May 5, 2001

COUNTING EARLY AMERICANS' GUNS

Michael Bellesiles' research on gun ownership in early America was never destined for obscurity.

Even before Arming America: The Origins of a National Gun Culture appeared last year, there was controversy over its central thesis. Bellesiles claims that guns were rare during the nation's formative years, and that America's gun culture is an invented tradition.

"The notion that a well-armed public buttressed the American dream," he writes, "would have appeared harebrained to most Americans before the Civil War."

The publisher pitched it as ammunition in the gun-control debate, with lavish blurbs such as: "This book will transform the modern gun debate" . . . "inescapable policy implications" . . . "a myth-busting tour de force."

It even won one of the three 2001 Bancroft prizes in American history.

Such prominence has brought Bellesiles' work intense, and increasing, scrutiny.

A significant strand in his argument is woven from probate records from colonial times through the period following the Civil War. He includes a table showing, for example, that for a sample of 40 counties from 1765 to 1790, only 14.7 percent of probate inventories listed firearms. Among more than 1,200 inventories for that period, he says, few are for women and none of those listed guns.

Last year James Lindgren, who teaches law at Northwestern University, and law student Justin Lee Heather analyzed several series of probate records. They found that from 50 percent to 73 percent of men's estates listed guns, and 6 percent to 38 percent of women's estates, depending on the series.

For one series of records from Providence, R.I., Bellesiles says that 90 of 186 inventories mention some form of gun, and more than half are evaluated as old and of poor quality.

Lindgren and Heather studied the Providence records, and by their count, 94 inventories of men's estates listed guns, not significantly different from Bellesiles' 90. But only 149, not 186, of the estates included records of personal property such as guns. And only 15 of 168 total guns were described as old or broken -- that's 9 percent, not "more than half."

Among 17 women whose inventories are in these records, only one had guns, but she had five.

Bellesiles says that gun ownership appears to be linked to prosperity, and Lindgren and Heather's analysis shows the same. The low levels of gun ownership Bellesiles found, if accurate, may reflect merely that few people could afford guns.

Still, these are remarkably large discrepancies, and they've begun to attract attention. The Wall Street Journal published a highly critical column April 9, citing the dispute over probate figures, among other problems.

Bellesiles' petulant letter in response sniffed that probate records are only a minor part of Arming America,accounting for just five paragraphs out of a 444-page book."

He is not, strictly speaking, correct about that. The index to his book lists eight references to "probate records" over 11 pages, including the statistical table I mentioned above, which is one of only 10 in the appendix. And at that, the index misses at least one important reference, a footnote citing the work of Alice Hanson Jones and other scholars who have studied probate records.

But the clear implication is that this is all a great deal of fuss about very little. And that's wrong too. The probate evidence is important to his thesis that guns were rare and unimportant in America's formative years. If it is significantly in error, that's like the clock that strikes 13; not only is it wrong in itself, it casts doubt on everything that came before.

Bellesiles also says in his letter to the Journal that in his tally he did not use the 919 inventories from 1774 included in Jones' 1977 book, American Colonial Wealth.

Curious, that. Arming America draws upon an article Bellesiles published in the Journal of American History in September 1996. For probate records from the years 1765 to 1790, the article included the same table and graph that appear in the book. In 1996, he wrote that he had studied more than 1,000 inventories and that "integrating Alice Hanson Jones's valuable probate compilation into this general study" revealed a startling distribution of guns in early America. How do you integrate records into a study without using them?

But the reason he now gives for excluding the Jones inventories is even curiouser. "Jones's (sample) is from 1774 to 1775, when the provincial governments were handing out guns, and thus may lead to an overcounting of firearms in private hands."

Leave out data from years that might disprove your thesis? A strange way to conduct research.

Bellesiles' actual data are not currently available to be checked; he has said he kept his tally on paper, and his notes were ruined when water pipes broke in the building where he had his office. But without them, no one can tell whether Bellesiles' study tolls the death knell of the Second Amendment right to keep and bear arms -- or is just striking 13.

(810 words)