by C. Leon Harris

            The short answer to the above question is, “Yes.” A slightly longer answer is that probably fewer than 4% of pension applicants made intentionally false statements about officers, engagements, and other details of importance to historians. The complete answer lies within a dark tale of dozens of worthy old soldiers robbed of their pensions and honor by dishonest agents, an incompetent District Attorney, and a feckless Pension Commissioner.


            The tale begins early in 1834 when the Pension Office of the War Department became aware of suspicious applications under the Pension Act of 1832 coming out of Lewis County in what is now West Virginia. The Secretary of War selected a newly-minted US District Attorney for the Western District of Virginia, Washington G. Singleton, to investigate the applications. Having left his young family in Winchester Va., Singleton arrived in Lewis County in June 1834 and spent next several weeks riding through the rugged Allegheny Plateau interviewing every applicant for and recipient of a pension. In the days before telephones and when most men were illiterate his visits were unexpected, so the men had no chance to review what they had said in their pension applications. Singleton swore them in and took statements of their ages and service for comparison with what had been claimed months before.

            Singleton reported to the Pension Commissioner on 88 pension applicants in Lewis County, and he was able to reach a conclusion in 71 of them. Out of those 71 cases he judged that only 8 applicants were entitled to pensions. The other 63 cases he judged to be fraudulent, usually because the applicant told him he had done less service than declared in his pension application, had never served as a soldier, or had served after the Revolutionary War. Most of them had applied because pension agents convinced them they were entitled and tricked them into putting their marks on false declarations.

            Singleton referred to the crooked pension agents as the Lewis Speculating Gentry. There were about ten of them – lawyers, justices of the peace, and other leading citizens. One had just been elected to the Virginia legislature. The Lewis Speculating Gentry also offered their services in other counties of what is now West Virginia, creating an impressive total of 131 known applications. Singleton made a judgment on 107 of these applications, deeming 97 to be fraudulent. Details of the Lewis Speculating Gentry and Singleton’s investigation are in my appendix to the pension application of David W. Sleeth (S6111). Transcriptions of this and the other pension applications cited here can be found at the Southern Campaigns of the American Revolution pension transcription site (

            Many of the declarations of service produced by the Lewis Speculating Gentry went on for pages in elaborate and convincing detail. Some of the authors could probably have made more money writing military histories rather than pension applications. Often after transcribing an application I was surprised to see in Singleton’s report that it was a complete fabrication. I had to wonder how many other pension applications had fooled me. Could pension applications be trusted at all?


            Singleton continued his investigation until September 1835 with only a month off to confer with the Pension Commissioner in Washington. He interviewed applicants throughout most of what is now West Virginia, far beyond the grasp of the Lewis Speculating Gentry. His conclusions in these cases were only slightly reassuring. Of the 264 cases not tainted by the Lewis Speculating Gentry Singleton made a judgment in 161, and in 91 of those judgments were that the application was fraudulent. It was an improvement over the record of the Lewis Speculating Gentry, but 56% fraud is still intolerable. If that proportion were representative of pension applications in general, they would be useless to historians.

            But there was a striking difference in Singleton’s reports on these supposedly fraudulent cases compared with the frauds perpetrated by the Lewis Speculating Gentry: most gave a narrative of service that agreed in great detail with what was claimed in their pension applications months before. This would hardly be expected if the pension applications had been fabricated.

            Why had Singleton judge these claims to be fraudulent? In a large number of cases he gave no reason. After admitting that he had been “unable to procure any evidence touching” the application of Thomas Smith (S15989), for example, Singleton nevertheless concluded that it was fraudulent. Singleton stated that John Brookover (S5300) had “but a small portion of mind,” but he nevertheless took his sworn testimony and concluded “he never done one particle of service in my opinion.” In this and many other cases, the allegation of fraud was based partly on “neighbourhood opinion,” which Singleton picked up along the way. Not surprisingly, such hearsay was not always reliable. In his report on Andrew Johnson (S15905) Singleton wrote, “there is not a man in the county who believes that he has the slightest claim to a pension,” yet Johnson later produced a supporting statement signed by more than a dozen neighbors. In this case and several others Singleton also relied on his and neighbors’ opinions that the applicant was too young to have served, but census records showed otherwise.

