Via Tommaso Dorigo of the CDF collaboration, the news that the Tevatron Electroweak Working Group has released a new analysis of combined CDF and D0 data with the most accurate result so far for the top quark mass: 172.5 +/- 2.3 Gev. Last summer this value was at 174.3 +/- 3.4 Gev (see a posting here), an improvement over the earlier value derived just using Run I data of 178.0 +/- 4.3 Gev.
The paper describing these results is available now here, and will soon be on the arXiv as hep-ex/0603039. This new result represents a determination of the top quark mass to 1.3% accuracy, and the paper claims that further Run II data should ultimately allow an accuracy of better than 1%.
For a talk about the significance of the top quark mass, see here.
Peter said “… the Tevatron Electroweak Working Group has released a new analysis of combined CDF and D0 data with the most accurate result so far for the top quark mass: 172.5 +/- 2.3 Gev …”.
I think that it should be noted that the “172.5 +/- 2.3 Gev” value applies only to one of three peaks in Fermilab t-quark data.
For example, the original 26 April 1994 paper FERMILAB-PUB-94/097-E contained a 26-event histogram for W+(3 or more)jets, without b-tags, that is Figure 65, and which (with an x for each event in each 10-GeV T-mass bin)looks like:
80
90
100
x
110
120
x
130
x
140
xxxxxxxx
150
xx
160
xxx
170
xxxx
180
xxxx
190
200
210
220
xx
230
240
In addition to the broad central peak around 170 GeV or so,
there is a small high peak around 220 to 230 GeV,
and a tall narrow low peak around 140 to 150 GeV.
In that paper, Fermilab dismissed the tall narrow low peak, saying “… We assume the mass combinations in the 140 to 150 GeV/c^2 bin represent a statistical fluctuation since their width is narrower than expected for a top signal. …”.
Nevertheless, the low and high peaks persist in later data for which they are not excluded by cuts etc, and in my opinion all three peaks should be taken into account in analyzing the T-quark data.
If it is assumed that the T-quark is a simple quark just like all the others, and if one if forced to pick only one of the three peaks and reject the rest, then perhaps the Fermilab consensus view that “the” T-quark mass is “172.5 +/- 2.3 Gev” might be justified (but the other two peaks would still be unexplained, as they do not appear to be accounted for by conventional background).
However, I do NOT think that the simple T-quark assumption is accurate.
Kent W. Staley, in his book “The Evidence for the Top Quark” (Cambridge 2004), mentions on pages 295-296:
“… hints that the simplest hypothesis that the top candidate events are just the ttbar events and SM background may not be entirely correct such hints ‘should be monitored carefully, as they may be offering us glimpses of new physics’ …
Even when experimenters find that they have acieved experimenters’ success, they look more closely to see the interesting flaw in their achievements – the discrepancy that will mean, not failure … but the possibility of some new success …”.
Staley’s book contains detailed references and context that I will omit here.
Possible clues about the three peaks might come from such sources as:
hep-ph/030713, by Froggatt, that studies the interrelationships of the T-quark, the Higgs, and the Vacuum; and
hep-ph/0311165 and hep-ph/9603293, by Yamawaki et al, that describe T-quark condensate models, including:
Nambu-Jona-Lasinio models with mt about 145 GeV
Kaluza-Klein (8-dimensional) model with m_t about 173 GeV
Bardeen-Hill-Lindner models with mt about 218 GeV.
Please see the papers themselves for details.
In deference to Peter’s wish to exclude discussion of non-standard physics models, I will stop here.
Tony Smith
http://www.valdostamuseum.org/hamsmith/
So is this just a refinement of the Standard Model or something interesting? I’m afraid I’m too much of a layman to tell.
Bourgeois Nerd,
In some sense, this is just a more precise determination of one of the standard model parameters. But it’s a very important one: the Higgs couples to fermions according to their mass, so it is by far most strongly coupled to the top quark. So, if you want to detect the Higgs, either directly or indirectly, knowing the top quark mass as accurately as possible is important.
How does the new top mass change limits on the Higgs mass and other things?
Taking these three values together makes the TQM in the range
[ 173.7 – 174.8 ] (or 174.25 +/- 0.55).
That’s the lowest of the three maxima and the highest of the three minima.
Although I suppose it doesn’t really work that way.
