Fall From Grace: Arming America and the Bellesiles Scandal (Part 2)Historians/History
II. THE BOOK
A. What Is a Gun Culture?
Arming America claims that we did not have a gun culture before the
Civil War, but that we have had one since then. There is an obvious
conceptual problem with this thesis: What would it mean to have—or not
have—a gun culture? It is hard to judge the truth of this claim without
deciding on what a gun culture is. Bellesiles gives us some hints of what he
means, but he never clearly states his criteria. This is an unfortunate way to
frame the inquiry. Cultural analysis is not an all-or-nothing proposition.
America had one form of gun culture in the late eighteenth century, it had
another form of gun culture in the late nineteenth century, and it has another
Although Bellesiles never defines what he means by having a gun
culture, he puts great store in owning guns, familiarity with guns, and the
prevalence of guns in popular culture—such as in magazines, television,
and movies. If having a gun culture requires gun-lover magazines and
violent film and television crime stories (or the contemporary equivalent),
then we have a gun culture today, but did not two centuries ago. If, instead,
having a gun culture means growing up in households with guns, learning
how to shoot them, widespread participation in military training where guns
are used, and using guns as a tool (such as for vermin control), then we
definitely had more of a gun culture in the eighteenth century than we do
An analogy to horse-riding might be helpful. If one examines
familiarity with horses and the use of horses, there was obviously much
more of a horse culture in the eighteenth century than there is today. But if
one measures a horse culture by the expressed sheer love of horses, the
romance of the cowboy on horseback, magazines about riding, and the
variety of games and competitions involving horses (racing, rodeos, polo,
off-track betting, newspaper odds, and so on), there is probably more of a
horse culture today—even though very few people ride. I would say that we
had more of a horse culture in early America, but it was different in kind:
Then, horses were more important as tools and as transportation, rather than
as objects of recreation, love, and fetishism.
It would be more accurate to say that we have a different form of gun
culture today than we did in the eighteenth century. It is not even obvious
how useful the concept of a gun culture is. It is more important to
understand the claims that give meaning to Bellesiles’s concept of a gun
culture—how many guns there were, what condition they were in, where
they were stored, who owned them, how much they cost, how accurate they
were, how they were used, and what they meant to their owners.
In perhaps the strongest part of the book, Bellesiles describes the
marketing savvy of Samuel Colt,43 who helped create the romance of the
gun with the advertising campaign for his revolver pistol in the two decades
before the Civil War. In the mid-nineteenth century, guns became mass produced,
much easier to load between shots, and more lethal. Bellesiles
also shows how the outlaws and legends of the American West—the James
Gang, Buffalo Bill, and many others—first learned their craft in the Civil
War and its precursor in Kansas. If Bellesiles had confined his argument to
describing a switch from simpler guns manufactured one at a time to more
sophisticated mass-produced guns, and from a gun culture in which guns
were a tool to one in which guns were an object of romance, then he
probably would have encountered little dispute.
What made the book such a sensation was his description of guns in the
seventeenth, eighteenth, and early nineteenth centuries. He claimed that
guns were exceptional rather than common, in poor condition even in
private hands, not stored in the home but rather in central armories, too
expensive to be owned outright by most men, and restricted by law to the
Protestant upper and middle classes. None of this is true.
B. How Common Was Gun Ownership?
The most contested portions of Arming America involve the book’s
most surprising claim, that guns were infrequently owned before the mid-
1800s. As I show below, the claim that colonial America did not have a gun
culture is questionable on the evidence of gun ownership alone. Compared
to the seventeenth and eighteenth centuries, it appears that guns are not as
commonly owned today. Whereas individual gun ownership in every
published (and unpublished) study of early probate records that I have
located (except Bellesiles’s) ranges from 40% to 79%; only 32.5% of
households today own a gun.44 This appears to be a much smaller
percentage than in early America—in part because the mean household size
in the late eighteenth century was six people,45 while today it is just under
two people.46 The prevailing estimate of 40% to 79% ownership differs
markedly from Bellesiles’s claim that only about 15% owned guns.47 In the
remainder of this Section, I explain why.
1. The Gun Censuses
Bellesiles bases his claims of low gun ownership primarily on probate
records and counts of guns at militia musters.48 He also discusses censuses
of all guns in private and public hands, but on closer examination, none of
these turns out to be a general census of all guns.
