It appears that even noted professors at Hahvihd with fancy vocabulary are not immune from logical errors.
Prof. Stephen Rosen is worried about Ahmadinejad's recent rhetorical escalation against Israel. He was worried that it might be a precursor to a chemical/biological weapons attack by Iran against Israel. So he checked some of his old research and found that, sure enough, when states use bio/chem weapons against their adversaries, it's preceded by dehumanization and rhetorical escalation!
In other words, he saw X and worried it might lead to Y. Sure enough, all instances of Y are preceded by X! Which of course says nothing about whether a) X causes Y, or b) whether or not there are 1000 Xs for every Y.
Ironically "silent evidence" is a key theme in Nassim Taleb's "The Black Swan", originally recommended to me by a former (current?) colleague in the Office of Net Assessment of Dr. Rosen's.
It would be interesting (and tedious and exhausting) to examine all instances of rhetorical escalation by states/armed groups, and see how often it actually leads to war crimes or the use of bio/chem weapons.
For what it's worth, I think Ahmadinejad's rhetoric is due to a) the assassination of Hezbollah's operations chief Imad Mughniyah, and b) his falling domestic approval ratings.