Neutrino mistakes wrong tracks and hints, - PowerPoint Presentation

1. Neutrino mistakeswrong tracks and hints, hopes and failures From: History of Neutrinos, Paris September 2018For: La Physique des Neutrinos, Paris September 2022

2. Why am I giving this talk?Probably because I’ve been putting out a free monthly ~100 line “long-baseline newsletter” about all things neutrino since 1992 [To join 2410 subscribers, send “subscribe” to maury.goodman@anl.gov or see http://www.hep.anl.gov/ndk/longbnews/]Though I didn’t identify “What we learn from our mistakes”, some found the stories interesting8 September 2022Maury Goodman, Argonne2

3. Context My original talk was in front of (mostly old) neutrino physicists Most of the original audience knew of many of these stories, which mostly involved reported new physics. Lots of jargon and details you’ve probably covered this week For each “story” I’ve added one or more CONTEXT slides, but I will present them quickly. This talk also jumps frequently from one topic to another PLEASE - Interruptions and questions during the talk are welcomed.8 September 2022Maury Goodman, Argonne3

4. Neutrino mistakeswrong tracks and hints, hopes and failuresFake nsHistory of NeutrinosParisSeptember 2018

5. Neutrino mistakeswrong tracks and hints, hopes and failuresHistory of NeutrinosParisSeptember 2018No!a=e2/c-5 points

6. The United States Post Office doesn’t want mistakes8 September 2022Maury Goodman, Argonne6Ben PickettInverted Jenny

7. Report of m e g(i.e. rumor)8 September 2022Maury Goodman, Argonne7

8. Context for m e g (1) In graduate school, I learned about the “conservation of muon number”.Later, the buzz was “no flavor changing neutral currents” 8 September 2022Maury Goodman, Argonne8

9. m e gWhen I was a graduate student at Fermilab in the mid 70’s, I heard a rumor that this process had been measured B ~ 10-8 at the SIN facility in Switzerland. Expected (now) B(m e g) = 5  10-48 [Dm221 (eV)2]2 sin2 q12 cos2 q12 = 6  10-57 I never heard a talk about this and a positive result was never published SIN published a limit The only confirmation of my memory is a discussion with Robert Shrock which led me to a footnote in a paper by Bjorken and Weinberg8 September 2022Maury Goodman, Argonne9

10. SIN Results8 September 2022Maury Goodman, Argonne10

11. SIN Results8 September 2022Maury Goodman, Argonne11

12. m e g…8 September 2022Maury Goodman, Argonne12

13. m e gWhen I was a graduate student in the mid 70’s, I heard a rumor that this process had been measured B ~ 10-8 at the SIN facility in Switzerland. Expected B(m e g) = 5  10-48 [Dm221 (eV)2]2 sin2 q12 cos2 q12 = 6  10-57 I never heard a talk about this and a positive result was never published SIN published a limit The only confirmation of my memory is a discussion with Robert Shrock which led me to a footnote in a paper by Bjorken and WeinbergThis rumor led to a series of lectures at FNAL by Robert Shrock. That is where I learned about neutrino oscillations.8 September 2022Maury Goodman, Argonne13

14. Outline8 September 2022Maury Goodman, Argonne14

15. “mistakes” considered for this talkSuperluminal ns17 keV n NuTeV anomalyTritium endpoint (-)m2Kolar eventsEarly atmospheric n lack of polarizationMINOS anti-n q23God’s mistaken grammarLabels for Dm2abPDG m(n) encodingWhich n is a particle?Karmen time anomalyTime variations in Troitsk mn2ITEP m(ne) = 30 eV in 19808 September 2022Maury Goodman, Argonne15SIN report of m →e gHigh y anomalyNuTeV Helium bag eventsKlapdor’s 0nbb signalLSND/eV “sterile” nsIMB limit on n oscillationsAlternating neutral currentsReines-Sobel n oscillationsVanucci PS191 oscillationsBNL 776 & 816 oscillationsBEBC oscillationsHPWF “super” trimuonsOscillations in BugeyMajoron emission in 0n2b PNL/USCSPT vs. V-A

16. What is a mistake?My a-posteriori definition – an experimental search for new n physics for which there was an apparent error of the first kind or error of the second kind.Or a consequential semantic issue.Possible causes:A statistical fluctuation?A systematic error?A wrong interpretation of good data?A theoretical misunderstanding?…Not included:Anything unrelated to ns(e.g. Dark Matter claims, PDG-cm …)Hardware problems(IMB bag collapse, Super-K tubes…)A wrong theory(Hot Dark Matter, SU(5)…)8 September 2022Maury Goodman, Argonne16

