"There are reasons to think that earthquakes just might not be predictable without knowing far more than we'll ever know about the stresses deep in the Earth," Vidale said.
Perhaps more surprisingly, even our understanding of what areas are at most risk of earthquakes is extremely limited, Northwestern's Stein told LiveScience. For example, no one expected that the section of fault that ruptured to cause Japan's horrific 9.1-magnitude Tohoku quake in March 2011 could result in a quake that large. The maximum was supposed to a magnitude 8, Stein said. Quakes are measured on a logarithmic scale, so a magnitude 9 quake has 10 times the amplitude and about 31 times more energy release than a magnitude 8, according to the U.S. Geological Survey (USGS).
The underestimate of the fault proved deadly, as Japanese seawalls were built under the assumption that extra-large quakes wouldn't produce any extra-large tsunamis. While a magnitude-8 quake might cause a 32-foot (10-meter) tsunami, Stein said, a magnitude-9 could cause a tsunami twice as tall. [Album: Monster Waves]
Japan is not the only spot that Earth's vibrations have been underestimated. Seismologists predicted less shaking in the 2010 Haiti earthquake than actually occurred. And a deadly magnitude-7.9 earthquake in Wenchaun, China, occurred in a spot previously rated as low-risk.
A big part of the problem, Stein said, is that the Earth moves on a different schedule than the human life span. Seismic records only stretch back 100 years, and human writing a few thousand years past that. Stein and his colleagues looked at the seismic records and more than 2,000 years of written records in northern China and found that, during that timeframe, no magnitude-7 or greater quake ever hit at the same place on a fault more than once.
"Every time there's a big earthquake, it's on an area that hasn't been active for 2,000 years," Stein said.
In other words, if Italian scientists are criminally liable for bad predictions, wouldn't all seismologists be just as criminal for their imperfect predictions?
"There's sort of a pattern here," Stein said. "Nobody knows how to do this very well. Countries have large programs to make hazards maps … these things are often big failures. Given that, the case for criminalizing it seems very small."
Could it happen here?
In the United States, the legal system would likely agree with Stein. According to Adam Kolber, a law professor at Brooklyn Law School, the Italian case would be very unlikely to go forward in the U.S.
First of all, for a manslaughter conviction, the experts would have to have what's called mens rea, Kolber told LiveScience. That means that they would have to be aware of a substantial and unjustifiable risk that their statements would cause someone's death.
Secondly, Kolber said, you'd have to prove that the statements directly caused someone's death.
"You have to find some particular person for whom if they were told there is a substantial risk of an earthquake that they would have left town or something like that, and that's going to be hard to show," Kolber said.
Finally, First Amendment freedom of speech rights might prevent prosecution.
"To the extent that they're giving their scientific opinions, there's a First Amendment interest in protecting the speech," Kolber said.
Closing down science communication
Scientists contacted by LiveScience said they weren't personally concerned about prosecution for sharing their scientific opinions with the public, though some said they worried about a chilling effect on scientific openness in Italy.