The trend is set in Bellesiles’s first count of guns in an American
community—the 1630 count of all the guns in the Massachusetts Bay
Colony of about 1000 people. Bellesiles’s account is quite specific: “In
1630 the Massachusetts Bay Company reported in their possession: ‘80
bastard musketts, . . .  Fowlinge peeces, . . . 10 Full musketts . . . .’
There were thus exactly one hundred firearms for use among seven towns
with a population of about one thousand.” 49 If you go to the pages of the
Records of Massachusetts Bay cited by Bellesiles, however, you find that
this list of guns was something quite different. It was not a list of guns
owned by freemen or the company “in their possession” in America, or
even a list of guns owned by the company in England. Rather, as stated on
page 2 of volume 1 of the original handwritten records, it is a list of
“Armes ffor 100 men” that the company wanted to ship over to America.50
On the previous page, page 1, there is a list of “Apparell ffor 100 men.” 51
The pages record their early plans for the trip, even before they got their
charter. They planned to have clothes and arms for each and every man.
This list of 100 guns for 100 men is no more an inventory of all the
guns for 1000 people actually in the Massachusetts Bay Colony than the list
of apparel for 100 men is a list of all the colony’s clothes. It is just not true
that the other 900 residents were unarmed nudists. On the contrary, the list
indicates that every man should be both clothed and armed.
Quite suspiciously, the date is wrong—Bellesiles cites the date of the
list as 1630, rather than 1628-1629 as in the original cited text.52 Had
Bellesiles listed the date correctly as 1629 (or 1628 in the old calendar),
careful scholars would have suspected that it was not a list of guns in the
Massachusetts Bay Colony, because the government and most of the people
of the colony did not come to America until 1630. If he had made only one
of two errors, either error would have been plain to a sophisticated reader.
By making two errors (both the substance and the date) rather than one,
they would both escape notice—unless someone checked the source (as did
Clayton Cramer originally).
Other sources confirm that gun ownership in Massachusetts Bay was
high. According to surviving probate records from Essex County,
Massachusetts, from 1636 to 1650, 71% of male estates owned guns, as did
25% of female estates.53 Somehow plans in England to arm each and every
man—100 guns for 100 men—are turned by Bellesiles into a nonexistent
census of guns actually “in their possession” in the colony, showing only
10% of the colonists as being armed—thus fitting his general claim that few
Americans were armed.
There are other “gun censuses” from which Bellesiles reports data.
Robert Churchill, who has analyzed them closely, describes problems with
one of them:
The Provincial Congress of Massachusetts conducted another
census in 1775. According to Bellesiles, the returns showed “that
there were 21,549 guns in the province of some 250,000 people.”
Here again, the records describe something different. The
Provincial Congress asked town officials and militia commanders
to “take an exact state of their numbers and equipments” of the
“ several companies of their regiments.” This was, in other words, a
census of the arms in the hands of the militia. The exact size of the
associated militia is not reported, but it is unlikely that it greatly
exceeded 30,000 men. Thus, 70 percent of the Massachusetts men
who joined the armed political movement to nullify the Coercive
Acts possessed arms.54
Contrary to Bellesiles’s claims, this was not a general gun census, but rather
a count of guns in the hands of the militia, which might also have excluded
many guns not suitable for militia service.
Churchill also describes similar problems with Bellesiles’s
characterization of federal gun censuses, such as the census of 1803.55
Although none of Bellesiles’s gun censuses turns out to be a gun census of
all military-style arms owned by each citizen, Robert Churchill has located
a few actual gun censuses of men in the 1770s in several Rhode Island and
New Hampshire towns. These few extant New England gun censuses
suggest that gun ownership was slightly higher than the percentages
generally observed in New England probate records.56
2. The Militia Counts
Bellesiles tells many stories of militia gunlessness. But these stories are
often unsupported by the sources that Bellesiles cites as evidence.57 Robert
Churchill offers the following example in his review of Arming America in
Reviews in American History: Bellesiles describes the problems that
Connecticut faced in its efforts to raise troops for an invasion of Canada in
1746. Bellesiles wrote that “Connecticut finally raised its six hundred
troops, 57 percent of whom did not have guns.” 58 Churchill discovered that
Bellesiles had switched the numbers around: “In the records he cites, . . .
