ABSTRACT Title of Dissertation: “THE REAL DISTANCE WAS GREAT ENOUGH”: REMAPPING A MULTIVALENT PLANTATION LANDSCAPE USING HISTORICAL GEOGRAPHIC INFORMATION SYSTEMS (HGIS) Benjamin Adam Skolnik, Doctor of Philosophy, 2019 Dissertation directed by: Professor Mark P. Leone, Department of Anthropology This dissertation uses the tools of historical Geographic Information Systems (hGIS) to locate and describe mid-nineteenth-century plantation landscapes in Talbot County, Maryland. The methodology described here combines historic maps, historic and modern aerial photography, LiDAR-derived elevation data, historic census data, and textual descriptions. It also uses them in conjunction with an ongoing archaeological research project at Wye House, the ancestral seat of the Lloyd family and site of enslavement of Frederick Douglass, near Easton, Maryland in order to further develop ways for archaeologists, historians, and other researchers to work with cartographic and spatial data in a digital framework. This methodology can be used across multiple scales to survey remotely individual sites or even entire counties for potential archaeological resources. Furthermore, it examines the autobiographies of Frederick Douglass, not only because he was a witness to these landscapes, but also because he can be read as a social theorist who addresses issues of race and racialized landscapes throughout his writings. Lastly, it uses these sources of data to consider Dell Upton’s spatial hypothesis regarding racialized plantation landscapes. Taken together, this study of mid-nineteenth-century Talbot County, Maryland represents one way to identify and recover lost sites of African American heritage that would otherwise remain lost. “THE REAL DISTANCE WAS GREAT ENOUGH”: REMAPPING A MULTIVALENT PLANTATION LANDSCAPE USING HISTORICAL GEOGRAPHIC INFORMATION SYSTEMS (HGIS) by Benjamin Adam Skolnik Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2019 Advisory Committee: Professor Mark P. Leone, Chair Professor George Hambrecht Dr. Matthew Palus Dr. Michael Roller Professor Mary Sies © Copyright by Benjamin Adam Skolnik 2019 Acknowledgements Whether they know it or not, many people have contributed to this dissertation in one way or another. I would first like to thank my advisor Dr. Mark P. Leone for (among many other things) giving me a way to think about the world around me and to engage with the past. I would also like to thank the other members of my advisory committee, Dr. George Hambrecht, Dr. Matthew Palus, Dr. Michael Roller, and Dr. Mary Sies, who have each in their own way given me a way to think about the material traces of the past and what to do with them. I also need to acknowledge Dr. Stephen Prince and Hasan Jackson who introduced me to LiDAR and taught me how to work with it and Dave Foreman and the Maryland Department of Natural Resources for providing access to the State’s LiDAR dataset. Graduate school is often a team sport, so I need to thank the graduate students at the University of Maryland who have made contributions to this dissertation in one way or another, including: Beth Pruitt, Kate Deeley, Stefan Woehlke, Tracy Jenkins, Amanda Tang, Jocelyn Knauf, Mary Furlong Minkoff, Adam Fracchia, Michael Roller, Courtney Singleton, Sarah Janesko, Janna Napoli, Kat Aben, Trish Markert, and Samantha Lee. Archaeological fieldwork cannot happen without a field crew and I need to thank the following University of Maryland Field Methods in Archaeology (ANTH496/ANTH696) Field School students, without whom I could not have ii completed the field work described in this dissertation: Aleksandr Agarunov, Elise Ansher, Bill Auchter, Angelica Barrall, Elizabeth Bernhardt, Elizabeth Berry, Natalie Berry, Emily Bokelman, Andrew Brown, Sophia Chang, Alison Curry, Sam Dennis, Paige Diamond, Drew Doherty, Daniel Dorsey, Jen Eliot, Shaun English, Michelle Fuentes, Clio Grillakis, Ian Guillermo, Matt Hager, Joseph Harden, Elizabeth Harford, Lauren Hicks, Brittany Hutchinson, Katie Hutchinson, Lauren Kauffman, Patrick Kim, Paula Maia, Edward McLaughlin, Andrea Milly, Darin Murray, Esther Nehrer, Richard Nyachiro, Cecelia O’Brien, Norma Oldfield, Dorie Phillips, Nicole Punzi, Juan Rivas, Kaitlin Schiele, Audrey Schaefer, Kasey Schumacher, Larry Seastrum, Maria Sharova, Sabrina Shirazi, Alyssa Snider, Breton Stailey, Marcella Stranieri, Max Isobel Taylorch, Alun Walpole, Laura Wright, Julia Torres Vasquez, Aley Villarreal, Rebecca Webster, and Duncan Wintermyer. University of Maryland undergraduates Zev Cossin, Ryan Elza and Justin Uehlein, and University of Maryland graduate John Blair all provided additional fieldwork. Likewise, archaeology is not complete once fieldwork is over, so I need to also thank the following University of Maryland Archaeology in Annapolis Laboratory students for cleaning, processing, and cataloging the archaeological collections generated by our excavations at Wye House and elsewhere: Jane Baden, Melissa Baley, Angie Barrall, Rachel Beiser, Elizabeth Berry, Natalie Berry, Emily Bokelman, Caroline Bruchman, Holly Buchanan, Sarah Buchanan, Dani Buffa, Stephanie Callahan, Katherine Calvert, Sophia Chang, Tiffany Cusick, David Cynman, Cameron Daisey, Paige Diamond, Devon DiComo, Ryan Elza, Shaun English, Kolade Fapohunda, Cristina Fifer, Joshua Finken, Brian Flemming, Raquel iii Fleskes, Brandon Gallagher, Mayra Garcia, John Garcia-Tobar, Evelyn Gibson, Jennifer Glode, Clio Grillakis, Ian Guillermo, Umai Habibah, Sam Hajarian, Lauren Hicks, Tyler Hicks, Anthony Hudzik, Lauren Hume, Moriah James, Claire Jamison, Katie Janota, Robert Juste, Nana Kennedy, Michelle Kreiner, Sarah Lebarron, Becca Lane, Emma Leibman, Hope Loiselle, Jasmine Mathis, Katie Mayer, Sara Miller, Winston Miller, Amanda Modica, Hayatt Mohamed, Lena Muldoon, Megan Najafali, Alexandrea Neville, Kristin Ngo, Thompson Noelle, Santiago Noguera, Christopher Pepe, Kayleigh Pinkett, Norma Quijano, Danadaree Rajapakshe, Chris Reavis, Sharon Ridge, Ashley Rivas, Juan Rivas, Taqwa Rushdan, Nina Scall, Audrey Schaefer, Larry Seastrum, Jillian Seay, Maria Sharova, Sabrina Shirazi, Rebecca Stein, Marcella Stranieri, Austin Tumas, Tim Vettel, Elliot Weis, Brandt Wootan. I also need to specifically thank Scott Oliver for his work processing the 2011-2014 Wye House collection. I need to highlight here the contributions of two University of Maryland undergraduate students, Clio Grillakis and Marcella Stranieri, both of whom started as field school students, but made additional contributions to the work detailed in these pages. This includes digitizing historic maps and censuses through a College of Behavioral & Social Sciences Emerging Scholars grant and a Future of Information Alliance Robert W. Deutsch Seed Grant. I also need to thank these organizations for their support in funding this work that would become parts of my dissertation. I would like to thank Mrs. Mary D. Tilghman and Richard and Beverly Tilghman, Diane and Leland Brendsel, Hope and Tony Harrington, Patrice Miller, and Peter Stifel for welcoming myself and others into their homes and sharing the iv historic sites for which they have been wonderful caretakers. I would also like to thank Ed Taylor and Jerry Wilmoth for their efforts in keeping Wye House and Myrtle Grove standing. Although largely not described in this volume, my involvement with Archaeology in Annapolis has included time working on The Hill in Easton, Maryland, and I need to acknowledge the help of Harriette and Eric Lowery, Priscilla Morris, Carlene Phoenix, John “Tenny” Sener, Dale Green, and Patrick Rogan. I need to thank my entire family for their support through this process, especially Mom and Dad and Carolyn and Gerry and Kevin, and especially for the countless hours of babysitting they have provided to allow me to finish this dissertation. And lastly, I would like to thank Jessie and Henry and Minna for everything they have done for me. I simply do not have the words to express how important they are to me or how much they contributed to these pages. v Table of Contents Acknowledgements ....................................................................................................... ii Table of Contents ......................................................................................................... vi List of Tables ............................................................................................................. viii List of Figures .............................................................................................................. ix Chapter 1: Introduction ................................................................................................. 1 Chapter 2: Background ................................................................................................. 8 Slavery and Heritage ................................................................................................. 8 Archaeology and Plantations .................................................................................. 29 Chapter 3: Theoretical Framework and Methodologies ............................................. 42 Theoretical Framework ........................................................................................... 42 Landscape Archaeology ...................................................................................... 42 Historical Geographic Information Systems (hGIS) ........................................... 50 Methodologies......................................................................................................... 55 Chapter 4: Historical Geographic Information Systems (hGIS) and Regional Studies ..................................................................................................................................... 62 Georeferencing the 1858 William H. Dilworth Map of Talbot County, Maryland 62 Spatializing the 1860 United States Census for Talbot County, Maryland ............ 91 Chapter 5: Mapping the Plantation: Using Airborne LiDAR and Historical GIS (hGIS) to Survey Extant Plantation Sites.................................................................. 127 LiDAR (Light Detection and Ranging) Survey .................................................... 127 William Paca House and Garden, Annapolis, Maryland ...................................... 135 Plantation Survey .................................................................................................. 143 Clay’s Hope (T-189) ......................................................................................... 153 Compton (T-146) .............................................................................................. 159 Hope House (T-90) ........................................................................................... 164 Myrtle Grove (T-53) ......................................................................................... 171 Otwell (T-164) .................................................................................................. 178 Ratcliffe Manor (T-42) ..................................................................................... 183 Sherwood Manor (T-244) ................................................................................. 190 The Anchorage (T-52) ...................................................................................... 195 The Wilderness (T-149) .................................................................................... 201 Wye House (T-54) ............................................................................................ 206 Discussion ............................................................................................................. 214 Chapter 6: Locating That Which Has Been Lost: Using Historical Geographic Information Systems (hGIS) to Locate Quarters at Wye House............................... 219 hGIS and LiDAR .................................................................................................. 219 Data Sources ......................................................................................................... 225 Previous Historical and Archaeological Research at Wye House ........................ 247 Revisiting Previous Archaeological Interpretations: the Wye House Corn Crib . 259 Locating Lost Structures: the 2-Story Quarter and the Brick Row Quarter ......... 266 Chapter 7: Mapping the Enslaved Landscape Through Frederick Douglass: Integrating Qualitative Data into Historical Geographic Information Systems (hGIS) ................................................................................................................................... 284 vi Frederick Douglass and Wye House ..................................................................... 284 Frederick Douglass and Talbot County, Maryland ............................................... 300 Remapping an Eighteenth-Century Garden .......................................................... 311 Chapter 8: Mapping the Enslaved Landscape Through Dell Upton: Integrating Qualitative Data into Historical Geographic Information Systems (hGIS) .............. 324 Dell Upton and Racialized Plantation Landscapes ............................................... 324 Modeling Control on the Landscape ..................................................................... 338 Remapping the Plantation Landscape ................................................................... 363 Chapter 9: Epilogue .................................................................................................. 373 Bibliography ............................................................................................................. 376 vii List of Tables Table 2.1……………………………………………………………………………..28 Table 5.1…………………………………………………………………………....150 viii List of Figures Figure 2.1 – Talbot County, Maryland ..........................................................................9 Figure 2.2 – Wye House (18TA314) ...........................................................................39 Figure 4.1 – 1858 Dilworth map of Talbot County, Maryland ....................................64 Figure 4.2 – 1858 Dilworth map showing Ground Control Points ..............................71 Figure 4.3 – 1858 Dilworth map showing locations of structures ...............................77 Figure 4.4 – 1858 Dilworth map showing farm boundaries ........................................78 Figure 4.5 – 1858 Dilworth map showing farm sizes ..................................................79 Figure 4.6 – 1858 Dilworth map showing two structures at Wye House ....................80 Figure 4.7 – 1858 Dilworth map showing district boundaries.....................................82 Figure 4.8 – 1858 Dilworth map showing coastlines, streams, lakes, and marshes ....84 Figure 4.9 – 1858 Dilworth map showing roads..........................................................85 Figure 4.10 – 1858 Dilworth map showing railroads ..................................................86 Figure 4.11 – 1858 Dilworth map showing kernel density of structures .....................88 Figure 4.12 – 1858 Dilworth map showing kernel density of roads ............................90 Figure 4.13 – Population distributions of enslaved individuals at Lloyd plantations ...........................................................................................105 Figure 4.14 – Linkage between 1858 Dilworth map and 1860 US Census ...............109 Figure 4.15 – 1858 Dilworth map showing locations of structures and residential structures ......................................................................................112 Figure 4.16 – 1858 Dilworth map showing heads of households matched in census .............................................................................................................118 Figure 4.17 – 1858 Dilworth map showing dot density map for matched households......................................................................................................120 Figure 4.18 – 1858 Dilworth map showing slave owners matched in census ...........124 Figure 4.19 – 1858 Dilworth map showing dot density map for matched enslaved..........................................................................................................125 Figure 5.1 – LiDAR data visualization ......................................................................133 Figure 5.2 – Paca House and Garden AA-657 ...........................................................136 Figure 5.3 – William Paca Garden, Annapolis, MD (AA-657) with golden mean superimposed .......................................................................................142 Figure 5.4 – 1847 US Coast Survey Map of St Michaels River, Third Haven Creek, Easton &c. ..........................................................................................145 Figure 5.5 – 1862 US Coast Survey Chesapeake Bay – From Magothy River to Choptank River ..........................................................................................146 Figure 5.6 – 1937 Aerial Survey U.S. Agriculture Stabilization and Conservation Service ....................................................................................147 Figure 5.7 – 10 Talbot County Plantations chosen for further hGIS analysis ...........151 Figure 5.8 – Clay’s Hope T-189 ................................................................................154 Figure 5.9 – Clay’s Hope hGIS geodatabase .............................................................155 Figure 5.10 – Clay’s Hope hGIS geodatabase, detail ................................................156 Figure 5.11 – Compton T-146 ...................................................................................160 Figure 5.12 – Compton hGIS geodatabase ................................................................161 Figure 5.13 – Compton hGIS geodatabase, detail .....................................................162 ix Figure 5.14 – Hope House T-90 ................................................................................165 Figure 5.15 – Hope House hGIS geodatabase ...........................................................166 Figure 5.16 – Hope House hGIS geodatabase, detail ................................................167 Figure 5.17 – Myrtle Grove T-53 ..............................................................................172 Figure 5.18 – Myrtle Grove hGIS geodatabase .........................................................173 Figure 5.19 – Myrtle Grove hGIS geodatabase, detail ..............................................174 Figure 5.20 – Otwell T-164 .......................................................................................178 Figure 5.21 – Otwell hGIS geodatabase ....................................................................179 Figure 5.22 – Otwell hGIS geodatabase, detail .........................................................180 Figure 5.23 – Ratcliffe Manor T-42 ...........................................................................183 Figure 5.24 – Ratcliffe Manor hGIS geodatabase .....................................................184 Figure 5.25 – Ratcliffe Manor hGIS geodatabase, detail ...........................................185 Figure 5.26 – Sherwood Manor T-244 ......................................................................191 Figure 5.27 – Sherwood Manor hGIS geodatabase ...................................................192 Figure 5.28 – Sherwood Manor hGIS geodatabase, detail ........................................193 Figure 5.29 – The Anchorage T-52 ............................................................................196 Figure 5.30 – The Anchorage hGIS geodatabase ......................................................197 Figure 5.31 – The Anchorage hGIS geodatabase, detail ...........................................198 Figure 5.32 – The Wilderness T-149 .........................................................................201 Figure 5.33 – The Wilderness hGIS geodatabase ......................................................202 Figure 5.34 – The Wilderness hGIS geodatabase, detail ...........................................203 Figure 5.35 – Wye House T-54..................................................................................206 Figure 5.36 – Wye House hGIS geodatabase ............................................................207 Figure 5.37 – Wye House hGIS geodatabase, detail .................................................208 Figure 5.38 – Wye House Henry Chandlee Forman map and LiDAR ......................213 Figure 6.1 – 1847 US Coast Survey, Wye House Detail ...........................................228 Figure 6.2 – 1862 US Coast Survey, Wye House Detail ...........................................229 Figure 6.3 – 1858 William Dilworth Map Wye House Detail ...................................230 Figure 6.4 – 1877 Lake, Stevenson, Griffing Composite from Atlas of Talbot and Dorchester Counties.....................................................................231 Figure 6.5 – 1877 Lake, Stevenson, Griffing Atlas Wye House Detail.....................232 Figure 6.6 – Dallin Aerial Surveys, Wye House, c. late 1920s .................................234 Figure 6.7 – Dallin Aerial Photograph Wye House, c. late 1920s, Georeferenced and warped ............................................................................236 Figure 6.8 – 1937 Aerial Imagery Index, U.S. Agriculture Stabilization and Conservation Service .....................................................................................238 Figure 6.9 – Henry Chandlee Forman map of Wye House, 1967 Georectified .........240 Figure 6.10 – Henry Chandlee Forman map of Wye House, 1967 Inset, Georectified....................................................................................................241 Figure 6.11 – 1784 Plat of Wye House, Traced in part by Henry Chandlee Forman ...........................................................................................................243 Figure 6.12 – LiDAR-derived elevation and slope, Wye House ...............................244 Figure 6.13 – LiDAR-derived elevation and slope, with structure outlines and labels from 1967 Forman map ................................................................245 Figure 6.14 – North Building Site Plan, Reproduced from Blair et al. 2009.............259 Figure 6.15 – Wye House, Corn Crib, HABS Survey, 1936 .....................................260 x Figure 6.16 – Dallin Aerial Surveys – Wye House, c. late 1920s, Detail of Corn Crib .......................................................................................................261 Figure 6.17 – 2011 Shovel Test Pit Survey with structure outlines from 1967 Forman map ..........................................................................................269 Figure 6.18 – Shovel Test Pit Q/R 3, 9 April, 2011, Brick Feature ...........................270 Figure 6.19 – Shovel Test Pit Y1, 8 April, 2011, Brick Feature ...............................271 Figure 6.20 – 2011-2014 Test Unit Map with LiDAR and estimated building outlines ...........................................................................................................272 Figure 6.21 – Test Unit 79 and 80, Robbed brick pier and robbed brick trench .......273 Figure 6.22 – Test Unit 66, Bottom of Excavation, East Profile, Brick Pier .............274 Figure 6.23 – Test Unit 73, Bottom of Excavation, North Profile, Stone Foundation .....................................................................................................275 