17. EpistemologyDefinition of epistemology : the study or a theory of the nature and grounds of knowledge especially with reference to its limits and validity OrHow do we know what we know?8 September 2022Maury Goodman, Argonne17

18. Some of my guiding principlesin evaluating experimental results There are an infinite number of tests of the null hypothesis There is no theory of systematic error You can’t prove anything in Physics The union of two confidence levels isn’t a confidence level. The commonly used 5s criterion is based on misunderstandings and is wrongly used8 September 2022Maury Goodman, Argonne18

19. 17 keV neutrino8 September 2022Maury Goodman, Argonne19

20. Context for beta decay searches Tritium decay has a large Q, 18.6 keV. A small neutrino mass (eV) affects the shape of the electron spectrum at the endpoint. A large mass (keV) would cause in a kink in the spectrum.8 September 2022Maury Goodman, Argonne20

21. 17 keVtimeline1985 Simpson kink at 1.5 keV in tritium decay; 18.6-1.5=17.1, P = 3%Phys. Rev. Lett 54 1891-18931985 various negative results P < 0.3%Phys. Rev. C32 2215-22161989 Hime & Simpson kink in 35S; 16.9 keV, P = 0.7%Phys. Rev D39, 18051991 Hime & Jelley 2 measurements in 35S; 17 keV, P=0.8% 8sPhys. Lett. B257 4411993 Mortara et al., definitive exclusion Phys. Rev. Lett. 70 3941993 Hime, Identifies scattering effects as likely responsiblePhys. Lett B299, 165-1738 September 2022Maury Goodman, Argonne21

22. 17 keVStuart Freedman8 September 2022Maury Goodman, Argonne22

23. 17 keV8 September 2022Maury Goodman, Argonne23Himes’ reinterpretation with “more complete” electron response function with “intermediate scattering”

24. 17 keVRandom commentThe last days of the 17 keV neutrino was contemporaneous with:The serious consideration of long-baseline experimentsThe beginning of my newsletter (May 1992)There was much more theoretical interest (>  5) in the possible existence of the mn = 17 keV n than atmospheric n oscillations, mn = 1-100 meV8 September 2022Maury Goodman, Argonne24

25. Klapdor’s neutrinoless double beta decay8 September 2022Maury Goodman, Argonne25

26. 8 September 20222626Decay rate is given by the golden rule and depends onan effective Majorana mass. It requires knowledgeof nuclear physics quantities.CP-phases: ±1Neutrino massesElements upperrow of MNS-matrixDecay Rates TPhase spaceMatrix elementMContext for 0nbb studies Maury Goodman, Argonne

27. Neutrinoless Double Beta DecayHeidelberg-Moscow Collaboration looking for 0nbb in 76GermaniumEur.Phys.J.A12:147-154,2001; 14 authors; T > 1.9 1025 y @90%CLMod.Phys.Lett.A16:2409-2420,2001; 4 authors; T= 1.5 -0.7 +16.8 1025 y @95%CL8 September 2022Maury Goodman, Argonne27

28. Neutrinoless Double Beta DecayEarly reactions8 September 2022Maury Goodman, Argonne28

29. Neutrinoless Double Beta DecayNewsletter storyAs a result, I got a quick email from John Beacom who didn’t believe the result, and said this didn’t meet the standards of my newsletter. I replied I didn’t have standards, I had deadlines.8 September 2022Maury Goodman, Argonne29

30. Neutrinoless Double Beta DecayNewsletter story-2I responded that my newsletter gave equal attention to the discovery of neutrino oscillations and a novel about a Neanderthal neutrino physicist. 8 September 2022Maury Goodman, Argonne30

31. Neutrinoless Double Beta DecayIrrelevant facts which affect believabilityResult published before there was a preprintPublished in a journal on which Klapdor was associatedSignificant fraction of collaboration didn’t sign the paperSignal failed the upside-down testThe only talk I heard from him was arrogantData wasn’t shared with all collaboratorsHe repeatedly touted this with the DAMA DM “discovery”Doug Michael’s view:“Even if it’s right, it’s wrong”It felt like a-posteriori analysis to me8 September 2022Maury Goodman, Argonne31