four of ten companies reported the state of their arms, and a fifth gave a
partial return. Of the 454 men covered by the returns, 371 (81.7 percent)
were armed.” 59 Of the five units reporting their arms, two were 100%
armed and the worst armed of the other three was 57% armed. It is hard to
know exactly what Bellesiles did, but he may just have seized on the
number of the worst armed unit and reported that number for all units, but
only after flipping it to 57% unarmed. By misleadingly counting the worst armed
unit as the entire company and flipping the results from armed to
unarmed, Bellesiles is able to make a very well-armed Connecticut militia
(82% armed) appear to be a mostly unarmed militia (43% armed).
Elsewhere, Churchill offers other instances. For example, Bellesiles
discusses a 1744 return of militia arms from Worcester County,
Massachusetts. He claims that four companies were “Intirely Deficient” 60
in their firearms, when all they lacked was ammunition.61
Consider another story of militia gunlessness told by Bellesiles:
When news of Lexington reached New Haven, Benedict Arnold
inspected his troops and found them largely unarmed. He
threatened to break into the town arsenal in order to arm his men,
but the town’s selectmen relented and opened the doors to his
militia, with Arnold supervising the distribution of Brown Besses.62
The source that Bellesiles cites tells a different tale: “In New Haven, the
enthusiasts were not thwarted, although Benedict Arnold had to threaten to
break open the powderhouse before town leaders supplied his volunteers
with ammunition.” 63 The striking story of Benedict Arnold’s men lacking
guns (as opposed to ammunition) and of Arnold himself distributing Brown
Besses appears to have been invented.64 Bellesiles then uses this story to
show that even the best-armed colonies such as Connecticut “faced a
shortage of firearms from the very first day of the conflict.” 65
Both of the last two examples show a persistent problem with
Bellesiles’s accounts—he repeatedly reports evidence of a lack of
ammunition as a lack of guns. Bellesiles thus creates the impression that the
sources he describes support his stories of gunlessness.
There are also serious methodological problems with Bellesiles’s main
militia arms data over time.66 Bellesiles presents his Massachusetts gun
militia data as if they were counts of all privately owned guns in
Massachusetts, which they were not.67
First, Bellesiles confuses absence from the annual muster with
gunlessness. If half of the adult men showed up at muster and they were
90% armed, Bellesiles would infer that only 45% of the adult male
population of the colony as a whole was armed. This would make sense
only if every man who did not appear at muster did not own a gun. One
would expect two sorts of people to fail to show up—older or sicker men,
who would be likely to have had substantial experience with guns earlier in
their lives, and wealthier men, who were both more likely to risk the fine
for skipping muster and more likely to own guns.
Further, Bellesiles confuses arms produced at militia musters with arms
owned. There were many guns that would have been suitable for shooting
birds (“ fowling pieces” ) or vermin, or for hunting larger animals, that
would not meet the standards of the day for battle muskets, which were
very heavy with extremely long barrels. It is somewhat akin to confusing an
M-16 with a shotgun. In addition, the average family size in the North was
six people in 1790.68 Households with more than one adult male might have
had only one gun or only one military-style gun, and, as a result, one or
more men in that household would show up unarmed in Bellesiles’s data.
Last, Bellesiles anachronistically compares gun ownership to the
general population, a fairly obvious interpretive “life cycle” error. With
average family sizes of six,69 most women and children would have lived in
a household with guns. By comparing gun ownership to the general
population, boys who would grow up to own guns as frequently as their
fathers are counted as not owning guns. Instead of comparing his
percentages to the number of households, he dilutes his percentages with
children, counting white male children who would grow up to own a gun as
nonowners.70 To take such an individualistic approach in the presence of
such large family sizes is the kind of anachronistic move that one would not
expect a historian to make. That would be like comparing home ownership
today to the general population and counting children who live in homes
owned by their parents as not being homeowners—or even worse,
computing fertility rates by including men and little children in the base.