Figure 6.24 – Test Units 79 and 80, Top of circular deposit .....................................277 Figure 6.25 – Test Units 79 and 80, Top of circular deposit, Plan view drawing .....278 Figure 6.26 – Test Unit 71, Trench with disarticulated brick rubble, possible foundation location ........................................................................................280 Figure 6.27 – Test Unit 84, Top of Level C, Brick rubble ........................................281 Figure 7.1 – Spatial Annotation: My Bondage, My Freedom, “A General Survey of the Slave Plantation”, Douglass, 1855 ..........................................287 Figure 7.2 – LiDAR-derived elevation and slope, Topographic relief of the Long Green ....................................................................................................290 Figure 7.3 – 1847 US Coast Survey, “Numerous other slave houses and huts” .......291 Figure 7.4 – 1847 US Coast Survey and 1784 Plat, Windmill ..................................293 Figure 7.5 – Detail from Dallin Aerial Photograph, c. late 1920s, Possibly the Sloop Sally Lloyd ...........................................................................................294 Figure 7.6 – 1858 Dilworth Map, Tuckahoe, Maryland and Douglass’s Birthplace .......................................................................................................301 Figure 7.7 – 1858 Dilworth Map, Hillsborough, Maryland and Douglass’s Birthplace .......................................................................................................304 Figure 7.8 – 1877 Lake, Stevenson, Griffing Atlas and 1858 Dilworth map, Mills near Douglass’s Birthplace ...................................................................306 Figure 7.9 – 1858 Dilworth map showing optimal route between Tapper’s Corner and Wye House ..................................................................................308 Figure 7.10 – Ratcliffe Manor, Central path, remnants of boxwood edging, and southwest façade .....................................................................................312 Figure 7.11 – LiDAR-derived elevation and slope, Racliffe Mnaor (T-42) ..............314 Figure 7.12 – LiDAR-derived elevation data, Ratcliffe Manor (T-42) .....................315 Figure 7.13 – LiDAR-derived elevation data, Ratcliffe Manor (T-42) .....................315 Figure 8.1 – 1858 Dilworth map, Distance to structures ...........................................345 Figure 8.2 – 1858 Dilworth map, Distance to roads ..................................................346 Figure 8.3 – 1858 Dilworth map, Distance to structures or roads .............................348 Figure 8.4 – 1858 Dilworth map, Distance to navigable waterways .........................349 Figure 8.5 – Visibility from Residential Structures shown on Dilworth map (1858) .............................................................................................................355 Figure 8.6 – Visibility from Residential Structures shown on Dilworth map (1858), Wye House detail ..............................................................................356 xi Figure 8.7 – Visibility from Roads shown on Dilworth map (1858) .........................357 Figure 8.8 – Visibility from Roads shown on Dilworth map (1858), Wye House detail ...........................................................................................358 Figure 8.9 – Visibility from Structures and Roads shown on Dilworth map (1858) .............................................................................................................359 Figure 8.10 – Visibility from Structures and Roads shown on Dilworth map (1858), Wye House detail ..............................................................................360 Figure 8.11 – Visibility from Navigable Water ........................................................361 xii Chapter 1: Introduction This dissertation is a dissertation in historical archaeology, broadly conceived. While it invokes some of the familiar tools and techniques of dirt-and-trowel archaeology, it is primarily concerned with new uses historical geographic research methods in order to produce new knowledge about American slavery and the plantation and their landscapes. The primary focus of this dissertation is the use of historical Geographic Information Systems (hGIS) in historical archaeology and the ways in which it can support and inform a research project in landscape archaeology. hGIS are capable of organizing, displaying, combining, and analyzing multiple historic and contemporary spatial datasets simultaneously, which allows a researcher to leverage the strengths of each dataset while overcoming many of their limitations. These datasets can be any dataset that records geographic or spatial information, either directly, or in some cases, indirectly. These include historic maps, photographs, censuses, tax lists, plats, textual descriptions, as well as contemporary satellite imagery, aerial photography, and digital elevation models. hGIS can be used across multiple scales of analysis ranging from specific sites or structures to individual plantations to county-wide, regional, national, or even international in scale. Focusing on Talbot County on Maryland’s Eastern Shore during the middle half of the nineteenth century, this dissertation will present a series of techniques that will provide a way for researchers to investigate the spatial distribution of historic phenomena from the combination of inherently spatial data and from historic datasets that previously did not have explicit spatial information. It will use these results as a 1 part of a traditional archaeological research project to locate, identify, and describe features which have disappeared from the contemporary spatial landscape. Lastly, this dissertation will attempt to recover and explore cultural understandings of space and place produced in the context of racialized plantation landscapes and held by the enslaved whose coerced labor powered the agricultural economy that was prevalent throughout the region at that time. Until recently (and in many cases it still is), the prevailing discourse on Talbot County’s antebellum plantation heritage, like elsewhere in the American South, is generally one that is white and upper-class, and revolves around themes of gentility and hospitality (Butler 2001; Eichstedt and Small 2002; Hoelscher 2003; Modlin 2008; Seaton 2001). This has left little or no room for a discussion of African American heritage in Talbot County, Maryland, or of their contributions to its history. It was not until 2011 that a statue honoring Frederick Douglass, abolitionist, orator, and formerly enslaved Talbot County resident, was installed in front of the Talbot County courthouse, perhaps tellingly, opposite another, earlier statue honoring the Talbot Boys, 84 local residents who fought for the Confederacy in the American Civil War (Campbell 2016; Witte 2011). This effort to recognize and honor the legacy of Douglass locally, as well as partnerships between Archaeology in Annapolis at the University of Maryland, professors and students from Morgan State University, and community groups like Historic Easton Inc. have attempted to at least partially address this silence. Collectively, the uses of hGIS presented here will be one attempt to produce a remapped, many-faceted version of the plantation landscape in Talbot County, 2 Maryland from just prior to the Civil War and emancipation that emphasizes the African American landscape. In the process, this dissertation asks, and attempts to answer several questions, including how we can document and recover African American enslaved landscapes on both a regional and a local scale, if the writings of Frederick Douglass can be used to support the enslaved African American worldview Dell Upton hypothesizes in “Black and White Landscapes in Virginia” (1984) and continued in “Imagining the Early Virginia Landscape” (1990), and whether we can use this worldview to highlight a different perspective of plantation landscapes. Overwhelmingly, historical research into the uses of space on plantations have focused on the landscape of the enslaver. This has been done for several reasons, which will be addressed, one of the contributions of the methodologies proposed here is to remap the plantation landscape from the perspective of the enslaved and emphasize a new way to talk about these spaces. While the main plantation houses and the larger plantation landscapes that stretch out from and surround them have been fairly well-described and documented, much less has been done to address plantation landscapes as centered on the quarters, fields, forests, roads, paths, rivers, creeks, and other built and natural geographic features that would have played a central role in the daily lives of those enslaved there. Large, well-known and well- visited plantation sites in the Chesapeake such as Mount Vernon, Monticello, Poplar Forest have been increasingly aware of this discrepancy, addressing it in research, interpretation, and programming (Chappell 1999; Heath and Gary 2012; Stanton 2012; Thompson 2019) ; however, these sites represent an extremely small fraction of 3 all plantation sites, either historically or surviving into the present in one form or another. As envisioned, this project is ultimately about recovering past configurations of space and place, both in their physical sense and in their cultural or mental sense. While it can, and does, document the formal plantation landscape centered on the main house, using the tools of hGIS, it seeks to reorient the plantation landscape toward the quarters and fields and recover these configurations to address it from the perspective of the enslaved. Recovering this perspective of the plantation landscape tells a story not adequately addressed by the dominant narrative of plantations as gentility and hospitality, nor generally addressed at large in the present. Through addressing the above research questions, this dissertation will develop and test methodologies in hGIS. Finding ways to integrate historical data with spatial data, both historic and modern, creates robust datasets that situate these data in time and space and creates new geographic and historic knowledge. Using the methodologies of landscape archaeology and GIS, this project generates new knowledge of nineteenth-century plantation landscapes on both the local and regional scale, characterizes these spaces in the present for surviving landscape elements, and provides a set of tools for other researchers to do the same at their sites. This information can be used in an academic research setting, in preservation and planning context to identify and protect archaeological resources, or as a part of a broader heritage project in memory and remembering the legacies of slavery. This dissertation also considers the importance of Frederick Douglass, not just as a historic tour guide to the places examined here, but also as a social theorist. The 4 institution of race-based slavery was supported by a constellation of other cultural practices, behaviors, “taken-for-granteds”, worldviews, and systems. Many have already discussed the role of these ideologies in maintaining this system of exploitative labor and racial identification. Recently, some archaeologists have turned to identifying ways in which individuals or entire groups may have escaped this cultural system, either momentarily or entirely (Fellows and Delle 2015; Orser and Funari 2001; Sayers 2014, Weik 1997). In his 2005 book The Archaeology of Liberty in an American Capital, historical archaeologist and critical theorist Mark Leone states that one of his goals in writing is to “ask whether groups have escaped the enveloping role of ideology. Who saw through it, and what is the evidence?” (Leone 2005:28-9). Similarly, through his writings, and especially in his autobiographies, Douglass makes it clear that he saw through the ideology of slavery. He writes that even as a child, he asked himself, “Why am I a slave? Why are some people slaves, and others masters? Was there ever a time this was not so? When did the relation commence?” (Douglass 1855:89). He locates slavery as a manmade institution, stressing it is not a natural one, and as such, argues that, “what man can make, man can unmake” (Douglass 1855:90). He continues on the next page, “This, to me, was knowledge; but it was a kind of knowledge which filled me with a burning hatred of slavery, increased my suffering, and left me without the means of breaking away from my bondage. Yet it was knowledge quite worth possessing” (Douglass 1855:91). Seemingly channeling twentieth-century French social theorist Michel Foucault on knowledge and power, Douglass reflects on one way in which one might 5 remove oneself from the condition of bondage, writing, “‘Very well,’ thought I; ‘knowledge unfits a child to be a slave.’ I instinctively assented to the proposition; and from that moment I understood the direct pathway from slavery to freedom” (Douglass 1855:146-147). Through a reading of Douglass, the conclusion is almost inescapable that he saw through, pierced, or found the holes in the net (depending on which theorist’s metaphor you use) of the governing ideologies of nineteenth-century America that supported a system of plantations and slavery. Not only did Douglass make clear the ideology that supported slavery, but he wrote and spoke about it prolifically throughout his adult life and used this knowledge to combat it. What remains to be explored by archaeologists studying plantations and their landscapes is the extent to which Douglass’s insights were commonplace and shared throughout enslaved communities on plantations throughout the South and what they did with this knowledge. This dissertation represents an attempt to understand some of the spatial components that go along with Douglass’s and Leone’s questions. The quote in the title, “The real distance was great enough” comes from Douglass’s 1855 My Bondage and My Freedom. He writes: To look at the map, and observe the proximity of Eastern Shore, Maryland, to Delaware and Pennsylvania, it may seem to the reader quite absurd, to regard the proposed escape as a formidable undertaking. But to understand, some one has said a man must stand under. The real distance was great enough, but the imagined distance was, to our ignorance, even greater. Every slaveholder seeks to impress his slave with a belief in the boundlessness of slave territory, and of his own almost illimitable power. We all had vague and indistinct notions of the geography of the country. The distance, however, is not the chief trouble. The nearer are the lines of a slave state and the borders of a free one, the greater the peril. (Douglass 1855:281) 6 Here, Douglass is measuring both geographic distance as well as a kind of social distance. He maps geographic space not in terms of miles or feet, but in terms of proximity to freedom. By using modern mapping technologies in conjunction with historic data, we can create new geographic, spatial, and historic knowledge. As a direct application, this knowledge can be used by archaeological research projects like Archaeology in Annapolis to find and recover features which have been lost and erased from the landscape. It can also go beyond the recovery of physical remains and assist us in an attempt to recover cultural understandings of space and place. In terms of American plantation-based slavery, this physical erasure from the landscape has been at times deliberate and at time neglectful and the recovery of this knowledge represents a serious reclamation of African American heritage. To borrow from Douglass, this is “knowledge quite worth possessing” (Douglass 1855:91). To this end, this dissertation asks and attempts to answer the following research questions: • How are mid-nineteenth-century plantation landscapes structured in terms of their layout, components, and geographic setting? • Can hGIS (historical Geographic Information Systems) and remote sensing be used to document surviving plantation landscapes? What do mid-nineteenth-century plantation landscapes look like in the present? • Can hGIS and remote sensing be used to identify non-extant components of these landscapes in the form of archaeological features? • Can the plantation landscape be remapped from the perspective of those enslaved on them? • Is the enslaved spatial worldview as described by Upton embedded within the writings of Frederick Douglass? 7 Chapter 2: Background Slavery and Heritage The study area of this dissertation is multi-scalar, but the common denominator is mid-nineteenth-century plantations in Talbot County, Maryland (Figure 2.1), their surrounding built and natural landscapes, and their remains that exist into the present. By necessity, this involves earlier, underlying seventeenth-and- eighteenth-century components to these landscapes, modifications made to them in the twentieth century, and the connections between these sites and to places outside of the county. Talbot County on Maryland’s Eastern Shore was first settled by English colonists in 1634 (Tilghman and Harrison 1915:1). The early economy of the county relied heavily on a system of monocrop agriculture to grow tobacco for export to the European market (Preston and Harrington 1983:59). Later, toward the end of the eighteenth century, the agricultural base of the county shifted to wheat (Preston and Harrington 1983:81); however throughout the eighteenth and much of the nineteenth centuries, the agricultural productivity of the county depended on the labor of enslaved Africans and African Americans. These people have largely been excluded from the traditional histories of the area. This is an area to which historical archaeology can contribute. Histories of Talbot County, Maryland have already been written. Some are good, some are not so good, and just like this dissertation are 8 products of their time and place. The goal of this dissertation is not to recite or reproduce another such work here. Rather, the goal of this dissertation is to use the tools of historical Geographic Information Systems in conjunction with those of an anthropologically-informed historical archaeology in order to create a different kind of history, one which is overdue and needed. Plantations are discrete economic units within a larger economic system designed to mass-produce an agricultural product or cash crop, usually for export to an international market. Plantations generally function through the use of a coerced or marginalized labor force, both in the past and in the present. Those discussed here in the context of the nineteenth-century American South operated through a system of race-based slavery; however, plantations can and do operate through wage-based labor as well. Far from being relics of the distant past, the legacies of this economic 9 model and the social conditions that accompany it continue to reverberate into the present. Agricultural systems based on enslaved labor can operate on many scales ranging from the multi-farm plantation as to small family farms. As an economic mode of production, this system required large amounts of cheap land and cheap labor in order to support those few at the top of the social and economic classes who legally owned both the land and labor used to work that land. The economic realities of this system help to explain both the diffuse settlement pattern observed throughout the colonial and antebellum American south and the adoption of a system of enslaved African and African American labor which developed from the late seventeenth century to supply the labor necessary to produce agricultural commodities for export such as tobacco, sugar, coffee, rice, indigo, corn, and wheat (Gottschalk 1945:219- 223; Russo and Russo 2012). In 1619, the first twenty Africans were brought to the English settlement at Jamestown. By 1650, enslaved Africans made up approximately 2% of the population of the Chesapeake (Taylor 2002:142). Initially, these forced migrants had many of the same rights as their white, indentured counterparts. However, toward the middle of the seventeenth century, racial legislation in many of the British colonies curtailed many of these rights and codified legal distinctions between black and white laborers. These include the right to bear arms, the right to intermarry, and the lengths and conditions of servitude—including lifetime servitude (Reich 1998:124-125). At first, it was generally believed that being Christian precluded one from bondage. There are recorded cases where African slaves were freed because of their acceptance of Christianity. However, a Virginia law passed in 1667 declared, that “the conferring of 10 baptisme doth not alter the condition of the person as to his bondage or freedome” (Reich 1998:125). Similar laws were passed in other colonies shortly thereafter. As the religious rationalizations for enslaving some individuals and against enslaving others were eliminated, racial ones slowly took their place (Reich 1998:125). Enslaved black labor slowly replaced indentured white labor in the Chesapeake toward the end of the seventeenth century (Epperson 2001; Taylor 2002:153). These racial justifications for slavery were firmly entrenched by the end of Bacon’s Rebellion in 1676 as class tensions between whites were diffused by the landed elite who successfully reoriented the primary social distinction not between rich and poor but between black and white (Taylor 2002:139-140), an underlying social division which continues into the present. Slavery was fundamental to the economy of colonial and antebellum America. In the seventeenth and eighteenth centuries, the economy of the English Chesapeake relied on the agricultural production of crops like tobacco, a process which is extremely land and labor intensive (Taylor 2002:142). While it took a Civil War and the adoption of the 13th, 14th, and 15th Amendments to the Constitution after the war, American slavery officially ended with the Emancipation Proclamation, issued in 1862 and taking effect on January 1st the following year, the underlying racial and economic justifications for slavery continued unabated, and even the legal interpretation of these Reconstruction Amendments was rolled back in the following years (Foner 1999). In many places, former slaves continued living on the same plantations doing the same work for their former owners but now in exchange for low wages (Royce 1993:1-3). Where formerly slave codes existed that limited the rights 11 of enslaved African Americans, Jim Crow laws were passed to limit the rights of free and newly freed African Americans (Tischauser 2012:1-16). Despite the achievements of the Civil Rights Movement of the mid-twentieth century, these racial fault lines continue into the present and their origins can be traced back to the plantations of the colonial Chesapeake and American South (Durant and Knottnerus 1999; Epperson 1990, 1999, 2001). Since the abolition of slavery, its visible memory has largely been erased from the plantation landscape; however, its larger legacies still exist in other forms to this day. After emancipation and the war, many of these plantations fell into a state of neglect and disrepair following the loss of their primary labor force. Quarters, barracks, and cabins were no longer needed and were either repurposed, torn down and salvaged, or allowed to fall down on their own. Starting