Note how much of the questionable research occurred in Vermont.

http://www.weeklystandard.com/Content/Public/Articles/000/000/000/911xuovl.asp

The Historian Who Couldn’t Shoot Straight
From the February 25, 2002 issue: The truth about Michael Bellesiles’s
“Arming America.”
by David Skinner
02/15/2002 6:00:00 PM

MICHAEL BELLESILES is a professor of history at Emory University. When
his “Arming America: The Origins of a National Gun Culture” appeared in 2000, it came wrapped in a yellow strip of paper printed with four blurbs–one from the Pulitzer Prize-winning historian Michael Kammen, who called the book “a classic
work of significant scholarship with inescapable policy implications.”

Kammen was right about the implications. Although a work of colonial
and pre-Civil War history, “Arming America” spoke directly to recent debates about gun control. Arguing that no American “gun culture” existed
before 1850 or so, Bellesiles marshaled a variety of sources to show
that guns were much rarer, significantly less useful, and far more regulated than previously believed. He recently told a reporter he is actually a
longtime gun enthusiast, but in his book’s introduction he took dead
aim at Charlton Heston and the NRA.

And why not? If no absolute, presumptive right to own a gun existed
back when the Second Amendment was written, then no such right
exists today.

That’s why, when the book first appeared, its reviews were practically
love letters. In a cover story for the
New York Times Sunday book-review section, Garry Wills said “Arming
America” had “dispersed the
darkness” by showing that privately owned firearms in America were
“barely in existence” before the Civil
War. In the Los Angeles Times, University of Colorado professor Fred
Anderson hailed Bellesiles’s
“intellectual rigor” and “thorough scholarship,” calling the book a
“brief against the myths that align freedom
with the gun.”

Even the more critical reviews proclaimed “Arming America” an
important, scholarly achievement. In the
New York Review of Books, Edmund S. Morgan declared that Bellesiles
“may have overstated his case, but
only a little. He has the facts.” Rutgers professor Jackson Lears wrote
a 6,700-word review for the New
Republic in which he took issue with the author’s cultural history but
assured readers Bellesiles had written
a “debunking counter-narrative the old-fashioned way, by means of
exhaustive research.” Wesleyan
University professor Richard Slotkin in the Atlantic Monthly called the
book a “groundbreaking study,”
praising its “stringent quantitative analysis.” In April 2001, Columbia
University gave “Arming America” the
Bancroft prize, the preeminent award for history writing.

Even as the celebrations continued, however, a handful of scholars
began to challenge Bellesiles’s
research. Questions of fact usually lost in the small-print of
appendices and endnotes have been dragged
into the light of day. And as a result, a much-praised book has been
exposed as sloppy, inaccurate, and
possibly fraudulent. Meanwhile Bellesiles’s career hangs in the
balance.

MUCH of the controversy has been generated by small skirmishes. Take,
for example, the misquotation of
the Militia Act of 1792 on page 230 of “Arming America,” where
Bellesiles makes it seem that Congress was
responsible for supplying militias with guns, instead of the members
bringing their own weapons. That’s not
a small matter: The exact provenance of the militias’ guns is a central
concern in “Arming America.” Caught
out, Bellesiles wrote in an article in the Organization of American
Historians newsletter that he’d accidentally
quoted the 1803 amendment to the Militia Act.

Unfortunately, the passage doesn’t match up with the 1803 text either.
And there’s another problem:
Bellesiles thanked a listserv editor named Ian Binnington for notifying
him of the mistake, while Binnington
says it was a man named Clayton Cramer who discovered the error.
Indeed, Binnington says Bellesiles had
to have known that, for he responded to Cramer’s message reporting the
error.

It’s hard to imagine that Bellesiles mistook Cramer for anyone else,
since Cramer was the first critic on the
scene after Bellesiles published the 1996 journal article that became
the basis for “Arming America.” Cramer
has posted on his website a three-hundred-page refutation of
Bellesiles, along with copies of historical
documents. “I’m one of those weird little people who look up
footnotes,” says Cramer, a software engineer
and a frequent contributor to Shotgun News.