            Singleton decided that the application of Edward Kearney (S15495) was fraudulent partly because “the opinion of his neighbours so far as I learnt it is highly unfavourable to him,” but also for the following reason: “he was rather too zealous in the cause. 18 months continuous service for a Militiaman, in three several tours, is in my opinion rather over doing the matter.” Yes, you read correctly: Singleton decided Kearney’s application was fraudulent because he volunteered to serve too much in the patriot cause.

            Singleton also judged some cases fraudulent because of the pension agent who drew them up. He had a particular animosity to Col. Joseph Johnson, labeling the application he wrote for Arthur Trader (S30169) to be a case of “Fraud Black & diabolical” for no apparent reason.

            For the Pension Commissioner, Col. James L. Edwards, the words “Fraud” or “Not entitled” on the cover of Singleton’s reports were sufficient to doom a claim to a pension. If Edwards read the entire reports there is no indication that they stirred any doubts about Singleton’s judgment. Edwards even allowed himself to be persuaded by Singleton that fighting Indians on the Virginia frontier during the Revolutionary War was not really military service, but a sort of volunteer neighborhood policing. About four dozen applicants were charged with fraud and denied pensions as a result, even though the act of 1832 and Edwards’s own regulations for administering it clearly provide for “Indian Spies” and volunteers of all kinds. This change in policy appears to have been ad hoc, since Edwards granted pensions to applicants who had fought Indians east of the Alleghenies and to those who defended their own neighborhoods from Loyalists in the Carolinas and elsewhere.


            With only Singleton’s conclusions and Edwards’s decisions to go by, one might think that fewer than half of pension applications could be trusted. Comparing the applicants’ narratives in Singleton’s reports with what they claimed in pension applications, however, leads to a different conclusion. After excluding the applications associated with the Lewis Speculating Gentry, there are 181 files complete enough for such a comparison. Of these 181 cases, 154 (85%) showed a close agreement between the narratives and the pension applications. Only 27 (15%) showed significant differences between the narratives and the applications and might therefore be considered fraudulent. Of these 27, however, 19 merely overstated the length of service, which is seldom of interest to historians. Officers’s names, engagements, and other details in these 27 narratives were the same as in the declarations. Only 8 of the 181 applications were clearly false, with the applicant telling Singleton he never served or served very differently from what was claimed in the application. I conclude, therefore, that 8 out of the 181 pension applications (4%) appear to have been total fabrications.


            Assuming that pension applicants and agents in western Virginia (except for the Lewis Speculating Gentry) were neither more nor less honest than those elsewhere, these results suggest that only 4% of pension applications contained deliberately false statements about events of interest to historians. The percentage may actually be lower, because all the fraudulent claims were for militia service, and none was for service in the Continental Line. Since about one-fourth of all pension applications are for Continental service, it is likely that the actual proportion of false pension applications is closer to 3%. Even this may be too high, because Singleton’s investigation occurred in a part of the country where service in the militia was over-represented.

            To put it another way, probably at least 96% of pension applicants stated their officers’s names, engagements, and other details of historical interest according to their best recollections after some 50 years. One hundred percent would be better than 96%, but a historian who sets such a high standard will be left with no records from the Revolutionary War except whatever videotapes and satellite imagery may survive

            Pension applications do have the disadvantage that they were written a half century or so after events. In spite of that limitation some historians have made unique contributions by mining the 20 thousand transcribed and fully searchable pension applications for information that is not available otherwise. The most successful approach appears to be to use individual applications for color, and many applications for consensus.

            An example that comes to mind is Hershel Parker’s paper on “John Butler’s ‘Want of Good Generalship’” in the January 2015 issue of Journal of the American Revolution Parker uses the quotations of individual soldiers to express their views of NC militia Gen. John Butler’s generalship in words that add both color and authenticity, judiciously avoiding relying on any one account as documentation. From a large number of accounts, however, Parker extracts a consensus view that is probably more honest than those previous histories and biographies. In this and probably many other cases, the pension applications, though vulnerable to fraud and the frailty of memory, are more reliable than anything found in conventional historical sources.


I am grateful to Charles B. Baxley for suggesting that I write this article, and for so much more that he has done for Revolutionary War researchers. Thanks to Will Graves for transcribing many of the pension applications used in this research and for keeping the transcription project running so smoothly.

Posted on November 25, 2016, links added on November 29, 2016.