32. Neutrinoless Double Beta DecayEven though few in the community “believed” it, Klapdor’s value became a benchmarkEXO Phys. Rev. Lett., 109, 032505 (2012).GERDA 2013KamLAND-Zen 20138 September 2022Maury Goodman, Argonne32

33. Superluminal neutrinos8 September 2022Maury Goodman, Argonne33We don’t allow neutrinos in here, said the bartender. A faster-than-light neutrino walks into a bar.NeutrinoWho’s There?Knock Knock

34. Context for time of flight studiesMINOSFNAL to SoudanOPERACERN to Gran Sasso8 September 2022Maury Goodman, Argonne34

35. Superluminal neutrinos August 2007 – MINOS superluminal preprint (published 2008) 22 Sep 2011 – OPERA preprint -- 6 s 23 Sep 2011 -- CERN seminar broadcast live on the web. PREPRINT 23 Sep 2011 – (original) CERN press release 17 Nov 2011 -- Revised preprint submitted for publication (not published) 25 Feb 2012 – Possible loose connector announced 8 September 2022Maury Goodman, Argonne35

36. 8 September 2022Maury Goodman, Argonne36Superluminal neutrinosMINOSIt is typical within the field of High Energy Physics that we have not read a majority of our own papers. (!)

37. Superluminal neutrinosOPERA: Oscillation Project with Emulsion Tracking ApparatusCNGS: CERN Neutrinos to Gran Sasso8 September 2022Maury Goodman, Argonne37

38. 8 September 2022Maury Goodman, Argonne38Superluminal neutrinosMy 2011 commentsNo neutrino’s velocity was measuredThree ways to make a mistake:Clocks (t)Surveying (distance)Beam physics assumptionsMy opinion – OPERA is wrongBut, if it (OPERA) is “right” we don’t know what “it” is, so we can’t test “it”

39. Superluminal neutrinosORIGINAL PRESS RELEASEGeneva, 23 September 2011. The OPERA1 experiment, which observes a neutrino beam from CERN2 730 km away at Italy’s INFN Gran Sasso Laboratory, will present new results in a seminar at CERN this afternoon at 16:00 CEST. The seminar will be webcast at http://webcast.cern.ch. Journalists wishing to ask questions may do so via twitter using the hash tag #nuquestions, or via the usual CERN press office channels.The OPERA result is based on the observation of over 15000 neutrino events measured at Gran Sasso, and appears to indicate that the neutrinos travel at a velocity 20 parts per million above the speed of light, nature’s cosmic speed limit. Given the potential far-reaching consequences of such a result, independent measurements are needed before the effect can either be refuted or firmly established. This is why the OPERA collaboration has decided to open the result to broader scrutiny. The collaboration’s result is available on the preprint server arxiv.org: http://arxiv.org/abs/1109.4897.The OPERA measurement is at odds with well-established laws of nature, though science frequently progresses by overthrowing the established paradigms. For this reason, many searches have been made for deviations from Einstein’s theory of relativity, so far not finding any such evidence. The strong constraints arising from these observations makes an interpretation of the OPERA measurement in terms of modification of Einstein’s theory unlikely, and give further strong reason to seek new independent measurements.“This result comes as a complete surprise,” said OPERA spokesperson, Antonio Ereditato of the University of Bern. “After many months of studies and cross checks we have not found any instrumental effect that could explain the result of the measurement. While OPERA researchers will continue their studies, we are also looking forward to independent measurements to fully assess the nature of this observation.” “When an experiment finds an apparently unbelievable result and can find no artefact of the measurement to account for it, it’s normal procedure to invite broader scrutiny, and this is exactly what the OPERA collaboration is doing, it’s good scientific practice,” said CERN Research Director Sergio Bertolucci. “If this measurement is confirmed, it might change our view of physics, but we need to be sure that there are no other, more mundane, explanations. That will require independent measurements.”In order to perform this study, the OPERA Collaboration teamed up with experts in metrology from CERN and other institutions to perform a series of high precision measurements of the distance between the source and the detector, and of the neutrinos’ time of flight. The distance between the origin of the neutrino beam and OPERA was measured with an uncertainty of 20 cm over the 730 km travel path. The neutrinos’ time of flight was determined with an accuracy of less than 10 nanoseconds by using sophisticated instruments including advanced GPS systems and atomic clocks. The time response of all elements of the CNGS beam line and of the OPERA detector has also been measured with great precision."We have established synchronization between CERN and Gran Sasso that gives us nanosecond accuracy, and we’ve measured the distance between the two sites to 20 centimetres,” said Dario Autiero, the CNRS researcher who will give this afternoon’s seminar. “Although our measurements have low systematic uncertainty and high statistical accuracy, and we place great confidence in our results, we’re looking forward to comparing them with those from other experiments."“The potential impact on science is too large to draw immediate conclusions or attempt physics interpretations. My first reaction is that the neutrino is still surprising us with its mysteries.” said Ereditato. “Today’s seminar is intended to invite scrutiny from the broader particle physics community.”The OPERA experiment was inaugurated in 2006, with the main goal of studying the rare transformation (oscillation) of muon neutrinos into tau neutrinos. One first such event was observed in 2010, proving the unique ability of the experiment in the detection of the elusive signal of tau neutrinos.8 September 2022Maury Goodman, Argonne39

