, when special-relativistic considerations do not apply. A special relativistic correction at a speed of 4Km/s (about 1.3/100000ths of light) cannot possibly amount to 20 ppm. Ajoykt (talk)

Please note - the criterion is verifiability, not truth. Though I agree this argument doesn't seem plausible, we are not editors of a scientific journal. MIT's Technology Review is certainly a reputable source on this subject, and if someone wants add the reference to this paper back in, there should not be anything wrong with that.--90.184.154.70 (talk) 21:23, 15 October 2011 (UTC)

MIT Technology Review is not a reliable source for fundamental physics papers. It is a newsletter, independent of the University, on latest technology issues. The problem is physicists may think the mistake in the objection so obvious they may not bother to refute it on arxiv or elsewhere. The "Contaldi" clock synchronization preprint stayed around for almost a month and was quoted and re-quoted in many articles before somebody published a scathing rebuttal (arXiv:1110.2909); the objection was known to be wrong well before that. Ajoykt (talk) 22:41, 15 October 2011 (UTC)

Newsletters counts as a reliable source. We have no business vetting papers based on whether they are "obvious mistakes" or not.--90.184.154.70 (talk) 01:47, 16 October 2011 (UTC)

That link has more on what kind of newsletter-blogs count as reliable. Here we have somebody who calls himself KentuckyFC (in other words, unsigned), reposting material from arxive.org. We don't know whether the blog is vetted by the editors. It is a news outlet reposting what is interesting on arxiv.org. How is this a credible source? We are not talking of a 'paper' here; we are talking of a posting to arxiv.org. Anybody can post there, as long as names are not hidden. Ajoykt (talk) 02:37, 16 October 2011 (UTC)

KFC used to post on this blog in his native dialect. Count Iblis (talk) 03:27, 16 October 2011 (UTC)

I've dropped the KFC source and added two other ones.[1]--90.184.154.70 (talk) 03:33, 16 October 2011 (UTC)

Agree with Ajoykt. Regardless of quality, plausibility and the fact that it's verifiable. This is not a paper, it has not been reviewed and therefore it has the same status as numerous other blogs/postings on this subject. Those aren't mentioned, why should this one be? Putting this in the article distracts visitors from the simple fact that should be communicated in the article: There is no scientific explanation for the observed anomaly. --Unsigned comment by User:131.155.81.156

That is certainly not a fact! In all likelihood there is a scientific explanation, likely a mundane one too. That you assume there is none just goes to show why it is so important that we stick to WP:V instead of our own conceptions of The Truth.--90.184.154.70 (talk) 02:57, 18 October 2011 (UTC)

The sentence "There is no scientific explanation . . . " can be interpreted two different ways: "There is, as of this moment, no scientific explanation . . ." or "There is, and will never be, a scientific explanation . . ." You are reading the meaning in the second sense. I agree with the meaning in the first sense. --Ajoykt (talk) 04:48, 19 October 2011 (UTC)

If you can verify that "There is, as of this moment, no scientific explanation..." then the way to "communicate" this is to write it explicitly in the article along with proper citations - not to withhold other verifiable information as to leave a particular impression that may or may not be true.--83.89.0.118 (talk) 05:02, 19 October 2011 (UTC)

Of course the intended meaning was "no scientific explanation as of this moment". Obviously there is no way to back that statement with a citation. The whole point is that there is no reliable source to cite from. At this moment. But let's return to the subject. The work by Van Elburg cannot be considered a reliable source because it has not been peer-reviewed, let alone been published in a scientific journal. If it's anything, it is original research. — Preceding unsigned comment added by 131.155.81.156 (talk) 13:09, 19 October 2011 (UTC)

Plenty of reliable sources has covered the van Elburg paper, including Bad Astronomy, New Scientist and Wired. Citing a reliable source is not original research.--83.89.0.118 (talk) 19:24, 19 October 2011 (UTC)

Bah, I had a similar idea, but I assumed that the synchronization is done in the rest frame of the satellite and that the Opera team had wrongly corrected for a 32 ns delay because of a software bug, a sign error will make this 64 ns instead of zero. I didn't bother to write up a preprint and submit to arXiv :( . Count Iblis (talk) 02:26, 16 October 2011 (UTC)

As a side note, GPS receivers are prepared to deal with the relativistic effects caused by Earth's gravitational field, and also by the high speed of the GPS satellites. This is detailed in the GPS ICD, and in basically all GPS textbooks (the most common, IMO, being Misra & Enge). I've seen mention of this objection before, but no mention by Van Elburg of why this correction, usually included in receiver firmware, was not already applied. I made a note of it in the section of the article, since I think most readers would be unaware of this. siafu (talk) 21:04, 17 October 2011 (UTC)

Added a ref to Misra & Enge for the skeptical. siafu (talk) 21:17, 17 October 2011 (UTC)

The following is copied from my talkpage:

Corrections in GPS

-- Conversation began on User talk:Siafu --

Hi. You added this sentence to the OPERA neutrino anomaly article. What is the source for this claim? Thanks.--90.184.154.70 (talk) 21:14, 17 October 2011 (UTC)

I added a ref to the article to Misra & Enge, but per my talk this is in the GPS ICD, and therefore in every GPS textbook ever. siafu (talk) 21:16, 17 October 2011 (UTC)

Boldly removed the sententence from the article again. No matter how many textbooks discuss generic relativistic corrections in GPS, it is not a comment on van Elburg's particular calculation; reading it as such is a WP:SYN.--90.184.154.70 (talk) 21:36, 17 October 2011 (UTC)

I don't see how it could be seen that way. Van Elburg is claiming that this particular correction wasn't made, and that that mistake accounts for the anomalous results. It just so happens that not making this correction would be extremely unusual-- that's not synthesis, it's verifiable fact, and I can provide plenty more references if you need them-- it is, in fact, essential to make this correction in order to obtain a position solution, since 3 ns ~ 1 meter in position error. I'm going to copy this whole conversation to the article talk page, also, since we shouldn't be picking over content here out of sight of the other editors. siafu (talk) 21:41, 17 October 2011 (UTC)

Your source says that GPS makes corrections for relativistic effects. Your source does not say that GPS makes corrections for the particular effect that van Elburg refers to. In may be that you know that the effects are identical - that's great. But you are not a reliable source.--90.184.154.70 (talk) 21:50, 17 October 2011 (UTC)

My source says that GPS makes corrections for the effects cited by van Elburg; other sources are more detailed, but even the same terminology is used. Moreover, the way the article presents it now makes it seem like this is a totally unexpected error source, when the actual fact is that GPS makes relativistic corrections, and the method for doing so has been specified since the late 70's. It's not synth to point this out, since no conclusion is being drawn, all that is being made clear is that not making the corrections would be the exception, and since this is a fact that the lay reader would be almost certainly ignorant of, it should be pointed out. siafu (talk) 21:54, 17 October 2011 (UTC)

The original formulation did draw the conclusion that the effect cited by van Elburg has been corrected for. There is no verifiable evidence for this. I've tried rephrasing it to reflect that it is a general statement about GPS, not about van Elburg's particular calculation.--90.184.154.70 (talk) 22:04, 17 October 2011 (UTC)

The formulation now concludes the effect cited by van Elburg is corrected for in those GPS systems which follow a common, credible textbook on GPS. We don't know whether the OPERA experimenters, 160 physicists total, used a GPS receiver device which did follow the text. But if you want to add that rider to the sentence, the onus is on you to show why such a statement would be credible. Ajoykt (talk) 22:11, 17 October 2011 (UTC)

Wrong. The WP:BURDEN of evidence lies with the editor who adds or restores material, in this case User:Siafu. I tried removing the sentence, but it was readded multiple times.--90.184.154.70 (talk) 22:18, 17 October 2011 (UTC)

I think the citation proves the formulation that exists now: that the effect mentioned by van Elburg, relativistic effects from satellite velocity, is corrected for in those GPS systems which follow a common, credible textbook on GPS. I see no reason to remove that part. Ajoykt (talk) 22:27, 17 October 2011 (UTC)

Van Elburg's objection is about the relativistic correction from GPS satellite velocity; he painstakingly even derives that velocity. I cannot confirm Siafu's claim the cited book has what he says it does, but assuming that is so, yes, the citation is directly relevant to the specific relativistic concern Van Elburg expressed. I also think even otherwise the SR correction doesn't amount to 20ppm at speeds of 4km/sec, but that is a different topic. Ajoykt (talk) 21:59, 17 October 2011 (UTC)

WP:SYN has a good example illustrating why exactly this type of statement is synthesis:

"The UN's stated objective is to maintain international peace and security, but since its creation there have been 160 wars throughout the world."

Both parts of the sentence may be reliably sourced, but here they have been combined to imply that the UN has failed to maintain world peace. If no reliable source has combined the material in this way, it is original research.

