Arming America

[Nuclear Unicorn]

4. That fact that outright fraud cannot be established does nothing to diminish the fact

the reviewing committee stated they could find no evidence to support Bellesiles' research and could not replicate his results.

Evaluating Table One is an exercise in frustration because it is almost impossible to tell
where Bellesiles got his information. His source note lists the names of forty counties, but
supplies no indication of the exact records used or their distribution over time. After
reviewing his skimpy documentation, we had the same question as Gloria Main: "Did no
editors or referees ever ask that he supply this basic information?"

Professor Bellesiles’ critics have charged him with claiming to work with records that do
not exist. The San Francisco issue has been widely discussed. His critics have charged
that he has fraudulently claimed to have read records that do not exist because the San
Francisco records he claims to have used were destroyed in 1906. He has responded by
saying that he at least thought he was using San Francisco records and has provided
examples of several files from Contra Costa County that contain references to San
Francisco.

If Professor Bellesiles did indeed read Contra Costa records believing they were from
San Francisco, then the issue could again be one of extremely sloppy documentation
rather than fraud. There are three aspects of this story, however, that raise doubts about
his veracity.

a. He didn’t accept the opportunity to go find the San Francisco records until a friend
suggested he may have found them in Contra Costa. So the idea that he had confused
the origins of the records seems to have come from outside. In addition, there is some
question as to whether the records he now cites could indeed be ones that he had read
in 1993.

...

b. The records he selected and photocopied from that Contra Costa archive were hardly
random, but explicitly chosen because they had the words “San Francisco” in them,
even though the records themselves clearly identify them as deriving from the Contra
Costa court.

c. The records he selected do not seem to provide the sort of information his project
requires. They may be California records. They may bear the name “San Francisco”
somewhere in the files, but they do not appear to be detailed inventories of personal
property. The Welsh inventory includes only livestock and wheat, and the Crippen
only livestock and a wagon. These do not seem to be appropriate sources for
determining either the presence or absence of guns.

The issue with Table 3 seems to be less the existence of these annual militia censuses,,
however, than with Professor Bellesiles’ claim that they represent an accounting of
"Private Gun Ownership in Massachusetts." As a number of scholars have pointed out,
such censuses were actually counts of guns brought to the annual muster. [Robert
Churchill review, AA 00382-AA 00390, James Lindgren review, AA 00339]

The criticism aimed at Professor Bellesiles’ handling of militia data is in some respects
different from those we encountered with probate records. The issue is not whether the
records exist. In most cases, Professor Bellesiles’ footnotes are both extensive and
accurate. The problems have to do with his reading and interpretations of these records.
Robert Churchill and others have noted both his failure to provide appropriate context
and his often misleading use of statistics. Still, there are similarities to the problems we
found with probate records. His lumping together of guns and ammunition in his
discussion of Benedict Arnold’s march on the powderhouse in 1776 is reminiscent of his
conflating of wills and inventories. His attempts to estimate ratios of guns to militia sizes
show the same lack of skill with quantification as his probate computations.

While this certainly constitutes
sloppy scholarship, it does not prove a deliberate attempt to mislead, however misleading
the result.

But in one respect, the failure to clearly identify his sources, does move into the realm of
“falsification,” which would constitute a violation of the Emory “Policies.” The
construction of this Table implies a consistent, comprehensive, and intelligible method of
gathering data. The reality seems quite the opposite.

The most egregious misrepresentation has to do with his handling of the more than 900
cases reported by Alice Hanson Jones. When critics pointed out that Jones’ data
disagreed with his, Bellesiles responded by explaining that he did NOT include Jones’s
data in his computations because her inventories, taken during the build-up to the
American revolution, showed a disproportionately high number of guns! Here is a clear
admission of misrepresentation, since the label on column one in Table One clearly says
"1765-1790." If Professor Bellesiles silently excluded data from the years 1774-1776, as
he asserts, precisely because they failed to show low numbers of guns, he has willingly
misrepresented the evidence. This, compounded with all the other inconsistencies in his
description of his method and sources and the fact that neither he nor anyone else has
been able to replicate any part of his data, suggest that there is a real discrepancy
between the research Professor Bellesiles did and his presentation of that research in
Table One.

Question 5. Did professor Bellesiles engage in "other serious deviations 'from accepted
practices in carrying out or reporting results from research'" with respect to probate
records or militia census records by:

(a) Failing to carefully document his findings;
(b) Failing to make available to others his sources, evidence, and data; or
(c) Misrepresenting evidence or the sources of evidence."

We have reached the conclusion with reference to clauses “a” through “c,” that Professor
Bellesiles contravened these professional norms, both as expressed in the Committee
charge and in the American Historical Association’s definition of scholarly “integrity,”
which includes “an awareness of one’s own bias and a readiness to follow sound method
and analysis wherever they may lead,” “disclosure of all significant qualifications of
one’s arguments,” careful documentation of findings and the responsibility to “thereafter
be prepared to make available to others their sources, evidence, and data,” and the
injunction that “historians must not misrepresent evidence or the sources of evidence.”

In summary, we find on Questions 1 and 2, that despite serious failures of and
carelessness in the gathering and presentation of archival records and the use of
quantitative analysis, we cannot speak of intentional fabrication or falsification. On
Question 3, we find that the strained character of Professor Bellesiles’ explanation raises
questions about his veracity with respect to his account of having consulted probate
records in San Francisco County. On Question 4, dealing with the construction of the
vital Table One, we find evidence of falsification. And on Question 5, which raises the
standard of professional historical scholarship, we find that Professor Bellesiles falls
short on all three counts.

Anyone who continues to believe Bellesiles does so as an article of faith. They WANT to believe him but they have no basis in fact for doing so.