in the late nineteenth and early twentieth centuries and closely related to the Colonial Revival movement, a body of literature began to develop describing many of these privately owned plantation houses, their gardens, their landscapes, the biographies of their owners and their families, and the historic events that transpired at these homes in the colonial and early federal past. Rather than address the system of enslaved labor which made these plantations possible, these sources reflect a strong bias toward narratives of beauty, dignity, and glory of the main house, its gardens, and the almost-exclusively white owners, as was the emphasis of the overall movement (Cangelosi 2013:lxxix-lxxxi; Hardwick 1997:21-33; Little 2012:73). Lydia Brandt has even argued that the Colonial Revival movement helped give shape to the Southern “Lost Cause” narrative (Brandt 2009). Many of these texts are organized geographically and produced by 12 local historians, architects, and historical societies. Some of these texts of relevance here include: Tales of Old Maryland: History and Romance on the Eastern Shore of Maryland (Shannahan 1907); Colonial Mansions of Maryland and Delaware (Hammond 1914); Homes of the Cavaliers (Scarborough 1930); Gardens of Colony and State: Gardens and Gardeners of the American Colonies and of the Republic Before 1840 (Lockwood 1934); Early Manor and Plantation Houses of Maryland (Forman 1934); Tidewater Maryland Architecture and Gardens (Forman 1956); Old Buildings, Gardens, and Furniture in Tidewater Maryland (Forman 1967); and Where Land and Water Intertwine: An Architectural History of Talbot County, Maryland (Weeks 1984). These sources can be seen as a product of the larger Colonial Revival movement. The collective discourse on plantations they were informed by and in turn helped to shape is very much in the same vein as that of Margaret Mitchell’s Gone with the Wind (1936) (Rahier and Hawkins 1999), a discourse that has very much embedded itself within the collective public consciousness. For a particularly good example of this literature and the kinds of historical discourse it embraces and reinforces, see William A. Kinney’s April 1954 article in National Geographic titled “Roving Maryland’s Cavallier Country”. It includes sites beyond the plantation and beyond Talbot County and presents a past that is nostalgic, quaint, hospitable, and overwhelmingly white. Until relatively recently, these narratives went largely unchallenged at these sites (Ruffins 2006). Furthermore, the legal framework behind cultural resource designation and management in this country further buttresses these narratives. In 1966, Congress passed the National Historical Preservation Act (NHPA), which among other things, 13 authorized the National Park Service (NPS) to “’expand and maintain [a] National Register of Historic Places,’ including properties of local, state, and national historical, archaeological, cultural, and architectural significance” (King 2004:22). The National Register’s website, which is maintained by the National Park Service proclaims across the homepage, “The National Register of Historic Places is the official list of the Nation’s historic places worthy of preservation.” (National Park Service 2018). Elsewhere on the website, it is reiterated that, “The National Register is the official federal list of districts, sites, buildings, structures, and objects significant in American history, architecture, archeology, engineering, and culture.” (National Park Service 1999). The National Register of Historic Places has become the de facto standard for historic preservation and historical significance in the United States, along with to lesser degrees similar state and local level historic registers. Across the country and especially in the Chesapeake, many plantation sites are included on the National Register. It is difficult to say exactly how many plantation sites there are on the National Register because the definition of what constitutes a plantation can be vague, but a cursory search of National Register of Historic Places nomination forms held online by the National Archives (https://catalog.archives.gov/search) for the word “plantation” returns well over 5,000 results and a search of just the title field of National Register nomination forms held by the National Park Service (https://npgallery.nps.gov/NRHP) returns 503 results. Primarily included on the Register under eligibility criterion C, which makes eligible those sites “that embody the distinctive characteristics of a type, period, or method of construction, or that represent the work of a master, or that possess high 14 artistic values, or that represent a significant and distinguishable entity whose components may lack individual distinction” (36 CFR 60.4), these main plantation houses are officially designated historic resources, with the backing of the United States federal government. Two issues that arise from this system of officially designating locations of heritage are relevant here. First, because of the language of this legislation, the way in which it has been implemented, and the documentation required to rise to the level of significance and prove integrity for inclusion in the National Register, emphasis is given to the main house or plantation building as the primary historic resource at the exclusion of any outlying ancillary structures or surrounding landscape elements which would have been instrumental to the operation of these plantations in their economic or social functions. Of the Talbot County, Maryland plantations that will be discussed in Chapter 5, most have several pages of text devoted to documenting the construction sequences of the main house, the sequence of families that owned the main house, and the specific architectural features that make this structure unique and significant for inclusion on the National Register of Historic Places, while sometimes only a small paragraph or two to include all of the surviving outbuildings and landscape elements, if any is to be given at all. This is not to say that this imbalance is deliberate or intentional, only that there is a structural imbalance concerning the documentation and inclusion of historic plantation resources for inclusion on the National Register of Historic Places. The effect of this is to privilege the slave owners who lived in these houses over the enslaved who worked on the plantation and shapes the discourse around these sites. A second issue to raise here with the National 15 Register stems from the first issue and is that in declaring these sites as historically significant, the story told by and at these sites becomes official, sanctioned by the power and authority of the National Park Service and in turn the federal government, thus further reinforcing these interpretations of plantation spaces. Despite (or perhaps in some cases because of) the structural problems one might identify with using the National Register of Historic Places to define heritage, surviving plantation sites across the Western Hemisphere have enormous potential as sites dedicated to the presentation and interpretation of slavery. Not only did slavery happen at these sites but their very existence is owed to enslaved labor. A discussion of these houses and their histories should theoretically be inseparable from a discussion of slavery because the two are so intertwined. Interpretation and discourse at and produced by these sites overwhelmingly focuses on the plantation owners at the expense of those enslaved there (Jackson 2008). However, as A.V. Seaton notes, heritage and memory is a flexible system of meaning, subject to conscious and unconscious influences by its users. Seaton writes, “…plantations and other [related] locations are polysemic signs, capable of many different interpretations, unless they are scripted, coded and marked to anchor meanings to slavery” (Seaton 2001:117). This observation underlies the most significant problem associated with the creation, perpetuation, or modification of heritage at these locations. Those individuals and groups that own these sites are primarily the ones that are in a position to script, code, and mark meanings at these locations. As is discussed by many of the authors below, these individuals and groups are primarily white, upper-class and by and perpetuate the erasure of slavery from these sites and in public memory by choosing not to 16 confront it or in doing so in a manner that perpetuates it nonetheless. Those with control of these sites have a vested interest in erasing the presence of slavery from the historical landscape and in doing so re-anchor the values transmitted by these places, including hospitality, gentility, and elegance and hinder, halt, or even reverse any attempt to address the history and legacies of slavery which still reverberate into the present. In introducing their book Slavery, Contested Heritage, and Thanatourism, Graham M.S. Dann and A.V. Seaton (2001) problematize heritage at sites of slavery and identify additional questions in need of resolution: Should such sites be memorialized? If so, what ethical issues have to be confronted and resolved? How should history be presented as a heritage subject? Who should control the forms of heritage development at dissonant sites? Whose past should be privileged? How, in pluralistic societies with a diverse ethnic mix, is it possible to narrate histories that include all constituent variants equitably? (Dann and Seaton 2001:25) Central to this critical evaluation of heritage is the post-modern constructionist theory of history as described by Richard Handler and Eric Gable in their 1997 book The New History in an Old Museum: Creating the Past at Colonial Williamsburg (Handler and Gable 1997:60-70). It is through their approach, a critique of the presentation of historical research at Colonial Williamsburg, that the role of researchers, scholars, interpreters, docents, authors, or other interlocutors of the past have in the creation of history is openly acknowledged. They, like others before them (perhaps most notable to archaeologists is Michael Shanks and Christopher Tilley’s 1987 book Re-Constructing Archaeology: Theory and Practice) challenge the traditional objectivist approach to history in which facts are held to be immutable and believed to be able to speak for themselves. As Handler and Gable show through their 17 work at Colonial Williamsburg, this conceptualization of history is extremely powerful at maintaining the status quo, especially when presented from a position of power. In doing so, power relations in the past are naturalized and carried into the present unchallenged (Handler and Gable 1997:220-235). As argued by the authors discussed below, the primary way history is “created” or otherwise experienced at these sites is through the model of tourism. These researchers have examined the ways in which slavery is discussed at a specific kind of plantation site, that of the plantation museum. Each of these studies approached their analysis differently, but arrive at the conclusion that slavery is silenced, marginalized, or in some way distorted for tourists—the primary consumers of history—at these sites. A textual analysis of 102 promotional brochures from plantations across the South conducted by David Butler (2001) revealed that site ownership is a significant factor in whether slavery is discussed at a particular museum and to what extent it is discussed there. Federally owned sites were the least represented in his sample (only 3% of the sample) but these sites were the most inclusive and thorough in their representation of slavery. Sites owned by a state (30% of the sample) were likely to feature research on slavery at that site; however, this did not always manifest itself in the interpretation at the site. Privately owned sites (43% of the sample) typically did not mention slavery; however, in talking with them, Butler found the owners were generally very interested in the history of their plantations, including slavery. Properties owned by societies (like historical societies) (10% of the sample) would refer to slavery at the site if the society’s financial resources permitted. At sites owned by foundations (typically dedicated to promoting 18 the legacy of a particular family or house) (13% of the sample), slavery was almost totally ignored as it was not compatible with the narrative they were presenting. For the most part, the presence of slavery at these sites was either non-existent or marginalized in some way (Butler 2001:168-169). As Butler discusses in his analysis, each of these groups that control access to these plantations and are responsible for the narratives that are presented there has different motivations, interests, and resources available to them that factor into the way they present and interpret their sites. Seaton (2001) continues this argument by pointing out that discourse is not created solely by individual historical interpreters or by museum organizations. In addition to the plantation museum owners, they identify three other groups, each with their own interests and goals, all of whom interact with each other through these sites to produce a discourse on heritage in what Seaton calls “The Heritage Force Field” (Seaton 2001:123-126). In addition to the Owners/Controllers of these sites, Seaton identifies the Subject Group, those whose heritage is being presented at these sites, Host Communities, those residents near the heritage site, and Visitor Groups, those people most likely to visit and consume the narrative being presented at the site. Because of the diverse make-up of these constituencies, the topography that makes up this Force Field changes from site to site and is not static in time but changes as discourse changes. One of the strengths of this model is that it posits that its participants are not inherently at conflict with each other, that the interests and goals of these groups can (and do) overlap. The least contentious heritage sites are those where the owners, subjects, hosts, and visitors are the same group or are groups with similar backgrounds, missions, and desired 19 outcomes. At the opposite end of the spectrum, the most contentious heritage sites are those where the controllers of the site represent, marginalize, or even ignore subject groups without their input using narratives that are foreign to the subjects of the interpretation. These are two extreme cases of heritage discourse formation at opposing ends of a spectrum that includes an infinite number of configurations and possibilities and is in a constant state of flux as discourse evolves in response to these and external factors (Seaton 2001:123-126). Fundamental to understanding and implementing the Heritage Force Field model is the fluid nature of what constitutes heritage and how it changes. The model emphasizes embedded power relations, individual agency, and the relationships and interactions between these parties and external factors. Seaton explains: The result of all these impacts over time is that heritage is never a stable, finally completed process but a constantly evolving process of accommodation, adjustment and contestation….The Force Field model seeks to depict how the agencies and social actors involved in heritage development, and the political and temporal environments in which they interact, may produce a more complex configuration of influences than the assumption of this two-cornered fight between truth and falsehood. (Seaton 2001:126) In utilizing this model, not only do we identify some of the ways in which heritage is created and where the conflicts in doing so can occur, but we also come to realize the temporal component of such a creation and can trace its development through time. This is the same process that Foucault embarks on in his genealogical analysis. In doing so, we recognize the arbitrariness of its present configuration (arbitrary not in that it is random, but in that it need not have turned out the way it has) and through this we can demystify it and begin to counter the narrative it presents as natural and timeless (Leone 2010:181-183; Rabinow 1984:59). 20 In their book Representations of Slavery: Race and Ideology in Southern Plantation Museums, Jennifer Eichstedt and Stephen Small (2002) further emphasize the ways in which discussions of slavery are systematically silenced or distorted at plantation museums. They visited 122 plantation sites across the country and found that every one of these 122 sites engaged in some form of silence or distortion. Furthermore, they characterize and classify the range of ways they encountered to include the symbolic annihilation of slavery, trivialization or deflection of the topic of slavery at the site, segregating discussions of slavery away from the primary plantation house or tour, or what they title “relative incorporation” of slavery within the narrative presented at the site. They classified symbolic annihilation as an interpretation that involved discussing slavery either not at all or fewer than four times during the tour. By not discussing slavery, it is effectively erased from historic discourse and heritage making at these sites. They classified the trivialization of slavery as an interpretation that tried to ameliorate the conditions of slavery at the site or presenting them as not being that bad. They classified deflection as an attempt to change the subject away from slavery so that it would not have to be discussed in any further detail. Segregated discussions of slavery took two forms. First, the physical location where slavery was mentioned was physically removed from the main interpretive framework of the plantation tour. There was frequently a set place to talk about slavery, and only there could it be discussed. This frequently took the form of outlying slave cabins that were difficult to get to or not readily visible from the rest of the plantation. The second form of segregation involved specific tours to address slavery at the site. While these tours generally addressed slavery in a much more 21 inclusive and comprehensive manner, the very fact that they were distinct tours separate from the main plantation tour places conversations about slavery in a confined space, separate from the rest of the historical narrative being created about the place. These tours are self-selecting in that generally only those people already interested in the topic of slavery will attend them and, in so existing, they establish the previously existing tour of the main house as the “normal” or “regular” tour. Lastly, they did not encounter relative incorporation at very many sites, but these sites address the historic realities of slavery within the context of the main plantation tour; however, this still happens within the context of the larger plantation heritage complex and unable to address some of the problems associated with it. These techniques for dismissing slavery can be seen to reside on a continuum of inclusion with symbolic annihilation at one end and only relative inclusion at the other. The final step on this scale would integrate African American history throughout the historic narrative being told at these places; however, at none of the sites they visited addressed African American heritage on equal footing as the distinctly Euro American heritage presented at the site. One of the significant findings Eichstedt and Small arrive at in their study is that at all of the plantations they visited the presentation and discussion of slavery and African American heritage was subordinated to the presentation and discussion of the great house and its white inhabitants. They conclude that while some sites do a better job at presenting the realities of slavery, it is probably not possible to do so in an adequate manner without first seriously challenging the dominant form of heritage 22 creation centered on plantation house museums as the discursive sites of narrative formation. They write: The mainstream plantation museum infrastructure creates and re-creates a discursive framework of the South as genteel, honorable, and romantic….This larger discursive framework of the South and the strategies used to contain and organize the discussion of slavery resonate with larger racial formations, as well as racial formation projects that elevate whites and whiteness and subjugate African Americans and other people of color. (Eichstedt and Small 2002:258) In trying to understand the ways in which the narrative regarding slavery is distorted, E. Arnold Modlin, Jr. (2008) conducted an analysis very similar to that of other authors presented here. He visited 52 plantation sites in North Carolina, and analyzed the frequency and manner in which docents discussed slavery while giving guided tours. His findings are in line with the other studies presented in this chapter, but his is worth repeating here because he detailed and categorized the specific ways in which tour givers at plantation museums ignore and deflect the issue of slavery. He identified myths that play across multiple scales, including national, regional, state, local, and site-specific levels and found that many individual myths can operate on multiple levels at the same time. These were used as narrative frameworks by docents and repeated to visitors while on tours. The specific tactics Modlin identifies are as follows: • Focusing on the Civil War; • Slavery was rare in the North; • Slavery did not exist in the North; • Slavery was only found on plantations; • Virginia as the birthplace of democracy; • Louisiana and the tragedy of the Civil War; • Georgia and Gone with the Wind; • North Carolina and religious settlers; • Slavery did not happen here; 23 • Slavery was somehow unique here; • Some blacks were slave owners; • The owner of the house did not own slaves; • Silence; • Little is known about the enslaved; • Life was hard for everyone; • Using the terms ‘servants’, ‘workers’, or ‘craftsmen’; • Slaves as family; • The good master; • The faithful slave; • The atypical slave. (Modlin 2008:277) Note how these specific narratives overlap and fit into many of the categories defined above by Eichstedt and Small. Modlin’s goal in identifying these tactics is to operationalize the specific ways in which slavery is silenced or marginalized at these sites. Once they are identified and defined, they can be challenged when encountered and a new discourse can begin to be formed. The end result of these kinds of historic interpretation encountered by these above authors at these sites of heritage formation is that the established narrative of early American history that comes from these plantations is one of whiteness (Eichstedt and Small 2002:140-145). Slavery and the African American experiences on these plantations are ignored, silenced, marginalized, segregated, or otherwise distorted and subjugated to the experiences of the white plantation owners and their families. This normative discourse largely valorizes the white slave owners and absolves them of their responsibility for slavery and at the same time eliminates the role of African American labor which was fundamental to the economic success of both the early Chesapeake and America. With this observation, the answer to the question, “Why revisit such a shameful chapter of American history?” becomes 24 apparent—because ignoring it or allowing this distortion to continue warps these cultural relations in the past, normalizes this distorted version of their origin, and perpetuates this distortion into the present. A critique leveled against tourism as an industry, and especially tourism at plantations and other sites of colonial exploitation, is that it perpetuates the same colonial economic model and power relations that it objectifies. In his “Ideology and Ideological State Apparatuses (Notes Toward an Investigation)”, Louis Althusser (1971) identifies a cluster of embedded social institutions that through their surface- level functioning act to perpetuate the ideologies that led to their creation, further reinforcing this ideology. These plantation museums are very much operating as what Althusser would recognize as Ideological State Apparatuses. In a book chapter titled “Supplanting the Planters: Hawking Heritage in Barbados”, Graham Dann and Robert Potter (2001) discuss how modern plantation museums operate metaphorically in the same manner as their colonial antecedents. Rather than producing an agricultural crop for export to the European market, these sites now export a nostalgic experience for a largely European-descended clientele. The great house still acts as the focal point of these landscapes and acts to mask the presence of an exploited working class, whose legal and economic subjugation enable these structures to exist in the first place (Dann and Potter 2001:75-77). They conclude with a quote from C. Palmer, “Colonialism was concerned with power, domination and control and with the superiority of one group over another through the perpetuation of inequality. These are all criticisms that at some time or another have been leveled at tourism” (orig. Palmer 1994:808). 