3. The Probate Records
The dispute over the probate records, which has been the primary topic
in the public debate for the last year, is essentially settled. Four scholars—
Robert Churchill in Reviews in American History,71 Randolph Roth in
William and Mary Quarterly,72 and Justin Heather and I in the William and
Mary Law Review73—confirm serious errors in Arming America and
confirm each other’s counts. Certainly, in most fields, that would settle the
matter (until new data surfaced). The only other scholars who questioned
our probate data were unable to explain their conclusions and have backed
away from them.
Probate inventories are appraised lists of assets at death. They were
used to disclose property available for creditors, to achieve any necessary
title-clearing, and to ensure a proper distribution of assets among the
members of the large families that prevailed in early America.74 In an article
forthcoming as of this writing,75 Justin Heather and I compare the relative
frequency of gun ownership in these inventories to the presence of other
commonly owned items. As for the methodology of drawing inferences
from probate records, we suggest that the ownership of any item of interest
should be compared to the ownership of other commonly owned items,
since probate inventories are inherently incomplete.76
Gun ownership was particularly high compared to ownership of other
common items. For example, in 813 itemized male inventories from Alice
Hanson Jones’s 1774 national database, 54% of estates listed guns,
compared to only 30% of estates listing any cash, 14% listing swords or
edge weapons, 25% listing Bibles, 62% listing any book, and 79% listing
Guns are thus more common than Bibles in several databases that
Heather and I examined. Further, guns are generally found in roughly as
many probate estates as books of any kind, a finding suggesting that guns,
like books, were very commonly owned by early American families. Based
on the 1774 probate records, the frequency of gun ownership (54% of male
estates, 50% of both male and female estates combined) was roughly
midway between the ownership of any coins or other money (about 30% of
male estates) and the ownership of clothes (about 79% of male estates).78 If
gun ownership really was about two-thirds of the level of clothes ownership
(and about five-thirds of the level of cash ownership), then gun ownership
was roughly as common as was generally thought to be the case before
Arming America was published. Contrary to Arming America’s claims
about probate inventories in seventeenth- and eighteenth-century America,
there were high numbers of guns, guns were much more common than
swords or other edge weapons, women in 1774 owned guns at a rate (18%)
higher than Bellesiles claimed men did in 1765-1790 (14.7%), and 83-91%
of gun-owning estates listed at least one gun that was not old or broken.79
For the probate data from Providence, Rhode Island (1678-1726),80
Bellesiles misclassified over 60% of the inventories he examined.81 He
repeatedly counted women as men, counted guns in about a hundred wills
that never existed, and claimed that the inventories evaluated more than half
of the guns as old or broken when fewer than 10% were so listed.82 Heather
and I found that nationally, for the 1765-1790 period, the average
percentage of estates listing guns that Bellesiles reported (14.7%) is not
mathematically possible given the regional averages he reported and known
minimum sample sizes.83
Bellesiles argued that guns were rarely listed in probate inventories—
according to him, only 14.2% of 1200 frontier inventories in the 1765-1790
period included guns, and 53% of the guns were explicitly listed as broken
or otherwise defective.84 To support this claim, Bellesiles has put a report
on his website that recounts frontier estates from Vermont, where four of
his six frontier counties are located. Bellesiles finds only forty-five estates
listing guns, missing seventy estates with guns altogether.85 Among his
forty-five estates, he also misses several pistols. Further, he misreports the
description of several guns, making them appear to be in worse condition
than they are listed.86 He misses all of the twenty-six gun estates in Windsor
County, even though Windsor County is in his sample.87 He misses every
gun estate in Rutland County from 1786 through 1790.88 He claims to count
records in the Gloucester County courthouse in Chelsea, Vermont, when
there is no Gloucester County or Gloucester County courthouse.89 The
courthouse in Chelsea, Vermont, is the Orange County courthouse, but
Bellesiles misses all five gun estates in its records during the period,
assuming these are supposed to be in his sample.90 Bellesiles gets one of the
locations of the Windsor County records wrong—there are none in the town
of Windsor.91 Last, fewer than 15% of the guns, not 53% as he lists for
frontier counties in 1765-1790, are listed as broken or defective.92
Bellesiles’s responses to criticisms of his probate data have been
inadequate. In the paperback edition of the book, he has quietly dropped all
of the originally challenged claims from Providence, Rhode Island, without
acknowledging his previously published errors.93
Justin Heather and I have analyzed part of Bellesiles’s nineteenth century
probate data and are finding the same disturbing pattern that exists
in Bellesiles’s data for the previous two centuries. In particular, in his Table
1,94 Bellesiles reports gun counts for forty counties, including San Francisco
County. In correspondence95 and in a report on his website from February
through early September, 2001, Bellesiles claimed to have examined the
San Francisco probate records at the San Francisco Superior Court.
