Cosmic axis begets cosmic controversy.

What's the fastest way to get an astrophysicist to write a paper? Publish a report claiming the universe has a special direction.

That's the way it seemed last week, when a flurry of new papers, none of them peer-reviewed, appeared on the Internet. The authors lambasted a recent study suggesting that the cosmos has a distinct axis, along which the polarization of radiation is rotated more than it is in other directions (SN: 4/26/97, p. 252). If correct, that finding could overturn long-cherished notions about the birth and evolution of the universe.

In the latest reports, astronomers using newer, higher-quality data say they have failed to replicate the controversial finding, while theorists assert that they have identified fatal flaws in the statistical analysis performed by the original study's coauthors, Borge Nodland of the University of Rochester (N.Y.) and John P. Ralston of the University of Kansas in Lawrence.

Astronomers John F.C. Wardle of Brandeis University in Waltham, Mass., Rick A. Perley of the National Radio Astronomy Observatory in Socorro, N.M., and Marshall H. Cohen of the California Institute of Technology in Pasadena analyzed 26 galaxies and quasars--far fewer than the 160 galaxies, used by Nodland and Ralston. However, the new data consist of high-resolution images from the Very Large Array radio telescope in Socorro and the W.M. Keck Telescope on Hawaii's Mauna Kea.

The researchers used an easy-to-see reference--the alignment of jets of radio waves emitted by many of the quasars and galaxies in their study--to measure the polarization angle of radiation. Because polarization is generally thought to start out perpendicular to the jet direction, any extra rotation, or twist, in the polarization of light should be readily apparent, Wardle and his colleagues assert.

Their new data "directly refute" the presence of cosmological rotation, the researchers write. They add that Nodland and Ralston, who had to rely on lower-resolution images, calculated net polarization by averaging the polarization of radio waves from several different regions within a galaxy, a technique that provides "a very blunt tool for searching for systematic [rotation]."

Ralston disagrees, arguing that averaging minimizes the confounding effect of a galaxy's magnetic field, which also twists the polarization of the radiation the galaxy emits. Nodland adds that the extra twist which he and Ralston found along a particular direction in space is a statistical effect apparent only as an average over a large sample of galaxies distributed evenly across the sky.

The team's statistical analysis has also come under fire. Daniel J. Eisenstein of the Institute for Advanced Study in Princeton, N.J., and Emory F. Bunn of Bates College in Lewiston, Maine, assert that Nodland and Ralston did not properly test whether the cosmic twist they found is real. Sean M. Carroll of the University of California, Santa Barbara and George B. Field of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Mass., come to a similar conclusion.

Ralston says, however, that the statistical comparison recommended by these researchers would fail to detect a genuine twist.