25 Returning to Handler and Gable’s The New History in an Old Museum, an additional hurdle to presenting slavery and plantations as tourism is found in their discussion of “good vibes” at Colonial Williamsburg. They discuss the importance of impression management at Colonial Williamsburg and the emphasis placed on guests enjoying themselves. This follows from Colonial Williamsburg being operated as a site of tourism and its connection to the hospitality industry that caters to these tourists. From interviews with employees, they report that just about the only thing somebody could get an employee fired for was for being rude to a visitor. Because of institutional pressures, Handler and Gable conclude that what ended up being most important to frontline interpreters, staff, and management, was that visitors left feeling good about themselves. The business logic argues that if people go to a tourist attraction and leave upset, they will probably not return and that is bad for business (Handler and Gable 1997:172-207). Butler emphasizes this point by observing that many tourists visit plantation sites to enact and experience a form of escapist fantasy (Butler 2001:171). This consumer behavior underlying the tourism model of heritage creation and consumption is fundamentally at odds with exploring an uncomfortable discourse that addresses slavery or otherwise attempts to reconfigure the narrative created and reinforced at these sites. The problem then arises of how to address controversial or unpleasant aspects of history in a museum setting without alienating the visitor, hurting ticket sales, and harming these sites’ ability to remain open and operate. Unfortunately, as evidenced by Handler and Gable, even large, internationally-known institutions like Colonial Williamsburg have trouble navigating this question. 26 While Modlin, Eichstedt and Small, Butler, and Handler and Gable have identified a range of methods used to silence, marginalize, and distort the presence, role, and legacy of slavery at plantation museums, and the ways in which historic museums struggle to talk about these uncomfortable topics, a significant gap remains unidentified and unaddressed in this body of research. Their datasets pull from plantations operated in the present as museums, however, there are significantly more eighteenth-and-nineteenth-century plantations that have survived into the present that are not operated as museums. These are primarily held as private, residential dwellings, closed to the public, and not operated under the museum or tourist model. As such, these sites have a fundamentally different relationship to discussions of slavery and heritage formation, and Seaton’s Heritage Force Field model looks very different at these sites. As a comparison between these two subsets of plantation sites, a preliminary study by Ryan O’Connor for Archaeology in Annapolis identified a subset of 48 of former plantation properties on the National Register in three Maryland counties, Anne Arundel, Queen Anne’s, and Talbot (Table 2.1). Of these (excluding those in the City of Annapolis), only three (Hancock’s Resolution, Sotterly, and possibly Reed’s Creek Farm) operate like the museums described above. The remaining 45 of these nationally significant historic properties, such as at Wye House, which will be discussed extensively throughout this dissertation, are the private homes of wealthy, predominantly white individuals. In a few rare cases, the properties are owned by descendants of the same families that originally built them; however, the majority are owned by individuals with no historic connection to the house. 27 NR Listing Name Construction Date AA-129 Hancock's Resolution 1785 AA-134 Evergreen 2nd & 3rd quarters of 1700s AA-136 Howard's Inheritance 1760 AA-138 Tulip Hill 1755, enlarged 1787-90 AA-141 Cedar Park 1697, expanded 1800 AA-151 Larkin's Hill Farm mid 1600s AA-152 Larkin's Hundred 1704 AA-153 Mary's Mount 1742 AA-155 Obligation 1743 AA-169 Primrose Hill 1760 AA-177, AA- 177A Bunker Hill 1820 AA-183 Belvoir 1690, expanded mid 1730s AA-191 Rosehill mid 1700s AA-228 Norman's Retreat 1811 1683, expanded 1720 & AA-229 Sudley 1750 AA-232 Gresham late 1700s AA-253 Portland Manor 1725 AA-257 Burrages End 1720-40 AA-268 Holly House 1667, enlarged 1730 AA-325 Whitehall 1764 AA-330 Sandy Point Farmhouse 1815 AA-485 Brice House 1766-73 AA-626 Hammond-Harwood House 1774 AA-628 Chase-Lloyd House 1769-74 Gov. William Paca House and AA-657 Garden 1763-65 AA-726 Upton Scott House 1762-65 QA-3 Kennersley 1785-98 QA-4 Bloomingdale 1792 QA-5 Reed's Creek Farm 1775 QA-7 Bowlingly 1733 QA-8 Thomas House 1798-1821 QA-9 Readbourne 1730, expanded 1791 QA-11 Content last quarter of 1700s QA-23 Chester Hall 1790s QA-90 Stratton 1790 QA-127 Wilton 1749-1770 QA-206 Legg's Dependence 1760-80 QA-224 Bachelor's Hope 1798-1815 T-48 Crooked Intention 1717-1753 T-52 The Anchorage 1720s T-53 Myrtle Grove 1734 28 T-54 Wye House 1781-84 T-90 Hope House 1740 T-146 Compton before 1794 T-149 The Wilderness 1780-90, expanded 1815 1720-30, expanded 1800- T-164 Otwell 10 T-189 Clay's Hope 1783-98 T-244 Sherwood Manor 2nd half of 1700s Table 2.1 So while the research presented here shows there is a significant problem with addressing slavery at plantation heritage sites, these sites are only a very small subset of this historic resource base. On these private sites, there is no public discussion of slavery. The end result is an overwhelming erasure of slavery from these former plantations, from their landscapes, and from the past. This is the discourse on plantations that historical archaeology has entered: picturesque, stately, genteel, hospitable, and white. Within many academic circles and in some notable museums, this seems to have been rejected; however, on both museum and privately owned sites, an honest, equal recognition of slavery is fundamentally lacking. (Ruffins 2006). Archaeology and Plantations It is also within, and frequently in direct reaction to this context that some historical archaeologists approach the plantation. As listed by Barbara Little in Historical Archaeology: Why the Past Matters (2007), the goals of historical archaeology include, “preserving and interpreting sites, supplementing and challenging the history we know through documents, reconstructing people’s ways of life, improving archaeological methods, and understanding modernization and 29 globalization” (Little 2007:22), and archaeologists working on plantations have addressed all of these goals. Like Little, whose definition has historical archaeologists in part supplementing and challenging traditional histories, Kenneth Feder frames the field and argues that, “giving voice to the voiceless of history—the poor, the oppressed, the disenfranchised—is among historical archaeology’s most important contributions to anthropology” (Feder 1994:16). This is perhaps archaeology’s single most important contribution to historical studies. In fact, one of the earliest areas or subfields to be defined within the field of historical archaeology was plantation archaeology, a study of their landscapes, and an engagement with the enslaved populations on them. Starting in the late 1960s and into the 1970s, Charles Fairbanks began excavating at Kingsley Plantation for the Florida Park Service, which had the intention of rebuilding a slave cabin for interpretive purposes. While areas of the plantation had been excavated previously, none had been excavated with the intention of uncovering the material culture of the enslaved Africans and African Americans whose coerced labor enabled the plantation to function as a viable economic unit (Fairbanks 1984:1-2). As explained by Fairbanks, one of the early goals of plantation archaeology was to “search for Africanisms among the material artifacts of those newly arrived slaves, evidence of adaptation in housing, dress, behavior to the new situation, and data on lifestyle” (Fairbanks 1984:2). “Africanisms,” or African material cultural survivals were hoped to exist archaeologically, even if historical evidence for them was sparse. Perhaps what practitioners were looking for were obvious African holdovers on the plantation, such as the circa 1750 round brick quarter at Keswick 30 Plantation in Midlothian, Virginia, which is twenty-five feet in diameter with a conical roof or the “African House” at Melrose Plantation in Natchitoches, Louisiana, built by a former slave sometime after 1778, which has a massive overhanging roof and porch. Both of these structures are unlike contemporary European or American vernacular building traditions and are unmistakably African in character (Jones 1985). While researchers found no evidence for these things or processes at Kingsley Plantation, this question, “what from Africa has survived in America?” coupled with the related question, “how has a uniquely African American culture been created in America?”, has subsequently guided several generations of historical archaeologists (Fennell 2007; Ferguson 1992; Moyer 2015; Samford 2007; Singleton 1985). These researchers have since excavated on plantations and on sites associated with race- based slavery not only in the American South but also throughout the larger British and European colonial world in an attempt to answer these two research questions and to address one of the issues raised regarding discourses surrounding slavery and its legacies earlier in this chapter. In recent decades, archaeologists have begun to uncover and recognize evidence of African worldviews, cosmologies, and spiritual practices on American sites (Cuddy and Leone 2008; Fennell 2007; Ferguson 1992; Leone 2005; McDavid 1997). While Ferguson’s work on the subject started with specific artifacts marked with potential African symbolism, others like Leone have expanded this search to include site plans and patterning of symbolic artifacts across sites while others like Fennell have begun to ask questions regarding the relationship between landscape and cognitive or cultural worldviews. 31 As hinted at by Fennell in his work on cosmologies and worldviews, the spatial arrangement of phenomena can be significant. Likewise, inherent within these plantation systems is a spatial dimension to their operation. As such, landscape studies have become an integral part of studying these plantations. They are comprised of agricultural fields, undeveloped forestland, buildings and work areas to support agricultural production, buildings and barnyards to house and feed animals, quarters to house the enslaved workers, manor houses wherein dwell the owners and their families, offices to organize and manage the entire system, and roads, rivers, and waterways connecting the various components. Given the plantation model of agricultural production, all of these are situated and organized in space to increase agricultural productivity and efficiency within a capitalist world-system (Lewis 1985:37). The plantation landscape itself is arranged in space according to a certain cultural logic, the logic of capitalism and the Georgian Order. This spatial dimension of landscape should be variable through time and space and sensitive to the types of agricultural activities, technologies, labor, and products on the plantation and capitalist system. Because this dissertation is a study in the archaeology of landscapes of slavery and plantations, a notable study to include here is Cheryl LaRoche’s work on the geographies of freedom and resistance and the Underground Railroad (LaRoche 2013). LaRoche examines the ways free Black communities in Illinois, Indiana, and Ohio interacted with and supported these landscapes of freedom. Her book specifically highlights the role of Black churches, cultural institutions, and iron furnaces in these clandestine landscapes based on movement. LaRoche writes: 32 The landscape is an intimate yet underexplored component of the Black experience, where danger lurked and freedom beaconed. Although the land was a site of disorientation, hardship, frostbite, and starvation, it also held crucial pathways out of slavery. Generations of escapees on the Underground Railroad turned to the sheltering anonymity of the land to conceal their journey. (LaRoche 2013:87) These landscapes, only temporarily occupied by those seeking their freedom, are similar to those more permanent landscapes of marronage described by Sayers (2014), Chowdhury (2014), Goucher and Agorsah (2011), and Weik (1997). These spaces where the enslaved or otherwise marginalized people could either temporarily or on a much more permanent basis claim their freedom show us the spatial and geographic limits and limitations of plantation slavery and capitalism. Furthermore, LaRoche reminds us that running away is only one of the methods available to the enslaved seeking their freedom. She also documents several other resistance strategies with the aim of removing oneself from bondage used by the enslaved in Maryland, including self-purchase, opportunities during military conflicts, revolts and revolutions, colonization efforts, or petitioning the court (2007). Inspired by folklorist Henry Glassie’s influential structural analysis of vernacular architecture in Virginia (Glassie 1975), James Deetz identifies a dramatic shift in the material culture of seventeenth-and-eighteenth-century America in his classic 1977 book, In Small Things Forgotten. Deetz correlates the rise of the Georgian Order in architecture and the rise of the larger concept of individualism as a fundamental transition in the historical American consciousness and describes how this shift in worldview through time is reflected in material culture. He specifically details changes in house plans, table settings, and gravestones as evidence of this 33 change and holds these analyses up as an example of how archaeological research into material culture can address much larger cultural changes. This new worldview (or some might call it an ideology) of the Georgian Order, that arises in the second half of the eighteenth century, embedded within the vernacular houses described by Deetz is the dominant worldview of modernity and capitalism (Leone 1988). While the Georgian Order can be seen embedded within the physical organization of these plantations, recent work in historical archaeology seeks to understand how these spaces were approached and appropriated by those whose exploited labor made them possible and whether they accepted the worldview implied by the Georgian Order. One influential contribution to archaeological theory relevant to this discussion is Mark Leone’s analysis of the William Paca garden in Annapolis, Maryland (Leone 1984). This 1984 chapter is significant not just for its analysis of a single site, but also for the conversation it has fostered within the field. Leone uses Marxian concepts of power, hierarchy, domination, and ideology to describe the conscious and unconscious reasons behind why Paca built his garden and how it functioned, not in the botanical sense, but in an ideological one. A significant inspiration for these studies has been the work of Michel Foucault, especially his work on discipline, surveillance, and observation (see Foucault 1977). In this interpretation, gardens like William Paca’s are implicitly conceived of and built by a group of largely white, land-owning males of European descent interested in either maintaining their own power and position in society or in an attempt to acquire this power and position which they do not yet have. Leone argues that the gardens 34 function by naturalizing and masking inequalities and social distance between those like Paca who build them and those for whom their message is intended. To make this connection between landscape and power explicit, Chandra Mukerji ties formal gardens and landscapes to the development of the European state in her 1997 book, Territorial ambitions and the gardens of Versailles. She postulates that formal gardens were used by the French monarchy at Versailles to consolidate and solidify both the symbolic and the real power and authority of the modern French state. As does Leone, she at least partially locates the authority of political actors through gardens and landscape. Through Foucault, Leone, and Mukerji, one can view formal landscapes and control over space as an attempt to secure or consolidate political, economic, social, or ideological power. Ultimately, this is a masking ideology, namely the naturalization of power and authority within the natural, and not human, world. The effect of this is to put social hierarchies beyond contest, “[f]or the order was natural and had always been so” (Leone 1984:34). Implicit within this understanding of plantations is the plantation system and capitalism operating as dominant ideology. It is understood that the power within the system resided at the top with the planter elite and that the plantation system was organized around the principles of the plantation within a larger capitalist system. Leone founded Archaeology in Annapolis (AiA), a partnership between the Department of Anthropology at the University of Maryland, College Park and the Historic Annapolis Foundation in 1981 in order to “[examine the] structure of power based on changing wealth” (Leone 2005:31). AiA’s extensive bibliography can be accessed online at www.aia.umd.edu/bib. Much of the early work produced by Mark 35 Leone and researchers with AiA focused on this dominant ideology, how it developed, and how it was maintained within eighteenth-and-nineteenth-century Annapolis. Leone identifies how the formal gardens and landscapes created by colonial elites replicated and reinforced this ideology (Leone 1984), how the Georgian Order and symmetry embedded within most eighteenth-and-early-nineteenth-century plantation landscapes mirrors the capitalist ideology (Leone 1988), and how colonial town plans are expressions of power (Leone and Hurry 1998). These studies have given researchers a way to examine how power, dominance, surveillance, and ideology function across plantation landscapes. What needs to be emphasized is that while the dominant ideology of modernity and capitalism is focused on here, it is but one worldview operating across these sites and landscapes. One question this dissertation hopes to continue in asking is how to recover these other alternative, non- dominant worldviews or ideologies on plantations. Since 2000, AiA has branched out from Annapolis and included a focus on plantation sites and other sites of African American history around the Chesapeake Bay and its tributaries. Because Leone’s chapter on the Paca garden has invited so much commentary from archaeologists, Dan Hicks has labeled it a “place for thinking,” a term he borrows from Leone in describing how the garden was used as a site for ideological transmission and replication, in that it is a site that scholars can and do frequently return to in order to apply or critique both familiar and new theoretical frameworks (Hicks 2005). Of those who have returned to the Paca garden to discuss Leone’s use 36 of dominant ideology as well as propose new ways of thinking about the archaeological record, two are presented here. The first is Ian Hodder, who argues that Leone’s top-down approach dismisses the possibility of agency for those who are subjected to the message of the garden (Hodder 1986; reprinted Hodder and Hutson 2003). He argues that Leone’s early use of the dominant ideology model as used in the 1984 chapter is at least only a partial understanding of the garden and similar spaces. He suggests instead that there is an interplay, or a negotiation, between dominant and various subverted ideologies that play out through the relationships between people, as well as with material culture, of which the garden and landscape are examples. He voices a skepticism regarding Leone’s assumption that all visitors would have translated the language of William Paca’s garden in the same way or even in the manner in which Paca intended it to be. For the purposes of this dissertation, the most significant point of Hodder’s critique is that it is possible to reject the dominant ideology, in favor of an alternative ideology, or at least accept it simultaneously with one of these alternative ideologies. The second critique of Leone’s work in the garden comes from Mary Beaudry, Lauren Cook, and Stephen Mrozowski (Beaudry et al. 1991). They argue that rather than understanding material culture and its meaning as being produced by a dominant ideology (top-down) or in a hyper-local context (bottom-up), there is a textual discourse in which material culture is a part. They argue for an ethnographic understanding of archaeological material culture, one in which they juxtapose the material culture evidence with written sources. They argue archaeologists should “[analyze] cultural texts, written or otherwise, from ‘the inside out,’” and by doing so, 37 this allows us to “reconstruct meaning in the active voice, in the multiple voices of the ‘silent majority’ whose past discourse through artifacts reveals they were not so inarticulate after all” (Beaudry et al. 1991:175). They see material culture as a language, with its own grammar, which can be recovered by the archaeologist, if they know how to read it. The reading of landscapes as texts can be seen throughout much of the work being done under the umbrella of landscape archaeology. When one considers who may contribute to and how one derives meaning from a landscape, this textual metaphor is integral to the use of landscape here in this dissertation. Importantly, in response to Hodder’s and Beaudry, Cook, and Mrozowski’s critiques, Leone has since refined his use of the dominant ideology. Given that the intended meanings of a text or a landscape are not guaranteed, he notes that, “the issue is that if people saw through ideology, there was the possibility of resistance, which could take many forms….So who saw and what did they see?” (Leone 2005:45). Since 2005, Archaeology in Annapolis has been excavating at Wye House, a plantation outside of Easton, Maryland, located at the confluence of the Wye River and the Miles River (Figure 2.2). Originally settled by Edward Lloyd I around 1658, Wye House has been occupied by the Lloyds and their descendants since that time and into the present. The current Wye House is actually the second or third main house to stand on this property, was built sometime between 1781 and 1784, is a seven-part wood frame house in the late Georgian/early Federal style, and is surrounded by a historic core and agricultural landscape that comprised the plantation in the late eighteenth and early nineteenth centuries. 