Cramer is a surprisingly good writer, with a master’s degree in
history, but, he says, “Most journals are not
interested in hearing from me.” Still, he has his backers. Don Hickey,
a historian who peer-reviewed
Bellesiles’s 1996 article, remarked recently on a historians’ Internet
discussion group: “When I checked
some of the sources that he claimed Bellesiles had misused, I concluded
that Cramer was right.” Another
scholar adds that “if Cramer says something is in the record, it’s in
there. If Michael Bellesiles says
something’s in there, it may or may not be.”

So, for example, in chapter three of “Arming America,” Bellesiles wrote
that colonial legislatures “strictly
regulated the storage of firearms, with weapons kept in some central
place, to be produced only in
emergencies or on muster day, or loaned to individuals living in
outlying areas. They were to remain the
property of the government.” Jump to page 472, where you’ll find a
formidable endnote listing nineteen
sources and forty-two citations in support. “I have only checked
seventeen of the nineteen sources,” says
Cramer in an e-mail. Those two “might, by some miracle, match his
claims. But the other seventeen aren’t
even close to being right.” To prove his point, Cramer has posted
copies of documents from eleven of the
seventeen sources on his website. Connecticut, for example, appears to
have supplied the militia only with
gun powder and ammunition, while the members were expected to have
their own guns.

Much of the argument about “Arming America” concerns probate evidence.
Bellesiles has expressed shock
that records he discussed for only a few paragraphs would garner so
much attention. But those paragraphs
aren’t exactly filler. The probate evidence should have taken months,
if not years, and much travel for
Bellesiles to gather, and none of his reviewers considered it
insignificant. Garry Wills discussed Bellesiles’s
probate numbers before any other finding. Edmund S. Morgan wrote, “The
evidence is overwhelming. First
of all are probate records.” And this material supplied Bellesiles’s
most startling empirical findings: From
1763 to 1790, only 14.7 percent of American men owned guns, and only
slowly over several decades did
that number rise to 32.5 percent in the late 1850s.

ENTER James Lindgren, a professor of probate law at Northwestern
University. Lindgren decided to look
into Bellesiles’s probate evidence after reading an online dispute
about the book’s plausibility. “I had no
idea there would be errors,” Lindgren says. He asked Bellesiles what
data he had, so he “could formulate a
reasonable request” for a sample. The “Arming America” author wrote to
Lindgren that he counted guns in
probate records by keeping a tally on legal pads, and that these legal
pads had been compromised by a
flood in his office. The revelation of Bellesiles’s counting method was
almost as damaging as the fact that he
couldn’t make his work available for replication. Making hash-marks on
a legal pad, as Ohio State historian
Randolph Roth softly puts it, “isn’t the state of the art.”

Lindgren and his co-author Justin Heather proceeded to test
Bellesiles’s probate findings against databases
compiled by other scholars and original inventories around the country.
If Bellesiles’s thesis of rising
ownership were correct, they should have found a percentage around or
even lower than 14.7. “Guns are
found,” Lindgren and Heather concluded after their examination, “in
about 50-73 percent of the male estates
in each of the seven databases.” Of course, different databases use
different criteria for counting, and
probate records are far from comprehensive in listing property. Add to
this variations in spelling,
bookkeeping, and legal custom, and you can explain a generous margin of
difference. But Bellesiles’s
national average for the years 1765 to 1790 is nowhere near the range
Lindgren and Heather found.

When Lindgren and Heather narrowed their focus and attempted to
replicate Bellesiles’s findings for
Providence, Rhode Island, for the years 1679 to 1726, they found 63
percent of adult male estates listed
guns, while Bellesiles found only 48 percent. Bellesiles claimed to
have counted 186 adult male wills when
there were only 149. “More than half of these guns are evaluated as old
or of poor quality,” wrote Bellesiles.
Lindgren and Heather found only 9 percent so listed–and their count
has since been verified by others.

As went Rhode Island, so did Vermont. Reading the 312 probate records
from 1770 to 1790 that Bellesiles
cites, Lindgren and Heather found that there were 289 male estates, 115
of which listed guns–for an
ownership rate of 40 percent. And this was a conservative count, says
Randolph Roth, who confirmed the
numbers on microfilm. So how did Bellesiles come up with his figure of
14.4 percent? On his
university-hosted website, Bellesiles lists names and gun descriptions
but does not reproduce the 312
probate inventories, so one can’t actually verify his count. “He won’t
share his data,” Roth says, not for the
first time.