Since no reliable source has combined van Elburg's calculation with your generic statement about GPS to imply that the effect van Elburg mentions has been corrected for, that implication is WP:SYN.--90.184.154.70 (talk) 22:40, 17 October 2011 (UTC)

The OPERA scientists combined the exact two things you talk of. It is there in the "Boo! Hiss! . . ." reference. Ajoykt (talk) 23:54, 17 October 2011 (UTC)

No, they have never referred to van Elburg or his calculation, so that can't be part of their combination. For User:Siafu's sentence to be valid, a reliable source literally has to go out and say something to the effect of: "I have read van Elburg's paper and the effect he mentions has been corrected for in OPERA's superluminal result." Anything less is WP:OR.--90.184.154.70 (talk) 00:07, 18 October 2011 (UTC)

The exact quote: "It's too early to say whether that's the final verdict on this story—the CERN scientists did claim to have accounted for such factors in their report." In that article "such factors" refers to the Van Elberg objection. That author thinks the combination holds, assuming OPERA people are accurate in what they say. Ajoykt (talk) 00:31, 18 October 2011 (UTC)

The OPERA scientists claimed to have accounted for factors that may or may not include the factor mentioned by van Elberg. That is undisputed. But no reliable source has claimed to have read van Elberg's paper and found that the factor he mentions is already corrected for by GPS.--90.184.154.70 (talk) 00:35, 18 October 2011 (UTC)

The sentence I included specifically avoided drawing any conclusions about van Elburg's supposition, it merely contextualized it by pointing out that it would be unusual for that to be the case. Because it would be. The UN quote is a bit of a canard here, since it shows an example of using a misleading statistic, specifically one that's incomplete, since the number of wars fought since the UN can only be useful when compared to some hypothetical number of wars that would have been fought without the UN. This is not like that at all, since it is not likely that the average reader (or, apparently, average PhD physicist) has any idea how GPS works it is not inappropriate to include a comment on that. WP:SYNTH is about synthesizing available information into a new conclusion. There is no conclusion here, just a note that normally these effects are already accounted for. For all I know, the operators of the receivers used for the OPERA experiment screwed it up and van Elburg is right... but this would be very unusual indeed. siafu (talk) 01:38, 18 October 2011 (UTC)

Also, I want to make it clear that this is not a radical revelation. The GPS ICD (i.e., the primary source) is available here, and the description of how to make the receiver-side relativistic correction is on page 106, and there is a brief discussion on page 25 of the SV-side relativistic correction. This document details the specifications of the GPS system; it is essentially the reference guide for GPS receiver designers. Every GPS textbook (i.e., the secondary source) touches on this topic, as it is an integral and essential part of how GPS works. Omitting it would be like omitting the fact that cars require working oxygen sensors to function properly-- it's not a synthesized new thought to point out that it is normally done. siafu (talk) 01:45, 18 October 2011 (UTC)

I doesn't really matter how many different ways you formulate it. No reliable source states that van Elburg's objection is accounted for by GPS. Period. A statement with no reliable sources has no place in Wikipedia, whether expressed or implied. Period.--90.184.154.70 (talk) 02:21, 18 October 2011 (UTC)

Which is fine, since we aren't actually saying that either. siafu (talk) 03:22, 18 October 2011 (UTC)

Okay, spelling it out: No reliable source states that "[the] effects [mentioned by van Elburg] are corrected for in normal GPS receiver operation". Period. A statement with no reliable sources has no place in Wikipedia, whether express or implied.--90.184.154.70 (talk) 03:35, 18 October 2011 (UTC)

The statement inserted is reliably sourced, and does not draw the conclusion you are so very concerned about; reliable sources do indicate that relativistic effects (read: all relativistic effects; there are not some mysterious other effects being invoked here) are corrected for in normal GPS operation. Reflexive identity does not require a source. siafu (talk) 13:50, 18 October 2011 (UTC)?

"Reflexive identity does not require a source." Says who?? This is exactly the problem: you infer an identity. That's great, much progress in science is made by inferring identities. However, that is not what Wikipedia is for! If the identity your statement relies on has no explicit, reliable source, then it doesn't belong in Wikipedia. This is not about me being 'very concerned' about anything, this is simply the rules. --The Artist Formerly Known as 90.184.154.70 (talk) 00:31, 19 October 2011 (UTC)

I'm familiar with policy. I am not familiar with this idea that we cannot recognize a noun in two different places as being the same thing, especially when it is used the same way in both contexts. I add again: all relativistic effects are accounted for in normal GPS operation. This is an important point that is well-known to any engineer or scientist familiar with GPS, and not a radical new idea (it's in the ICD!). It would be dishonest to present the case as if it is somehow likely that the OPERA team "failed" to take these accounts into effect, as failing to do something implies an extra step that was just not taken. It's just not possible to just fail to do that-- the GPS receiver would have to be malfunctioning in a dramatic fashion (64 ns = 19.2 m of position error!) and this would have to go completely unnoticed. Maybe it happened, I'm not concluding that it didn't and certainly not using wikipedia's voice to do so, but the context is both important and relevant. The argument these are not the same "relativistic effects", frankly, is simply specious; if we were talking about cars, it would be WP:SYNTH to connect "cars" with "cars", or to contextualize a statement about cars to point out that they usually run on the ground, that they usually have tires, or that they usually require fuel to run. siafu (talk) 01:22, 19 October 2011 (UTC)

"... if we were talking about cars, it would be WP:SYNTH to connect 'cars' with 'cars', or to contextualize a statement about cars to point out that they usually run on the ground, that they usually have tires, or that they usually require fuel to run."   You. Are. Right. Each of those examples require sources just as much as the present one. It is not our business to "connect" anything. Wikipedia is not a directory of information that you and I consider to be "obvious". It is a directory of verifiable information. If that is "specious", then Wikipedia is "specious".--83.89.0.118 (talk) 02:21, 19 October 2011 (UTC)

But you are claiming more - you are saying to write the above statement about cars, one needs to find a source that has all three statements in one place; it is not enough to just cite separate sources for each part. Why stop that logic at sentences? Why not apply it to paras? It would then not be enough to cite sources for each sentence in a para. To avoid problems with implications, one would have to find a source that has all the sentences of the para. We can carry the argument further to sections (should be present in entirety in some citation), articles, and the whole of Wikipedia. The question of "implications" is subjective. There isn't much point just the three of us arguing this, since it looks our stands on the subjective part are clear. If somebody else weights in, we can revisit this. --Ajoykt (talk) 05:18, 19 October 2011 (UTC)

I am not claiming anything, I am citing the rules: "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources." What is implied is a matter of WP:CONSENSUS.--83.89.0.118 (talk) 05:54, 19 October 2011 (UTC)

Relativistic effects interfering with synchronization of the two distant clocks can be an issue, if the OPERA team made a mistake, or if the GPS system for clock synchronization contains a bug. Both are viable possiblilities; note that the GPS system for clock synchronization has never been tested before to this accuracy. On of the very first reactions to the OPERA results from a rival team was in fact that GPS is not reliable enough for accurate clock synchronization over large distances.

The distance measuements using GPS satellites is well established, that's not an issue. But now, you need to also know the position of one of the GPS satellites in, say, the LNGS rest frame. The signal of that GPS satellite is detected by both the LNGS and CERN clocks as a synchronization signal. Clearly, if you know where that satellite was at the moment the signal was sent according to the LNGS clock, you know when that same signal arrives at the CERN clock, according to the LNGS clock. While there is an automatic procedure for this within the GPS system, it has never been tested to within 100 ns accuracy over large distances, so one cannot exclude any bugs being in this system. Count Iblis (talk) 16:22, 19 October 2011 (UTC)

While I agree with most of your comments, it is not correct to say that the GPS timing system has not been tested to 100 ns accuracy. As I mentioned before, 3 ns is ~ 1 meter of position error, since GPS is based on lightspeed travel time delay. GPS receivers are commonly used by the geophysics, seismology, and surveying communities to accuracies down to the level of a few mm, and in order to do this very fine clock synchronization is required. Even with double differencing, an advanced technique used to eliminate clock error, a stable frequency standard is required on the satellites (they were each launched with 4 atomic clocks, some cesium, some rhubidium). I don't want to extend the claims in the article to say that van Elburg is simply wrong because of this, but it's absolutely true that he's making an extraordinary claim. siafu (talk) 18:57, 19 October 2011 (UTC)

arXiv categories

Van Elburg's paper is in the "general physics" category in arXiv. This is where the moderators dump things which are of "no plausible interest" anywhere else, including stuff that is wrong but is formatted correctly and is from somebody with some physics background. Note the main articles on OPERA are in the "High Energy Physics" category.--Ajoykt (talk) 19:38, 18 October 2011 (UTC)

Yes, but then quite a few of the HEP-PH papers are also wrong, even some writtten by big experts. Take e.g. this paper. I was planning to this paper for a homework assignment asking students to explain where they went wrong, but unfortunately this spoilsport paper appeared  :( . Count Iblis (talk) 20:22, 18 October 2011 (UTC)

This section on an unverified claim should be deleted as it is based upon a complete misunderstanding of the experimental set up. In other words the claim is based upon a failure to correctly understand the experiment and how it was performed. Upon reading the paper it is clear that it is simply nonsense and doesn't deserve to be cited on Wikipedia. --72.64.36.237 (talk) 15:33, 19 October 2011 (UTC)

If you can find more than one WP:RELIABLE source to back that up, I am not opposed to deleting the section. But until then, it is not our place to judge.--83.89.0.118 (talk) 19:28, 19 October 2011 (UTC)

It is our place to judge. We have to judge whether to include the claim or not. I agree with 72.64.36.237, the claim looks pretty bogus to me and arXiv is not necessarily a reliable source. Martin Hogbin (talk) 22:42, 19 October 2011 (UTC)

No, it isn't. We are not editors of a scientific journal. The claim is verifiable and it is notable and these are the only criteria for inclusion in Wikipedia. Whether or not you agree that it looks bogus is irrelevant.--83.89.0.118 (talk) 22:52, 19 October 2011 (UTC)

Deletion

Given that OPERA has now responded, I am inclined to remove the section until/unless van Elburg comes up with a scathing rebuttal. --83.89.0.118 (talk) 22:40, 19 October 2011 (UTC)

Done.--83.89.0.118 (talk) 22:42, 19 October 2011 (UTC)

I agree with removing the whole thing. The guy has had plenty of time to check on the GPS stuff - he should have had an answer at the ready. --Ajoykt (talk) 22:49, 19 October 2011 (UTC)

The Van Elburg objection

This discussion has been archived pending any further developments.--83.89.0.118 (talk) 22:42, 19 October 2011 (UTC)

I did just re-add a small section on this, with the follow rationale:

• The fact that van Elburg made an objection is
  • verifiable (I sourced the section heavily).
  • notable (mainly based on substantial media coverage).
• Due to media coverage, this article has received increased traffic
• Without having anything about van Elburg, readers might wonder/try to re-add and not visit talk page.
• Readers additionally may have difficulty finding OPERA's response (I sure did when looking for a cite).
• By briefly mentioning the objection, and then stating and providing a source discrediting it, readers of the article are better informed about whole subject.
• Repeating for emphasis, I think that readers should be aware the theory has been discredited.
  • They might not see that without being able to find a source here linking to the statements that discredit.