38 Prior to AiA’s engagement with Wye House, much of what researchers know about Wye House comes from fairly typical historical sources. These include plantation account books, ledgers, letters, maps, plats, tax assessments, insurance policies, photographs, and other historical research. However, perhaps one of the 39 most significant sources of information about Wye House during the early nineteenth century comes from Wye House’s (perhaps Talbot County’s and maybe even the State of Maryland’s, depending on how one calculates such honorifics) most famous resident—Frederick Douglass. Douglass was enslaved as a child for several years at Wye House and in each of his autobiographies he writes about his experiences on and memories of the plantation. These textual accounts not only provide another description of the plantation and its buildings and landscapes but also provide this description from the perspective of someone who was enslaved there—a rare opportunity for researchers addressing the experiences of American slavery. Despite this wealth of historic data, and especially despite the vivid descriptions written by Douglass, the story that comes out of Wye House is that of plantations elsewhere— gentility, hospitality, and whiteness. As evidence of this, as recently as the summer of 2012, the magazine What’s Up? Eastern Shore published an article on Wye House titled “A Place Called Wye House” (McNulty 2012), featuring an elegant font punctuated with sweeping cursive accents, beautiful landscapes, inviting photographs of the house, greenhouse, orange trees, and grounds, rustic ruins and cemeteries, and, if one looks closely, what is probably intended to be a Confederate naval jack, differentiated from its better-known battle flag cousin only by the shade of blue used in its cross, laid against a gravestone. While archaeology features in this article, the overall impression of Wye House is of these overarching themes of gentility, hospitality, and whiteness. Despite the inclusion of archaeology, this is a genre of publication that has been produced and reproduced since the late nineteenth century. It is against this entrenched narrative that Archaeology in Annapolis came to Wye 40 House in 2005 at the invitation of its owner to begin an archaeological research project focusing not on Wye House and the generations of Lloyds that have lived there since 1658, but the other buildings on the plantation, the surrounding landscapes, and the heretofore anonymous, enslaved Africans and African Americans who also lived here and called it their home. 41 Chapter 3: Theoretical Framework and Methodologies Theoretical Framework Landscape Archaeology One of the major crucibles for the contemporary American concept of race were the eighteenth-and-nineteenth-century plantations which operated on the basis of a racialized regime of enslaved labor in order to produce an agricultural product to sell on the international market (Epperson 1990, 2001; John 1999). Plantation landscapes were—and still are—racialized landscapes (Durant and Knottnerus 1999). As it has been generally practiced by historical archaeologists, plantation archaeology is an archaeology ultimately concerned with the problem of slavery and the origins of race in America. Any attempt at dismantling, disempowering, or otherwise confronting race as a force in the twenty-first century requires a grounding in the history and development of its discourse; part of this agenda is a deconstruction of whiteness, blackness, and their origins (Epperson 2001). My understanding of the racialized nature of plantation landscapes has its own origins in Dell Upton’s treatment of the subject in an article titled “White and Black Landscapes in Eighteenth-Century Virginia” (Upton 1984). In it, his central thesis is that there was not one eighteenth-century landscape in colonial America, but at least two, operating simultaneously and coterminously, at times independently of each other and at times in direct dialog with each other. Because a system of racially-based slavery was a social reality with profound social, and in turn spatial implications, those of European descent, privileged with legal and economic advantages, and those 42 of African descent, restricted by these same systems, had fundamentally different ways of seeing, being in, thinking about, and moving through the landscape. Upton concludes that the plantation landscape was designed to be met as a physical expression of an idea about hierarchy and how the world is supposed to function (Upton 1984:71), but that the Africans and African Americans enslaved on them were not a part of this intended audience. As such, despite his previous statement regarding the ideological nature of the plantation and its landscape, the plantation was not intended to function as ideology in the specific case of the enslaved. In the absence of ideological persuasion, power and violence were the coercive forces that were used to maintain the institution of slavery. Upton provides a theoretical framework for integrating a top-down study of ideology and power on these sites that is also capable of acknowledging the ways the targets of those forces may or may not have accepted or rejected them. In his article, Upton’s work has two major aims. First, he addresses how plantation landscapes would have been encountered and negotiated by social actors. Because of the dichotomous black/white racialization of eighteenth-century plantation contexts, he lays out a dichotomous view of these spaces. Continuing the metaphor of social action as theatrical or dramaturgical performance established by Erving Goffman in his 1959 book The Presentation of Self in Everyday Life and best introduced to studies of the colonial Chesapeake by historian Rhys Isaac in his Pulitzer Prize-winning The Transformation of Virginia: 1740-1790 (1982), Upton likens these plantations and their landscapes as stages on which social action are enacted. Plantations were places where the plantation elite could figuratively and 43 literally elevate themselves above their white social inferiors as well as the enslaved individuals they owned, reinforcing this difference by controlling physical arrangements of space as well as controlling access to these spaces. Simultaneously, Upton describes how the enslaved Africans and African Americans who negotiated these spaces were able to access the same spaces used for social posturing through side doors and passageways and enter even the most formal spaces unimpeded and unchallenged by these restrictions. From this stems Upton’s second aim, which is to argue that the formality and structure of eighteenth-century plantations were not meant for, or directed at the enslaved community; his most forceful point is that ideology was not necessary to maintain the black/white racial and social hierarchies. He concludes: When it was experienced as intended, [the landscape] could be a powerful and intense ideological statement. But the duties and personal experiences of slaves circumvented this experience. Blacks were not drawn into the social posturing of gentry society, and whites did not expect them to be. The elements of raw power replaced those of ideological persuasion. (Upton 1984:71) With this conclusion, Upton’s work provides an important modification to Leone’s treatment of landscape and ideology and shows us where the latter’s original 1984 interpretation of the Paca Garden falls short. Upton’s treatment of racialized landscapes helps us to consider the possibility of a multivocal plantation landscape. Upton’s key point to consider here is that social position affects the ways in which one moves through and experiences the built landscape. Upton continues parts of his argument in a later chapter, titled “Imagining the Early Virginia Landscape” (Upton 1990). His hypothesis here is that from the perspective of the enslaved, the plantation landscape was ultimately about control. There were areas like the main 44 house where the plantation owners were able to exercise the most control, areas where the enslaved were generally allowed a larger degree of autonomy such as yards and gardens, and areas largely outside the control of the plantation owner such as the woods surrounding the plantation. Upton argues that because of the nature of slavery, the means used to enforce it, and the skills necessary to negotiate it, the enslaved came to see their landscapes as a series of loci wherein they are either supposed to be under the control of somebody else or they possess varying degrees of freedom and autonomy. Upton cites accounts provided by travelers to the Chesapeake complaining how difficult it was to get directions from the enslaved they encountered along their route. Rather than describing the travel route in relation to a traveler traversing an open network of space, their directions included specific landmarks and these landmarks’ owners. Rather than relaying the path the traveler was to take, the direction-giving enslaved highlighted points along that path. Upton’s argument is that as far as these enslaved direction-givers are concerned those are the salient pieces of geographic information that underpin their mental map of their local world. These two different understandings of the landscape, how to give directions, and how one is to move through it are fundamentally different. Upton concludes from the level of frustration caused by these directions that caused the recipients to record them for posterity that they represent a culturally foreign, “alien and, to them, irrational mode of thought” (Upton 1990:75). He observes this frustration comes from both small farmers and gentry planters alike, indicating this difference in spatial conception breaks out along racial or ethnic lines and not class lines. These travelers 45 shared a common European-derived white conception of space and navigation, reinforced through their experiences in their social positions and the ways they use and think about space. Upton calls this mental template of space dynamic, flexible, continuous, and articulated. He describes it as a “network of flexible relations that bound both the natural and social works [which] could be manipulated from many points within it” (Upton 1990:80). This insight into separate but coterminous cultural systems for perceiving space should act as a gateway to illustrate the potential for recovering cultural logics that operate within and further reinforce racialized plantation landscapes in the Chesapeake. Rather than being viewed as messy, this multivocality should be embraced and explored. While the data Upton draws upon is limited and anecdotal, he provides a testable hypothesis and one could use his chapter as a theoretical guide for a discussion of other kinds of landscapes of control, including modern prisons, hospitals, schools, militaries, factories, or other locations where control of some form is exerted over subject bodies. The usefulness of Upton’s two pieces on racialized landscapes is that it reminds us that not only are landscapes the products of cultural frameworks of space, but they in turn surround us and influence the way in which we act in that space. In expanding this interaction between landscape as cultural framework and landscape as lived experience, it is useful to include a discussion of Christopher Tilley’s A Phenomenology of Landscape (1994). Tilley explains that movement through space can be thought of as a kind of language, calling this movement a pedestrian speech act. Tilley writes, “Pedestrian ‘speech acts’ may be likened to the speech acts of language. Walking is a process of appropriation of the topographical system, as 46 speaking is an appropriation of language. It is a spatial acting out of place, as the speech act is an acoustic acting out of language” (Tilley 1994:28). He argues that central to the experience of landscapes is the physical act of being in them, that is, walking through them. Through walking, portions of the landscape are experienced and portions are bypassed. Through habitual repetition of the enactment of these spatial rules, a narrative unfolds and a “biography of place” develops. The way people move through and interact with their physical or built environments in turn affects cultural understandings of space and vice versa. On the plantation, movement of the enslaved through the landscape was carefully regulated (Bankole 1999; Upton 1984). As such, these “biographies of place” are inscribed with the same power relations that define the cultural topography of the plantation landscape. Through Tilley and Upton, one can connect cultural frameworks governing space, the navigation of space through routine activities, personal biography, and geography. This connection is central to make the transition between plantation landscapes as dominant ideology and plantation landscapes as dynamic and creative places with multiple poles that can be used to mitigate or even subvert those ideologies. Our challenge here is to be able to recover elements of these past landscapes. As has been alluded to thus far, many landscape and plantation studies deal with concepts of power, control, resistance, and surveillance. Since 1981, much work has been produced by Mark Leone and his Archaeology in Annapolis collaborators regarding the replication and naturalization of ideology on eighteenth-and-nineteenth- century colonial sites. This includes the layout and construction of formal gardens 47 (Leone 1984; Leone and Shackel 1990), considerations of observation and surveillance (Leone and Hurry 1998), and the replication of ideology through material culture (Leone and Shackel 1987; Shackel 1993). Matthews and Palus examine the ideologies that govern landscapes in the more recent past and into the present as well as their historical trajectories (Matthews and Palus 2007). While much of this work is centered within the cities of Annapolis and Baltimore, their arguments can just as easily be extended out onto the plantations and farms that surrounded these cities. Since 2002, AiA has turned its attention eastward to Maryland’s Eastern Shore, first to William Paca’s country estate on Wye Island (Babiarz et al. 2008), then to Wye House (see Chapter 6 for a description of Archaeology in Annapolis’ excavations at Wye House), and then to a free African American community in Easton known as the Hill (Jenkins 2015; Jenkins and Skolnik 2013). There are two complimentary theoretical and methodological frameworks through which the problem of recovering and interpreting these racial plantation landscapes can be approached. The first framework is landscape archaeology, which offers a way of thinking about space in the archaeological record and can overcome the limitations frequently encountered by archaeological sites and their boundaries (Deetz 1990). This framework offers both objective and subjective ways of looking at space and place and the ways humans understand them culturally. This provides a way to conceptualize specific buildings, whole plantations, or even entire regional landscapes comprised of hundreds of individual plantations and farms as cultural artifacts. In the 1990s, several edited volumes of landscape archaeology studies were produced, covering a breadth of methodologies, subjects, and geographic areas. These 48 include Landscape Archaeology: Reading and Interpreting the American Historical Landscape (Yamin and Metheny 1996), The Archaeology of Garden and Field (Miller and Gleason 1994), Earth Patterns: Essays in Landscape Archaeology (Kelso and Most 1990). In the years since these volumes were published, archaeologists working under the umbrella of landscape archaeology have been successful in examining the areas outside of test units and excavated structures (Fowles 2010; Johnson 2012; Joseph 2004; Smith 2014). The application of landscape archaeology to this project seems especially fitting in that the field has been heavily influenced and refined by work on American plantation contexts. As landscape archaeologists took to interpreting three-dimensional space, they have also begun to consider the ways landscape elements from the past come to us in the present. Archaeology as a discipline has traditionally concerned itself with excavation through layers and depth at delineated sites or at specific structures. Stratigraphic excavation is one of the key methodological tools at the archaeologist’s disposal. This approach has worked well for traditional archaeological investigations but when it comes to examining something as large and encompassing as an entire landscape, some have called for approaching the archaeological record as a kind of surface into which are impressions and survivals of the past (Bailey 2007; Malinsky- Buller et al. 2011). This metaphor continues to include the physical ground surface into which past configurations of space can be found. The focus of these studies then is the extent and characteristics of these surfaces and not necessarily its physical depth, but its time depth. 49 The examination of this surface, that is, the horizontal record of human activity at any given moment, can operate on several different scales. Archaeologists can examine spatial patterning within individual excavation units, throughout sites, or across entire regions. Each level of analysis can reveal different details about different aspects of the landscape. Negotiating these scales of analysis can be a way to explore the ways in which people approach and have been affected by space. Historical Geographic Information Systems (hGIS) The second framework through which we can recover and understand past configurations of space is Geographic Information Systems (GIS). A GIS is “a computer-based system to aid in the collection, maintenance, storage, analysis, output, and distribution of spatial data and information” (Bolstad 2008:559). Drawing from research in the fields of geography and computer science, GIS enables a researcher to work with multiple overlapping spatial datasets simultaneously to collect, analyze, and produce new geographic knowledge. Increases in computing power and decreases in computing cost since the 1980s have led to a revolution in mapping the physical and human landscape through the use of computers (Lock 2003). Significant advances in geographic research enable the researcher to study large areas while retaining the high resolution that is necessary when studying features on a human scale. Different layers of data representing vastly different datasets in time and space can be superimposed on one another and displayed simultaneously allowing the user to create and query new geographic knowledge. The application of GIS to historic landscape studies has fundamentally changed the way researchers have approached space and the past (McCoy and Ladefoged 2009). 50 Archaeology as a discipline is not unfamiliar with the use of GIS. It allows researchers to map and record archaeological sites and excavations, as well as conduct some kinds of spatial site interpretation and analysis. For an overview of case studies in the uses of GIS in archaeology, see Wiseman and El-Baz 2007, Conolly and Lake 2006, Wheatley and Gillings 2002, Wescott and Brandon 2000, Gillings et al. 1999. Another way in which GIS can be incorporated into the archaeological toolkit is through what has been referred to as historical GIS. Historic GIS (or hGIS) is a subset of GIS research that uses historical documents and datasets, many of which are spatial in nature, to conduct historical research (Knowles 2002; Knowles and Hillier 2008; Martí-Henneberg 2011). Historical GIS presents a theoretical and methodological toolkit for the archaeologist to combine the historical question of “when?” with the geographic question of “where?” While hGIS can be very broad in terms of its uses and techniques, Anne Kelly Knowles provides a working definition for us to proceed. Paraphrasing, she identifies the following four characteristics of historical GIS scholarship: 1. Geographical questions drive a significant part of the historical inquiry. 2. Geographical information provides a good share of the historical evidence. 