Repeated inquiries to the San Francisco Superior Court have all yielded a
version of the same answer: They do not have the probate records that
Bellesiles claimed to have counted there because they were destroyed in the
1906 San Francisco earthquake and fire.
Representatives of the History Center at the San Francisco Public
Library, the Bancroft Library of the University of California, the Sutro
Library, the Family History Center Libraries, and the California
Genealogical Society agree that they know of no surviving runs of San
Francisco probate inventories for the years Bellesiles claimed to have
counted—1849-1850 and 1858-1859—because (as most note) they were
destroyed in 1906.96 Kathy Beals, an author who has written a book on pre-
1906 San Francisco probate records,97 reports that a list of the names of
those who left wills from the 1850s exists, but that there are no known runs
of inventories or property lists.98 A few scraps of other probate records exist
from 1880 through 1905, but nothing of substance before 1880.99 Rick
Sherman, the Research Director of the California Genealogical Society in
Oakland, California, confirmed the unanimous belief that such records do
not exist.100 Bellesiles has repeatedly stated that he used only complete runs
of inventories, not a few inventories discovered here or there, as did Alice
Hanson Jones in her study of New York probate records.101
In January 2002, Bellesiles publicly claimed on Emory’s Academic
Exchange to have located some of the long-lost San Francisco inventories
from the 1850s in the Contra Costa County History Center in Martinez,
California.102 Bellesiles claimed that the staff did not even know that they
had any probate inventories, even though, as the staff points out, they are
part of the core of the collection.103 He also supplied copies of these
supposed San Francisco inventories to journalists. I have reviewed these
documents and the original files from which they were copied; there is
nothing in them to suggest that they are San Francisco County estates.
Several documents that Bellesiles copied clearly reveal themselves to be
Contra Costa County estates. The staff of the History Center has reviewed
Bellesiles’s claims carefully and concluded that every estate he found was a
Contra Costa County estate. I have confirmed their conclusion. In the
original files, there are well over a hundred documents establishing that
these are Contra Costa estates. There are many petitions to and orders of the
Contra Costa County Probate Court. There is not one petition to or order of
the San Francisco Probate Court. Further, the staff casts serious doubt on
Bellesiles’s claim to have done substantial work in their archives before
recently.104 Emory University’s history department was so embarrassed by
Bellesiles’s claims that it sent a letter apologizing to the Contra Costa
County History Center for Bellesiles’s comments.105
Neither part of Arming America’s study of seventeenth- and eighteenth-century
probate data is replicable, nor is Bellesiles’s study of probate data
from the 1840s and 1850s. In terms of paragraphs, the probate study is only
a small part of the book—about twelve paragraphs in the text discuss the
probate evidence, plus textual footnotes and the entire page of data in Table
1.106 Yet it is the most dramatic and potentially persuasive evidence he
offers. The probate data are the only data purporting to show systematic
changes in gun ownership over long periods of time (1765-1859), a crucial
part of Arming America’s central claim that gun ownership was very low in
the seventeenth and eighteenth centuries and grew gradually in the few
decades before the Civil War. Further, the probate data are by far the most
important evidence purporting to show that guns in private hands were
mostly in poor working condition.
Moreover, it would not be proper simply to omit a discussion of probate
data now that it is clear that they undercut the conclusion of Arming
America—that would be the suppression of contrary evidence. One may
speculate what the book might have been without the probate data, but it is
not possible to ignore the fact that this important body of evidence exists.
The patterns in the actual probate data from colonial America are
potentially devastating to Arming America’s central arguments. The fact
that gun ownership was much higher in the seventeenth and eighteenth
centuries than Bellesiles claims it was on the eve of the Civil War renders
the main story in Arming America incoherent. If guns were already more
common in the eighteenth century than Bellesiles says they were on the eve
of the Civil War, then his narrative of how we got from low gun ownership
to high gun ownership collapses into the opposite story of a shift from high
gun ownership to somewhat lower gun ownership.
