One of the questions I answered a few weeks ago was “How do you know where to dig?” That was my short answer. Since then, many people have specifically asked us, “How did you know where to find these quarters that you’re excavating this summer?”
This is my long answer.
Before we can know where to dig, we must first ask ourselves what we are looking for. Central to Archaeology in Annapolis’s ongoing research project at Wye House has been a focus on those enslaved people who lived and worked on the plantation. In the seven years we have worked at Wye, we have uncovered many work buildings and excavated portions of the Orangery (Greenhouse). While these buildings tell us what it was like to work on a plantation such as Wye, these locations generally aren’t the homes of the enslaved persons we’re trying to find out about. While we did excavate portions of a quarter attached to the Greenhouse, this quarter could have only housed a small fraction of the approximately 150 enslaved persons who lived at Wye. Therefore, part of our research design is to find the quarters where the rest of these people would have lived, cooked, gardened, relaxed, and slept.
This summer at Wye House, we have integrated several sources of data in order to figure out where to dig for these quarters. These include historic aerial photographs, a tracing of a now-lost historic map, modern aerial and satellite imagery, aerial LiDAR data (LiDAR is a mapping technique which uses pulses of light to create a highly accurate topographic map), and descriptions of the plantation by Frederick Douglass from his autobiographies. Individually, each of these sources offers tantalizing clues as to the locations of the structures; however, none offer enough detail on their own to show us where these quarters are located. Once we combine these separate data sources in innovative ways, we can generate new spatial knowledge which has helped guide us to these quarters. Integral to this process is the use of GIS (Geographic Information Systems).
GIS is used by many different fields for many different purposes. A mind-boggling array of spatial analyses can be done on computers by analysts using a host of different software programs. For us, its primary use was to tie together multiple sources of historical data, many of which are spatial in nature. Because the historic aerial photograph, the tracing of the historic map, the LiDAR dataset, and modern aerial imagery all show the arrangement of things in space and on the landscape, we can use GIS to stack these layers on top of one another and impose a common coordinate system such as Latitude and Longitude or the Maryland State Plane. Unfortunately, it’s a little more complex than just opening all of our layers at once and nicely asking the computer to tell us where the quarters are.
The modern aerial imagery is specially formatted so that any single pixel in the image is tied to a specific geographic coordinate. For example, if you were to find your house on one of these images, click on the pixel in the image that represents your front porch, and write down the coordinates that are generated, they would be the exact same coordinates you would get if you stood on your front porch with a GPS unit. Luckily for us, a significant portion of the landscape at Wye hasn’t changed very much. Structures, intersections, and shorelines seen in our historic photograph and map are still there. So by using a modern aerial image of Wye House as a base layer, we can begin creating a database of stacked layers containing common points.
With the historic map, this is as simple as creating pairs of shared points on the map that are still present today. For example, the locations of the corners of the main house, the greenhouse, a few other outbuildings, the wharf , road intersections, and the shoreline are known both on the map and in real life. Once enough points are matched up, the GIS program and the computer can fit the historic map to the modern base map and assign a coordinate to every pixel on our historic map. This process is known as geo-rectification or geo-registration. On the historic map, there are two structures of interest to us this summer, one is labeled ‘2 Story Quarter’ and the other as ‘Brick Row Quarter’. It’s the coordinates of these buildings that we want to be able to figure out. Unfortunately, this map was created partially from a tracing of a map drawn hundreds of years ago and partially from memory. It isn’t exactly the most accurate map to work with. Because of this, we need to supplement the location of the quarters as suggested by this map with other sources.
The historic aerial photograph can also be geo-rectified or geo-registered; however, it must undergo an additional process before we can determine the real-world coordinates of any point within it. Because it was taken from an oblique angle and not directly above the plantation, we must orthorectify it before we can use it. Orthorectification is a process that essentially remove the effect of perspective from our photograph. To orthorectify our photograph, we must create a series of common points just like with geo-rectification. However, when the computer tries to fit the historic photograph to our base map, it will have to warp and distort it to achieve a good fit. In the days before computers and GIS, this process was known as ‘rubber sheeting’ because it was done using an image printed on a rubber sheet. As you can imagine, stretching and twisting a photograph printed on a piece of rubber so that it replicates a top-down view will make your image look pretty funny, however, once we have this top-down view, we can determine the coordinates for all of the points in our photograph. If one looks very carefully at the areas in this photograph that the historic map suggests are the quarters we’re looking for, you won’t see a pair of quarters but rather what looks like a pile of rubble and debris on the surface. This would suggest that by the 1920s or 1930s when this photograph was taken, these quarters no longer existed above the ground.
Lastly, the historic map and historic photograph were stacked on top of a topographic map created using LiDAR data. We used this to further narrow down the locations of these structures by identifying topographic features on the landscape which represent the possible remains of these quarters. By this point, we have a pretty good idea of where to look for the building.
In the spring before the field season began, we spent two days conducting a shovel test survey on the two suspected locations. One of the first things we found was brick. LOTS of brick. And mortar. LOTS of mortar. We also found significant quantities of glass, ceramics, animal bones, oyster shells, nails, and other artifacts unlike anything we’ve found at Wye so far. This would suggest that people were not just working nearby but living there as well. So we are satisfied that we found a pair of buildings right where we thought we’d find a 2 Story Quarter and a Brick Row Quarter. And the artifacts we’re recovering from these buildings seem to suggest a domestic space and not one devoted to agricultural or industrial work.
So does this mean we have actually found the 2 Story Quarter and the Brick Row Quarter? I would like to think so; however, that’s what our excavations this summer are for. Once we expose more of these two structures and begin analyzing the artifacts in our lab we’ll have a much better idea of whether or not we have found a pair of historic quarters at Wye House.