I very much appreciated the Report by the committee of outside experts for Emory University, who did a mostly excellent job in sifting through evidence in an extraordinarily difficult situation. In general, the Committee are to be commended for their care and fairness. The Report, however, made a few minor errors. I would like to point out one set of them implicated recently by a comment on this thread about my discussion of Robert Churchill’s work on the state of arms of the 1746 Connecticut militia.
A. THE EMORY REPORT DISCUSSES THE 1746 CT MILITIA
Pages 14-15 of the Emory Report point in part to a discussion of historian Robert Churchill’s findings about the state of arms for the 1746 Connecticut militia, which are described on pages 2205-2206 of my Yale Law Journal review of Arming America. I reported on what Robert Churchill had told me about the records and what he had written in the September 2001 Reviews in American History.
I reported what I understood to be Churchill’s four conclusions:
1. Arming America falsely reported the CT militia as being 43% armed.
2. The records actually showed it to be 82% armed.
3. Arming America reported the state of arms of the worst-armed unit as the state of arms for the entire militia.
4. Arming America flipped the number of that worst armed unit from 57% armed to 57% unarmed.
My Yale review presented the last two points as somewhat speculative after saying “It is hard to know exactly what Bellesiles did, but he may just have [done 3 & 4 above] . . . .”
The Emory Report confirmed the first three of these four claims I reported, though whether the second claim is 81% or 82% depends on how one counts. But about the fourth claim, the report said, at p. 14-15:
[BEGIN QUOTE]
"In Arming America, he [Bellesiles] wrote that in preparation for an assault on Canada, "Connecticut finally raised its six hundred troops, 57 percent of whom did not have guns." [Arming America, p. 141] Our assistant confirmed the problems other scholars had originally noted. The primary sources Bellesiles cites confirm that in extant reports from company captains 368 of 456 men (80.7%) were armed. Our hunch is that Bellesiles skimmed the surface of these sources, relying instead on a passage from Harold Selesky, War and Society in Colonial Connecticut, "The volunteers in 1746 were not vagrants, although a few were `very poor,' and many enlisted without a blanket or a gun. In some companies as many as fifty-seven of the hundred men lacked a firearm." [New Haven and London: Yale University Press, 1990, 91.] Significantly, Selesky's "some companies" became for Bellesiles the whole. Selesky's own source is letter from Josiah Starr to Jonathan Law, August 6, 1746, which reported that he had a "full Compliment of Men Inlisted: all Clothed, & all of yt Company, Except fifty Seven provided, with a good Gun and are all Ready to move on first Notice." (Appendix B: 2. Militia Returns) Curiously, James Lindgren, too, seems to have misread Starr's letter. In his Yale Law Review article he claimed that Bellesiles had reversed the numbers, reporting 57% armed as 57% unarmed. In this case, Bellesiles' number was right."
[END QUOTE]
The Emory Report also includes a longer discussion in its attached memo from its research assistant (pp. 33-34):
[BEGIN QUOTE]
"Bellesiles claims that, when Connecticut attempted to build up its militia in preparation for an invasion of Canada, the colony managed with no small difficulty to raise 600 troops "57 percent of whom did not have guns (Arming America, 141)." James Lindgren, following the lead of Robert Churchill, has countered that the Connecticut militia was 81.7% armed, that Bellesiles considered only the least well-armed company out of five who made a return, and that he misread the one source he did use. According to Lindgren, 57% of the least well-armed militia company had guns; 43% were unarmed.
"The second part of Lindgren's contention is inaccurate. The inadequately armed company was under the command of Capt. Josiah Starr. In his 6 August 1746 letter to Gov. Jonathan Law, Starr wrote "I have Called my Company together & find I have my full Compliment of Men Inlisted: all Clothes; & all of ye Company, Except fifty Seven, provided with a Good Gun... (Connecticut Historical Society Collections, vol. 13, 276) "Since the six hundred men Connecticut was trying to raise were divided into six companies, it seems safe to assume that Starr's "full Compliment" consisted of one hundred men (Public Records of Connecticut, vol. 9, 211-14). If all "Except fifty Seven" were armed, then 57% were, indeed, unarmed.
"Lindgren is correct, however, in his assertion that Capt. Starr's company was exceptional. The Collections of the Connecticut Historical Society include four full and one partial return in the form of letters. Capt. James Church informed the governor that he had one hundred men, "But as to our Equipment we are not Compleet we want about twenty fire arms." Capt. Elisha Williams described his company as "compleat, well Cloathed, and furnished with Guns." Capt. Elisha Hall wrote "four men are not Cloathed & the rest will be ready to Embarque according to Order & Equipt as to Arms & Cloathing." Hall does not specify that his company was "complete," but neither does he mention any deficiencies in terms of the number of his men. Capt. Samuel Talcott's return was incomplete. He had enlisted 31 men, but "they want five guns and 3 Blankits have but two of them Hangers otherwise Expect they will be acquiped." Presumably by "hangers," Talcott meant that two of his men wavered in their commitment to enlist (OED, "hanger"). Talcott's First Lieutenant, William Smithson, also made a return which listed nineteen armed and six unarmed men (Connecticut Historical Society Collections, vol. 13, 269-76) .