Also potentially devastating to the arguments in Arming America are
the conditions of guns in probate records. In every database Heather and I
have looked at (including the databases Bellesiles cites in Arming America),
at least 83% of estates with guns have guns that are not listed as old or in
poor working condition.107 A more coherent story would have been that
America went from fairly ineffective guns to fairly effective mass-produced
guns, but that is not Bellesiles’s main story; more to the point, such a story
would have been largely uncontroversial.
The importance of the probate data is suggested in the reviews and
press accounts: the New York Times (“Mr. Bellesiles’s principal
evidence” ),108 the Washington Post (Bellesiles’s “freshest and most
interesting source” ),109 the New York Review of Books (“ The evidence is
overwhelming. First of all are probate records.” ),110 the New Republic
(“ [T]he core of his argument depends on statistics: government censuses of
militia members and a sample of probate records . . . .” ),111 and Reason
(Bellesiles’s “main proof for the absence of firearms” ).112
Bellesiles himself emphasized probate records when he summarized his
argument in a November 3, 1997, interview with the Emory Report:
“ ‘Contrary to the popular image, few people in the United States owned
guns prior to the 1850s,’ Bellesiles said. ‘Probate and militia records make
clear that only between a tenth and a quarter of adult white males owned
In articles on Arming America both in law reviews and especially in the
popular press, Bellesiles’s evidence from probate records was the single
most commonly mentioned source of quantitative evidence supporting his
thesis. Scholars have quickly made use of Bellesiles’s undercounts of guns
in probate records to support their views of the Second Amendment.114
Thus, while the probate data are discussed on only about thirteen pages
in the book,115 they are recognized by some reviewers as the single most
important class of evidence among the many classes of evidence that
Bellesiles discusses. Admittedly, others put more weight on this evidence
than does Bellesiles. Not surprisingly, Bellesiles and his supporters are now
claiming that the probate data are relatively unimportant.116 Yet without the
probate data, his book runs the risk of falling into the genre that Bellesiles
has called “dueling quotations.” 117 One cannot just wish the probate data
away; it points strongly against the main narrative of Arming America.
43. BELLESILES, supra note 3, at 377-83.
44. This results from my analysis of the March 2001 release of the National Opinion
Research Center’s General Social Survey, 2000 [hereinafter 2000 NORC GSS]. The data are also
available at Nat’l Opinion Research Ctr., General Social Survey, at http://www.icpsr.umich.edu/
GSS/ (last visited Apr. 8, 2002). According to the survey, 32.5% of households owned any gun,
19.7% owned a rifle, 18.6% owned a shotgun, and 19.7% owned a pistol or revolver. 2000 NORC
GSS, supra. Only 1.2% of respondents refused to respond to the question. Id.
45. Inter-Univ. Consortium for Political & Soc. Research (ICPSR), Census Data for the Year
1790, http://fisher.lib.virginia.edu/cgi-local/censusbin/census/cen.pl?year=790 (last visited Aug.
46. 2000 NORC GSS, supra note 44.
47. BELLESILES, supra note 3, at 445 tbl.1.
48. Bellesiles emphasized probate records when he summarized his argument in a November
3, 1997, interview with the Emory Report: “‘Contrary to the popular image, few people in the
United States owned guns prior to the 1850s,’ Bellesiles said. ‘Probate and militia records make
clear that only between a tenth and a quarter of adult white males owned firearms.’” Michael
Terrazas, Bellesiles Lays Blame for U.S. Gun Culture at the Feet of Samuel Colt, EMORY REP.,
Nov. 3, 1997, http://www.emory.edu/EMORY_REPORT/erarchive/1997/November/ernovember.
49. BELLESILES, supra note 3, at 63.
50. 1 RECORDS OF THE GOVERNOR AND COMPANY OF THE MASSACHUSETTS BAY IN NEW
ENGLAND 26 (Nathaniel B. Shurtleff ed., AMS Press 1968) (1853).
51. 1 id. at 23-24.
52. 1 id. at 25-26.