40. Gedanken historyAfter the press release, this made worldwide front page news.But…Suppose OPERA had the same seminar but CERN had not issued a press release…Two weeks later there would have been an article in the science section of the New York TimesThe scientific story would have been the same.The worldwide fuss would not have been the same.8 September 2022Maury Goodman, Argonne40

41. IMB neutrino oscillation limitusingupward stopping atmospheric n8 September 2022Maury Goodman, Argonne41

42. Context for neutrino oscillations(1) P = sin2(2q) sin2(1.27 Dm2L/E)Dm2= |m22-m12| (eV2)sin2(2q) is the strength of the mixing; sin2(2q)=0 is “no oscillations”En is the neutrino energy (GeV) (log(Ep) at accelerator)L is the distance from the source to the detector (km)P is the probability of oscillation. Results usually shown on a parameter space plot, (sin2(2q) vs. Dm2).You measure things related to L and En.For atmospheric neutrinos, zenith angle is a proxy for L.8 September 2022Maury Goodman, Argonne42updown

43. IMB neutrino oscillation limitIn 1992, IMB published a neutrino oscillation limit based on the ratio of upward-going stopping m from atmospheric n to upward going m.This is where we now think it is. 8 September 2022Maury Goodman, Argonne43

44. IMB A retraction was never really published, but the limit was apparently quite sensitive to structure functions.1999 Cosmic ray conference abstract:8 September 2022Maury Goodman, Argonne44

45. LSND andeV sterile neutrinos8 September 2022Maury Goodman, Argonne45

46. Semantic interlude Sterile neutrinos, with m ~ eV, have been a suggested solution of four so-called short-baseline neutrino anomalies: LSND, Cr source for Ga experiments, reactor anomaly, MiniBooNE.Those have motivated a series of experiments, including the Fermilab short-baseline programMore generally, sterile neutrinos with other masses (from keV to the GUT scale), have been suggested for a variety of theoretical reasons.8 September 2022Maury Goodman, Argonne46

47. Context for sterile neutrinos (1) 8 September 2022Maury Goodman, Argonne47 Width of Z Nn Zt+t-,e+e-,m+m-qq̅, nmn̅m,nen̅e,ntn̅t,?? SLAC 2.80.6 LEP 3.29 0.17 Now 2.994 0.012 (1g events)So,… if you measure more light neutrinos, they must be sterile, i.e. don’t couple to the Z with the weak interaction.

48. Context for sterile neutrinos (2) There are a series of experimental anomalies that (by themselves) could be interpreted as due to nm ne oscillations with Dm2 ~ 1 eV2 LSND Chromium Source measurements for Gallium solar n detectors Reactor neutrino flux normalization MiniBooNE (at Fermilab)With 3 mn, there are only 2 independent Dm2LSND8 September 2022Maury Goodman, Argonne48

49. Light sterile neutrinosPartial TimelimeNeutrino 1994; 8 events B = 0.9Nucl. Phys. B (Proc. Suppl.) 38 229-234 19951st Paper 16.4+9.7-8.9 ± 3.3 excess Phys. Rev. Lett. 75 2650-2653 1995Hill paper with limit 5 events B=6.2Phys. Rev. Lett. 75 2654-2657 19952007 MiniBooNE Results Inconsistent with Existence of "Sterile" Neutrinoshttps://www.aps.org/publications/apsnews/200706/miniboone.cfm2006 Gallium Anomaly 0.79+0.09−0.10 expected rate with souceJ. N. Abdurashitov et al.,Phys.Rev.C 73(2006) 0458052011 Reactor neutrino anomaly Phys.Rev.D83:073006,20118 September 2022Maury Goodman, Argonne49LSND was built to look for what we now call q12