3. The bulk of evidence, or the evidence that provides the study's key analytical framework, is structured and analyzed within one or more databases that record both location and time. 4. Historical arguments are presented in maps as well as in text, graphs, tables, and pictorial images; maps serve to show patterns of change over time. (Knowles 2008:7) What defines hGIS as a distinct field within GIS studies is its use of the tools of GIS to answer questions about the past. As will be shown, this requires a specific subset of the tools available to GIS researchers and a specific way of thinking about space in the past. This allows researchers to apply modern statistical analyses to 51 historic datasets, digitize and overlay historic maps, and spatialize historic datasets that were not originally spatial in nature. As the field of archaeology—and landscape archaeology especially—is interested in the intersections of time and space, this represents an important theoretical and methodological development. In the approximately 20 or 25 years since GIS has become accessible to archaeologists, this ability to overlay multiple historic and contemporary spatial datasets to create new geographic and historic knowledge has been extremely fruitful for researchers, especially in the areas of the history of land use, spatial economy, reconstructing past landscapes, and mapping the development of infrastructure (Knowles 2008:8-16). Some have gone so far as to declare that “GIS is a major invention for archaeology perhaps comparable to radio-carbon dating in its importance” (Neustupný 1995:134). While it is yet unseen if GIS will be as significant to our understanding of the past as radio-carbon dating, I believe it certainly has the potential to do for spatial questions what radio-carbon dating did for temporal ones. As Knowles observes, one significant product of hGIS research is the map. While maps are useful for documenting the spatial distribution of phenomena, they, like the Paca Garden as discussed by Leone, are pieces of material culture with embedded ideological origins and functions. As useful as GIS is as a tool for investigating spatial phenomenon, its uncritical use has the potential to reinforce the ideologies and assumptions embedded within it. Some have pointed out the connection between GIS and quantitative, scientific, positivist modes of knowledge production, and these are worth repeating here. 52 One such warning about the uncritical use of the map comes from J.B. Harley’s chapter titled “Maps, knowledge, and power” published in Denis Cosgrove and Stephen Daniel’s Iconography of Landscape (1988). Harley makes three important arguments that help us recognize the ideological position of the map and to help guide us as we in turn use maps to address ideology embedded within a landscape. First, he makes explicit that maps are a type of language or text with an author or authors, biases, and motivations behind their creation. This locates maps and mapping in the same kinds of discourses as described for texts and material culture by Beaudry and others that has informed this dissertation on landscape. Every historic map used here in this dissertation was created by a person or group of people for a specific reason and that reason dictates which features are and are not included on the map and how they are depicted. Likewise, every map produced for this dissertation is the same. Second, he notes that embedded within and communicated by maps is symbolic knowledge, allowing for the replication of political power through maps. Third, he argues maps are inextricably linked with knowledge and power. He explicitly invokes the work of Foucault on knowledge and power, including cartography as surveillance and “a way of presenting one’s own values in the guise of scientific disinterestedness” (Poster 1982:118-19; cited in Harley 1988:279). Harley links maps and map production with empire, nation state building, colonialism, private property, power, and cartographic “fact.” This cartographic fact, which asserts the map is an accurate and direct representation of reality is one of the underlying assumptions behind the European conceptualization of space as described by Upton and is at least partially responsible for the amount of power held by maps. 53 In response to observations such as Harley’s, some researchers have questioned the way they use maps. An early assessment of the use of GIS in archaeological research observes that, “In its least harmful form, the indiscriminate use of GIS…may result in the slick, but repetitious, confirmation of otherwise obvious relationships. In the worst case, it might involve the unwitting exposition of an environmentally or functionally determinist analytical viewpoint” (Gaffney et al. 1995:211). Like argued by Harley, GIS, like the map, is an ideological technology produced by a certain set of cultural practices and frameworks which, in turn, through the use of GIS further reinforce those frameworks. Its indiscriminate and acritical use can and does reinforce the implicit assumptions that are a precursor to using GIS as an analytical framework. In response to this point, since the 1990s, the fields of critical, participatory, and feminist GIS have attempted to understand the ways in which maps and GIS promote a specific worldview, provide alternative ways of dealing with spatial information that attempts to ameliorate or circumvent this problem, or use the tools of GIS to further specific social causes (Elwood 2008; Kwan 2002; Pavlovskaya 2018; Schuurman 1999; Wilson 2017). These critiques urge us to be clear in how we are using GIS in our research, the kinds of data we use, the kinds of analyses we run, the conclusions we draw from it, and the audiences and purposes of our work. It should be noted here that there are many similarities between space as conceptualized through maps and mapping in a traditional GIS and the articulated, dynamic mental templates employed by eighteenth-century white Virginians as outlined by Upton. If historical archaeology is ultimately a study of the origins of the 54 modern world as Charles Orser suggests (Orser 1996), the implication is that our contemporary frameworks of space and geography come to us through eighteenth- and-nineteenth-century plantations and through Deetz’s Georgian Order. This is not to say that there are no other conceptions of space operating in the present, in fact, future research could look for these alternatives in the present as a way to understand how similar alternatives could have functioned in the past. However, with the omnipresent Google Maps, GPS, GIS, and other spatially-aware digital mapping platforms, it would appear that the spatial logic implicit within Upton’s assessment of plantation society is the dominant form of spatial conceptualization in the present. Methodologies As previously mentioned, this dissertation uses the mapping tools and techniques of GIS within the framework of landscape archaeology to better understand the plantation landscape. This examination of the plantation landscape can occur at several different scales of analysis. In the abstract sense, one of the products of this dissertation is a descriptive map of mid-nineteenth-century Talbot County, Maryland, created by combining multiple historic and modern datasets within a GIS framework. This map is actually comprised of a series of different historic maps, historic aerial photographs, modern aerial photographs, textual descriptions, census data, and LiDAR-derived elevation data, all of which can be turned on or off and stacked in any combination, depending on the immediate research need. The map can be zoomed or panned or rotated as needed. It can be annotated, drawn on, and edited when new information needs to be added to it. To call this document a map is a bit 55 misleading because maps have traditionally been fixed, static, and two-dimensional, but that is what it is at its core. In Chapter 4, this map is used in an attempt to understand county-wide demographics and population distribution. In the case of Talbot County, Maryland, an 1858 map created by William Dilworth (Dilworth 1858) enables us to bridge the geographic map and historic census returns and create a spatial join between the spatial information contained in the map and the demographic data contained in the 1860 United States Census. While the 1858 Dilworth map represents a remarkable dataset for understanding the spatial distribution of property and households across Talbot County, it does not contain any demographic data. Furthermore, while the 1860 U.S. Census represents a remarkable dataset for understanding demographics, wealth distributions, household composition, slave ownership, etc., it does not contain spatial information that would enable the researcher to map any of these attributes. On the map, property (farm) boundaries are indicated along with the first initial and last name of its owner. On the census, each individual is recorded with a first name and a last name. Once these two datasets have been digitized, the names of the property owners listed on the map can be matched to the names of heads of household on the census. Once this link has been established, the names and data from the census will be joined to spatial boundaries on the map through the names on the map. The social landscape as recorded in the 1860 U.S. Census is therefore mapped onto the physical landscape of Talbot County as recorded in the 1858 Dilworth map. Other features can be extracted from the Dilworth map including transportation networks, spatial distribution of plantations and structures, and 56 hydrology. To this, additional modern information can be added including elevation. The Dilworth map was chosen here for several reasons. First, it represents one of the earliest complete maps of the county with enough detail to allow this kind of analysis. Second, a U.S. Census was taken within a short time of its creation, allowing for continuity between the two datasets. Third, it dates to the height of the plantation system in the American south, just prior to the American Civil War and the abolition of slavery. The Dilworth map was digitized by University of Maryland undergraduate Clio Grillakis as a part of an Emerging Scholars Program grant awarded to Archaeology in Annapolis in the Fall of 2012 by the College of Behavioral and Social Sciences at the University of Maryland. Furthermore, Grillakis and University of Maryland undergraduate Marcella Stranieri transcribed the 1860 U.S. Census for Talbot County, Maryland as a part of a Future of Information Alliance Robert W. Deutsch Foundation Seed Grant awarded in 2013. The approach used here is similar to that taken by other researchers. Ian N. Gregory and Humphrey R. Southall have undertaken a similar exercise mapping the historic demographics of Gloucestershire, England (Gregory and Southall 2002). They have digitized Ordnance Survey maps created between 1881 and 1931, which happen to be the same dataset used by W.G. Hoskins in his early and influential historic landscape study (Hoskins 1955), and joined to these the returns from the British census. The data from each political division is given a spatial position and boundary. Similarly, Don DeBats has transcribed the 1860 US Census for Alexandria, Virginia and the 1870 US Census for Newport, Kentucky and linked them to period maps as well as election results and individual ballots (DeBats 2015). This ability 57 enables the researchers to visualize demographic trends through time as well as in space and is a model for the work presented in Chapter 4. To the Dilworth map and 1860 US Census, additional historic datasets can be added. With this dataset of the physical and cultural landscape, a different map of Talbot County can be made. In Chapter 5, this hGIS framework will be combined with a detailed digital topographic elevation model derived from a LiDAR (Light Detection and Ranging) survey of Talbot County. A methodology for mapping physical plantation landscapes in the present on the scale of individual plantations has already been developed by Harmon, Leone, Prince, and Synder (Harmon et al. 2006). They used LiDAR, a laser- based mapping technology that operates on the same principles as RADAR or SONAR, to create a detailed digital elevation model (DEM) of two contemporary plantation landscapes in order to approximate their historic configuration. They then use the DEM to characterize the extant landscape features and identify potential archaeological features. While Harmon and others would remind us that these remotely-sensed datasets record the landscape in the present, these landscapes still operate as palimpsests (Bailey 2007; Harmon et al. 2006). In this manner, the physical landscape itself is a complicated archaeological artifact, one with many overlapping and/or incomplete layers, created in the past, encountered and reworked throughout time, which can be engaged with by the archaeologist. Working from the list of National Register of Historic Places plantation in three Maryland counties compiled by University of Maryland undergraduate Ryan O’Connor and presented in Chapter 2 as Table 2.1, a sample of ten extant plantations in Talbot County has been chosen for analysis using the technique detailed by 58 Harmon and others as well as the recommended analysis they propose for future work in their conclusion (Table 5.1). These ten plantations were chosen because they appear to have intact historic cores, historical documentation is readily available for them, and they appear on the kinds of spatial datasets that enable this kind of analysis. For each of these, a LiDAR-derived DEM has been created and combined with historic maps using GIS to identify and characterize extant plantation landscape elements that appear to have historic depth. It is well documented that many of these formal landscapes were built according to a set of geometric rules. From this analysis, we can identify the rules that were used to design and layout these landscapes. In Chapter 6, the methodology developed by Harmon and others (2006) and refined in Chapter 5 will be further refined with additional hGIS research in order to locate two structures at Wye House believed to be quarters for the enslaved. Whereas Chapter 5 represents a broad but shallow survey of a regional plantation landscape, Chapter 6 represents a deeper dive into one specific site using additional historic maps and research, aerial imagery, and previous archaeological work. This represents a proof of concept and case study for the incorporation of hGIS into ongoing archaeological research projects and a useful tool for archaeological prospecting. In Chapter 7 and Chapter 8, hGIS will be used in conjunction with Frederick Douglass’s eyewitness accounts of slavery in Talbot County and Dell Upton’s theoretical treatment of these plantation landscapes. Aware of some of the critiques of GIS research, we must recognize that this project in historic mapping documents only part of the plantation landscape. These tools by themselves are largely quantitative tools based in mathematics, statistics, and computer science and are largely unable to 59 address Upton’s multivalent racialized plantation landscapes. One way to test Upton’s hypothesis is by incorporating textual analysis from Douglass’s autobiographies. Whereas it is extremely difficult to recover something as abstract as mental maps or worldviews from the archaeological record, we can use Douglass’s three autobiographies to help guide us through the historic terrain mapped in Chapters 4 and 5. Douglass has been chosen here for his prolific writing, the circumstances surrounding his enslavement, and the previous archaeological and historical work conducted by Archaeology in Annapolis connected to Douglass. The beginnings of his autobiographies are all interwoven with spatial references as he recollects his childhood and adolescence, during which he was enslaved on several plantations in Talbot County, Maryland. This spatial analysis is much different from the spatial analyses in the chapters that precede it. A textual spatial annotation of these works can be used in conjunction with Upton to understand the ways Douglass conceptualized his enslaved landscape in one part of the antebellum Chesapeake. This data comes from the scenes he recalls from the plantations, his views on slavery, and the ways in which Douglass refers to places throughout his autobiographies. His spatial references can be added to the growing digital historic map of Talbot County. Using Upton and Douglass as our guides, they may be able to show us the ways in which the formal plantation landscape was not directed at or for the enslaved on the plantation, the ways in which they were able to see through and combat ideology, and that the experiences of being enslaved shape the way one views and negotiates space. Inspired by Upton’s description of the enslaved spatial worldview, that is, seeing the landscape in terms of loci of control, a map or a model can be generated 60 showing the spaces wherein this control is more or less present. Here, we can turn to Douglass’s description of this landscape and use him as a guide, learning from the ways he saw this landscape. Proximity to and visibility of plantation houses represent areas of higher control over the enslaved whereas the rivers and streams, ravines, and forests represent areas of lower control. Using GIS, we will model a map of proximity and visibility to create a map that can help identify those areas within the county wherein the landscape can affect control of space. Together, these methodologies in landscape archaeology as detailed in the following chapters represent an attempt to use hGIS and mapping technologies in conjunction with an ongoing archaeological research project to map and present traditional understandings of plantation landscapes alongside an alternate plantation landscape. 61 Chapter 4: Historical Geographic Information Systems (hGIS) and Regional Studies Georeferencing the 1858 William H. Dilworth Map of Talbot County, Maryland Integral to this dissertation is the use of Geographic Information Systems (GIS), which enables the archaeologist to work with multiple spatial datasets simultaneously to collect, analyze, and produce new geographic knowledge. This research will replicate, develop, and test methodologies in historical GIS (hGIS). The ability to overlay multiple historic and contemporary spatial datasets to create new geographic and historic knowledge has been extremely fruitful for researchers, especially in the areas of the history of land use, spatial economy, reconstructing past landscapes, and mapping the development of infrastructure (Knowles 2002:xvi-xx; Knowles and Hillier 2008:8-16). Archaeologists, long accustomed to borrowing from other disciplines, have readily adopted GIS as a methodological toolkit for addressing spatial issues (Conolly and Lake 2006; Gillings et al. 1999; Wescott and Brandon 2000; Wheatley and Gillings 2002; Wiseman and El-Baz 2007). One of the most powerful tools of hGIS is the ability to georeference historic data sources. Georeferencing is the processes by which maps, aerial or satellite images, or other spatial datasets are translated, rotated, and scaled, and warped and distorted if needed, to assign real world coordinates, and can then be displayed along with other spatial datasets from the same area. This is usually accomplished through the use of an established coordinate system either on a global scale like latitude and longitude, a more local scale like the State Plane system 62 established in the United States, or a site-specific grid system commonly found on individual archaeological sites. Doing so allows researchers to map structures and other cultural features depicted on these maps in their real-world positions in an attempt to identify where in the real world those structures used to be and where their archaeological remains may still be found. Structures can be placed on digital models of the terrain and viewshed or visibility analyses can be conducted to determined what can be seen from those structures and from where they are visible. Characteristics of the landscape like elevation and slope can be computed and visualized, human alterations to the landscape can be identified, abandoned roads relocated, and even buildings that have long since disappeared from the landscape can be located from shallow depressions or other clues left in the ground surface (Harmon et al. 2006). From data like this, we can assess the physical topography of these landscapes on a scale and at a resolution not previously possible. Several edited volumes collecting hGIS studies have been published, including Toward spatial humanities: historical GIS and spatial history (Gregory and Geddes 2014), Historical GIS research in Canada (Fortin and Bonnell 2014), Placing history: how maps, spatial data, and GIS are changing historical scholarship (Knowles and Hillier 2008), Historical GIS: technologies, methodologies and scholarship (Gregory and Ell 2007), A place in history: a guide to using GIS in historical research (Gregory 2003), and Past time, past place: GIS for history (Knowles 2002). One of the cornerstones of historical GIS research are historic maps. Maps are inherently spatial documents that record some kind of cultural and/or physical configuration of space at a specific time. They simultaneously answer for us two of 63 the questions archaeologists ask most: “Where?” and “When?” Underpinning this chapter on hGIS and mid-nineteenth-century plantation landscapes in Talbot County, Maryland is William H. Dilworth’s 1858 Map of Talbot County, Maryland; with farm limits, referred to hereafter simply as the Dilworth map (Dilworth 1858) (Figure 4.1). 64 Surveyed by Dilworth and engraved by Rae Smith of New York, the published map measures 45 inches by 45 inches square, or approximately 14 square feet. On April 20, 1858, three items appeared in The Easton Star regarding Mr. Dilworth and his map. The paper reports that this mapping project was undertaken at the suggestion of the Board of Trustees of the Maryland Agricultural Society for the Eastern Shore, and will show “all the election and school districts, roads, streams, towns, churches, mills, stores, houses, (with the owner’s names opposite,) and approximate farm lines” (Easton Star, 20 April 1858:2). Here, William Dilworth lists his title as “Civil Engineer and Surveyor”. Somebody writing to the editor under the pseudonym “A Talbot Farmer” hoped that the map would provide a sense of pride in the “accurate and minute delineation of all the beauties and advantages in the remarkable geography of this county, and the location of each man’s possessions, and habitation” (Easton Star, 20 April 1858:1). The map was originally intended to be backed with muslin, varnished, and hung on rollers so that it could be hung from a wall. Dilworth suggests in The Easton Star his map would make “a handsome ornament for the Office, Parlor and Library” (Easton Star, 20 April 1858:2). Coincidentally, A Talbot Farmer makes the following observation: “As has been intimated heretofore, we ourselves, do not know how fair is our heritage, and this map will teach us, as as [sic] well as strangers, what we possess in old Talbot” (Easton Star, 20 April 1858:1). It is in this same sentiment that we turn to Dilworth’s map. This map was chosen for this analysis for several reasons. First, it represents one of the earliest detailed maps of the county that shows features relevant to these kinds of hGIS analyses. This is useful here because the age of it shows us the county 65 in one of its earliest and most complete depictions and the scale at which it is mapped shows us detail on the local level. Its inclusion of primary structures in the county as well as approximate property boundaries enables us to undertake the below analysis. Second, there exists a U.S. Census taken within two years of this map’s creation. This will become incredibly significant below when we attempt to embed demographic data into the map. And third, it was created just prior to the American Civil War, at the height of the antebellum system of plantation slavery. This is not meant to imply that this system was either stable or mature; an argument could probably be made that the introduction of the twin technologies of the railroad and the mechanization of agriculture had the potential to radically alter this landscape had slavery continued unabated beyond the 1860s. Ultimately, the system of plantation slavery which had been widespread in Talbot County for approximately 150 years when this map was produced has left its mark on the landscape and is recorded in this map. Like many historic maps, the Dilworth map of Talbot County, Maryland is a fascinating document. This one depicts the county at the height of the antebellum plantation system and in the process of mapping the political and administrative boundaries of the county, Dilworth also maps many other features that are useful to historical researchers. His map depicts the larger physical and social landscape within the county and the features found on it enable us to reconstruct several aspects of the larger regional landscape as it existed in 1858. These include political boundaries, roads, railroads, coastlines, rivers, property owners, property boundaries, and the primary structures on each property, and together this represents a remarkable dataset for understanding the spatial distribution of historic human occupation and the 66 economic organization in Talbot County in the mid-nineteenth century. The Dilworth map is a unique snapshot of the regional landscape as it existed at a moment in time more than 150 years ago. The most significant omissions from the Dilworth map are the names of the people living in the county who were not the owners of large agricultural properties and the locations of those buildings on each property that were not the main residential dwelling. These omissions will be addressed through the further use of hGIS in this and the following two chapters. The first step in working with this map in a hGIS framework is digitizing