53. In the earliest years of those estates, 1636-1650, Justin Heather and I counted sixty-one
probate inventories—all but two of which were sufficiently itemized to be used. Fully 25% of the
eight female inventories had guns. Among the fifty-one itemized male inventories, 71% contained
guns. Lindgren & Heather, supra note 24 (manuscript at 66 n.178) (citing 1 PROBATE RECORDS
OF ESSEX COUNTY, MASSACHUSETTS, 1635-1664, at 3-130 (George Dow ed., 1916)).
54. Churchill, supra note 39, at 333 (citations omitted).
55. Id. at 333-34.
56. Robert H. Churchill, Gun Ownership in Early America as Reflected in Manuscript Militia
Returns (Sept. 2001) (unpublished manuscript, on file with author).
57. The most extensive work on this problem has been done by Robert Churchill. See
Churchill, supra note 39; Churchill, supra note 56.
58. BELLESILES, supra note 3, at 141.
59. Churchill, supra note 39, at 333 (citation omitted).
60. BELLESILES, supra note 3, at 150; infra note 61.
61. Posting of Robert H. Churchill, email@example.com, to H-OIEAHC@h-net.msu.edu
(Sept. 19, 2001) (copy on file with author). Churchill wrote:
Bellesiles cites a 1744 militia return from Worcester County, Massachusetts. He claims
that 8 of 21 companies that “filed a report on their firearms” reported that they were
“ entirely deficient.” In the original document the colonel of the regiment reported the
state of the arms and ammunition of each company. He noted that four of the
companies were “entirely deficient as to arms.” He reported the other four as “entirely
deficient as to ammunition.” Bellesiles has thus altered the language in the original to
advance his thesis of gun scarcity.
Id. (citation omitted).
62. BELLESILES, supra note 3, at 181.
63. HAROLD E. SELESKY, WAR AND SOCIETY IN COLONIAL CONNECTICUT 228-29 (1990);
see Churchill, supra note 61 (discussing this source).
64. Churchill, supra note 61.
65. BELLESILES, supra note 3, at 181.
66. See id. at 447 tbl.3.
67. Churchill, supra note 61.
68. See supra note 45 and accompanying text.
69. See supra note 45 and accompanying text.
70. See BELLESILES, supra note 3, at 447 tbl.3.
71. Churchill, supra note 39.
72. Roth, supra note 41.
73. Lindgren & Heather, supra note 24.
74. For more on probate inventories, see 3 JONES, supra note 23, at 1847-60; and McGaw,
supra note 22, at 339-43.
75. Lindgren & Heather, supra note 24.
76. Id. (manuscript at 16-21 & tbl.2, 28-29 tbls.3-4).
77. Id. (manuscript at 28 tbl.3).
78. Id. (manuscript at 25, 28 tbl.3).
79. Compare id. (manuscript at 25 & n.62, 28 tbl.3, 42 tbl.8, 49), with BELLESILES, supra
note 3, at 445 tbl.1.
80. 6, 7 & 16 EARLY RECORDS OF THE TOWN OF PROVIDENCE (Horatio Rogers et al. eds.,
Providence, Snow & Farnham City Printers 1894-1901).
81. Lindgren & Heather, supra note 24 (manuscript at 48-49 & nn.84-94).
83. Id. (manuscript at 51-54 & nn.105-13).
84. BELLESILES, supra note 3, at 13, 266-67, 445 tbl.1. This statement appears to be false. A
preliminary analysis of complete data from four of his six frontier counties and partial data from
the other two counties suggests that fewer than 15% of 1765-1790 frontier estates with guns list
only old, broken, or dysfunctional guns, and fewer than 15% of the guns listed are old or
dysfunctional. See James Lindgren & Justin Heather, Vermont Data File, 1770-90 (Feb. 1, 2002)
(unpublished data, on file with author).
85. Compare Michael A. Bellesiles, Vermont Probate Records, 1770-1790 (Oct. 12, 2001), at
http://www.emory.edu/HISTORY/BELLESILES/, with infra Appendix, Section L (listing
86. See infra Appendix, Section K.
87. See infra Appendix, Section L.
88. See infra Appendix, Section L (collecting data from book II of the Rutland District
manuscript probate records).