50. Light sterile neutrinosIssuesInconsistency of 0.3%, 3% and 20% signalsLSND Decay in flight signalKarmen’s limit better than its sensitivityMost analyses presented based on 2-ns (qme)Cosmological limits on Nneff & Smn MiniBooNE’s low energy excess3+1 vs 3+2 vs 3+3Limits from MINOS+, NOvA, Ice-Cube, …Inconsistency of ne appearance and nm disappearanceWhat if the “Best Fit” is a bad fitValue of Dm2 for low energy excess8 September 2022Maury Goodman, Argonne50

51. Light sterile neutrinosThe best fit is a bad fitWhy physicist disagree:If the data doesn’t agree with the null hypothesis or the alternative hypothesis, some say you need more data. I say you need more hypotheses8 September 2022Maury Goodman, Argonne51

52. Light sterile neutrinosJoe Lykken & evidenc8 September 2022Maury Goodman, Argonne52http://news.fnal.gov/2018/06/big-boost-for-fermilabs-short-baseline-neutrino-experiments/

53. Light sterile neutrinosIn my opinioneV sterile ns suggested by an interpretation of LSND have been ruled out for a long timeThe SBL anomalies are realThey may have interesting or uninteresting explanations.If you don’t know what you are looking for…You might find itYou might not find itBut you cannot logically rule it out8 September 2022Maury Goodman, Argonne53

54. Sterile neutrinos2018-2022Reactor neutrino anomaly disappears with a modified n flux from the 235U chain.MicroBooNE was built to test if the MiniBooNE low energy excess (LEE) was due to photons or electrons. It measured no excess. They won’t say that this result rules out a sterile n interpretation of MiniBooNE.With new data, BEST still sees a Gallium anomalyNeutrino 4Theorists have been active trying to think of ways that would give an excess in MiniBooNE but not MicroBooNE I’ve not seen any reevaluation of the backgrounds and systematic errors in MiniBooNE which might cause the excess. (S/B is small)8 September 2022Maury Goodman, Argonne54

55. God’s mistake?8 September 2022Maury Goodman, Argonne55

56. Context for n parameter measurement (1) 8 September 2022Maury Goodman, Argonne56= UPDG 2022sin2(θ12)  = 0.307  0.013   Δm221  = (7.53  0.18)  10-5 eV2   sin2(θ23)  = 0.545 ±0.021Δm232 = 2.45  0.03  10-3 eV2 sin2(θ13) = 0.0220  0.0007    If masses are hierarchicalmheavy = 48 meVmlight = 9 meVbut S(mn) < 87 meV1 meV (milli eV)= 10-3 eV

57. Context for n parameter measurement (2) 8 September 2022Maury Goodman, Argonne57n sourcesSolar n (made by nuclear fusion in the sun) .1-10 MeVAtmospheric n .1-50 GeVAccelerator n 1-100 GeVSupernova n 10-50 MeV Cosmological remnant n 3o (330/cm3)Reactor n 1-6 MeV Energetic Astrophysical Systems 1-100 TeVGeophysical 0.2 MeV Nuclear Weapon tests

58. 58Intelligent Design of Neutrino Parameters? (~2005) (from S. Wojcicki) The optimum choice for Dm221? Such as to give resonant transition (MSW effect) in the middle of solar energy spectrum -, Dm221 = 8.2 x 10-5 eV2The optimum choice for sinq12? Big enough for oscillations to be seen in KamLAND - ~0.8The optimum choice for Dm232? Such as to give full oscillation in the middle of the range of possible distances that atmospheric n’s travel to get to the detector - Dm232 = 2.3 x 10-3 eV2 The optimum choice for sinq23? Big enough so that oscillations could be seen easily - q23 ~ p/4The optimum choice for sinq13? Small enough so as not to confuse interpretation of the above - q13 < 100But the acid test - will q13 be big enough to see CP violation and determine mass hierarchy? 8 September 2022Maury Goodman, Argonne