ALTHOUGH his complaints are as devastating as anyone’s, Roth has some
nice things to say about the
author of “Arming America.” Calling him “a gifted writer,” he says he
still thinks highly of Bellesiles’s earlier
book on Ethan Allen. Indeed, Roth first became involved in the
controversy when Bellesiles wrote him to ask
for help in answering his critics. He wrote back, offering to go to
Vermont to check probate records there and
asking for a list of the records that had been consulted. Bellesiles
didn’t take Roth up on the offer and
provided no lists.

Since then, Roth has investigated and found previously unreported
errors in “Arming America” concerning
homicide rates. In an article for the William and Mary Quarterly–the
latest issue of which is a much-touted
showdown between Bellesiles and his critics–Roth notes Bellesiles’s
claim that “In forty-six years, Plymouth
colony’s courts heard five cases of assault and not a single homicide.”
The cited source, Records of the
Colony of New Plymouth in New England, actually covers fifty-eight
years, and the endnote in “Arming
America” doesn’t say which forty-six years Bellesiles has in mind.
Nonetheless, Roth found mention in those
records of eleven murders and four suspicious deaths from 1633 to 1691.
(My own examination confirmed
Roth’s count of eleven cases, though not all were clearly murders.)
Answering Roth in the William and Mary
Quarterly, Bellesiles partly acknowledges the error, but then quickly
retreats to his more general argument
that murder was nonetheless rare during those years. But, as Roth and
others point out, “Every error in
representation and reporting tends in the same direction.”

WORSE was yet to come. On page 353, “Arming America” reads, “During
Vermont’s frontier period, from
1760 to 1790, there were five reported murders (excluding those deaths
in the American Revolution), and
three of those were politically motivated.” The endnote refers the
reader to records at the county courthouse
in Rutland, Vermont. But, Roth says, there are no records at the
courthouse before 1778, which is when
Vermont came into existence–and the court records from 1782 to 1790
have been missing for a hundred
years.

Gay Johnson, the clerk of the Superior Court in Rutland County,
Vermont, confirmed that they lack the

volumes for 1782 to 1790. Although the court does not have a complete
index for all their old records,
correspondence between her office and another scholar mentions that
those volumes are not in the court’s
holdings. (She added that if anyone should know, it’s Randolph Roth:
“He’s done more work here in the
twenty-five years I’ve been here than anyone. . . . He’s been to every
inch of this building.”)

This is not the only time Bellesiles has named records that don’t
exist. One of the counties cited as a part of
Bellesiles’s national sample is San Francisco. In correspondence with
Lindgren and on his website,
Bellesiles reported that he looked at records for 1849 to 1850 and 1858
to 1859 at the San Francisco
Superior Court. But the court said it has no such records; the entire
archive before 1860 was destroyed in
the 1906 earthquake. Bellesiles later changed his story, several times
actually, before claiming that he had
seen those San Francisco records at the Contra Costa County Historical
Society instead. But the society
itself has denied this, posting on its website the fact that it has
only Contra Costa probate records.

MORE THAN anything else, the missing California records, discovered by
Lindgren in July and reported in
September simultaneously by Melissa Seckora in National Review Online
and David Mehegan in the
Boston Globe, threw the academic dispute about “Arming America” into
high gear. Other publications picked
up the story as some scholars and book reviewers began expressing
doubt–including Books & Culture
editor John Wilson, who recently retracted the positive review he’d
written. In November, Emory University
demanded a full accounting of Bellesiles.

Still, others are standing by him. In December the New York Times
reported that Columbia University dean
Jonathan Cole had distributed “documents detailing Mr. Bellesiles’s
mistakes” to the Bancroft prize jurors,
but spokesman James Devitt insists the university still believes it was
correct to award the prize to Bellesiles.
Controversy, Devitt explains, is common in scholarly debate and there
is “nothing new” in this controversy.
Asked who the jurors are, Devitt insists that information is “private,”
but that they are people who “definitely
have an expertise in these areas.”