89. Bellesiles, supra note 85.
90. Lindgren & Heather, supra note 84.
91. Bellesiles, supra note 85.
92. Compare BELLESILES, supra note 3, at 266-67, with Lindgren & Heather, supra note 84.
93. Compare MICHAEL BELLESILES, ARMING AMERICA: THE ORIGINS OF A NATIONAL GUN
CULTURE 109-10 (Vintage Books 2001) (2000), with BELLESILES, supra note 3, at 109-10.
94. BELLESILES, supra note 3, at 445 tbl.1.
95. In correspondence with me on November 30, 2000, Michael Bellesiles wrote that he
examined the records for San Francisco at the San Francisco Superior Court, a claim repeated in
an essay on using probate records that was on his website from February 2001 through mid-
September 2001. E-mail from Michael Bellesiles to author (Nov. 30, 2000) (on file with author).
96. Telephone Interviews with various librarians, History Center at the San Francisco Public
Library, Bancroft Library of the University of California, Sutro Library, and Family History
Center Libraries, and with Rick Sherman, Research Director, California Genealogical Society
(July 7, 2001 through Sept. 10, 2001); E-mail from Rick Sherman, Research Director, California
Genealogical Society to author (July 9, 2001) (on file with author).
97. KATHY BEALS, SAN FRANCISCO PROBATE INDEX, 1880-1906: A PARTIAL
98. E-mail from Kathy Beals to author (July 10, 2001) (on file with author); E-mail from
Kathy Beals to author (July 11, 2001) (on file with author).
99. E-mail from Kathy Beals to author, supra note 98.
100. E-mail from Rick Sherman to author, supra note 96.
101. See Odyssey with Gretchen Helfrich (WBEZ radio broadcast, Jan. 16, 2001),
http://www.WBEZ.org/services/ram/od/od-010116.ram; Posting of Michael A. Bellesiles,
firstname.lastname@example.org, to H-OIEAHC@h-net.msu.edu (Jan. 9, 2001) (copy on file with author).
102. Michael Bellesiles, Emory Academic Exchange (Jan. 22, 2002), at
http://www.emory.edu/ACAD_EXCHANGE/2002/decjan/whatsnew.html; see Betty Maffei,
Notes on Supposed San Francisco Records in the Contra Costa County Historical Society History
Center, at http://www.cocohistory.com/frm-news.html (last updated Jan. 27, 2002).
103. Maffei, supra note 102.
105. See Ron Grossman, Emory Can Wait No Longer: Historian Is Under Investigation, CHI.
TRIB., Feb. 13, 2002, at C5 (describing the apology from James Melton, chair of Emory’s history
106. BELLESILES, supra note 3, at 13, 74, 79-80, 109-10, 148-49, 262, 266-67, 386, 445 tbl.1.
107. See Lindgren & Heather, supra note 24 (manuscript at 25, 28 tbl.3, 42 tbl.8, 49);
Lindgren & Heather, supra note 84.
108. Ramirez, supra note 29.
109. John Whiteclay Chambers II, Lock and Load, WASH. POST, Oct. 29, 2000, at X2.
110. Morgan, supra note 36, at 30.
111. Jackson Lears, The Shooting Game, NEW REPUBLIC, Jan. 22, 2001, at 30, 32.
112. Malcolm, supra note 38, at 48.
113. Terrazas, supra note 48.
114. See, e.g., Michael C. Dorf, What Does the Second Amendment Mean Today?, 76 CHI.-
KENT L. REV. 291, 312 (2000); Robert E. Shalhope, To Keep and Bear Arms in the Early
Republic, 16 CONST. COMMENT. 269, 274 (1999); Koren Wai Wong-Ervin, The Second
Amendment and the Incorporation Conundrum: Towards a Workable Jurisprudence, 50
HASTINGS L.J. 177, 184-85 (1998).
115. See supra note 106 and accompanying text.
116. See Michael A. Bellesiles, Arms and the Ancestors, WALL ST. J., Apr. 4, 2001, at A25;
Kevin R. Hardwick, Colloquy, CHRON. HIGHER EDUC., Feb. 22, 2002, at http://chronicle.com/
colloquy/2002/guns/183.htm; Posting of Chris Waldrep, email@example.com, to H-LAW@
h-net.msu.edu (Dec. 12, 2001) (copy on file with author); Posting of Jack Rakove to H-LAW@
h-net.msu.edu (Apr. 18, 2001) (copy on file with author).
comments powered by Disqus