March 5, 2004  


How Many Physicists Does it Take to Write a Paper?

By Kate Metropolis

SMAISMRMILMEPOETALEUMIBVNENUGTTAVIRAS must rank as one of history’s shortest scientific publications. In 1610, it caused a sensation among the prominent scientists of Europe. It came, after all, just months after the same author had announced that four satellites were orbiting Jupiter. Publishing his result in the form of an anagram allowed Galileo to claim priority for his latest discovery without revealing what it was to his competitors.

Three months later, Galileo went public and decoded the anagram: Altissimum Planetan Tergeminum Observavi—I have observed the farthest planet [then thought to be Saturn] as a triple sun.

Yet, two years later, after turning his telescope on Saturn again, Galileo had doubts about his conclusion that Saturn was a triple star. "Looking at Saturn within these last few days, I found it alone, without its accustomed stars, perfectly round. ...How can this be? Are the two smaller stars consumed like spots on the sun? ...Or was the appearance a fraud and illusion? I cannot resolve so new, so strange, so unexpected a change. The shortness of time, the weakness of my intellect, the terror of being mistaken, have greatly confounded me."

Were Galileo to spring out of a wormhole from seventeenth century Tuscany into last week’s meeting of the BABAR collaboration, he would have found that while the scale of endeavor has changed dramatically—600 physicists working together instead of one—the goals of publishing scientific results have remained essentially unchanged. You want to be first. And you want to be right.

As of February 17, 2004, BABAR physicists had completed 46 physics analyses since January 1, 2003, bringing the total for the lifetime of the collaboration so far to 80. "There are a lot more in the pipeline," Jeff Richman (BBR), Physics Analysis Coordinator, said confidently.

Belle, the rival collaboration at Japan’s B-Factory, completed 29 analyses during the same period, for a lifetime total to date of 83.

How much should one make of those numbers? "What’s important," said Pat Burchat (Stanford), "is not the number of papers, but what’s being published." A Stanford Physics Professor, Burchat is both a past chair of the BABAR Publications Board and has served as BABAR Physics Analysis Coordinator. "What fraction of the data is actually being used? What’s the quality of the analyses: Is it something novel, or just a repetition of what’s already been done?"

About 150 BABAR physics analyses are underway at the moment. Before they are submitted to one of the peer-reviewed journals, often Physical Review Le.ers, they undergo what Burchat described as a high level of scrutiny. (See Cahn interview for details, page 3.)

The scrutinizers on the three-person committees that review every putative publication are there to ask critical questions and to give constructive feedback to the authors. "Some people are known for a high rate of suggestions," said Burchat, "but when we ask someone to serve on a review committee, we are looking for a high rate of useful suggestions."

The process is especially valuable for graduate student authors, Burchat pointed out, particularly those whose advisors aren’t able to spend a lot of time at SLAC because of teaching responsibilities at their home institutions.

The size of the enterprise is reflected in the author list, which runs to several pages in Physical Review Letters. "BABAR has a system of distributed responsibility," said Burchat. "The people without whom we’d never even get the data—the run coordinators and the detector subsystem managers who are on call 24/7, the data processors—are fully entitled to be listed as authors, whether they’ve even had time read the paper."

There’s even a protocol for when a person is added to or taken off the author list. Usually someone has to be on the experiment for at least a year before they’re added, because that seems a reasonable length of time for them to have made a real contribution to the experiment. After someone leaves BABAR, they’re usually entitled to stay on the author list for at least a year.

What happens, though, when a physicist is trying to get tenure and his or her name is buried in an author list several pages long? In experimental high energy physics, according to Burchat, it is not papers but letters of recommendation that allow others to evaluate an individual’s contributions.

Although carefully checking BABAR results takes months, Galileo would probably not consider that a long time. His result that the Earth is not the immovable center of the Universe was, after all, not fully accepted for 360 years.


The Stanford Linear Accelerator Center is managed by Stanford University for the US Department of Energy

Last update Tuesday March 02, 2004 by Emily Ball