MethodsThe project makes two primary interventions in presenting this spatial history of the Army in the Reconstruction South. The first is in figuring out where soldiers were stationed, how many were stationed there, and how long they remained. The second is in proposing rough estimates for the effective reach and accessibility of the U.S. military, what we call Zones of Occupation and Zones of Access.
To obtain the data that is the foundation of this website, Downs worked through the individual reports listed here. In the vast majority of cases, he based his data on the final monthly report. Thus months should be read as the end of the month. June 1865 refers, then, to June 30, 1865, not June 1, 1865. When final monthly reports were missing, he used numbers from the nearest trimonthly report. When a report was missing, he utilized the next month's listing of the previous month's troop level.
When data conflicted, he utilized the most detailed records available, from post to subdistrict to district to department to division. When similarly detailed reports gave conflicting numbers, he utilized the type of report that had the longest run. In the vast majority of cases, numbers from different reports either exactly or nearly matched.
When the records indicated that a company or group of companies of a specified number of men were divided among a group of posts but did not differentiate individual staffing levels, he divided the number of men by the number of posts. If a record indicated that a company totaling 76 men were spread among four towns, he counted each town with 19 men.
If he only was able to obtain total number of companies without a specified number of people in it, he counted that as the equivalent of 50 men for purposes of our plotting. If there only was a record of a detachment, with no numbers given, he counted that as the equivalent of 15 men for purposes of our plotting on the map.
Nevertheless, any reader should be skeptical of any individual number. While this represents, by far, the most detailed and accurate record of troops in the former Confederate states after surrender, there are numerous ways in which records could conflict or errors could be preserved. Therefore, readers should utilize the numbers with caution. Undoubtedly reports exist that would contradict or contribute to some of these numbers. If you find such a report, please email email@example.com
Among the likely sources of error are:
1) Conflict between different level reports.
2) Clerical and arithmetic error. Reports often did not align perfectly even when written by the same officer covering the same unit. Reports frequently gave previous month troop levels that did not actually match the previous month's report.
3) Illegible handwriting.
4) Missing reports. There are many reports that are missing or perhaps are stored elsewhere. No one should take a gap in troop coverage on the map as ironclad proof that no soldiers were there. In fact there are times when it is very likely, from earlier and later reports, that troops were present for months when we lack reports.
5) Insufficiently detailed reports. Some reports listed every detachment of troops at every station, but many solely reported company locations based on the presence of the company commander. Without question there are many places where detachments operated that were not reflected in these reports. It is impossible to say for certain, for example, if the spread of coverage in Georgia in the late summer and early fall of 1865 actually represents anomalously wide coverage or simply anomalously precise reporting.
6) Troop movements within the month. For the purposes of this map, we captured the places where troops reported being based at the end of the month. This captures permanent and semi-permanent outposts, but it does not capture detachments that moved out into the countryside and either returned by the end of the month or were reported at their home base even if they were in fact not home.
7) Place names that were impossible to reconstruct. In some states, especially Florida and Texas, some place names were not possible to find on contemporary maps. In other places, multiple towns had the same name. In those cases, we utilized ones that had record of existing in the Civil War era. While this provided the most-reliable method for noting the placement, it is impossible to be certain that every location is accurately placed.
8) Human error. While Downs double checked reports from the field with the separately filed outpost reports-with the assistance of Will Hickox-it is possible that human error led to additional errors.
Zones of Occupation and Zones of Access:
Knowing where troops were located at any given time is only the starting point for understanding the bounds of federal occupation in the Reconstruction South. The zone of federal occupation shifted over time, sometimes slowly, sometimes rapidly in response to decisions in Washington and events on the ground. To express the potential extent of federal intervention in the South, we mapped the distance that federal troops could potentially travel in a single day, calling this the federal zone of occupation. We also visualized the rough estimates of the places from which freedpeople might reasonably travel to make a complaint. Judging that most complaints were made by those within a day's journey, we made estimates of these regions for every U.S. military encampment, calling this area the freedpeople's zone of access. Mapping these boundaries required a number of assumptions, resulting in a map that is imprecise and a simplification at best but which we still believe helpful and revelatory for thinking about occupation after the Confederacy's surrender.
We assumed that U.S. infantry would march up to eighteen miles per day, and that cavalry would ride for thirty. Though forces could and did move at a faster clip, particularly before the Confederacy's surrender, our assumptions were based on more modest distances traveled in the South. We imposed a six hour time limit on travel from any location where troops were stationed, using any mode of transportation. If they arrived at a rail line during their journey, we assumed that they would be able to board a train and take it, at 20 miles per hour, as far as they were required to, disembarking at any point along the way. We assumed that embarking on a train would not be immediate; though the wait times at rail depots surely varied tremendously, we assigned a uniform wait time of two hours. The zone of occupation would extend as far as the soldiers could travel on the train in the six allotted hours.
Zones of access, the places from which freedmen might reasonably travel to a military outpost, were more constrained. Freedmen rarely had access to horses. They certainly had limited access, if any, to rails in order to complain about local whites. So the zones of access end up being simple circles with a radius of 18 miles, a six hour walk into town to lodge a complaint. Distance depended not simply upon the extension of technologies of travel, as Charles Paullin and John K. Wright taught us with their famous maps of the rate of travel from New York, but upon access to those technologies.
Our work on the spatial history of Reconstruction draws not just on John Wright's early work on relative geographies, but on recent writing on space, mobility, and power across disciplines and methodologies. Here we would especially note two contributions: Lauren Benton's excellent book on law and geography among European Empires, A Search for Sovereignty (New York: Cambridge University Press, 2010); and William G. Thomas III, Kaci Nash, and Rob Shepard's work on the geography of the railroad, the Civil War, and what they call "funnel zones." This work will appear as "Places of Exchange: An Analysis of Human and Material Flows in Civil War Alexandria, Virginia," in 2016 in the pages of Civil War History.
These assumptions, we admit, are subjective, based on our reading of the army and its willingness and ability to influence affairs in the South. Some are indeed false; for example, we assume that the variegated southern terrain did not result in any variations in travel speed, simply because modeling those variations would be too difficult. For lack of annual data on rail lines, we use William G. Thomas III's railroad data for 1870 for all months and years in this model. Yet in the end we believe that the resulting visualizations are an aid to conceptualizing political and military space in the contested, post-slavery South.