I'll let Dan Clore do most of the work here. This is not a direct refutation of Norberg, but someone else, so some of the remarks will make no sense, but takes care of the general question:
http://www.geocities.com/SoHo/9879/jamesd.html
"In actual fact, Herman and Chomsky do give all of these sources in their lengthier treatment of the same subject in After the Cataclysm, Volume II of The Political Economy of Human Rights [PEHR II]. There, far from vague references to "academics [Herman and Chomsky are] embarrassed to name", we find full citations. The Far Eastern Economic Review published a series of articles by its Southeast Asia correspondent, Nayan Chanda (in particular, one in the May 1977 issue). He concluded that the number of executions was "possibly thousands", not the much greater numbers being bandied about at the time. Far from being fired for this heresy, he has since become the Deputy Editor of the journal.
Nor was the work published in the London Economist that of a "communist", although it is in fact a letter (and is also used by Nayan Chanda in his studies in FEER). In fact, far from being written by a communist, the letter (March 26, 1977) was written by W.J. Sampson, a statistician and economist, who had worked for the CIA-puppet government of Lon Nol (which the Khmer Rouge overthrew) in the office of central statistics, and had published several technical reports on the economy of Cambodia. He based his conclusions on interviews with refugees and other eyewitnesses. He concluded that slayings (executions) numbered "in the hundreds or thousands rather than in hundreds of thousands" with a much larger death toll from illness and starvation.
As for the Melbourne Journal of Politics, it published analyses by noted Cambodia scholar Ben Kiernan, whom James Donald has frequently cited as a credible source. Checking his later work in book form, we find that it is fully consistent with Herman and Chomsky's use of it. Perhaps this explains why Kiernan has both aided Herman and Chomsky in their work on Cambodia, and come to their defense when others have made charges similar to those of James Donald.
Donald's argument thus falls flat on its face when subjected to a cursory examination. The very worst "lie" he has found, is that Herman and Chomsky did not take care to specify that one reference to the London Economist was to a letter rather than an article -- when they did take care to point this out when they gave a full citation of it in the very next paragraph of the same review!"