After taking another look at the Stephen Davis piece, I realized he goes off the rails even before the out-of-context quote I addressed earlier. As his standard for what constitutes a war crime, he cites the
Nuremburg Charter, which states that the tribunals created by that charter have jurisdiction over three classes of crimes, including:
(b) WAR CRIMES: namely, violations of the laws or customs of war. Such violations shall include, but not be limited to, murder, ill-treatment or deportation to slave labor or for any other purpose of civilian population of or in occupied territory, murder or ill-treatment of prisoners of war or persons on the seas, killing of hostages, plunder of public or private property, wanton destruction of cities, towns or villages, or devastation not justified by military necessity;
This is not a good source for determining what is a war crime because the list of war crimes is intended to clarify the jurisdiction of the tribunals, not to specify standards for conviction. The standard for conviction was whether the accused violated the laws or customs of war as those were understood at the time the alleged crimes were committed.
For example, Davis claims that when Sherman's army built huts using wood belonging to civilians, that was "plunder" and hence a war crime. But I don't believe that seizing supplies for ones troops has ever been considered to be a war crime. The "plunder" considered by the Nuremburg tribunals consisted of seizing stuff and shipping it to Germany.
The writers of the Nuremburg Charter didn't spell this out because it was the job of the judges to figure out what was and was not a war crime. All the writers of the Charter were doing was ensuring that
if Germans had seized stuff in violation of the laws of war, the tribunals had the authority to punish the offenders.
In short, despite his reference to Nuremburg, Davis isn't applying the standards used by the Nuremburg tribunals, or any other recognized statement of international law, when he declares Sherman a war criminal.