Not trying to step on toes here, but I think briefly listing both sides is better than having nothing. Although the belief wasn't held for the same length of time, I think this would be analogous to removing any mention of a flat earth from the project. That was a bigger belief, so I bet that it gets a lot more coverage on the project than this does, but I think the blurb is defensible.Not planning on revisiting so please place a {{talkback}} on my page if you disagree. Thanks!--Policy Reformer(c) 09:47, 20 October 2011 (UTC)

I agree, although the theory was not that reliably sourced in the first place and it has now been definitively discredited it was a theory that received some attention, from the media and OPERA. For the reasons stated above I think we should have something on it.

Maybe we should have a section on proposed (and notable in some way) but discredited theories. Martin Hogbin (talk) 14:30, 20 October 2011 (UTC)

On that front, I can think of the Contaldi objection (GR-based objection on clock synchronization). It too got a lot of press, before being rebutted on arXiv: arXiv:1109.6160 (Contaldi) and arXiv:1110.2909 (rebuttal). I think the media is missing the point work done by committees, while slow and inefficient, usually has few obvious errors. --Ajoykt (talk) 14:49, 20 October 2011 (UTC)

Martin's suggestion is a good one. List the discredited theories - the media coverage is notable.-- cheers, Michael C. Price ^talk 14:59, 23 October 2011 (UTC)

The van Elburg paper is a spurious argument and is not relevant to the discussion. His argument is that the procedure used to make the measurement is wrong because it uses GPS. His claim is an argument that the use of the GPS system to make the measurement is incorrect. That is to say he is saying that GPS doesn't do the measurements correctly. The answer to his objection is that the GPS system does the measurements correctly. The point of this experiment is that when the best measurement techniques are used, the result is that the TOF for neutrinos is less than calculated for light. It seems to me that this this argument is not going to be resolved in the way that this section tries to do so. The section should be deleted as the objection goes to the accuracy of GPS and that is beyond the scope of the article to resolve. — Preceding unsigned comment added by 71.251.177.155 (talk) 13:18, 24 October 2011 (UTC)

We are well aware of this. As the posters above note, not everyone may be as enlightened as you.--83.89.0.118 (talk) 17:19, 24 October 2011 (UTC)

Article title

I'm not sure this is the best possible title for this article. I think it should either be sufficiently descriptive (e.g. OPERA superluminal neutrino anomaly or OPERA neutrino speed anomaly or OPERA neutrino time-of-flight anomaly) or as simple as possible (OPERA anomaly). (Personally, I'd prefer the former.) The OPERA experiment being about neutrinos, I don't think the current title is any more informative than just “OPERA anomaly”, and needlessly longer. What do you think? ― A. di M.​_plé​^dréachtaí 17:49, 11 October 2011 (UTC)

It sounds reasonable to me to move to OPERA anomaly. After the move it would probably useful to then redirect the current name to that page. Polyamorph (talk) 17:52, 11 October 2011 (UTC)

From a search result point of view, it is better to have the words "faster than light" somewhere in the title. That is what really interests people in the article, I presume. Ajoykt (talk) 03:57, 12 October 2011 (UTC)

We can have redirects that contains that string, if you like. Polyamorph (talk) 07:44, 12 October 2011 (UTC)

Yes. OPERA anomaly as the title with redirects for "faster than light" and from the current title sounds good. Ajoykt (talk) 15:41, 12 October 2011 (UTC)

The use of the word anomaly is invalid here. There is no anomaly. It is a properly performed experiment. The fact that the result doesn't fit the kind of result that the physics community wants is not relevant. The result is what it is. The proper title should reflect the actual result which was that the measurement showed that the measured neutrino TOF was less than the calculated TOF for light waves in a vacuum.72.64.36.237 (talk) 15:39, 19 October 2011 (UTC)

I think that you must not know what an anomaly is. Both the OPERA team itself, as well as secondary sources (including Cohen-Glashow, whose paper is the only one currently accepted for peer review), refers to it as an anomaly. Here, the word of a Nobel laurate has greater weight than yours or mine.--83.89.0.118 (talk) 20:12, 19 October 2011 (UTC)

The result is anomalous with what is predicted by theory. That doesn't mean the experiment is poorly performed (usually that is so, but credible sources say otherwise for this case). I don't think the term reflects badly on the OPERA team. Anyway this is a word they coined. --Ajoykt (talk) 21:25, 19 October 2011 (UTC)

The ICARUS negative on Cohen-Glashow emissions

Hmmm, ICARUS is claiming the OPERA results are wrong, without even bothering to measure the speed of neutrinos. They assume the C-G theory (though it has no experimental backing since no one has ever seen anything superluminal, other than OPERA), and then they declare neutrinos have got to go well below the speed OPERA reports. I guess we will now have a rash of media reports writing off the OPERA results. Ajoykt (talk) 03:29, 18 October 2011 (UTC)

To be fair, they only seem to be saying that their results "refute a superluminal interpretation of the OPERA result according to the Cohen and Glashow prediction". I.e. they only refute the OPERA result to the extent that C-G theory is right.--90.184.154.70 (talk) 04:17, 18 October 2011 (UTC)

They claim a smaller (stricter) value for 'delta' for the CNGS beam. Since delta is the excess velocity for neutrino over light, that is effectively claiming CNGS neutrinos go slower than light (or at least no faster than 40 parts in a billion). This part of their claim is not qualified by C&G's theory being right. They consistently use the word "must": "superluminal charge-less neutrinos must emit radiation," "these analogs to [C-G emission] must appear," They also use words like "hereby alleged anomaly" to refer to the OPERA result. Ajoykt (talk) 04:27, 18 October 2011 (UTC)

It is misleading to say that there is no experimental support for C-G theory. Thing is that there exists a successful Standard Model (extended to take into account finite neutrino masses and mixings). The interactions with particles and neutrinos are correctly described by this model, where precision tests are possible, the agreement with theory and experiment is phenomenal.

The prediction that superluminal neutrinos must emit Cherencov radiation is then almost proven. The Standard Model Lagrangian is, after all, what you are left with if you were to integrate out the unkown high energy physics. So whatever you postulate of what lies beyond the Standard Model, as far as low energy physics is concerned, you must end up with the known Standard Model, otherwise you would fail to predict what we already know to be true.

This then means that you can only add to the Lagrangian tiny renormalizable terms allowing for Lorenz invariance breaking. The amplitude for Cherenkov radiation production is then largely determined by the known terms in the Lagrangian, the tiny terms only allow the neutrinos to travel faster than light, but they don't contribute to the radiation to leading order. To prevent Cherenkov radiation from being emitted, you must somehow make the effect for the tiny terms very large, but that is impossible. You therefore would have to assume that these terms aren't tiny at all but then that would mess up the correct predictions of the Standard Model. So, you need to add more terms to make sure all the known predictions are correct and yet that neutrinos don't emit Cherenkov radiation.

But doing that would amount to postulating a brand new theory without any experimental evidence at all, other than the assumption that the OPERA results are correct. The C-G theory, in contrast, is what you would expect based on current knowledge. Count Iblis (talk) 15:48, 18 October 2011 (UTC)

Predictions are proven by experiments, not by the confidence levels of theorists. The C-G theory is unproven. After all, more than a century ago, that clocks should tick at the same rate in every frame was also "almost proven." The point is one can refute an experimental result only by pointing out what is concretely wrong with it, or by trying to, and failing to, replicate the results. The entire C-G exercise seems only to prove how more radical a change the OPERA results, if true, presage for existing physics. That fact cannot refute the experiment itself. I am not saying OPERA results are right, but the ICARUS result is no refutation, despite their wording. Wording we have to include in Wikipedia. Ajoykt (talk) 15:57, 18 October 2011 (UTC)

But note that Einstein developed relativity based on the, at that time, known experimental knowledge, no one had ever observed time dilation. His theory was well accepted long before rigorous and precise tests of relativity were available, simply because it offered a much better explanation for known facts.