In Dr. Tait’s presentation, he suggests the mass of the top can tell physicists something about electroweak symmetry breaking. I know a tiny bit about the Higgs mechanism giving mass to the W’s and Z, and obviously if the top couples most strongly to the Higgs, there’s some kind of circumstantial relationship. But beyond the mere fact that the vector bosons and the top all are mighty heavy, weigh about the same, and hence have similar couplings to the Higgs, what deeper principles are being revealed?
Ah, I also had a prediction for the top quark, but it was exactly at 175 Gev (188 u), hmm. (well, it could be a prediction for top+charm meson).
Seriously, what matters is not 175 or 173, even Tony’s 145 peak. What matters is why it is at the electroweak symmetry breaking scale.
Dear Bourgeois Nerd,
the top quark mass is important because it is important for the existence of supersymmetry and the minimal supersymmetric standard model.
Thomas Larsson: yes, the Higgs mass in the minimal supersymmetric standard model depends sensitively on the top and stop masses.
More generally, the exact top quark mass is important because it affects string theory phenomenology. Click the PDF file that Peter Woit linked in the last sentence of this article – with the word significance – and go to the middle of the PDF file.
Of course, without supersymmetry, the top quark mass is not too significant, and the Higgs mass is independent of the top quark mass.
See my blog for more details about this issue.
Best wishes
Lubos
This blog post is indeed remarkable. Lubos writes:
“Its mass used to be thought to be 178 GeV. It was lowered to 175 GeV and now to 172 GeV, with error margin around 2 GeV. [..]
At roughly the same time, Faraggi included some loop corrections which made the agreement worse because the physical mass he obtained [using string theory] was 192 – 200 GeV. [..]
this story is an example that string theory can predict properties of particle physics and it can even give us the right predictions.”
It is encouraging that superstring-theorists tried (try?) to predict particle properties. But the Faraggi result was obviously *falsified* by the experiment.
Dear stan,
you forgot to quote my sentence about the crackpots who form the core of the readership of Peter Woit’s blog. It was inconvenient for you, was not it? The tree level calculation from 1991 was confirmed by experiments, the loop-corrected calculation from 1995 was falsified by experiments. At any rate, you should better be careful in saying that the loop-corrected prediction was falsified because the owner of this blog does not like those who disagree with his thesis that string theory etc. is “not even wrong”.
Best
Lubos
So, string calculations do not need loop corrections?
Luboš Motl: you forgot to quote my sentence about the crackpots who form the core of the readership of Peter Woit’s blog.
Well, you’re here all the time. Sounds like a case of the crackpot calling the crackkettle black.
Alejandro Rivero, your dumb questions convinced me that I must simply put you in the very same group as Nigel Cook, Tony Smith, Danny Ross Lunsford, Quantoken, and many others. Sorry but it will probably be hard for you to reverse this policy.
The full calculations of course include loops. But the tree level calculations don’t. Saying that tree level calculations are worthless, which is what you are essentially doing, is incredibly stupid. Most of the key calculations in physics, including those in the discovery of the electroweak theory, were tree level calculations.
The full calculation needs us to clarify many more extra details, and it is completely conceivable that the 1995 loop calculation is simply further away from the correct answer than the 1991 tree level estimate. Experimentally, it is clearly the case.
Best
Lubos
I definitely recommend Lubos’s blog entry
http://motls.blogspot.com/2006/03/peter-woit-recommends-supersymmetry.html
It really has everything:
1. delusional thinking about string theory: “string models predict the exact masses of all particles”.
2. bizarro ideas about the scientific method: see his discussion of the Faraggi “prediction” of the quark mass.
3. lunatic rants about string theory skeptics, e.g. yours truly.
As for his comments about the top mass being not significant if there is no supersymmetry, this is just misinformation.
In the standard model itself, precision electroweak measurements depend both on the top quark mass and on the Higgs mass. Knowing the top quark mass, these measurements give significant constraints on the Higgs mass. See page 6 of the Tait talk I quoted. I don’t know of a source for updated numbers on these constraints using the latest top mass.
In the MSSM, at tree level the Higgs mass has to be less than the Z mass, which is wrong. Loop effects allow a higher Higgs mass, and the one-loop calculation strongly depends on the top mass. One can push up the Higgs mass, but doing this requires increasing degrees of fine-tuning. People are already uncomfortable with the degree of fine-tuning required to get the Higgs mass above the LEP experimental lower bound of 114 Gev. See page 12 of the Tait talk.