…[research assistant computes by one method that (80.7%) were armed]
"If we assume, as Lindgren apparently did, that Talcott's two "hangers" decided to opt out of military service and his five men lacking guns were "acquiped" before the expedition left for Canada, then 81.7% of the militia was armed with 454 men were carrying 371 guns."
"Bellesiles's mistake might be understandable if he relied solely on Selesky's War and Society in Colonial Connecticut which he cites in his footnote along with selections from Records of Colonial Connecticut and Collections of the Connecticut Historical Society. Selesky states: "In some companies as many as fifty-seven of the hundred men lacked a firearm (Selesky, 91)." This is, of course, quite true albeit a little misleading. Bellesiles's statement that 57% of the Connecticut militia as whole was unarmed is, however, wrong."
[END QUOTE]
B. MINOR ERRORS IN THE EMORY REPORT
There are a number of minor errors in this part of the Emory Report.
1. The report says: “Curiously, James Lindgren, too, seems to have misread Starr's letter.” (p. 15)
The research assistant’s memo says: “Lindgren is correct, however, in his assertion that Capt. Starr's company was exceptional.” (p. 33)
Yet I never read, nor misread, nor cited, nor miscited Starr’s letter--nor did I ever mention Capt. Starr or his letter. In my Yale review, I was reporting on what Robert Churchill wrote and what I understood him to say to me. I cite Churchill as the authority in this section of the Yale review and cite no other primary or secondary evidence for my contentions, never mentioning Capt. Starr.
2. The assistant’s memo says, “Presumably by ‘hangers,’ Talcott meant that two of his men wavered in their commitment to enlist . . . . If we assume, as Lindgren apparently did, that Talcott's two "hangers" decided to opt out of military service . . . .”
There are two errors in this claim in the Emory Report. First, I did not read the word “hangers” in the CT records, because I did not read the original CT records, nor did I cite them or mention “hangers.” I explicitly relied on my reading and understanding of Churchill’s claims, whom I cited as my only source.
Second, a “hanger” is a sword, not a person—as some of the committee members may have known but missed in their review of their research assistant’s memo. The assistant cites the OED, but somehow misses the proper definition; a “hanger” is “A kind of short sword . . . .” This is not a recent discovery on my part; hangers were counted as bladed weapons in probate counts in my William & Mary Law Review article.
3. Robert Churchill informs me by email (no, I haven’t checked) that the research assistant’s assumption that Hall’s company was full is slightly in error. According to Churchill, the unit that the research assistant assumes had 100 men actually had 98 men in the detailed records he examined.
4. The research assistant’s memo (and to a lesser extent the Report itself) speculate that Bellesiles may have just used Selesky as his only source and then misread him, even though they note that Bellesiles cites the original records as well. But on another page of Arming America (AA, p. 101), Bellesiles describes the range of the state of arms for individual units in New England in 1746, saying that the highest level of arms were from Hartford and Providence, yet even these two well armed “towns reported those entirely unarmed numbering between 16 percent and 54 percent” (AA, p. 101). It is clear from this passage that Bellesiles looked (or claimed to have looked) at the reports for individual units of CT militia to determine which units had the highest counts, so the Emory Report’s speculation appears to be false on how Bellesiles could have made the error that Churchill identified, I described, and the Emory Report confirmed. Precisely how Bellesiles happened to make the error that we all agreed happened is something that my Yale review got half wrong, but the Emory Report’s possible explanation turns out to be very probably wrong as well, since Bellesiles claimed on p. 101 to have looked at the state of arms in individual CT units.
5. On this next point I am quibbling about misleading language rather than pointing out an actual error on the part of the authors of the Emory Report. The authors of the Emory Report are not confused on this point, but their chosen language would confuse most readers. Referring to fact that 57% were unarmed in the worst armed unit, the report stated, “In this case, Bellesiles' number was right.”
As the Emory Report makes clear, however, this number is not the right number for the state of arms of the Connecticut militia, which is what Bellesiles said the 57% applied to. It is the right number for the state of the worst armed unit, which is something Arming America offered no numbers for. So Bellesiles's number is not right as applied to what he applied it to (as the Emory Report makes clearer in other passages quoted above).