So, I would say that if OPERA is proven correct, that would be similar to experimental tests of special relativity falsifying it against all expectations, and eventually proving correct the old fashioned aether theory. Count Iblis (talk) 16:31, 18 October 2011 (UTC)

I completely agree with Ajoykt. What you, I and the ICARUS experiment "would say" is completely irrelevant. Until OPERA produces a slower-than-light result, nothing has been refuted.--83.89.0.118 (talk) 01:13, 19 October 2011 (UTC)

B.t.w., you can, of course, criticise the ICARUS measurement of neutrino speed for making asummptions that are "in small print" about assuming a that a certain aspect of the theory is correct, but in practice all measurements depend on assuming some theory. In case of null results that is used to rule out new physics, this is obviously potentially a big problem. Compare e.g. the experimental results that are used to set limits to the mass and charge of the photon. They obviously depend on assuming a theory that describes the properties of massive or charged photons in a sensitive way, as pointed out here w.r.t. massive photons and for the charge of the photon, there isn't a consistent theory at all.

What makes the case of superluminal neutrinos different is that the Cherenkov radiation depends primarily on the known terms in the Standard Model Lagrangian, so it's to a high degree model independent. Count Iblis (talk) 17:24, 18 October 2011 (UTC)

If OPERA is proven right, it would be just like Michelson-Morley. A new theory which has the old theory as its limit or special case would be needed. C&G haven't proven such theories are impossible - they have ruled out only a specific set. My core objection is to the claim these things "refute" the OPERA experiment. They don't. I don't know why the papers use that word. Theories don't refute experiments. ICARUS, though an experimental result, doesn't refute OPERA. Yes, all measurements depend on some theory. But if we already have a measurement in place, we can't refute that measurement by measuring something else hypothesized to correlate. A hypothesis very elegant, plausible and theoretically strong, but since it gets into faster-than-light territory, not proven. Ajoykt (talk) 18:07, 18 October 2011 (UTC)

As expected, blogoland has lit up with fireworks celebrating the demise of the OPERA superluminal result. Our article needs to stay more neutral than that. The anomaly is still an anomaly, and looks likely to stay that way till late next year, when new data from TK-2 and MINOS would be out; if there is a systematic error, it is likely so subtle we may not catch it. The "reinterpretation" of existing data from MINOS, expected in 4 months, probably would just reiterate their previous conclusion - not enough accuracy to be sure. Ajoykt (talk) —Preceding undated comment added 18:23, 18 October 2011 (UTC).

Note on sources

The Cohen/Glashow paper will be published in Physical Review Letters. See newscientis, and PRL. --D.H (talk) 15:19, 19 October 2011 (UTC)

Good catch. I was a bit incredulous that this would get published so soon.--Louiedog (talk) 15:28, 19 October 2011 (UTC)

Me too. An average review time in PRL is few months (and people with much more obvious subjects often wait much longer, up to a year including resubmissions). But for this paper it took them less than a month to accept. To me, the paper's arguments are not that obvious to be published in such a rush as there already exist other experts opinions. User1344 (talk) 06:43, 25 October 2011 (UTC)

I could be wrong, but in the "Energy decay through pair bremsstrahlung" section, shouldn't the stated 40 ppb value be 0.4 ppb? 91.155.92.202 (talk) 19:14, 19 October 2011 (UTC)

$\delta =4\times 10^{-8}=0.00000004=40/1000000000$. --D.H (talk) 20:13, 19 October 2011 (UTC)

BBC documentary, possible source?

If we have any brits among us, BBC has supposedly just aired a one-hour documentary about the OPERA anomaly: http://www.bbc.co.uk/programmes/b016bys2 --83.89.0.118 (talk) 02:27, 20 October 2011 (UTC)

MINOS

First of all, Ellis et al. arXiv:0805.0253 is used because of its citation of the MINOS result (not of the other new things discussed in that paper) therefore in this respect it is a valid secondary source (besides, Physical Review D is of course much more reliable than Nature news..)
Second: MINOS arXiv:0706.0437 gave results using two confidence levels (on page 5): The first with 1.8σ (=68% confidence level) gave the relation between neutrino and light speed of:

$(v-c)/c=5.1\pm 2.9\times 10^{-5}$

In numbers: Because $(v-c)/c=\delta$ we have the neutrino velocity $v=\delta c+c=1.000051(\pm 29)c$, thus a range between 1.000023c to 1.000080c faster than light. In the same page 5, MINOS gave also the result at the more significant 99% confidence level (this range is also cited in Ellis et al. page 1):

$-2.4\times 10^{-5}<(v-c)/c<12.6\times 10^{-5}$

So, this includes the complete range, including the uncertainties partly by the detector's exact distance etc. (the latter source of uncertainty is also mentioned in the nature article). In numbers: We have a range between v=0.999976c (slower than light) to 1.000126c (faster than light). So we have two possibilities: We write the range in the article, or the numbers (now I've written the numbers). PS: An interesting side effect (of course not mentioned in the article) is, that even the OPERA superluminal measurement lies within the MINOS-range. --D.H (talk) 20:28, 22 October 2011 (UTC)

Yes, OPERA matches MINOS in many ways. But MINOS also used a statistical calculation. The OPERA replication in the next two weeks or so should change the status quo considerably. If they repeat their results, even though there could still be distance or time measurement errors, it will still be a shock, Cohen-Glashow notwithstanding. If not, the story will be over. --Ajoykt (talk) 21:51, 22 October 2011 (UTC)

The Net Advance of Physics

It is a credible source - at least ScienceMag (AAAS's magazine - along with Nature News the most credible on science news) says so: http://www.insidecancer.org/downloads/insideCancer_Science.pdf --Ajoykt (talk) 21:56, 22 October 2011 (UTC)

Look again at the neutrino site, where we still must read:

PLEASE NOTE: The Net Advance of Physics does not take a position on the existence and/or explanation of this phenomenon. Also, the editor unfortunately lacks the time and expertise to read every paper in detail and with full comprehension, so some may be poorly classified. Suggestions for improvements are welcome. SUPERLUMINAL NEUTRINOS

The neutrino arrangement is surely useful, though clearly not reliable under those circumstances.... I also cannot see, what exactly could be taken from it for our article. It's an topic-ordered list, and most of the papers are not commented, or only described by one sentence. Some could use it as a backdoor for the inclusion of dozens of non-peer reviewed preprints in this article.... --D.H (talk) 22:22, 22 October 2011 (UTC)

Where are you going to find a secondary source where the editor has read the papers in detail and with full comprehension? They (there are two editors) have that sentence in there, comparing their work to a formal (MIT) publication. Yes there are only a few comments, but where comments do exist, that website is useful as a secondary source. The summaries are reliable. As of now, I am ok with the article as is. But in a few weeks, OPERA, I believe (my opinion based on knowledge of statistics, not a 'Wikipedia'ble fact) will reproduce their result. At that time, we are going to have another deluge on our hands. --Ajoykt (talk) 23:15, 22 October 2011 (UTC)

Secondary sources

Regarding arXiv-preprints in general: Including non-peer reviewed primary sources, without peer reviewed primary or secondary sources, is problematic in the light of WP:Notable, WP:Secondary, WP:Undue, WP:OR and WP:SYNTH. Now I'v looked through some (science, nature, newscientist etc.) secondary sources given in the article, and searched for references to the arxiv-preprints currently included in our article. Here is the result:

• Cited: Coleman&Glashow: arXiv:1109.6562, Icarus: arXiv:1110.3763, Elburg: arXiv:1110.2685.
• Not cited: Ehrlich: arXiv:1110.0736, Broda: arXiv:1110.0644, Alicki: arXiv:1109.5727, Winter: arXiv:1110.0424

That doesn't mean that I agree or disagree with those articles, but without secondary source there is no sufficient justification to include the non-cited papers here.--D.H (talk) 19:44, 19 October 2011 (UTC)

Though I think each of those papers add valuable information, I have to agree with you on a principal level - the references to them should probably be removed for now.--83.89.0.118 (talk) 20:02, 19 October 2011 (UTC)

Done. --D.H (talk) 20:23, 19 October 2011 (UTC)

Mark what needs secondary sources. Don't arbitrarily remove stuff. It takes time to dig up the secondary sources. They do exist. Ajoykt (talk) 20:31, 19 October 2011 (UTC)

Thanks for your effort, Ajoykt. It seems that only the Ehrlich-Ref is still missing. PS: I now see, that "83.89.0.118" marked them. --D.H (talk) 20:49, 19 October 2011 (UTC)

The secondary source for Erhlich seems a bit dubious. It looks like a personal homepage, only hosted at MIT.--83.89.0.118 (talk) 20:59, 19 October 2011 (UTC)

Well, the page does have the MIT "Laboratory of Nuclear Science" logo on it. As for credibility, all it does is organize the OPERA articles and provide brief comments. The overall website has just too much information to be one person's work, let alone a harried MIT grad student's. Note they claim to have been established in 1995, and have extensive archives dating back to 2006. See http://web.mit.edu/redingtn/www/netadv/ --Ajoykt (talk) 05:10, 20 October 2011 (UTC)

The site The Net Advance of Physics cannot be used as a secondary source, since it is mostly a resource list. The author even wrights:

PLEASE NOTE: The Net Advance of Physics does not take a position on the existence and/or explanation of this phenomenon. Also, the editor unfortunately lacks the time and expertise to read every paper in detail and with full comprehension, so some may be poorly classified. (The Net Advance of Physics).