Wesleyan University historian Richard Slotkin says he hasn’t followed
the controversy, but he would still
describe “Arming America” as “groundbreaking”–although he adds, “A
groundbreaking study is one that
opens up a subject in a way or from an angle that hasn’t been done
before.” Jackson Lears says, “I would
not have characterized his research as exhaustive,” if he had the
review to do over again. But then again,
Lears goes on, “I was basically reviewing it as a cultural historian.”
Not that he doesn’t think all of this is
serious: “If all of these charges can be shown to be true, there may be
reason to expect the Bancroft prize to
reevaluate its decision.”

Carl Bogus, associate professor at Roger Williams University School of
Law, strikes a different note. A
prominent gun-control activist and one of Bellesiles’s earliest
defenders, Bogus faced off with Joyce Lee
Malcolm in the Texas Law Review. (Malcolm was one of the first
scholarly critics of “Arming America,” and in
her January 2001 review of the book for Reason magazine, she raised the
first general warning that the
book was myth-making, not myth-busting.) Bogus now sounds less
combative. “The nub of the most serious
charges is that he has concocted data or deliberately distorted data,”
Bogus says in a phone interview.

“The critics,” he adds, “have made a powerful prima facie case”–but “I
am a lawyer, and I’ve seen many
cases evaporate when you hear the other side.” He believes however that
the burden of proof is now on
Bellesiles. “I consider Michael Bellesiles a friend, and maybe that’s
one of the reasons I’m willing to grant
him a little extra time here. And I hope he prevails. But the truth
will out.” Bogus agrees he has a
responsibility to reach a judgment on the matter. “I guess it is
because I feel some responsibility as a
participant in this debate and someone on Bellesiles’s team.”

BELLESILES’S essay in the William and Mary Quarterly is not the first
time he’s responded to his critics.
Last month, in the Chronicle of Higher Education, he referred to them
as a “jihad of technical nitpickers.”
Around the same time, he told the Atlanta Journal-Constitution the 14
percent ownership he found for the
period of 1765-90 should be “more in the neighborhood of 22-24
percent,” which would be a surprising and
major correction to his earlier numbers, but is still well below what
the existing literature suggests is likely.

Many scholars seem to be reserving judgment until they’ve read the
William and Mary Quarterly forum.
Officials at Emory University have now read the forum and recently
announced an investigation into
Bellesiles’s research. Out of the four essays on Bellesiles’s book,
only one is positively disposed–and then
only on theoretical grounds concerning interpretation of the Second
Amendment, something Bellesiles gave
little attention to in “Arming America.”

The article by University of Colorado historian Gloria Main slams
Bellesiles for his “failure to lay out his
critical methods for perusal” and for ignoring her own count of guns in
six counties of Maryland in 1650 to
1720. Main had concluded about 76 percent of male estates contained
guns. Inexplicably, Bellesiles found
7 percent. Ira Gruber, a history professor at Rice University, picks up
the argument on the militia end:
“Bellesiles’s treatment of the militia is much like that of guns; he
regularly uses evidence in a partial or
imprecise way.” In his own essay, Roth concludes that “Arming
America”‘s claims about gun ownership are
“not supported by the sources Bellesiles cites, the sources he does not
cite, or by the data he presents.”

But without the Lindgren and Heather paper or the arguments of Clayton
Cramer, the William and Mary
Quarterly issue can hardly be called a point-by-point debate between
Bellesiles and his critics. Several
major issues go unaddressed, particularly the missing California
records. Roth mentions the mythical
Vermont court records, but he does not come straight out and accuse
Bellesiles of fabricating a source.

Even Bellesiles’s response is hardly definitive. Much of it is
self-dramatizing and tangential. He spends nine
pages pondering the relative importance of probate records, before
condescending to address a selection of
criticisms. He does not volunteer explanations for apparently
nonexistent sources; except for the
male-female mix-up, he stands by his Rhode Island counts; he seizes on
small ambiguities to explain
mountain-sized differences. Basically, he is unrepentant.

What may be most revealing, however, is the mood of Bellesiles’s essay.
He opens by quoting an old joke
about the last words of a man going before the firing line: “I am
honored by all this attention.” And he closes
with a quotation from Jacob Burckhardt: “Such indeed is the importance
of the subject, that it still calls for
fresh investigation, and may be studied with advantage from the most
varied points of view. Meanwhile, we
are content if a patient hearing be granted us.”

The first quotation shows evidence of self-conscious candor. The
second, unfortunately, more than hints at prevarication.

David Skinner is an assistant managing editor at The Weekly Standard.