Fermilab's CDF collaboration has been in the spotlight for a few weeks, thanks to an analysis reporting an unexplained population of muons in its data set. Various science news outlets—including New Scientist, Discover, and Nature—and the blogosphere covered the story including speculation that the muons could be the product of a new particle, which yielded the catch phrase "the ghosts in the machine." Now theorists and phenomenologists are taking a crack at it. Speculation aside, the CDF collaboration is not making any specific claims about interpretation of the data but is set on solving the mystery, and have offered their analysis to the DZero collaboration, also based at Fermilab's Tevatron, for comparison.
The original paper popped up on the physics preprint arXiv on October 29, and scientists were waiting in the wings to discuss the results as soon as they could. The CDF detector, which collects collisions of protons and anti-protons, found a population of muons for which the analysts could not identify a source. Not only did the paper admit being unable to find the source, but declined to speculate on possibilities. Such an unexplained spray of particles could be background noise, a systematic error, or it could be a sign of a new particle.
This week, speculation about the muon source has jumped out of the blogosphere and into the arXiv, a place for physicists to post scholarly articles prior to them being accepted for publication in a peer-reviewed journal. At least three papers appeared on the arXiv this week addressing the CDF results (the blog Not Even Wrong discusses a few others); one postulating that the muons are the decay product of a new particle. However, this is merely theory aligning itself with one possibility, and offers no greater insight into the actual data. A paper by Matthew J. Strassler of Rutgers University ends with this philosophical paragraph: "…the challenges that this analysis faces are useful as a springboard for discussion…Opening our minds regarding the possible signatures that nature might provide, and finding new techniques for expanding the range of reasonable searches at hadron colliders, is surely beneficial for the field."
CDF analysts have been working on the current analysis for more than a year, but can't explain the surplus of the muons. When the group submitted the paper to internal peer review, it underwent another six months of intense scrutiny by the collaboration. The analysis reached what CDF co-spokesperson Robert Roser calls "a plateau of understanding," at which point the collaboration agreed they could do no more in the short term.
And this is where CDF received attention it didn't expect. The particle physics community hotly debates the questions of whether or not to publish a scientific paper that does not offer an explanation of its findings. Some argue, for various reasons, that a collaboration should only publish analysis in which they can fully explain the data. Others support publishing results with unanswered questions to welcome outside assistance. Due to the inconclusive nature of the paper, about one third of the collaboration members chose to omit their names from the authors list. Roser says it's not uncommon to have a few members decline to sign a collaboration paper, but this percentage is fairly large. Rumors suggested that the collaborators left their names off this CDF paper because of the paper's results, rather than because of its inconclusive nature.
In reality, the CDF paper's claims aren't controversial, because it hasn't made any claims at all. It reports an unexplained occurrence, but makes no assumptions about the cause. "As a collaboration we're not claiming any new physics. We did not feel that there was a clear cut case for any of that and we want to be responsible," says co-spokesperson Jacobo Konigsberg.
But the spokespersons happily agree that they welcome the speculations of others. "It's part of science," says Roser. "Scientists see an effect and try to understand it. They'll try to fit it to preferred models. But if the speculation turns out to be wrong, the analysis is still right.
"This isn't the end of story," continues Roser. The publication of the paper doesn't end the analysis, but marks the beginning of what could be years of work. DZero might decide to do a similar analysis, offering a comparison study. If DZero could replicate the data set, it would almost certainly rule out the possibility that the muons are merely a phenomenon of the CDF apparatus.
The CDF representatives also emphasize that further data collection could not have assisted the analysis. Neither, they say, might a more energetic particle collider such as the LHC. At this point, they cannot predict where they will find a solution to the problem. Besides the DZero comparison, Konigsberg says CDF might repeat the experiment and analysis with other triggers. "Triggers" are selection mechanisms for weeding out the most interesting data from collisions given that there is too much data to record it all, but they might change what the data sets look like. "We are beginning to consider which other data sets we should look at to see if…this effect shows up or not," he says.
"If we knew exactly what to do next and how to do it, we would have," says Roser. "Will we get to the bottom of this in a few months? Probably not. Now we have to effectively peel back the next layer of the onion. There are still things that we can do ourselves, they're just longer term projects. We'll keep digging deeper, break the analysis into smaller pieces and examine each piece. Really try to understand what potential effect can explain this."