For example, the Ehrlich paper is listed under specific topics, yet without mentioning on whether Ehrlich's position on the topic is affirmative of critical... (using the term "the editor" even suggests that it is indeed only one person who wrote that). However, I still think it's valuable, so will move the link into the external link section, though I will remove the sources (Ehrlich, Alicki, Broda, Besida) from the article text, until one finds better sources. --D.H (talk) 08:32, 21 October 2011 (UTC)

To quote Wikipedia policy: "Deciding whether primary, secondary or tertiary sources are appropriate on any given occasion is a matter of good editorial judgment and common sense, and should be discussed on article talk pages." This isn't as final as "remove everything with only primary sources." In this case, the primary sources are credible authors, not having any direct stake in what they are writing about. These are valid scientific papers, though not peer-reviewed; not crank stuff. As for the MIT website, I am not sure what you mean "written by one person." Almost all secondary sources are. That is not a personal website, however. — Preceding unsigned comment added by Ajoykt (talk • contribs) 14:51, 21 October 2011 (UTC)

Well, it would be problematic when there were a lack of secondary sources: But that's not at all the case, since we have plenty of them: Nature (journal), The Register, Science (journal), New Scientist, Physics World, PhysOrg, Scientific American, LiveScience, all of them being cited in our article. And of course, it's only a matter of time when all valuable papers are cited in those papers, so that we can cite them as well. So it's not necessary to be impatient, and to start to include new research papers only because we think they are valuable or not. Regarding the MIT-website: I simply said that the editor himself added a note, that he didn't even read all papers. And the papers are simply listed and ordered, mostly without comment. As a list it is surely valuable, but since the paper's content are not being discussed, it cannot be used for describing the content of individual papers. --D.H (talk) 15:23, 21 October 2011 (UTC)

Except for Nature News (the journal typically is full of primary sources), none of the rest are quite credible when it comes to either accuracy or coverage bias. And where the MIT-website does offer comments, it should be allowed as a secondary source (granted that is not true for the Ehrlich paper, but it is true for the rebuttal to Contaldi). Note we are not taking a position on whether the papers are right; we are just stating their positions. — Preceding unsigned comment added by Ajoykt (talk • contribs) 19:01, 21 October 2011 (UTC)

Also the "beam instability" section didn't reflect the description of the The Register article, so I rewrote it and moved it to the "other explanations" section. And based on the recent nature article, I've included the "extra-dimensions" model. --D.H (talk) 10:23, 21 October 2011 (UTC)

Amelino/Smolin paper

We have an editor who consistently[2] [3] [4] [5] [6] [7] is adding a statement based solely on a primary research paper posted on arXiv without peer-review. Additionally, the user refuses to discuss the issue here, on the relevant talk page. Should this statement be retained or removed from the article?--83.89.0.118 (talk) 15:40, 21 October 2011 (UTC)

I recommend the editor who included the Amelino/Smolin paper (arXiv:1110.0521), to look at WP:Secondary. Although it is interesting and written by well-known authors, their paper is not discussed and assessed anywhere by a reputable secondary source, so we have to wait until such sources exist. If none is provided, the reference to that paper should be deleted,. --D.H (talk) 11:46, 21 October 2011 (UTC)

What is a secondary source? From WP:Secondary:

Whether a source is primary or secondary depends on context. A book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences. A book review too can be an opinion, summary or scholarly review. Policy: Wikipedia articles usually rely on material from secondary sources. Articles may make analytic or evaluative claims only if these have been published by a reliable secondary source.

So what does this mean in our case? A few examples:

• We have OPERA arXiv:1109.4897, this is a primary source - though we can mention it in the article because it is cited by secondary sources like "Scientific American" etc...
• Now we have the Coleman/Glashow paper. It mentions the OPERA paper, so one could think that it is a secondary source. But this is not the reason why we cite it in our article: It is about their application of the vacuum Cherenkov effect on Neutrinos - and this is original research, making it a primary source. Why can we include it? Well, because it is mentioned by other reputable secondary sources, such as "nature" etc.
• Then we have the Amelino/Smolin paper arXiv:1110.0521. The same case as above: Those are reputable authors citing both OPERA and Coleman/Glashow, fine. But the reason why we talk about it, is their theory, that (deformed) doubly special relativity doesn't imply cherenkov radiation. However, no reputable secondary sources (non-peer reviewed preprints doesn't fall into this category) cites this specific paper, so we cannot include it. At least for now, but we of course can include it as soon as one of Nature (journal), The Register, Science (journal), New Scientist, Physics World, PhysOrg, Scientific American, LiveScience etc.. cites it. So many secondary sources should be sufficient to write an article... --D.H (talk) 15:59, 21 October 2011 (UTC)

WP:SECONDARY refers to Secondary source from which I quote:

Unlike in the humanities, scientific and medical peer reviewed sources are not generally considered secondary unless they are a review or a meta-analysis.

This can be assumed to be at least as true for non-peer reviewed sources. A secondary source must never convey original ideas or findings, which the Amelino/Smolin paper does. On this ground alone, I do not believe it has a place in this article.--83.89.0.118 (talk) 16:19, 21 October 2011 (UTC)

I would be a little careful in blanket-disallowing this paper, as it constitutes a commentary on the original paper and letters pertaining to it, with reference to Lorentz invariance. Of course the application of the mathematics is, presumably, new. Further the paper is itself cited in two other papers, whether these constitute primary, secondary or tertiary sources depends on exactly what they cite, and what they are being used as a source for. It is a good idea not to get too hung up on the distinction between primary secondary and tertiary. For example I would consider the paper a good source for "X and Y claim that a deformed Lorentz invariance can meet both the observations of the experiment and the requirement ..." but not good for "a deformed Lorentz invariance can meet both the observations of the experiment and the requirement ..." (unless one considers the mathematics to be so trivial as to be obvious to any reasonable reader). Rich Farmbrough, 17:04, 21 October 2011 (UTC).

It's really not about the quality of the Amelino/Smolin paper, it's only about the recognition by other sources, outside the preprint world of arxiv... Or do you want, that we do this for 10, 20, or 30 other arxiv-preprints? Many of them are quite good, but it's simply not our task to evaluate all those papers. So why not waiting a little more - I think it's only a matter of time when some of the secondary sources mentioned above cite this paper, then we can and even should include it; this is exactly what WP:Secondary is good for.. --D.H (talk) 17:42, 21 October 2011 (UTC)

The math is new and it is the whole essence of the statement being made by the editor. Because it is new, and haven't been critically examined by other sources, it is WP:PRIMARY and shouldn't be admissible here.--83.89.0.118 (talk) 18:01, 21 October 2011 (UTC)

There is more to this matter than just primary/secondary/tertiary. Wikipedia is not a newspaper. Most of the froth in this article will not even be a historical footnote in a few months or years. The article is a victim of WP:Recentism. Why all the haste? Given the amount of churn, we should be conservative in what gets added.

There is also the problem of WP:REDFLAG. While we do a good job of keeping crackpot musing out, the fact is that the subject is, all around, extraordinary, requiring extraordinary evidence. The OPERA team’s hearing in Wikipedia is already special dispensation. After all, the paper hasn’t even been published. The article has remained largely on the tacit agreement that those people must know what they’re doing given their positions. It is no surprise that a huge flurry of “papers” would arrive, hastily constructed speculations drawn up in the template of a research paper, from any physicist with an idea to toss out there. Most of these fail the WP:REDFLAG test (as does, technically, the OPERA result).

While mention in Scientific American or The Register or many of these other sources qualifies as secondary sourcing, what is the real importance of what they mention? Those editors are in no better (and often worse) position to vet what they mention than the editors for this Wikipedia article. They need to sell news (or advertising attached to it), and in cases like this, their decisions are also hasty. A little sensationalism sells.

Realistically as Wikipedia editors, how loosely or tightly we regulate content is up to our collective discretion within the framework of Wikipedia policies and guidelines. I argue that it behooves us to apply Wikipedia standards more strictly here, rather than less. Anyone can search the Web for the latest speculations. They don’t belong in an encyclopedia. Strebe (talk) 18:57, 21 October 2011 (UTC)

The problem with this definition of secondary sources is we end up following the sensationalism of the press. We report exactly what they report, since what they don't report, we can't cover. Nature News is credible, but it has had just 2 or 3 articles on OPERA. Look at one thing they say: "only a small minority of articles on arXiv are about objections to the experiment." Does the rest of our article reflect that? --Ajoykt (talk) 19:10, 21 October 2011 (UTC)

In principle this is all correct. We should set tight restrictions for the inclusion of our secondary sources to avoid sensationalism (which is fully in accordance with WP:Secondary and WP:Recentism). This probably means: Everything in the "Debates on the experiment" except the papers directly mentioned by nature (journal), or maybe also Scientific American and Science (journal) which have some credibility, should then be deleted - I have no problem with that. Then we only have to wait for really important news, such as the publication of additional papers in peer reviewed journals. --D.H (talk) 19:52, 21 October 2011 (UTC)

I would be ok removing pretty much most of the "Debates on the experiment", and using a summmary of the latest Nature News article in its place. The more than a year old results, Supernova and Fermilab, can probably be kept. I think Nature News did capture the state of the issue succinctly: neither refuted nor confirmed. --Ajoykt (talk) 20:08, 21 October 2011 (UTC)

Yes, that sounds very good. --D.H (talk) 20:15, 21 October 2011 (UTC)

Agreed. The fact that there is a burst of speculative activity and debate needs to be mentioned, but each argument does not. Strebe (talk) 20:47, 21 October 2011 (UTC)