59. 59And still?By 2011 we learned that q13 was as large as could be imagined in 2006How about the remaining parameters so that the “Intelligent Design” arguments can get longer (2012)?d ~ 3p/2 to most quickly determines the hierarchy to get large CP violation & answer the CP violation questionThe inverted mass order, so we can tell Dirac/Majorana & maybe beta decay endpoint Majorana, which seems to be more interesting so that some of our theorists will be happy (seesaw, etc.)8 September 2022Maury Goodman, Argonne

60. 60In 2022 it appears:By 2011 we learned that q13 was as large as could be imagined in 2006How about the remaining parameters so that the “Intelligent Design” arguments can get longer (2012)?d ~ 3p/2 to most quickly determines the hierarchy to get large CP violation & answer the CP violation questionThe inverted mass order, so we can tell Dirac/Majorana & maybe beta decay endpoint Majorana, which seems to be more interesting so that some of our theorists will be happy (seesaw, etc.)?  ?8 September 2022Maury Goodman, Argonne

61. Errors of the 2nd kindwhen you fail to reject the null hypothesisSolar neutrinosIt took a long time to accept neutrino oscillations accounted for the “solar neutrino problem”.nt appearance in Super-KCriticized when they 1st reported a 2.3 s nt appearance.[Higgs]Failure to accept normal n mass order I think neutrinoless double beta decay get more funding if we now think the mass order is inverted than if we think the mass order is normal.8 September 2022Maury Goodman, Argonne61

62. Lessons?8 September 2022Maury Goodman, Argonne62

63. Scientific inductionand HEP’s most common mistakeNull hypothesis (The data can be understood without new physics)Alternative hypothesis (A particular new effect)A test statisticA chance probability ( x s effect or y% CL limit)Error of the first kind (incorrect signal)Error of the second kind (incorrect limit)But all of this is only valid if the hypotheses and statistic are specified a-prioriAnd we do a-posteriori analysis all the time – we have too!I cringe when I hear colleagues justify 5s because “I’ve seen so many 3s effects go away.” An x s effect with a-posteriori and a-priori hypothesis are calculated in exactly the same way. The meaning is totally different.8 September 2022Maury Goodman, Argonne63

64. Things to think about (1) How can an outsider judge whether a result is based on a-priori or a-posteriori text? Importance of well-defined a-priori consistency checks Julien Lesgourges at Neutrino 2022 (29.09): “The meaning of s between cosmology and particle physics is not really the same.”In contrast, I would say, “The meaning of s between cosmology and particle physics and any other field is the same.” 8 September 2022Maury Goodman, Argonne64

65. Things to think about (2)Tests that one might perform are paradigm dependent and hence time dependent. A result which 40 years ago might have been interpreted as evidence for SPT interactions instead of V-A, might today be called evidence for sterile ns. Role of critics Should wrong results be withdrawn? Contradicted? Is ignoring them ok? At what point should a surprising result be publicized/published? What is the role of referees in a hard-to-believe result? Is it fair that a highly publicized wrong result becomes a benchmark and gets many citations?8 September 2022Maury Goodman, Argonne65

66. Things to think about (3) The roles of secrecy and rumors within experimental collaborations Considerations when a result is not published by the whole collaboration? The role of extraneous factors in believability Role of critics Physicists are naturally skeptical We more often ignore than actively criticize results we don’t believe Some “active” skeptics have not fared well- Morrison (solar nus), Miyake (Atmospheric nus)+ Stu Friedman (17 keV)Productive mistakes vs. unproductive mistakesWhen should we publicize outside the community (i.e. press release) regarding a result unexpected within the community8 September 2022Maury Goodman, Argonne66

67. Summary and Conclusions?8 September 2022Maury Goodman, Argonne67

68. “The great teacher, failure is” … Yoda“By seeking and blundering we learn.” ― Johann Wolfgang von Goethe8 September 2022Maury Goodman, Argonne68

69. Our field (particle physics) does a poor job ofPresenting statistical arguments in a consistent way.In particular distinguishing between a “x” s effect calculated from an a-priori test and an “x” s effect calculated from an a-posteriori testExplaining to ourselves and others how we conclude anything based on whatever combination of data, theory and instinct that we use. NeverthelessWe seem to do an excellent collective job of taking seriously results which get vindicated and being skeptical of results which do not Calling a result a mistake has a connotation of criticism. In a scientific sense, I do not criticize the vast majority of these reported results. 8 September 2022Maury Goodman, Argonne69

