Arming America: How “the Controversy Arose”

As I described yesterday, in 2002 Emory University asked three outside scholars to investigate charges of “failures of scholarly care and integrity” against Michael Bellesiles, author of Arming America.

Those scholars were academic heavyweights: Stanley N. Katz of Princeton, Hanna H. Gray of the University of Chicago, and Laurel Thatcher Ulrich of Harvard. They had the assistance of a graduate student who visited archives in Massachusetts, checked other sources, and reran calculations.

That committee filed their report (P.D.F. download) in July. Emory University released it in October. On the same day, Bellesiles resigned.

In his interview last year with Daniel Gullotta for the Age of Jackson podcast, Bellesiles made some comments about that report and other criticism of his book. I decided to assess those remarks against the historical record.

Bellesiles told Gullotta:

The controversy arose because seventeen years ago, there was a flood in Bowden Hall at Emory University in Atlanta, which severely damaged the offices of numerous professors in the history and philosophy departments, including mine. Most of the original notes for my book Arming America were destroyed in that flood. And within days, opponents of the book picked up on this loss to argue that I had never conducted the research supporting three paragraphs in the book that concern probate records.

The sprinkler-pipe flood happened in April 2000, nineteen (not seventeen) years before this conversation. Arming America was published in early September 2000, so “opponents of the book” couldn’t have responded to the flood “within days” because the book didn’t yet exist. But of course we may not recall exact details of a difficult time.

Here’s the sequence of events as best I can recreate it. Bellesiles published a paper on gun ownership in early America in the Journal of American History in 1996. Its evidence included travel accounts and probate inventories. Clayton E. Cramer, a graduate student with whom Bellesiles had corresponded about gun laws, then wrote to the journal listing other travel accounts that contradicted the paper’s findings. Bellesiles replied by dismissing Cramer’s criticism as politically motivated.

Meanwhile, Bellesiles had agreed with the Knopf division of Random House to publish what became Arming America. The July 1999 Economist reported on the upcoming book. In December, Charlton Heston, president of the National Rifle Association, sniped at Bellesiles’s work. The editing and production process on the book must have also begun in 1999. That sprinkler pipe burst in April 2000, making news only in the Emory community. In that same month, the New York Times reported on Bellesiles’s intriguing conclusions.

Arming America was officially published in September 2000, receiving prominent and mostly positive reviews in the mainstream press. As early as 30 August, Prof. James Lindgren of Northwestern University wrote to Bellesiles with questions about his research since he’d been working on the same questions using probate inventories. On 19 September, Bellesiles sent Lindgren an email saying, among other things, that the office flood had destroyed his notes. That appears to have been the first link between the burst pipe and the probate data, and it came from Bellesiles himself. (Subsequently, the Emory committee found, Bellesiles made a “disavowal” of some other statements in those 30 August and 19 September emails to Lindgren.)

The first public mention of that flood’s effect on the debate that I’ve found was a draft of Lindgren and Justin Lee Heather’s essay “Counting Guns in Early America” dated 28 December. Some critics of the book were indeed skeptical of Bellesiles’s explanation about the loss of his probate data—some had to be convinced there even was an office flood. But Lindgren and others accepted, if only for argument, that Bellesiles had indeed counted probate records on yellow pads as he described and included that in their analyses of his work. That was sloppy technique and the numbers still didn’t add up, they said.

But that aspect of the book wasn’t where the “controversy arose” first. Cramer had objected to Bellesiles’s conclusions back in 1997. After the book appeared, Cramer expanded on his criticism, finding more omitted and distorted sources. As a software engineer, he used his expertise with computers to set up webpages sharing those findings. Unfortunately for the appearance of political leanings, Cramer located his pages within the website of the Golden Gate United National Rifle Association, making it easy for Bellesiles and his defenders to dismiss the complaints.

Cramer, as a graduate student in California, didn’t have the resources to try to replicate most of Bellesiles’s probate research in the east. But he found plenty of other details in Arming America to criticize. Lindgren and his team had already worked in some of those probate archives, so they could analyze what data Bellesiles reported and find discrepancies. Eventually formal reviews in scholarly journals voiced more doubts, though most didn’t appear until late 2001 or 2002, after Arming America had received the Bancroft Prize.

I’ve always been struck by how Lindgren’s critique carried much more weight than Cramer’s. According to Bellesiles in his interview with Gullotta:

Now, I think the reason they picked on the probate records is because those are the most obscure of all the materials I use, that pretty much require you to go to the individual archives in order to examine them. It’s not something that could easily be verified by going to a good research university library.

Except that Cramer found a lot wrong with Arming America by “going to a good research university library.” Bellesiles’s ongoing emphasis on the book’s small section about probate inventories gives the false impression that no one had found other problems with the book.

There are better explanations of why Lindgren’s criticism got more traction within the academic world than Cramer’s. Lindgren was a professor at Northwestern. Cramer was a graduate student at Sonoma State University. Lindgren wasn’t a proponent of gun ownership in contemporary America while Cramer was. Lindgren’s argument rested mostly (but not wholly) on numbers. Cramer’s critique was largely about words, which can seem more open to interpretation. But isn’t quoting words out of context just as inaccurate as reporting a false count of wills?

Whatever the reason, we can see that Emory University gave more weight to the Lindgren critique. All five of the questions it tasked the outside committee with examining involved “probate records” of some sort. Furthermore, the committee noted that its mandate covered “ONLY” those questions. (In Appendix B, Part 3, the research assistant did address discrepancies with travel narratives that Lindgren had noted, but disagreed with parts of his assessment.)

TOMORROW: Bellesiles’s comments on the committee’s conclusions.