Well, that is a shame, since Wikipedia is in an unique position to inform the lay-person (which is the purpose of an encyclopedia) while this topic is still interesting. Without it, people will have to rely solely on sensationalist news accounts or wait years until the academic process catches up with the experimental result. WP:RECENTISM isn't really a hard-and-fast guideline against new knowledge, it actually acknowledges that "up-to-date information on breaking news events, vetted and counter-vetted by enthusiastic volunteer editors, is something that no other encyclopedia can offer." As long as we maintain a long-term perspective, using credible news sources shouldn't have to be a problem. For instance, the Cohen-Glashow theory is likely to stay notable for a long time, given that it is directly derived from the Standard Model, and any explanation of the OPERA anomaly eventually has to be reconciled with it. Speculative papers about Lorentz symmetry deformation, less so.--83.89.0.118 (talk) 04:09, 22 October 2011 (UTC)

I don’t think we disagree here. Strebe (talk) 07:45, 22 October 2011 (UTC)

My point being that the same goes for, say, the news source discussions of the SN1987a claim, which have currently been removed from the article.--83.89.0.118 (talk) 10:45, 22 October 2011 (UTC)

Which SN1987a claim are you referring to? (btw: I have re-included some general information about it]. --D.H (talk) 13:53, 22 October 2011 (UTC)

Just the general notion of how SN1987a relates to the OPERA measurement. Again, I would be opposed to using pre-OPERA sources for this, because they don't establish the relevance to the subject of the article, but I'll leave it for someone else to bitch about.--83.89.0.118 (talk) 14:13, 22 October 2011 (UTC)

I disagree with including arXiv material based on our ideas of what there is credible. First, most of us are not qualified to judge. Second, the experts among us likely don't have the time to vet these papers thoroughly. Third, bypassing these issues by looking only at the credibility of the author is troublesome. Theoretical physicists are human, and, like the rest of us, have their biases, need their funding, believe in their own theories, and would like an experimentalist to test their theories at the highest priority. Note how C&G changed their wording under peer-review pressure: they no longer "refute" the experiment, only "significantly challenge" it. We cannot account for all these issues. As for the examples you cite, we do now have C&G, and not the Lorentz deformation (or whatever). We don't have Contaldi, van Elburg and myriad others. If you can find reliable sources (not NYTimes) for the SuperNova stuff, go ahead and add it back. Trouble I had adding it back was the reports on both sides ("SN mismatches OPERA" against "it is in a different energy range") were only half credible. I would take Nature News and Science as credible sources, not general news outlets. --Ajoykt (talk) 15:33, 22 October 2011 (UTC)

That's my point, I don't see the need to exclude 'general news outlets', as long as what they write isn't blatantly contradicted by other reliable sources. I wouldn't use The New York Post or Daily Mirror, but what is wrong in using New Scientist, or New York Times, for that matter, if what they write isn't at odds with other sources?--83.89.0.118 (talk) 00:55, 23 October 2011 (UTC)

Why not just use the other sources directly? If we can't use the other sources because they are primary sources, then often all we have is a primary source and a secondary source repeating it because it is sensational. That is okay for news reporting, but doesn't seem apt for an encyclopedia. If the primary source were a peer-reviewed paper, this would probably work, but not if the primary source is arXiv. --Ajoykt (talk) 01:00, 23 October 2011 (UTC)

Well, you're implying that e.g. New Scientist isn't doing any editorial filtering at all. I don't think that is accurate. They make mistakes, but far most of the indiscriminate stream of speculative arXiv papers hasn't made it into the news.--83.89.0.118 (talk) 01:34, 23 October 2011 (UTC)

The standards of general news outlets are different from that of an encyclopedia. They do, and the public expects them to, trade off accuracy for speed. A non-peer-reviewed paper quoted by a general news outlet is not credible enough to be a reference in an encyclopedia. We shouldn't reprint news here. --Ajoykt (talk) 22:08, 23 October 2011 (UTC)

So, this is the classical inclusionist-deletionist divide. We're not going to reach any sort of agreement.--83.89.0.118 (talk) 04:08, 24 October 2011 (UTC)

Hi. First of all, I'd like to express my protest against those who deleted my contribution in 1st place, without prior discussing. Why you didnt want to discuss your decision before applying? You must restore my contribution first, then discuss the possibility that it should be deleted. Besides, your group destroys my contribution in such a way that it could be undone only manually and also it creates difficulties in analyzing your actions. Now about the Smolin etal paper. 1110.0521 is a secondary source because it is a research paper which analyses some prior results and works. You know exactly that it will be published (although, Arxiv is itself a reliable source because it is a e-media source and it is premoderated, and some renown scientists, like Grigori Perelman, just put their work on Arxiv without further "publishing."). If you dont trust three authors from INFN, University of Wroclaw and Perimeter Institute for Theoretical Physics, then I have doubts about your neutrality. Besides, you are making double standards because the CG paper is not officially published yet either. To conclude: you must restore my contribution first, then discuss the possibility that it should be deleted. Otherwise it is an offence to my work and work of those 3 authors. User1344 (talk) 07:00, 23 October 2011 (UTC)

The burden of proof is solely on you. Besides, you were pointed to this talk page section multiple times.--83.89.0.118 (talk) 07:58, 23 October 2011 (UTC)

User 83.89.0.118, you deleted my original contribution within 4 minutes since it was posted. You claimed that the Arxiv paper can't be regarded as a secondary source, without any attempt to prove your statement. On the other hand, you left intact other contributions on the same page which referred to Arxiv's sources. Can you explain why you did that? — Preceding unsigned comment added by User1344 (talk • contribs) 08:24, 23 October 2011 (UTC)

Because the other contributions were backed up with secondary sources, e.g. articles from New Scientist. Adding a link to the primary source is fine as long as one or more secondary sources are listed alongside it.--83.89.0.118 (talk) 08:39, 23 October 2011 (UTC)

What do you mean "backed up"? Those were the news websites, they cant be classified as the secondary sources because the latter are defined as those which "rely on primary sources for their material, making analytic or evaluative claims about them." Those news websites didnt contain any analysis or evaluation, did they? On the other hand, the paper 1110.0521 did contain the analysis of the Cohen-Glashow's paper (also at Arxiv), as can be easily judged from its abstract and content, therefore, why does it need additional "back-up"? User1344 (talk) 08:58, 23 October 2011 (UTC)

1) The work a journalist and his or her editor does in unison is evaluative. 2) The claim that you cited with regards to Lorentz deformation etc. does not appear in Cohen-Glasgow's paper, i.e. it is original thought. To the extent that a source conveys original thought, it is a primary source, which must be backed up by a secondary source.--83.89.0.118 (talk) 10:05, 23 October 2011 (UTC)

(1) No, news websites are news websites, they cant contain the analysis and evaluation one can rely upon in Wikipedia. Otherwise every 12-yrs old journalist or reporter will start to teach scientists how to do physics and decide what is wrong and what isn't, and the whole division into primary and secondary sources won't make sense anymore, (2) you are not only wrong - you seem to be ignorant in the modern beyond-relativity physics. The Cohen-Glashow paper is based on the beyond-relativity theory developed by Coleman and Glashow himself some time ago. That theory presumes that the Lorentz symmetry is broken by the presence of the preferred reference frame, and it has not been confirmed by experiment yet. Moreover, it is not the only theory available at this moment hence the Cohen-Glashow arguments are not universally applicable at most (actually, they merely indicate that the Glashow theory couldnt explain the anomaly if it existed). The paper 1110.0521 gives the example where the Cohen-Glashow arguments fail, namely, when the Lorentz sym is deformed rather than broken. Moreover, the paper analyses also the original OPERA report as well as some preceding work. User1344 (talk) 12:31, 23 October 2011 (UTC)

1) "news websites ... cant contain the analysis and evaluation one can rely upon in Wikipedia" According to which guideline or policy? 2) Calling anyone who disagrees with you ignorant, is not generally considered a winning strategy for building WP:CONSENSUS.--83.89.0.118 (talk) 14:33, 23 October 2011 (UTC)

Listen, in an attempt to justify your highly controversial actions towards my contribution you have ended up proving the point that one page written on some news webpage by sobebody who's not even an expert in the fied is more "secondary" or "reliable" than many-page article written by three university professors. Don't you find your logic a bit twisted? Now, you are pretending to be offended and try to deviate the main topic of discussion. I dont want to build anything with you because I have doubts both in your competence and neutrality, sorry. Ok, I've already spent lots of time lecturing you so let me ask you a binary-answer question: User 83.89.0.118, do you admit that the paper Arxiv:1110.0521 qualifies as a secondary source? User1344 (talk) 05:36, 24 October 2011 (UTC)

Listen, I don't really care about you. As you were told on AIV, arXiv is not a reliable source, and your position is weak. That is your problem - not mine.--83.89.0.118 (talk) 06:35, 24 October 2011 (UTC)

"...arXiv is not a reliable source" According to which guideline or policy? Show me the specific rule, not another obscure sentence. Hello. This message is being sent to inform you that there is currently a discussion at Wikipedia:Dispute resolution noticeboard regarding an issue with which you may have been involved. The thread is "OPERA neutrino anomaly". Thank you. User1344 (talk) 07:32, 24 October 2011 (UTC)

According to WP:CONSENSUS. Now please stop rambling.--83.89.0.118 (talk) 08:53, 24 October 2011 (UTC)

Consensus in our case means that you restore my contribution as proposed by D.H in his Revision as of 18:17, 21 October 2011. If you have any problems with that - let's continue on Wikipedia:Dispute resolution noticeboard page. Now please stop posting meaningless information here. User1344 (talk) 11:18, 24 October 2011 (UTC)

To quote your own posting on the dispute board: "This is my 5th report to Wiki admins since 21 October 2011. Starting from 10:57, 21 October 2011 group of users (before it were users D.H and 83.89.0.118, now it is also User Ajoykt, Revision 20:46, 21 October 2011) persistently continues to delete my contributions on the OPERA neutrino anomaly page . . ."