70. AcknowledgementsI am solely responsible for all mistakes in this talk about mistakesFor other material, I wish to thank Evgeny Akhmedov, Zelimir Djurcic John Losecco Naba Mondal, Jurgen Reichenbacher Jack Schneps Phil Schreiner Robert Shrock Hank Sobel Daniel Vignaud Cosmas Zachos…8 September 2022Maury Goodman, Argonne70

71. Merci de votre attention8 September 2022Maury Goodman, Argonne71

72. Vanucci PS191 oscillationsHPWF “super” trimuonsn grammarKarmen time anomalyWhich n is a particle?Labels for Dm2abSPT vs. V-AMajoron emission in 0n2b PNL/USCBEBC oscillations8 September 2022Maury Goodman, Argonne72Other “mistakes”High y anomalyNuTeV Helium bag eventsReines-Sobel n oscillationsITEP m(ne) = 30 eVTritium endpoint (-)m2Time variations in Troitsk mn2NuTeV anomalyEarly atmospheric n lack of polarizationPDG m(n) encodingBNL 776 & 816 oscillationsOscillations in Bugey

73. MINOS n̅ q238 September 2022Maury Goodman, Argonne73MINOS analyzed n and n̅ separately and conducted a blind analysis for both of them. When the numbers looked different, MINOS invented an a-posteriori test and quoted the difference as 2% chance P or about 2.4 s.MINOS then got more n̅ running and the discrepancy disappeared.PRL 110 251801 2013PRL 107 021801 2011

74. Alternating neutral currentsObservation of Neutrino Like Interactions Without Muon or Electron in the Gargamelle Neutrino ExperimentPhys. Lett. B46 (1973) 138-140Nucl. Phys. B73 (1974) 1-228 September 2022Maury Goodman, Argonne74From D. Haidt & A. Pullia, “The Weak Neutral Current, Discovery and impact” Rivista del Nuovo Cimento, 2013

75. Kolar Events5 events seen in KGF which were consistent with a vertex in airNever contradictedI’m not aware anyone triedBackground difficult to estimate8 September 2022Maury Goodman, Argonne75

76. High y anomalyHPW Unexpected y distributions (y = Ehad/En) in FNAL E1 - low x2 further papersNot seen CCFRContradicted by CHARM @ CERN8 September 2022Maury Goodman, Argonne76

77. NuTEV Helium bag events3 events with a background of 0.04 appeared to vertex in a Helium bag in front of NuTEV.Kinematics didn’t match the alternative hypothesis (decay of a supersymmetric particle.)Int. J. Mod. Phys. A16S1B, 761, 2001. 8 September 2022Maury Goodman, Argonne77

78. NuTeV anomalyMeasureFollow-up experiment not approved8 September 2022Maury Goodman, Argonne78

79. Reines Sobel n oscillationsSingle detector reactor experimentCompared CC/NC ratesH. Sobel “It has been a while, but if I remember correctly one of the cross sections we were using changed significantly.8 September 2022Maury Goodman, Argonne79

80. 30 eV ITEP m(n)From Boehm & Vogel, “The physics of massive neutrinos”Fit to 30.0 ± 1.9 evBut8 September 2022Maury Goodman, Argonne80

81. Tritium negative m2Mainz m2 = -3.7 ±5.3fit ±3.1syst eV2/c4Physics Letters B 460 (1999) 219-226Troitskm2 = -1.9 ±3.4fit ±2.2syst eV2/c4Physics Letters B 460 (1999) 227-235And several others, leading to speculation then about tachyons.“As was discussed at the KATRIN inauguration a few days ago, this is now thought to possibly have been due to inadequate inclusion of the effects of the fact that the tritium diatomic molecules have rotational and vibrational excitations and the decays populate excited states of the resultant tritium-Helium-3 diatomic molecules.” R. SchrockTroitsk also saw a possible periodicitry of the step position with a period of a half year.8 September 2022Maury Goodman, Argonne81

82. Lack of polarization in atmospheric n calculationsThe original atmospheric neutrino flux calculations used for the “ratio of ratios” did not take into account the fact that the muon is polarized. Cf Gaisser Stanev & Barr, Phys. Rev D. 38 (1988) 85. with Gaisser et al., Phys. Lett. B 214 (1988) 147.8 September 2022Maury Goodman, Argonne82