the document so that it can be loaded into a computer and given spatial references that allow us to display and analyze it with other co-located spatial sources. Generally, this is accomplished by scanning the historic paper or other textile map and storing it as a digital file. When working with visual or spatial data in a digital format, it is important to understand the way in which this data is stored and processed in a computer. There are two ways to represent digital graphics or spatial information and both types are used here. A raster image and a vector image are the two different approaches for recording visual information and the difference between them is important when manipulating this data or when working with multiple spatial datasets. Raster images are probably most familiar and consist of a grid of pixels (the word pixel is a portmanteau of the words “picture” and “element”), each with its own color value or intensity at each specific location. Digital photographs are stored in this manner and are a good example of the raster image format. As one zooms into a digital image, the individual pixels become visible and the maximum resolution of the dataset has been reached. Data that is continuous over a given area, that is, data that 67 has a value at every location like digital photographs, satellite images, or scanned historic maps are best recorded and displayed as raster data. In contrast to raster images, the other kind of data representation used for storing digital visual or spatial information is the vector image. Instead of storing spatial information in a grid of pixels, vector images are composed of discrete points, lines, and polygons. Rather than storing them as individual pixels in a grid, vector images are comprised of individual features that are stored as a series of instructions for drawing the vector image, including the locations and size of the features to be drawn. This data representation is best used for discrete data, that is, data that consists of specific features with specific boundaries or specific locations (ArcGIS Help Library). Online mapping services such as Google Maps store their data in this manner so that the roads, cities, and other geographic features can be drawn and displayed at many different scales without pixelating and can be stored at much lower costs. Depending on the type of data being displayed, both data representations— raster and vector—are useful and necessary. When working with historic maps, the scanned maps are best stored and displayed as raster data because we want to be able to view them as an image, consisting of specific colors displayed in a grid at each point in the map. However, once the features seen on these maps are traced or recorded digitally, these digital map layers are stored as vector data because we are interested specifically in the location, size, and direction of these discrete features such as structures that can be stored and displayed as points, roads which can be 68 stored and displayed as a series of line segments, and property boundaries which can be stored and displayed as polygons. Scanning Dilworth map on a digital scanner converts the physical paper map into a digital raster image. The map is stored as a series of pixels, each with its own location and color value. For a set of continuous digital data like a map or photograph, this is the best way to represent the original dataset so that a human observer can make sense of it. The copy of the Dilworth map used in this dissertation is archived at the Library of Congress’s Geography and Map Division and the scanned digital map file was produced and provided by the same. While it may be easier for a human researcher to interpret scanned raster representations of historic maps, it is much more difficult for computers to make the same sense of these kinds of datasets. One goal in working with these spatial documents in a computer-based GIS is to enable the computer to carry out computations and spatial analyses automatically at a speed and precision human analysts are not capable of performing. Digitizing historic maps to work within a GIS framework is perhaps the most important step of this process, as it is a form of digital data collection. This involves digitally tracing over and pinpointing every boundary, river, road segment, and structure seen on the map. In addition, associated data such as names or other attributes provided by the author of the map can be entered and stored in association with the digitized geometry they represent. This is generally the most time consuming part of map digitization and as such, many have called for the creation of online digitalize map distribution platforms to allow others to take advantage of already digitized datasets and to minimize the amount of duplicated 69 work. These include the National Historical Geographic Information System (www.nhgis.org), the Historical GIS Clearinghouse and Forum (www.aag.org/cs/projects_and_programs/historical_gis_clearinghouse/hgis_projects_ programs), and the Atlas of Historical County Boundaries (www.publications.newberry.org/ahcbp). At least at this stage of development, automatic feature detection software is not capable of understanding and automatically extracting features from scanned historic map raster inputs as well as their human counterparts. As such, the tedious and time-consuming process of digitally tracing the features in these historic maps falls to human researchers. The vector representations of the features seen on the 1858 Dilworth map used here were digitized in part from a grant awarded by the Emerging Scholars Program through the University of Maryland’s College of Behavioral and Social Sciences. The program is designed to give undergraduate students an opportunity to work closely with and be mentored by College faculty. Awarded in 2012, this grant was used to train University of Maryland undergraduate Clio Grillakis in the use of hGIS. Under the supervision of Dr. Mark Leone as well as this author, Grillakis was taught many of the relevant computer-based GIS skills required to digitize this historic map and used them to digitally trace the relevant features depicted on the map which will be discussed below. The map was georeferenced by this author with assistance from Grillakis in ESRI’s ArcMap GIS software program. In order to accomplish this task, a collection of known points must be identified first on the historic map and then in a known coordinate system, usually through a modern map or satellite imagery of the same 70 location. These pairs of points are known as ground control points or GCPs. A total of 311 ground control points were used to establish common linkages between the historic map and a basemap comprised of contemporary satellite imagery (Figure 4.2). Ground control points were selected from locations of known position that could 71 be identified in both datasets, consisting of stable landmarks whose locations are known in both datasets. These generally take the form of building corners, features of the road and railroad networks such as intersections and turns, or corners and intersections of property boundaries that have not changed and can be identified in both the historic and modern datasets. Less stable features such as coastline or riverine features were avoided as it is more difficult to ensure the spatial continuity of those features. Mathematically speaking, only two ground control points are required to establish a transformation between a historic map and modern spatial dataset. The first point can only establish a translation between the two coordinate systems. The second point establishes the correct rotation between the coordinate systems and establishes scale the scale along the axis between the two points. Adding a third point not along this axis establishes scale in the axis perpendicular to the first. Any additional ground control points beyond these first three, especially when distributed around the map will provide the researcher with a better understanding of the original map’s accuracy. Once a satisfactory number of ground control points has been created, the historic map can be placed or transformed into the new geographic coordinate system, placing the map in digital space. There are several methods to transform an image, but generally the process involves calculating the average transformation required to move each ground control point from the original coordinate system of the raster image of the historic map to the new coordinate system of the target workspace and then performing that transformation for each point in the raster dataset (Bolstad 2008:146-154). 72 There are several potential sources of error of which to be aware when georeferencing historic maps. Error can exist inherent to the original cartographic depiction as drawn by the original map maker, it can be introduced as the paper or other surface on which the map is printed or drawn warps, stretches, shrinks, tears, or is mended over time, or error can be introduced by the georeferencer choosing inaccurate or incorrect ground control points. The accuracy of this georeferencing process can be calculated and expressed as a Root Mean Square value (RMS). The RMS of a map transformation can be calculated by determining the average (mean) distance of all transformed points from their theoretical researcher-assigned positions. Here, distance is expressed as an absolute value and orientation or direction is ignored. This results in a value that represents the average displacement between any point in our transformed raster image and its theoretical real-world location. The smaller this RMS value, the greater the accuracy of our transformation (ArcGIS Help Library). Essentially, it is a measurement of how well the two patterns of ground control points match each other. If done perfectly then the spacing and relative shape of one set of points in the dataset to be georeferenced will match exactly the spacing and relative shape of the other set of points in the dataset with the target coordinate system and the RMS value will equal zero. There is no established method within archaeology for determining appropriate levels of RMS error as acceptable amounts of error varies by application. If georeferencing at the scale of a single house lot or a city block, an RMS value of a few feet can be acceptable; working at the scale of entire counties, an RMS value of thousands of feet or even more can be acceptable. Given that our use of georeferencing here involves working with one and two 73 hundred year old pieces of paper with hand-drawn notations and are at scales that encompass entire plantations or even counties, and that this use will be followed by an archaeological survey, we only require that it get us close to the suspected locations of archaeological features in order to be useful and worthwhile. The total RMS error of a first order polynomial (affine) transformation of the Dilworth map is 797.88 feet. That is, the coordinates of any given point in our transformed map will, on average, be 797.88 feet from its real-word position. This transformation will only translate, scale, and rotate the historic Dilworth map in order to best fit the control points. This map is accurate enough on a county-wide scale to allow us to use this transformation and the first-order transformation allows us to retain the original geometry of Dilworth’s outlines and shapes. While 797.88 feet of error seems unacceptably large, we can contextualize this number. The scanned resolution of the Dilworth map is approximately 470 pixels per mile. This works out to approximately 0.089 pixels per foot or 1 pixel every 11.22 feet. At this scale, 797.88 feet of error is only approximately 71.024 pixels on the map, only slightly more than one half of one percent of the map’s width. A comparison of the real positions of our control points with their positions in our transformed historic map revels that the displacement of ground control points within any local region is relatively uniform, that is, in a similar direction by a similar amount. This would indicate that the relative placement of nearby features on Dilworth’s map is fairly precise. The errors seem to be relatively small in any specific location and only accumulate when trying to include large areas. 74 Other, more complex geometric transformations can also be used to attempt to account for the spatial inaccuracies within the original historic map. As a comparison, a second order polynomial transformation reduces the RMS error to 727.85 feet and a third order polynomial transformation reduces it to 651.74 feet. Each of these transformations introduces mathematical complexity to the resulting map at the expense of fidelity to the original depictions of geographic features on the map. It is also possible to create a transformation that warps and stretches the Dilworth map to correspond precisely with each ground control point. In the pre-digital era, this technique was known as rubber sheeting, as the map or aerial would be printed on a physical rubber sheet and stretched and pinned in place (Estes et al. 1977; Scollar 1975). This reduces the RMS error to zero as each point is assumed to be precisely where it is supposed to be in the real world and distorts the rest of the map to conform to these points. Unfortunately, each of these increasingly more complex transformations distorts the shape, geometry, and depiction originally created by the original cartographer. Deciding which to use is a balancing act between preserving cartographic fidelity and the accuracy of the corresponding transformation. Because we are interested in the mid-nineteenth-century landscape as depicted in this map by its creator and the errors seem to be minimal at the local level, our preferred transformation should minimize these distortions at the expense of a few feet of accuracy. One of the reasons for working with these maps in a GIS is that it automates many of the kinds of analyses that we may want to conduct, but first, the features in the map need to be digitized in order for us to be able to work with it in a computer. 75 The number of each type of point feature such as structures can easily counted, the lengths of the line segments that comprise features such as roads or streams can be added, and the areas of the polygons that make up individual properties or the county can be combined. Dilworth’s map shows 1,169 structures on 882 separate properties within Talbot County (Figure 4.3). The legend of the map labels these simply as “Houses”, even though many are labeled directly on the map otherwise as mills, post offices, or stores. There are an additional 18 structures shown within the inset of Easton on the map. Dilworth also displays farm boundaries, which roughly correspond to property lines (Figure 4.4). The average (mean) property size is 199.31 acres and the standard deviation is 194.75 acres. There are 9 properties of 1,000 acres or more, with the largest being “Col. Edward Lloyds Woodland” at 2,100.02 acres. There are 57 properties of 500 acres or more (Figure 4.5). Of these, 26 are owned by members of just 6 families; the Lloyds (9 properties, 7,542.80 acres), Goldsboroughs (5 properties, 3,603.81 acres), Hughletts (4 properties, 4,116.20 acres), Tilghmans (4 properties, 2,397.25 acres), Rogers (2 properties, 1,337.92 acres), and Arringdales (2 properties, 1,198.362 acres) families. From these parcels larger than 500 acres, it should be clear the vast amounts of property and wealth in the county concentrated in the hands of only a few families. While it should be emphasized that acres of land held does not translate directly into social power or wealth and that familial connections go beyond just the surname of listed property owners, this map of property ownership should give us a rough sketch of the topography of power and wealth in a county that relies on land ownership and agricultural production. As 76 revealed in this snapshot, we see a concentration of wealth in the hands of a relatively small number of families and individuals. Through the mapped property boundaries and social connections shown through this map, we are able to quantify the distribution of a type of wealth (in the form of land) throughout the county. 77 The structures depicted on the map appear to represent the primary structures on each property. It should be noted that the 1,169 structures indicated by Dilworth do not represent the total building stock of Talbot County in 1858. Frederick Douglass, having been there three decades prior, describes in his autobiographies the 78 multitude of buildings at Wye House as “having the appearance of a country village”. For example, at Wye House, a property owned in 1858 by Edward Lloyd VI, (listed as Col. E. Lloyd on the map), Dilworth only shows two structures where historical and archaeological research have shown there to have been dozens. One of these two 79 shown at the end of the road or driveway on the map is certainly Wye House proper, originally built sometime between 1781 and 1784, and the other is probably the structure known today as the Red Overseer’s House (Figure 4.6). See both Chapter 5 and Chapter 6 for a more in-depth discussion of the built environment at Wye House. 80 In addition to sharing the locations of structures, Dilworth also includes labels for many of them. This attribute or label is extremely useful as we analyze this landscape. Most properties are labeled with the name of the property owner in the format [First initial, period, space, last name]. Sometimes a middle initial is added as well. In a few cases where the property owner is recently deceased, the map indicates the property is owned by the “Heirs of [deceased property owner]”. In other cases, the primary structure on a property is labeled as a store, church, school, post office, or other non-residential building. These non-residential properties provide a look into the non-agricultural parts of this landscape. There are 680 unique property labels on the map, which we will use below to incorporate additional demographic information by way of the 1860 US Census into this spatial dataset. Dilworth’s map also shows the boundaries between the five districts in the county; Chapel, Easton, Trappe, St. Michaels, and Bay Hundred (Figure 4.7). According to the United States Census Bureau, Talbot County has an area of 268.54 square miles or 171,866 acres. The area of Talbot County in the digitized Dilworth map as measured in GIS is 276.20 square miles or 176,770 acres. Dilworth himself calculates the area of the county on the map and lists it as 179,175 acres, which is equivalent to 279.96 square miles. These figures are roughly similar and, given the dendritic nature of the shoreline along the county’s many rivers and its frontage along the Chesapeake Bay as well as the constant dueling forces of soil erosion and deposition, a certain margin of difference is probably to be expected. In mapping coastal Chesapeake counties, it can be difficult to delineate the exact boundaries between the Bay itself and the mouths of its tidal tributaries. In 81 many regards, these wide, slowly-flowing, tidal rivers are just extensions of the Bay, snaking inland. Dilworth shades the Chesapeake Bay and these navigable tributaries as a single dendritic feature and does not differentiate between them. He depicts the smaller, presumably non-navigable rivers, creeks, and streams that feed into it as 82 discrete lines. Counting the coastline that fronts both the bay and these larger tributaries, Dilworth draws 585.94 miles of coastline in Talbot County. In addition, there are 151.27 miles of these smaller rivers, streams, and creeks drawn within the county. Many of them serve as boundaries between adjacent properties. These small waterways can also have industrial uses. Dilworth shows thirteen ponds, lakes, or inland tidal basins comprising a total of 226.39 acres. Of these, seven are probably artificially-dammed millponds, created by damming a creek or stream and using the newly created water reservoir to power a mill. Many of the mills shown on the map are located at these dammed mill-ponds. There are seven springs shown on the map. Dilworth’s map shows 5933.47 acres of marsh, primarily located along the coastline of the county (Figure 4.8). As shown in the main frame of the map, there are 458.18 miles of roads in the county (Figure 4.9). Dilworth’s own estimate is that there are only 281 miles of road in the county. This seems to be a fairly large discrepancy but may be accounted for if Dilworth did not count private driveways in his tally which are undifferentiated from public roads on his map and/or if he did not count the road network within Easton, which is depicted in more detail in an inset at the edge of the map. These roads account for an additional 6.71 miles of road within the boundaries of Easton proper. Connecting this land-based transportation network with the maritime one on the rivers and the Bay are three landings on the Choptank River. In the 1850s, a new transportation technology was beginning to make its way into the county. With its southern terminus on the outskirts of the town of Oxford, Dilworth notes that the Maryland & Delaware Railroad is currently under 83 construction in 1858 during his survey. It runs northeasterly through Easton and enters Queen Anne’s County near Hillsboro. As shown on the map, there are 21.64 miles of railroads in the county (Figure 4.10). 84 The digitized version of the 1858 Dilworth map of Talbot County not only allows us to count and measure these features but we can also use spatial statistics to further describe the county and the settlement patterns within it. 85 The geographic center of Talbot County is located at a point approximately one-half mile south of St. Michaels Road, and approximately one quarter mile west of Easton Parkway. The geographic center of a feature (in this case, the polygon created by tracing the boundary of Talbot County from the Dilworth map) can be calculated 86 by finding the center of gravity of the polygon. This is an average of the X coordinates and Y coordinates for each of those ranges within the polygon and if the polygon were physically cut out, this is the point at which it would balance (ArcGIS Help Library). The actual location of this central point can vary widely depending on whether or not water bodies like tidal rivers are included in the calculation and as land is lost or redeposited to erosion over time. For this calculation, only those areas indicated by Dilworth as being inside one of the county’s five election districts were included for consideration. Of the 1,169 structures seen on Dilworth’s map, the central feature (median) is Ratcliffe Manor, located two miles west south west from the Court House in Easton and the mean location of these structures is located less than half a mile across Dixon Creek from Ratcliffe Manor. One characteristic of the plantation landscape we can analyze here is the distribution of plantation structures across the county. Using the points representing each of our main structures, we can calculate the density of these structures across the county. This is calculated by applying a search radius around each point to be considered, specifying a cell size for a raster-based analysis, adding the number of overlapping search radii at each cell location, and displaying the results (ArcGIS Help Library). Density maps of these structures reveal the highest concentrations of structures, especially in Easton, St. Michaels, and Trappe. Slightly lower concentrations exist in Royal Oak, Oxford, and Wittman (Figure 4.11). It should be emphasized again that the structures identified by Dilworth are not the total sum of the building stock in the county in 1858. Rather, these should be viewed as the primary structures, the major residence on each property. 