If you want your stuff in, you will need to find more people to agree with you. --Ajoykt (talk) 15:30, 24 October 2011 (UTC)

If science-related issues were decided by majority then we would be still discussing how many elephants are needed to carry the flat earth... Four university professors who wrote the paper, and another one who has already cited it, obviously think it's a reliable source. Do they have to call some of those commercial news website to send a journalist and advertise their paper to get a secondary-source mention? What else they must do, buy Google adwords and banner-shows? Then it's going to be a trade fair not physics. User1344 (talk) 04:42, 25 October 2011 (UTC)

Let me repeat something I said earlier. I disagree with including arXiv material based on our ideas of what there is credible. First, most of us are not qualified to judge. Second, the experts among us likely don't have the time to vet these papers thoroughly. Third, bypassing these issues by looking only at the credibility of the author is troublesome. Theoretical physicists are human, and, like the rest of us, have their biases, need their funding, believe in their own theories, and would like an experimentalist to test their theories at the highest priority. Note how C&G changed their wording under peer-review pressure: they no longer "refute" the experiment, only "significantly challenge" it. We cannot account for all these issues. --Ajoykt (talk) 05:17, 25 October 2011 (UTC)

Let me repeat something too. I disagree with excluding sources based on research articles, unless they are obviously unsuitable. The paper you were deleting gives the important information C&G didnt mention explicitly, namely that the CG arguments are not universal but based on the earlier theory by Coleman & Glashow which hasn't been confirmed on its own. The paper also shows the limitations of the CG arguments. Thus, your group is de facto depriving a reader of awareness of other opinions on the subject matter which might create a wrong impression that the consensus in the physics community has been reached. This casts some doubts on your neutrality. Also, if you are "not qualified to judge" then don't touch other people contributions so rapidly, and let the time to do the trial. User1344 (talk) 06:24, 25 October 2011 (UTC)

Who said you are qualified to judge? Besides, why are you even here if you have so many doubts about the policies of Wikipedia?--83.89.0.118 (talk) 07:42, 25 October 2011 (UTC)

You lost the point, I'm not judging anyone because I'm not deleting other people contributions. Neither I'm going against Wiki policies - I'm actually following them: according to this, Self-published expert sources may be considered reliable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications. This is precisely the case I'm talking about, no? User1344 (talk) 16:13, 25 October 2011 (UTC)

You are judging everyone by thinking that you are somehow above the WP:BRD cycle. "Deleting other people's contributions" is part of the normal editorial process on Wikipedia.--83.89.0.118 (talk) 16:55, 25 October 2011 (UTC)

Nope, you are wrong, I don't think that way. I was overreacting sometimes but in essence I was trying to protect my contribution from self-appointed censors who enjoys deleting over creating. My contribution didn't do any harm here, it was informing readers of other points of views, and it is supported by Wiki policies. Thus, I was increasing an amount of awareness of a Wiki reader. Your actions, no matter how "justified" they look, are decreasing the awareness which contradicts to main ideas of Wikipedia. User1344 (talk) 05:56, 26 October 2011 (UTC)

You don't define the main ideas of Wikipedia. The guidelines do. And the guidelines say that: "if the information in question is really worth reporting, someone else will probably have done so."[8]--83.89.0.118 (talk) 08:03, 26 October 2011 (UTC)

Again, this guidelines comes in contradiction with the one I mentioned before. Our discussion is in a loop. User1344 (talk) 06:29, 27 October 2011 (UTC)

The article does mention C&G's theory may not be universally valid. Also, my stated stand has been, and continues to be, that the OPERA results have been neither confirmed nor refuted. If I believed OPERA results had been refuted, I would not bother to spend time on this article. As to the details of the objections to C&G's theory, the consensus here (see the other discussion: Questions for a wider audience) is that we should not allow non-peer-reviewed papers as primary sources. I don't agree with your argument the Smolin paper is a secondary source; it goes well beyond evaluating the OPERA results. I do agree Smolin is a credible physicist, but as I mentioned before, that doesn't by itself guarantee a paper he co-authored is accurate - elements other than lack of expertise can influence physicists to make outlandish statements in their pre-reviewed papers. We cannot peer review papers here, so we stick to the policy of not allowing non-peer-reviewed publications in. The C&G paper is journal-published; references to the pre-review paper have been removed; and the article, I believe, does convey the idea C&G is not the last word on the subject. --Ajoykt (talk) 16:44, 25 October 2011 (UTC)

the article, I believe, does convey the idea C&G is not the last word on the subject. No, article is still confusing - the sentence Andrew Cohen and Sheldon Lee Glashow have predicted superluminal neutrinos would radiate electrons and positrons and lose energy looks too decisive. The page forgets to mention that their prediction is based on some particular theory which is currently not confirmed by experiment and thus can't be regarded as a 100%-reliable argument on its own. Further, the response of Dario Autiero is given without reference and doesnt look convincing. User1344 (talk) 05:56, 26 October 2011 (UTC)

Now we have a convincing source, the director of research at CERN, no less, pointing out the central problem with the C&G refutation/signficant-challenge: that theory cannot override experiment. We also have C&G's alternate possibilities they had to include to get through peer review. --Ajoykt (talk) 02:38, 29 October 2011 (UTC)

At last someone of CERN bosses has managed to find his dentures and tell us that 2+2=4. User1344 (talk) 13:04, 29 October 2011 (UTC)

Let me repeat something too. I disagree with excluding sources based on research articles, unless they are obviously unsuitable. The paper you were deleting gives the important information C&G didnt mention explicitly, namely that the CG arguments are not universal but based on the earlier theory by Coleman & Glashow which hasn't been confirmed on its own. The paper also shows the limitations of the CG arguments. Thus, your group is de facto depriving a reader of awareness of other opinions on the subject matter which might create a wrong impression that the consensus in the physics community has been reached. This casts some doubts on your neutrality. Also, if you are "not qualified to judge" then don't touch other people contributions so rapidly, and let the time to do the trial. User1344 (talk) 06:24, 25 October 2011 (UTC)

The physics community itself hasn’t decided yet which papers are “suitable” and which are not. It’s fantasy to expect an encyclopædia to. Wikipedia isn’t for airing prognostications. Your disagreement is not with the editors of this article; it is with Wikipedia policy. The relevant policies have been cited and recited. Please give it a rest. When the physics community decides what arguments are “reasonable” then those arguments can be added into the article. If the article “might create the wrong opinion that a consensus in the physics has been reached” then it suffices to note in the article that no consensus has been reached. This is not hard stuff. The only controversy ought to be over how stringently Wikipedia policy and guidelines should be adhered to, not over whether we should make up new rules for your benefit. Strebe (talk) 07:46, 25 October 2011 (UTC)

Sorry but you are misinterpreting my arguments. I'm not going against Wiki policies - I'm following them: according to this, Self-published expert sources may be considered reliable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications. This is precisely the case I'm talking about. User1344 (talk) 15:57, 25 October 2011 (UTC)

P.S. Moreover, their paper has been already cited by another expert. All of these people produce unreliable sources? User1344 (talk) 15:57, 25 October 2011 (UTC)

I did not misinterpret your arguments; you are now bringing in a new argument. Your new argument noted here is reasonable, but you mysteriously left out the caveat, Take care when using such sources: if the information in question is really worth reporting, someone else will probably have done so. I also note that the use of may in Self-published expert sources may be considered reliable, which signals editorial discretion. Further, concerning work in the relevant field, everybody with an arXiv article up considers ‘him’self an ‘expert’ in the field, but meanwhile nobody really knows what’s going on. And that’s the real problem: Nobody really knows what’s going on. Physicists themselves are in disarray. That is what is important to this topic, not any specific interpretation or rebuttal—at least until physicists themselves have a chance to critique each other’s work. Meanwhile please respect the need to keep the article editorially cohesive and free of speculation. Strebe (talk) 19:37, 25 October 2011 (UTC)

There's no caveat in the sentence you cited. "Take care" means "take care" (who said that me as an editor didnt when choosing their paper?), and the authors of the paper are regarded as experts by the physics community, as you can easily check using Google and Wiki searches. About the rest of your post - my contribution is not about rescuing anybody from disarray. It is only about informing a reader about all views on the subject matter which do not contradict to Wiki policies. User1344 (talk) 05:56, 26 October 2011 (UTC)

In summary, your position appears to be that any material that cannot be clearly shown to violate Wikipedia guidelines for inclusion ought to be included at the discretion of the person who included it. My position is that your criterion is necessary but insufficient. Articles need editorial cohesion and balance, as determined by editors working on the article, and resolved by consensus. The many guidelines Wikipedia uses to balance these needs have been quoted many times, but they do not seem to suit you, so you have ignored them in pressing your point.

“Take care” does not mean you as an independently-acting editor; it means editors working collectively on the article, which of course you know, so your comment is in bad faith. You have not tried to “[inform] a reader about all views”; you have tried to inform readers about a particular view, which of course you know, so that comment is in bad faith also.