83. PDG m(n) encodingMainz & Troitsk reported measurements/limits on the “mass of the electron neutrino”. That’s like saying “the hole that the electron went through in the two slit experiment”.An experiment at LEP looked at the kinematics for t→3pn & t→5pn.The event with the largest effective mass from pions puts an upper limit on the mass of the neutrino. This was reported as the mass of the nt. It is actually an upper limit on the mass of the lightest neutrino.The right way to think of this was described in Shrock Physics Letters 96B p159 (1980). The language of the PDG’s RPP was cleaned up in 2003. But some of the “limits” are wrong, though irrelevant.8 September 2022Maury Goodman, Argonne83

84. Other fleeting neutrino oscillation reportsBugey 2 detector 3 s effect for sin22q = 0.2, Dm2 = 0.2 eV2, reported at Neutrino 1984CERN PS191 was mentioned to me but I could only find published limitsTwo contradictory (different L/E) positive results from BNL involving low energy electron excesses in a nm beam*BNL 776 – 23 ne (17) seen compared to 13.1 expectedBNL 816 – 110 ne seen compared to 53 expectedBoth reported at Neutrino 1988*This led me to predict (orally), when MiniBooNE was proposed that their search for a signal of low energy electrons in a nm beam would probably be positive, since every neutrino experiment had an excess of low energy electrons. This also may be why Bob Bernstein said that if MiniBooNE refuted LSND, everyone would believe it, but if they confirmed LSND, nobody would believe it.8 September 2022Maury Goodman, Argonne84

85. Exp 1 HPWF “super” 3mHigh energy multiple muons from E1A/E310 were called “super” events at FNAL seminars Never published?8 September 2022Maury Goodman, Argonne85

86. n grammarA hyphen is sometimes omitted when talking about a short-baseline neutrino program or the Long-Baseline Neutrino Facility. I usually describe thr rule as follows:If I do an experiment with a long baseline, long is an adjective and baseline is a noun. There is no ambiguity and no hyphen is needed.  If I do a long-baseline neutrino experiment, or work on a short-baseline neutrino program, it is the baseline that is long or short and not the neutrino, the experiment or the program. The hyphen removes this ambiguity.  In fact, a long-baseline experiment will likely take a long time, and be a long long-baseline neutrino experiment.  And if Ken Long from Imperial College ever builds a neutrino factory, that will be Long's long long-baseline neutrino experiment.)8 September 2022Maury Goodman, Argonne86

87. Karmen time anamolySpeculation of a slowly moving massive particle produced in the beam stop.b = 0.0283 ± 28 eventsPhys. Lett. B348 19 1995.Ruled out at CERN by Daum et alPhys.Rev.Lett.85:1815-1818,2000J. Reichenbacher thesis showed that beam-correlated neutrons caused the time-anomaly. 8 September 2022Maury Goodman, Argonne87

88. Which neutrino is a particle?ne, nm and nt are not particles. They are flavor eigenstates. Particles are solutions of Schrodinger’s equation in free space.8 September 2022Maury Goodman, Argonne88

89. n labelsDm212  m12 – m22Dm221  m22 – m12Dm213  m12 – m32Dm231  m32 – m12Dm223  m22 – m32Dm232  m32 – m22Dm221 + Dm232 + Dm213 = 08 September 2022Maury Goodman, Argonne89Some people mistakenly always match q12 with Dm212 , etc.q12, q13, q23 are labels, Dm2jk are ordered (sign)

90. SPT vs V-AS&T explanations of b decay of 6He, BNL 1953, 1955There were also some early neutrino experiment results on neutral currents that didn’t match expectations.(I was referred to some conference proceedings, but I couldn’t find anything published.)These issues led to “Adler’s Army” which was many postdocs at Princeton & elsewhere studying the possibility of SPT weak interactions in addition or instead of V-A.e.g. PRD11, 1043 (1975) & PRD10, 2216 (1974)8 September 2022Maury Goodman, Argonne90Budagov et al., Phys. Lett. 29B p525 1969.And other data that agreed betterP. Schreiner and F. von Hippel, PRL 30 p 339 1973At first I thought this had something to do with single pion CC data, and I found:

91. MajoronAIP Conf. Proc. DPF meeting SLC UT, Jan 1987Double Beta decay evidence for Majoron Phys.Lett. B192 (1987) 460-462Evidence againstPhys.Lett. B198 (1987) 253-254Aregument against evidence against8 September 2022Maury Goodman, Argonne91

92. BEBC beam dumpA. De Rjula et al Nuclear Physics B168 (1980) 54—68Could be ne →ntNot published?8 September 2022Maury Goodman, Argonne92