By analogy, if I were to say falsely that 79% of the probate inventories contained guns, when it turned out that 79% of the probate inventories contained clothes and only 54% had guns, I couldn’t defend by saying that the 79% “number was right.” Yet this is what the Emory Report does by saying that Bellesiles’s number is right, when it is not right for what Bellesiles applied it to—the state of arms for the entire CT militia. In short, the Emory Report said that Bellesiles incorrectly claimed that the 1746 CT militia was 57% unarmed, when Bellesiles should have said that it was 81-82% armed, but the 57% number would have been “right” if he had applied it to the worst armed unit--but his book didn’t apply it to just that one unit.
C. SUMMARY AND CONCLUSIONS
1. The Emory Report confirmed 3 of the 4 claims that I made about Churchill’s research on the 1746 CT militia. Arming America falsely reported the CT militia as being 43% armed, when it was 81-82% armed. The Emory Report rejected as erroneous part of the speculative explanation that I offered for this state of affairs. The Emory Report confirmed my and Churchill’s speculation that Bellesiles had used the state of arms of the worst armed unit as the count for the entire CT militia, but corrected the false statement in my Yale review that Bellesiles had flipped the state of arms in one unit from 57% armed to 57% unarmed.
2. In rejecting part of my story (which was based on my understanding of Churchill’s research) of how Bellesiles probably came to commit the error that we all agree he made, the Emory Report offered their own speculative explanation, which also turns out to be very probably wrong.
3. The Emory Report makes several small errors in describing the 1746 CT militia. It considers a “hanger” a person, when a hanger was a “sword,” an error that slightly affects their count of the state of arms. They assume that a unit has its full complement of 100 men, when Churchill reports that the background records show that the unit had only 98 men, another apparent mistake that slightly affects their count.
Most significantly for me, the Report claims that I misread documents that I never read, mentioned, cited, or claimed to have read, even attributing to me the research assistant’s own ignorance about the meaning of the word “hanger.” If I had counted the original 1746 CT records, I would have cited them, rather than citing just Churchill.
4. How did the error about the supposed flipping of the number from 57% armed to 57% unarmed arise? I got the incorrect idea that Bellesiles had flipped the number from my discussions with Churchill when we tried to figure out how Bellesiles had made the error that Churchill found, I reported, and the Emory Report confirmed. Either Churchill misspoke or (perhaps more likely) I misunderstood the details of his evidence. I then sent Churchill my draft review and specifically asked him to check the portions where I explicitly relied on him and cited him. For whatever reasons, Churchill did not see the error in the part of my manuscript where I described his research.
5. A Correction. On the last page of Justin Heather and my William & Mary Law Review article, we wrote: “Everybody makes mistakes (certainly we do).” When I read the Emory Report on its release at the end of October, I immediately contacted Robert Churchill to try to determine what happened. After a few days in which we discussed the matter and he reviewed the records, I immediately posted a very brief errata attached to the end of the copy of my Yale review article on my Northwestern website. I also sent the same copy to Glenn Reynolds, which he posted on his site. I also checked the link on HNN to the pdf version of my article and it pointed to my website version (with the errata).
When Churchill recounted his data in October, he changed his total from 82% to 81%, so I made that (unnecessary) change as well. To correct my Yale article, other than switching “57%” to “43%,” all I needed to delete is parts of two sentences. No other rewriting was necessary, since the point of that section of the review was that Bellesiles had reported substantially false data on the state of 1746 militia in CT, which the Emory Report confirmed. Here is my Oct. 29, 2002 Errata:
ERRATA: page 2206:
Line 5: Change “57%” to “43%”.
Line 7: Replace “,” with “.” and delete the rest of the sentence.
Lines 9-10: Omit “and flipping the results from armed to unarmed”.
Line 11: Change “2” to “1”.
6. What this Means. Given the secrecy in which the Emory Committee worked and the shortness of time, it is not surprising that they would make some small, innocent errors, most of which would have been caught in a more open scholarly process. Neither is it surprising that some of Arming America’s critics would make an occasional error—though it is instructive that even while pointing out this particular error involving the 1746 CT counts, the expert Emory Committee themselves actually made several more errors than the one they pointed out. I have to admit that I was a bit taken aback to read the Emory Report’s claims that I had misread specific documents that I never read, mentioned, or cited. But I just attributed these and other relatively minor Committee errors to their isolation, the time pressures they were under, and their prior unfamiliarity with the details of Arming America.
James Lindgren
Professor of Law
Northwestern University
Copyright by James Lindgren, 2002.
Not for republication except by permission.
Quotations from the Emory Report are as presented in Google’s HTML version.
by James Lindgren on December 20, 2002 at 2:35 AM