Now if in truth you want to inform readers about all views and are willing to make a list of all “notable” hypotheses, with hyperlinks to separate articles about each, or to the papers that describe them, then I (at least) would support the contribution. Just randomly stuffing in this one or that one because there is advocacy for it? Not so welcome. That violates the greater need for balance. Strebe (talk) 08:00, 26 October 2011 (UTC)

You have not tried to “[inform] a reader about all views”; you have tried to inform readers about a particular view You are right here but let me correct myself. I'm not supposed to know about all views, it's simply impossible. Instead, I have the right to add the one(s) I beware of. Other editors can add other views (in fact, they have done so, as you can see from the page so I dont have to repeat their work), and they are most welcome too. I am not going to delete their contributions, like other editors did. User1344 (talk) 06:29, 27 October 2011 (UTC)

a list of all “notable” hypotheses, with hyperlinks to separate articles about each, or to the papers that describe them. I was actually thinking of doing so when posted my first contribution but now it does not make sense because those works are based on primary sources (research papers at Arxiv), and there is no warranty that the above-mentioned editors won't delete this my contribution too. I know already their reasons: "if the information in question is really worth reporting, someone else will probably have done so. Not mentioned in news = no 2ndary sources = g'bye." User1344 (talk) 06:33, 27 October 2011 (UTC)

Certainly there never is any “warranty”, but such lists are common in articles. There is even a special kind of article that is just a stand-alone list. I don’t think anyone would consider a list controversial as long as it is JUST a list, without expounding on any of the items in it. Good-faith edits get deleted because they destroy the neutrality of an article through selective inclusion/exclusion and because they suffer from various shades of reliability in sourcing. A list doesn’t cause those problems; a list just aggregates what exists without implying any particular state of knowledge and consensus about the topic. Strebe (talk) 18:58, 27 October 2011 (UTC)

A list, as any other contribution, must obey the Wiki policies and guidelines. According to what I was told many times here, the current Wiki policies do not allow usage of Arxiv-based sources without secondary ones. User1344 (talk) 05:20, 28 October 2011 (UTC)

Questions for a wider community

1. Are arXiv e-print papers valid sources for a scientific Wikipedia article? 2. Are New York Times, New Scientist and similar popular but journalistically credible news outlets valid sources for a scientific Wikipedia article? 3. Should we lean toward including newsworthy items when the available secondary source is journalistically credible, but the primary source is a non-peer-reviewed paper?--83.89.0.118 (talk) 09:41, 25 October 2011 (UTC)

1. Should we include arXiv eprints as secondary sources, where they do not mirror a published paper?

2. Should we include NYTimes, New Scientist, and similar popular, journalistically credible, sources as secondary sources, where the primary source is a non peer-reviewed paper?

3. Should we lean toward including newsworthy and that way notable items when the secondary sources are journalistically credible, but the primary sources not credible when it comes to physics expertise and acceptance by the mainstream establishment?

The article, as of now, is structured assuming the answers to all three are "NO." But I am not sure what the consensus, as of now, is. Would like to hear the yes/no answers from everybody. I don't believe this issue is going away. I suspect strongly OPERA will reproduce their results without statistics in a few weeks, and MINOS will come up with another inconclusive result in a few months. Somewhere along the way, the press will report fringe theories, since physicists' initial shrill "refutation" of OPERA is going to weaken their position down the road. I am not against reporting fringe theories identified as such, but am going to vote neutral on all three questions. --Ajoykt (talk) 18:27, 24 October 2011 (UTC)

No, no and no.

1. If the paper isn't published then it's not even a primary source, and arxiv is certainly not a journal with a high standard of peer review.
2. Peer review is the standard here, and though pop science pubs and news pubs have certain standards, they aren't up to par with say Science or Nature. If this turns out to be "something" then it will be published in the correct places.
3. Journalistic credibility is necessary but not sufficient for inclusion, and a non-peer reviewed primary source simply isn't a good place to start.

N^oformation ^Talk 18:38, 24 October 2011 (UTC)

As per my comments here, I favor a strongly conservative editorial policy at least until the churn subsides. (1) No. (2) No. (3) No. Strebe (talk) 18:45, 24 October 2011 (UTC)

No, no, no. --D.H (talk) 18:55, 24 October 2011 (UTC)

I lean towards [no, yes, no], but I bow for the consensus, which seems to be [no, no, no].--83.89.0.118 (talk) 01:39, 25 October 2011 (UTC)

I would go with sometimes - for example it is without doubt fine (in principle) to say "Bloggs and Bloggs reported an excess of hadrons in their e-print<arXvi cite>, but their paper has yet to appear in a per reviewed journal." or "Bloggs and Bloggs reported an excess of hadrons in their e-print<arXvi cite>, but this claim was removed in the final Nature article.<Nature cite>" . Rich Farmbrough, 12:35, 25 October 2011 (UTC).

Citing primary sources is not fine (in principle), no matter how the citation is formulated. In principle, Wikipedia is not a compilation of quotes that may or may not be true. However, you may not agree with the principle. That would really be outside the scope of this talk page, though.--83.89.0.118 (talk) 13:44, 25 October 2011 (UTC)

2 poses an interesting question. Anyone can post nonsense on the arxiv and non-science publications can pick anything on arxiv up and make a big deal out of it without knowing that it's nonsense. This would seem to indicate that nothing reported in popular non-science publications should be included unless it's referring to a peer-reviewed journal article. Problem is: this is a different standard than for the rest of wikipedia. Anyone can post a video of a cat, a stinging rebuke of a political candidate, and it become news citable in wikipedia once the news has gotten hold of it. The press is very good at accurately telling when a subject is popular and what it's popular perception is, but generally terrible at getting science right. Does this mean we need an explicitly different standard for science?--Louiedog (talk) 14:02, 25 October 2011 (UTC)

That is an interesting perspective. Perhaps rather than a policy change, an essay on the subject would be beneficial. N^oformation ^Talk 07:23, 27 October 2011 (UTC)

It’s not so much that we need a different standard for science, as that we need to interpret Wikipedia guidelines more or less strictly according to the needs of the article. This topic being fast-moving, controversial, and a magnet for pet theories, the editors need to be far more selective than those of an article about a movie from 1976 that few watched even when it was new. The notability of some of the material people are trying to add to this article would be fine in other fields and in other circumstances. It’s just not fine here, now, where it’s controversial, competing with dozens of other candidates of similar notability, and hasn’t even been critiqued by peers. Strebe (talk) 19:05, 25 October 2011 (UTC)

Sources are not divided into reliable | not-reliable, but rather they fall along a spectrum, the spectrum is different for different subjects, and for different purposes within a subject. The nature of a spectrum is that there are no hard distinctions between different portions--it's not just the existence of more than 2 classes, but there being no distinct classes, although the extremes are clearly separate. Each case must be considered individually, though cases tend to repeat. The nature of primary sources in the sense Wikipedia uses it is that the sources need interpretation, and cannot be used as they stand, and that is actually similar in all subjects. Legal documents need to be placed within legal and historical processes, and to do so requires contextual explanation, with the full understanding requiring very broad cultural and philosophical concepts. Scientific data requires interpretation also, and to be placed within the context of other data, of theory, and of the general philosophic and cultural concepts of scientific explanation.

In a sense, the truly primary scientific data is the instrument readings or other observations themselves--the laboratory notebooks, not the summarized and processed data reported in a paper, which is usually a number of steps removed, and can even be regarded as secondary, for an instrument reading is meaningless except with the understanding of how it was obtained, and casual chain between the phenomena themselves and the readings on the instruments. (Sometimes the actual direct data is reported--but even that is obtained through instruments; the classic (though outdated) example is astronomical photographs, though even they require the external knowledge of the time and place of observation and the sensitivity of the emulsion.

What is reported in a scientific paper is rarely just the instrument readings, but, to some degree, the interpretation of them. Normally also, the relationship of the paper to other papers and theoretical concepts is discussed also, and this is a secondary discussion. (In the primary paper of key interest here, of course the key factor of theoretical discussion is that the authors stated they were not going to discuss it.) Sometimes the discussion of relatedwork can be quite extensive. The indexes recognize this: both medline and web of science classify as "reviews" papers with over a certain number of citation, on the basis that they intrinsically provide a review of the related literature.

For the purposes of Wikipedia, the necessary sources are the ones that the basis for an understanding of the subject. Normally this will be secondary literature, because in a general encyclopedia it is necessary to be very clear about the context. Normally does not mean always: for biological articles, the primary papers are freely used as references; for taxonomic articles, often they are the only references. Naturally, when there is interpretative work also, we report in as well.

As a related issue discussed above also, I think an article on science must if possible also include references form the general literature. If our purpose is to increase knowledge, we must increase it at the different levels that people understand. DGG ( talk ) 16:39, 26 October 2011 (UTC)

As Wikipedia editors, we require some general guidelines. Then we can make exceptions on a case-by-case basis per WP:CONSENSUS.--83.89.0.118 (talk) 07:00, 27 October 2011 (UTC)

ArXiv preprints are self-published sources — they are not refereed or subject to much editorial control beyond a cursory check by the moderators for whether they meet some very minimal standards of scholarliness (formatted like a paper and have a nonempty references section) and are on-topic. So per WP:SPS they can be accepted as reliable sources only if they're by someone already recognized as an expert in that particular field. —David Eppstein (talk) 22:10, 27 October 2011 (UTC)

The second part of the quote says we should include "expert" self-published papers very carefully. Experts, after all, have their agendas; peer review is what really makes a paper more agenda-free. For this topic, I think we need to stick to the highest standards within what Wikipedia policy allows. --Ajoykt (talk) 02:47, 28 October 2011 (UTC)

There is another perspective we haven't touched upon much which is WP:NOTABILITY. Arxiv preprints that haven't been cited by a reputable source can't be said to be notable, even by the standards that apply to non-science Wikipedia articles.--83.89.0.118 (talk) 13:21, 28 October 2011 (UTC)