Although making liquid nitrogen-temperature superconductors is easy enough that high school science projects already feature them, researchers still have little idea how the new ceramic oxides work and therefore little guidance for improving them. At the International Workshop on Novel Mechanisms of Superconductivity, held from 22 to 26 June in Berkeley, California, theorists reviewed a host of competing explanations of how these materials come by their remarkable properties, but they could not, get far in sifting through the candidates for the best one. One cause of the unsettled situation is that theorists have not yet pushed their models far enough to make many specific predictions about physical properties and therefore to provide a reason to choose one theory over another. But experimental data for comparison with theory are lacking, too. For example, experimentalists are just now succeeding in being able to grow single crystals and thin films of the ceramic oxide superconductors, whose properties were shown at the workshop to be highly anisotropic. Measurements already made on the polycrystalline sintered material available up to now are difficult to interpret and therefore need to be repeated on good-quality crystals and films, where the variation of properties with crystallographic orientation can be mapped out. Given the high level of Japanese activity in the field, it was surprising that no researchers from industrial laboratories in Japan presented their findings at the workshop. In the light of a budding international competition in commercializing superconductors, some American scientists interpreted the absence as an attempt to protect proprietary advances. A more pleasant surprise was the attendance of a delegation of six Soviet scientists, although one of the fathers of superconductivity theory, Vitaly Ginzburg of the P.N.Lebedev Institute of Physics in Moscow, who was expected, did not come.