Lubos Motl: Saying that tree level calculations are worthless, which is what you are essentially doing, is incredibly stupid.
Saying that he said this or even implied it when he did not is incredibly stupid.
Dear Peter Woit,
1. the statement that a particular background in string theory predicts the masses of all massive particles ia a mathematical fact, and only people who are un-educated in theoretical physics misunderstand why
2. a prediction means a calculation of a certain observable that will be observed in the future, which is what Faraggi (and many other people) did. His calculation could have been imperfect but it was definitely a prediction, and once again, only complete ignorants like you and your crackpot fans are unable (or unwilling?) to understand this fact
3. the lousy intellectual quality of your society of cranks is being proved by virtually every posting you have ever made, and almost every comment that your crackpots fans ever added to your postings
4. the Higgs mass itself can’t be affected or deduced theoretically from the top quark mass if you have no supersymmetry simply because there are quadratic divergences correcting the Higgs mass that are not cancelled automatically unless one has SUSY, but can be freely cancelled by (nonSUSY) counterterms. The value of the top quark mass influences various high precision measurements but does not, in any way, decide about the existence of the theory if the theory is non-supersymmetric; any value is equally OK theoretically
5. you exactly explained why top and stop are important for MSSM – because their large one-loop corrections make the predicted Higgs mass plausible. This fact is the main reason why studying the top with a better accuracy is a scientifically justifiable enterprise. The large one-loop top corrections are in no sense “fine-tuning” because the top Yukawa coupling is not a small number. On the contrary, it is experimentally proved to be around 1/sqrt(2), so your whining about “fine-tuning” in the one-loop corrections is just another, 9325th evidence of your complete ignorance about particle physics.
I recommend you to avoid commenting on things that you have no idea about, such as particle physics, string theory, cosmology, and many other fields.
Best
Lubos
It could be worthwhile to quote also here the paper Lubos brought into focus: http://ccdb4fs.kek.jp/cgi-bin/img_index?9110375
I am not sure if it is the first estimate running down to the 175 range; I have seen some this range quoted in other ocassions, but also in higgs estimates, so memory is fuzzy here. Perhaps Tony can remember better, as top is in his focus.
Lubos,
You’re missing the main point in the non-supersymmetric case. High-precision electroweak measurements depend on the Higgs mass, so they indirectly are measuring it. This measurement is highly sensitive to the value of the top mass. For latest results, including the latest top mass, see
http://lepewwg.web.cern.ch/LEPEWWG/plots/winter2006/
For the fine-tuning I’m talking about in the supersymmetric case, see the discussion at
http://golem.ph.utexas.edu/~distler/blog/archives/000336.html
and maybe you can write in to the author to inform him how ignorant he is on this issue.
It is perhaps remarkable that Tait’s slides still carry a plot from a 1993 article: the one in page 14 comes from page 9 of http://arxiv.org/abs/hep-ph/9311269 . Incidentally it can be noticed that it is still adjusting to a top quarks mass of 150 GeV (is it the second discontinous line counting from the bottom left corner?). Those interested on predictions should note also figure 5 in page 29 of this preprint.
On a different matter, I was thinking about aplying to the graduate program of Harvard. Do you know any professor there able to write a recommendation letter?
Lubos is such a crackpot. According to Lubos, (Lubos Theorem 1)measuring the mass of the top quark is a scientifically important endeavor iff supersymmetry is true.
(Lubos Theorem 2) For any theory X, If you cite a reference that talks about X, then you believe everything X says is true. (contrapositive of Thm2: If you do not believe everything that X says is true, then you do not cite a reference about X)
Maybe Peter and Lubos should remain confronational with each other, for “information” that was just supplied by that “energy” expenditure. 🙂
A story dedicated to Yellow Michael.
An anthropic string theorist, a non-anthropic string theorist, and a physicist were discussing the experimental status of string theory.
Anthropic string theorist:
– String theory is in complete disagreement with experiments, but that is OK because God created the universe to be compatible with human life.
Non-anthropic string theorist:
– String theory is in complete disagreement with experiments, but that is OK because Witten has won the Fields medal.
Physicist:
– String theory is in complete disagreement with experiments.
Hi all,
wow there is a lot to learn by reading these comments… People who trust recent advances in theoretical physics and people who eat sausage should not ask what there is inside. (One might rightly add people who blindly believe in experimental errors – but that’s another story).
