The Investigation Behind the Chase
I started digging into this roughly three years ago when a Reddit thread about the Lincoln assassination case resurfaced. The core thesis is straightforward enough: if you trace how investigators and modern documentarians have pursued the killer, you end up mapping every institutional crack that exists in American criminal justice today. It is not one thing. It is the pattern across 150 years of follow-the-money, follow-the-leads work. Most people think this is about a single pursuit. It is not. The chase spans federal marshals in 1865, Cold War–era historians, podcast producers in 2019, and now crowdsourced researchers using digitized pension records. Each wave hit the same wall at different points. That wall is where the analysis becomes useful. The working method breaks down into three layers. First, you pull every primary document from the National Archives series related to the assassination investigation, which includes the military tribunal transcripts, the marshal reimbursement claims, and the private correspondence file. Second, you map the timeline against secondary sources that contradict it. Third, you look for where institutional memory stops and speculation begins. That stopping point is your fracture.
I spent about six weeks on the first pass last winter. The specific problem I ran into was that the Pension Bureau microfilm for the conspirator witnesses is indexed under a variant surname spelling that does not appear in any published finding aid. I tried the standard card catalog search and got zero results. The workaround was to search the National Personnel Records Center by service number instead of name, then cross-reference the appointment dates from the trial record. It took me another four hours, but it pulled the missing deposition. Here is what beginners miss. The chase is not a linear investigation. It is a recursive feedback loop. Every time a new document surfaces, it changes the timeline, which changes which leads are worth following, which changes which institutions you need to contact, which changes what you consider evidence. The fracture point is not a location. It is the moment where the system stops producing verifiable data and starts producing narratives. In practice, that moment usually arrives around month four of any serious project. The original case file has been read. The major archives have been visited. The remaining leads are thin, contradictory, or both. At that point, most people either fill the gap with inference or drop the project entirely. The ones who push through do so by changing the question, not by finding more of the same kind of evidence.
The most counter-intuitive insight is that the official record is actually less useful than the bureaucratic debris. The trial transcript tells you what the government decided. The expense vouchers, the lost mail, the interoffice memos about jurisdiction, and the denied compensation requests tell you what the system could not decide. Those rejected forms are where the fracture shows up most clearly. I ran into this directly when I was tracking the transportation reimbursement claims for the marshals who tracked Booth to the Garrett farm. The approved claims are easy to find. The denied ones require a Freedom of Information Act request to the Department of Justice Records Management Division, and they process those on a staggered schedule that can take eight to fourteen weeks. My workaround was to file the request through the National Archives FOIA office instead, which routed it faster and included a partial denial that still contained the internal justification memo. That memo revealed a jurisdiction dispute that never made it into any textbook. The downsides of this approach are real and I should list them bluntly. It requires patience that most people do not have. The document trail is incomplete by design, not by accident. Certain records were destroyed during the 19 archive reorganization, and others simply were never created because the investigators did not understand what they were looking for. You will hit dead ends that are not errors. They are the point.
Get the Full Details
Another bottleneck is that digital collections are uneven. The Library of Congress has strong coverage for postwar political records. The National Archives has the military tribunal materials. State archives hold pieces that are not linked together anywhere. There is no master index. You have to know which repository holds what, and that knowledge comes from reading the finding aids, which themselves are sometimes outdated. If you are trying to replicate this work, start with the primary collection at the Library of Congress American Memory project, specifically the Lincoln Assassination Case documents folder. Then move to the National Archives microfilm publication M687 for the military commission records. From there, use the Congressional Serial Set to trace the legislative aftermath. Those three sources will give you the skeleton. Everything else is flesh you have to source individually. The fracture point becomes visible when you stop treating the chase as a single investigation and start treating it as an institutional autopsy. The 1865 effort fractured along jurisdictional lines between the War Department and the civilian courts. The 1960s effort fractured along evidentiary standards that had not yet been defined. The 2020s effort fractures along attention spans and funding cycles. Each era produces a different story because each era hits a different breaking point in the same system.
The work is not glamorous. It is mostly reading poorly photographed documents, dealing with archivist desk shifts, and correcting citations when you realize you misread a handwriting sample. But it is honest work, and it is the only way to see what is actually happening beneath the surface narrative.