Just a remark: I accept non-experimentalists being speculative about background fluctuations in principle – the charm meson started off as one at the very beginning, and so did other important discoveries in particle physics as well as other disciplines. But claiming that there are three peaks in the top quark mass distributions CDF and D0 isolate, based on a plot of a very background-rich sample isolated with a tenth of the statistics available today, is plain nuts.
Ah, and the top mass did not “use to be 178, then 175, then 172 GeV”. CDF measured it at 174 GeV in 1994 (!) in the “evidence” paper, one year before we decided to go observational. Later, more precise determinations have brought the value around a bit, but world averages have stayed in the 171-178 GeV range, and always within 1-1.5 sigma from 174.
Tommaso Dorigo said, about my comment: “… claiming that there are three peaks in the top quark mass distributions CDF and D0 isolate, based on a plot of a very background-rich sample isolated with a tenth of the statistics available today, is plain nuts. …”.
I agree. However, that is NOT what I said. I did specifically mention only one plot (1994 CDF untagged semileptonic), but I also went on to say: “… the low and high peaks persist in later data …”.
I did not give specific references to the later data for space and exposition reasons, but, since the question has been raised, here are some of them (note – of the three peaks (low, central, and high), the central peak is the one around 173 GeV):
the 1997 D0 untagged semileptonic histogram at hep-ex/9703008 showing all three peaks;
the tagged CDF semileptonic events in hep-ex/9801014 showing the low peak;
the tagged D0 semileptonic events in hep-ex/9801025 showing all three peaks;
the CDF dileptonic events in hep-ex/9802017 showing the low and central peaks;
individual D0 dileptonic events described in the thesis of Erich Varnes at http://wwwd0.fnal.gov/publications_talks/thesis/thesis.html including the following events:
Run 84676 Event 12814 (2-jet analysis gives low peak)
Run 58796 Event 417 (low peak)
Run 90422 Event 26920 (low peak)
Run 88295 Event 30317 (low peak)
Run 84395 Event 15530 (3-jet analysis gives high peak);
individual D0 dileptonic events described in hep-ex/9808029 using the matrix-element weighting algorithm that “… is an extension of the weight proposed in [R.H. Dalitz and G.R. Goldstein, Phys. Rev. D45, 1531 (1992)] …”, including:
e mu #1 (low peak)
e e #1 (low peak)
mu mu (showing low and high peaks)
which events were also described using the neutrino weighting algorithm, with these results:
e mu #1 (low peak)
e e #1 (low peak)
mu mu (high peak)
I hope that the above makes it clear that a reasonable analysis of later data shows that all three peaks do continue to show up.
Of course, you can argue about interpretations, and indeed (as Fermilab has done) argue that since the low and high peaks involve fewer events, they should be thrown out.
However, it seems to me that the 3-peak point of view is also reasonably arguable, and that work on theory and future data analysis should include work on all three possible peaks instead of designing cuts etc that exclude the high and low peaks from data analysis.
Tony Smith
http://www.valdostamuseum.org/hamsmith/
PS – Please note that the above is only an outline indicating some of the later data, and that extended discussion of relevant subtleties (see my web site) is beyond the scope of blog comment discussion. I would like to think that the above is extensive enough to refute the blog characterizations of me as “just plain nuts” (by Tommaso Dorigo) and “moronic crackpot” (by Lubos Motl).
Lubos Motl Said,
1. the statement that a particular background in string theory predicts the masses of all massive particles ia a mathematical fact, and only people who are un-educated in theoretical physics misunderstand why
2. a prediction means a calculation of a certain observable that will be observed in the future, which is what Faraggi (and many other people) did. His calculation could have been imperfect but it was definitely a prediction, and once again, only complete ignorants like you and your crackpot fans are unable (or unwilling?) to understand this fact
3. the lousy intellectual quality of your society of cranks is being proved by virtually every posting you have ever made, and almost every comment that your crackpots fans ever added to your postings
[…]
Excuseme my complete ignorance and my obvious lack of brilliance (i am not a Fields medallist and never will become one) Dr. Motl, but i am unable to understand above 3 points (really i am unable to understand the rest of points also, but this is not a journal :-).
3. How can one measure the crank level of any posting here? I am just curious! Please do not cite Baez ranks or similar, since none of them has been experimentally proven! From a historical point of view, i also find interesting all past genius of physics (including Isaac Newton) was named crancks by their then colleagues. Is not this a lesson for all of us?
Moreover, whereas i disagree with some comments posted in this blog, i agree with many others and, by now, the net balance is clearly positive.
2. I think that none manual on epistemology or experimental methods (i know) agree with you.
A prediction does not mean a calculation of a certain observable that will be observed in the future, as you are saying. A clear example is computation of size of hidden dimensions on string theory. One can calculate that “observable” and obtain any result from zero to infinite.
Yes, the mainstream claims that hidden dimensions “are” of Planck scale but arguments are easily rebated from chaos (microstructure from sub-Planck scale spacetime modifies correlations and the physics over the Planck [There is a good QM example worked in Nature by Zurek and methods can be adapted to a generalized string theory formulation in Liouville space similar to that of Nanopoulos and coworkers]).
Not forget also that in certain cosmological-string models the hidden dimensions are infinite in size (FAPP).
Therefore, any measure of the size of any hidden dimension (if any) and agreement with string theory (any value from zero to infinite) is not a prediction.
1. You claim it is a mathematical fact that a particular background in string theory predicts the masses of all massive particles.
Perhaps i am un-educated (in your own words) but, sincerely, I fail to see that from both a physical and a mathematical point.
See the point from set theory. If you are claiming that string theory is a collection of different sub-theories/models each one with diferent “parameters” (background), then if a particular background in string theory “predicts” -this is not a correct word by point 1- the masses of all massive particles, then many other backgrounds do not and offer us wrong results. That is also a mathematical fact you forgot cite above.
It is also interesting to note that if X string models are compatible (the word “predict” is as said incorrect) with values of masses, and Y are NOT compatible. The ratio X/Y is either very close to or equals to zero. Therefore, from a Landscape point of view, string theory is not compatible with masses.
But if you are claiming that only one of those backgrounds correspond to the “real” (physical) string theory (with the rest of a pure mathematical interest), then your above phrase cannot be correct. Instead of a particular background in string theory… you would say just “string theory” (meaning with its only physical background, THE background, that nobody knows) …
—
Juan R.
Center for CANONICAL |SCIENCE)
Lead, yes it is true that interesting information appears even if with the pityful waste of confrontational energy. All this issue of SuGra Electroweak Breaking, from the plot in Tait’s slides, is worthwhile to be contemplated. I was unaware of the trick of generating a phase transition by just running down the constants in such a way that the uncolored mass (the higgs) hits the change of sign.
Of course a collateral benefit of this confrontational show is that all the participant can hope get more visibility for their own papers. I expect the silencious espectators will consider it and jump to the arena some day.
A note on crackpots & data: Just last night I discussed this iossue with my two sons (one a computer scientist who refused fellowship & got MS in less than a year; the other, a construction supervisor who got his degree & went into industry) & both agreed with me that the test for crackpots was whether or not their stuff worked. I now work in data acquisition & analysis (so I understand Tony Smith’s concerns) and any theory I have must fit the facts & predict future behavior of equipment. My computer scientist son said he thought this throwing of “crackpot” at people & attacking their intelligence is a result of the scarcity of positions available & the fight for funding (one reason he left academia). I think it’s very sad. Was Einstein right: is being a plumber the best option (or electrician or carpenter)? And finally, Lubos, of course Woit is right about Higgs mass by way of the top and precision electroweak measurements.
An update about the prediction of top mass Lubos was proud of. It was bases in the techique of infrared quasi-fixed points for the top quark yukawa coupling, a technique that was resuscited in the eraly nineties and at that time involved SuGra and string based models. But it comes from ten years before; most people quotes a paper of C.T Hill of 1981, Phys Rev D 24 p 691, which in turn attributes the idea to Pendleton and Ross in some paper written down at the end of 1980 and published as B. Pendleton and G.G. Ross, Phys Lett 98B p 291 (1981). The paper of Hill does not claim string or susy inspirations, and I can not tell of P&R now because it is an offline paper. But here goes one half of the string-susy-fundamentation of top mass.
The another half is the trick for Radiative “infrared” Electroweak Breaking used in Sugra. It is that because the Higgs is uncolored but the top is not, the RG equations can be adjusted in such a way that as one run them down from M_GUT to low energy, at some instant a critical point “a la Curie” is crossed and the Higgs field gets the negative mass square it needs to break electroweak symmetry. I can not tell if this idea is original of SUSY theories; Tait seems to imply it, and I can not find it argued in other kind of papers. In fact it is a very elegant idea.