87 The distribution of these structures across the county can be described using a spatial statistic known as the Nearest Neighbor Index. This index or ratio is expressed as the observed mean distance divided by the expected mean distance. Any ratio less than one is indicative of a trend toward clustering and any ratio larger than one is 88 indicative of a trend toward dispersion. A ratio equal to one would be a randomly distributed set of points. A ratio of zero would be all of the points stacked on top of each other. As mapped by Dilworth, the average distance between any structure and its closest neighbor is 1810.50 feet. If these structures were distributed randomly across the county, the expected nearest neighbor should be 2443.01 feet. Therefore, the nearest neighbor ratio is 0.74 with a z-score of -16.93 and a p value of 0.00. Largely because of the dense urban groupings of structures in Easton, St. Michaels, Oxford, Royal Oak, and Trappe, the distribution of structures as seen in the Dilworth map is strongly clustered. Removing the urban structures in Easton, St. Michaels, Oxford, Royal Oak, and Trappe from analysis and recomputing the nearest neighbor index for the remaining rural structures still reveals a statistically significant clustering of structures in the Dilworth map. The new nearest neighbor ratio is 0.78 with a z-score of -13.56 and a p value of 0.00. Even when removing the densely clustered urban structures, the distribution of structures throughout the rest of the county still exhibits clustering. Rather than being deliberately placed far apart or even at random, it appears that there is a preference for clustering structures in the county. Examining the map, this pattern appears throughout, especially along the coastline, along roads, and at road intersections. Similarly, a density map of the road network in the county reveals a similar pattern, with the highest concentrations of roads being in Easton and St. Michaels were the town grids account for a significant amount of the local road network (Figure 4.12). The average distance of these structures to a road is 1218.80 feet. It 89 should be noted that in some places, it appears Dilworth has drawn what are private plantation driveways as roads and in other places left these driveways off of his map. For example, the driveway at Wye House is drawn and depicted on Dilworth’s map 90 as part of the county road network whereas the one at Gross Coate, a Tilghman plantation just across Lloyd Creek from Wye House is not. Spatializing the 1860 United States Census for Talbot County, Maryland While the use of spatial statistics allows us to describe and analyze this dataset, our use of the 1858 Dilworth map of Talbot County, Maryland does not have to stop at a descriptive analysis of the physical and cultural landscapes depicted by it. We can start with this digitized version of the Dilworth map and use it to build a spatial and demographic database which combines the map with another contemporary historic dataset—the 1860 United States Census (United States Bureau of the Census 1864). Historians, archaeologists, and other past-minded researchers have long turned to historic United States Census returns in order to conduct historic research. Conducted every ten years as mandated by the United States Constitution, the federal Census and its related datasets provide researchers with a wealth of demographic data that can be analyzed aggregately or even used to locate specific individuals. The ultimate goal here of using the 1860 United States Census returns for Talbot County, Maryland is to be able to embed the demographically-rich census data into the spatially-rich map data, resulting in a more robust historical spatial demographic dataset. In 1860, Schedule 1 of the United States Census lists the “Free Inhabitants” in each district within the county. This census records the names, ages, sexes, occupations, place of birth, and value of both personal and real estate of each free inhabitant, assigns a unique dwelling and family number to each dwelling and 91 family, and describes several other characteristics such as if they were recently married, in school, illiterate, or disabled. Schedule 2 of the 1860 US Census lists the “Slave Inhabitants” in each district. Grouped together by their owners, this census does not list individual names of the enslaved. There are some instances this information was recorded elsewhere such as the plantation-level censuses recorded by the Lloyds at Wye House between 1770 and 1834, but not on a county-wide scale. This enslaved United States Census only records the age, sex, and race of those enslaved in the county in 1860 as well as provides the numbers of enslaved who have been a fugitive, were manumitted, and the number of slave houses, all aggregated by slave owner. The 1860 US Census returns for Talbot County, Maryland (both Schedules I and II) were transcribed by University of Maryland undergraduate students Clio Grillakis and Marcella Stranieri as an additional part of the Freedom of Information Alliance Robert W. Deutsch Foundation Seed Grant awarded in 2013. Before continuing, it is important to note that the 1860 United States Census is a product of its time and specific cultural circumstances. As such, some of the language used below to discuss this dataset is not preferred in the present (or even in many cases acceptable); however, given this terminology and categories were the ones used to collect this data in 1860, it is difficult to discuss this dataset without understanding the historical context in which it was initially collected. Aware of this issue of discussing social datasets collected in the past using outdated and sometimes offensive terminology, the United States Census Bureau currently includes the 92 following disclaimer with many of its historic data products in the form of “A Note on Language” located at the bottom of its Publication webpages: Census statistics date back to 1790 and reflect the growth and change of the United States. Past census reports contain some terms that today’s readers may consider obsolete and inappropriate. As part of our goal to be open and transparent with the public, we are improving access to all Census Bureau original publications and statistics, which serve as a guide to the nation's history. (United States Census Bureau 2018) Metadata regarding the collection of the Census and the definition of its data fields can be found in a document prepared by the Census Office for their Marshals titled, Eighth Census, U.S. Instructions, &c. (United States Bureau of the Census 1860). Within this document is a copy of the Act of Congress authorizing the taking of this Census as well as specific instructions to the Marshals tasked with carrying out the enumeration on the local level. These instructions include guidelines for hiring and managing assistants, handling subdivisions of districts, transmitting returns, numbering pages, caring for pages, as well as detailed definitions of each attribute to be recorded. These instructions help us better understand each piece of data collected by making it clear what the census taker is recording with each notation. For Talbot County in 1860, the Marshal responsible for conducting the census was N. E. Nichols, himself appearing on line 9 of page one of the county’s census returns as a resident of Easton. His name also appears again in the enslaved census as the owner of a 35-year-old enslaved black male. While being split between two population schedules on the 1860 Census, some overall demographic information can be calculated. The attributes shared between the two censuses allows us to calculate total population, and distributions of age, sex, and race. It also allows us to compute these attributes cross tabulated. 93 As recorded on June 1st, 1860, there were 11,071 free persons (74.83%) and 3,724 enslaved persons (25.17%) living in the county for a total population of 14,795 people. Sex was recorded as a nominal variable as either M for Male or F for Female. Of these 14,795 individuals, there were 7,452 males (50.37%) and 7,341 females (49.62%). This proportion remains close when breaking down sex by race. Of free persons, there were 5,563 males (50.25%) and 5,506 females (49.7%) and of enslaved persons, there were 1,889 males (50.73%) and 1,835 females (49.27%). Race was recorded as a nominal variable as either W for White, B for Black, or M for Mulatto under the column in the census titled “Color”. In most cases in the census, W or White is not actually recorded on the census form. Rather, for these cases, the attribute is left blank and whiteness is to be assumed by default. The other option for the census taker under Color is Ind. or Indian. In Talbot County in 1860, there are no individuals labeled or identified as Indian. The instructions given to census takers regarding recording Color make it clear that, “it is very desirable to have these directions carefully observed.” The census records 8,106 white persons (54.79% of the total population), all of who were recorded as free. It records 5,946 black persons (40.19% of the total population), 2,743 who were recorded as free (18.54% of the total population or 46.13% of the black population) and 3,203 who were recorded as enslaved (21.65% of the total population or 53.87% of the black population). It also records 743 persons (5.02% of the total population) listed as mulatto, 222 of which are listed as free (1.50% of the total population or 29.88% of the mulatto population) and 521 of which are listed as enslaved (3.52% of the total population or 70.12% of the mulatto population). 94 The age of individuals was recorded as a numerical variable in the format of a number of years. The age recorded is “the specific age of each person at his or her last birth day previous to the 1st day of June, [1860]”, that is, their whole age in years. In those cases where the individual was under a year old, their age was recorded as a fraction of months out of a year (1/12 for one month, 6/12 for six months, etc.). In cases where the exact age of an individual could not be determined, the best approximation was used. The mean age of a person living in Talbot County in 1860 was 22.89 years old and the standard deviation was 18.00 years. For those free persons, the average age is just a little higher at 23.78 years with a standard deviation of 18.44 years. There were three cases of free persons with illegible or unrecorded ages. For those enslaved persons, the average age is just a little lower at 20.26 years with a standard deviation of 16.35 years. Schedule 1, the free schedule of the 1860 United States Census, records several additional attributes not recorded by Schedule 2, the enslaved schedule. In addition to the name of the recorded individual, these include Dwelling House and Family Number, both numerical variables that identify which structure or family the individual lives in or belongs to, Occupation, a nominal variable that describes the occupation of the individual, Real Estate and Personal Estate, both numerical variables that record the value of the assets held by an individual, Place of Birth, a nominal variable recording the state or country where the individual was born, as well as several other variables such as whether they were married, at school, illiterate, or disabled. 95 As recorded in the 1860 Census, there were 2,024 separate dwelling houses in Talbot County, almost double the 1,169 total structures shown by Dilworth on his map. A dwelling house is defined as “a separate tenement, inhabited or uninhabited, and may contain one or more families under one roof” (United States Bureau of the Census 1860:13). The average (mean) number of inhabitants in these dwellings is 5.47 and the standard deviation is 2.85. The largest number of inhabitants recorded in one dwelling is 22 and there are 35 uninhabited dwellings with no recorded residents. In most instances, the family field is synonymous with the dwelling field. A family is defined as “either one person living separately and alone in a house, or a part of a house, and providing for him or herself, or several persons living together in a house, or part of a house, upon one common means of support and separately from others in similar circumstances” (United States Bureau of the Census 1860:13). There are 2,048 separate families recorded in the census. The census also records the profession, trade, or occupation of each free inhabitant of the county. However, the most common response for this attribute is blank. 7,731 out of 11,071 free inhabitants (69.83%) had no occupation or profession listed for them. While the instructions given to the census taker indicate that “the employment of every person over 15, having an occupation, should be asked and recorded. In every case insert the kind of labor and nature of apprenticeship” (United States Bureau of the Census 1860:15, italics in original), many have no such notation. Of those recorded professions, the most commonly listed are farmer (1,110 cases), farm manager (33 cases), laborer (943 cases), farm laborer (128 cases), farm hand (42 cases), servant (210 cases), carpenter (79 cases), house carpenter (19 cases), ship 96 carpenter (8 cases), sailor (72 cases), mariner (34 cases), oysterman (15 cases), merchant (67 cases), shoe maker (36 cases), teacher (36 cases), sewing (29 cases), tailor (23 cases), seamstress (12 cases), clerk (28 cases), manager (26 cases), blacksmith (28 cases), physician (17 cases), miller (14 cases), wheelwright (13 cases), and lawyer (10 cases). Other, rarer occupations include Surgeon, Lieutenant, and Captain of the Navy, watch maker, pump maker, dentist, machinist, barber, plasterer, editor, druggist, gardener, teacher of music, plumber, hatter, comb maker, gas filler, photographer, coroner, lantern keeper, and Deputy Marshal (N.E. Nichols, the census taker himself). Collectively, this listing of occupations provides a cross-section of the economy of the county in 1860. It highlights the agricultural focus of the county as well as some of the other kinds of jobs and industries needed to support this economy. The total sum of Real Estate held by individuals residing in Talbot County as recorded in the 1860 US Census is $6,848,264. This represents wealth in land and in buildings. This number was ascertained by “personal inquiry of each head of a family” and is “the value as given by the respondent” (United States Bureau of the Census 1860:15). This figure includes property within Talbot County as well as property located elsewhere but held by individuals residing in Talbot County. The average (mean) amount of property held by a family is $3,343.88; however, this distribution is heavily skewed. Only 395 families (19.87%) own more property than this average whereas 1,108 (54.10%) own no real estate at all. The total sum of Personal Estate in Talbot County as recorded by the 1860 US Census is $5,225,769. This figure represents “the value of all the property, possessions, or wealth of each individual which is not embraced in [Real Estate].” 97 This includes “the value of bonds, mortgages, notes, slaves, live stock, plate, jewels, or furniture” (United States Bureau of the Census 1860:15). As with Real Estate, this figure is an estimation made by the respondent. The average (mean) amount of personal property held by a family is $2,554.14 and the mean held by an individual is $472.32. Again, this figure is heavily skewed toward the wealthy. 321 families (15.67%) own more property than this average while 605 families (29.54%) own no personal property at all. 1,041 (9.40%) individuals own more than this average while 9,243 individuals (83.49%) own no personal property at all. The Census also records the place of birth of each free inhabitant of the county in 1860. Where applicable, the census records the state of their birth or in cases where not applicable, the country. As is to be expected, the majority of residents of Talbot County (10,406 individuals or 93.99%) were born in the State of Maryland. As is also expected, most of the rest were born in nearby states like Delaware (266 individuals), Pennsylvania (108), New Jersey (33), Virginia (32), Washington, D.C. (19), New York (12), North Carolina (6), Ohio (4), and Kentucky (3). Fewer come from states farther away like Massachusetts (10), Indiana (5), Illinois (4), Maine (3), Mississippi (3). There is also a small but notable immigrant population in the county in 1860. 61 residents were born in Ireland, 54 were born in Germanic states, 17 were born in England, 3 were born in Scotland, 2 were born in Holland, and one individual was born in France, Mexico, and China each. The census records that 133 residents of Talbot County were married “within the year previous to June 1 [1860]” (United States Bureau of the Census 1860:16). It also records that 1,601 residents “who are or have been in educational institutions, or 98 who have been receiving stated instruction in any manner within the year; those whose education has been limited to Sunday schools are not to be included.” (United States Bureau of the Census 1860:16). It indicates that 2,246 “persons over 20 years…cannot read and write” (United States Bureau of the Census 1860:16). The census taker also recorded which residents were disabled using the following designations: 10 were “blind”, 3 were “deaf”, 1 was “dumb”, 5 were “deaf and dumb”, 9 were “idiotic”, and 5 were “insane”. The data recorded in Schedule 2 but not in Schedule 1 includes four additional attributes: recording fugitive slaves, manumitted slaves, disabled slaves, and slave houses. Of the 3,724 enslaved persons recorded in this census, only four (only one- tenth of one percent) were recorded as having been fugitives, that is, “escaped within the year [ending June 1, 1860, and have not been returned to their owners” (United States Bureau of the Census 1860:18). This number seems extremely low and the statistic is almost certainly underreported. The four recorded as fugitive were a 20- year-old black male, a 19-year-old black male, a 15-year-old black male, and a 22- year-old mulatto male. Of the 3,724 enslaved persons recorded in this census, 95 were recorded as having been manumitted within the year ending June 1, 1860. 54 were female (56.84%) and 41 were male (43.16%). The average age of those manumitted was 20.58 years. The average age for those manumitted females was 21.12 years and the average age for those manumitted males was 18.70 years. 81 (85.26%) of those manumitted were black and 14 (14.74%) were mulatto. 99 Like Schedule 1, disabilities were recorded among the 3,724 enslaved persons on the enslaved schedule. Four were recorded as being “blind”, seven “idiotic”, two “insane”, and one “dumb”. There are 145 separate slave quarters recorded. Each of these are “separate tenements…occupied by slaves” (United States Bureau of the Census 1860:20). The instructions to census takers states, “we wish by this column to learn the number of occupied houses, the abode of slaves, belonging to each slaveholder.” It would appear that at least some of the demographic information collected by census takers is directly relevant to this investigation. Only 92 of the 503 slave holders listed have separate slave quarters. 71 of these 92 only have one separate quarter, 13 have two quarters, two have three quarters, two have four quarters, one has six quarters, and one has seven quarters. Only Edward Lloyd VI of Wye House has more than seven slave quarters with 21 recorded separate dwellings for his enslaved laborers From the names of the slaveholders listed in Schedule 2, we can begin to understand the economic and social topography of slaveholding in Talbot County on the eve of the Civil War. Here, there are listed 503 individuals (3.40% of the total population) as owning the 3,724 enslaved laborers recorded in the county (25.17% of the total population). The average (mean) number of enslaved per slaveholder is 7.43 with a standard deviation of 19.94. The distribution of these numbers within the county is highly skewed. 115 individuals (0.78%) own only one slave, 310 (2.10%) between 1 and 5, and 408 (2.76%) between 1 and 10. Only 96 (0.65%) own more than 10 slaves, 31 (0.21%) more than 20 slaves, 6 (0.04%) more than 50 slaves, and only one individual (0.007% of the county’s total population)—Edward Lloyd VI—owns 100 more than 65 enslaved laborers. He is listed as owning 366 slaves, a number supported by internal censuses kept by the Lloyds through 1834. The majority of individuals in Talbot County do not own any slaves (14,292 or 96.60%). As with property, a small number of families own a significant proportion of the enslaved population. As might be predicted, the same families that own significant amounts of land also enslave large numbers of laborers to farm these lands. As recorded in the 1860 census, the Lloyds own 479 enslaved laborers, the Goldsboroughs own 167, the Hughletts own 126, the Tilghmans own 92, the Rogers own 61, and the Arringdales own 41. To this list, we should add the Lowes who own 61 enslaved people, the Martins, the Lounds, and the Skinners who own 58 enslaved people each, the Harrisons who own 53 enslaved people, the Hollidays who own 52 enslaved people, and the Dawsons who own 51 enslaved people. Together, these thirteen families enslave 1,357 people or 36.44% of the enslaved population of Talbot County, Maryland in 1860. While Schedule 2 of the 1860 United States Census does not provide the names of those enslaved in the county, in some instances we are able to recover that information. Leaving the federal census of 1860 for a moment, we can turn to a different kind of census. From Wye House, Edward Lloyd VI maintained extensive records of his plantation. Among these are plantation specific internal slave censuses that, unlike the federal censuses, record the names of those enslaved. In addition to providing us the names of those Africans and African Americans enslaved, these kinds of plantation-specific censuses kept at Wye can be used to reconstruct the demographics and circumstances surrounding the creation of an African American 101 culture in the context of this plantation. From 1770, these censuses provide a series of demographic snapshots of those enslaved on this plantation and provide a wealth of information as the population grows and changes through time. Beth Pruitt, working through Archaeology in Annapolis, has digitized many of these censuses and made them available online as a searchable database People of Wye House (http://aia.umd.edu/wyehouse). The following analysis was conducted by this author and a version of it appears on the People of Wye House website. Other historians have already worked with historic plantation census data like this one from Wye House. Lorena Walsh, working with decades of internal census data at Carter’s Grove in James City County, Virginia, has been able to piece together the creation and track the development of the enslaved African and African American community on that plantation (Walsh 1997). As with the federal enslaved census, these internal censuses record both ages and genders of those individuals enslaved on the plantation. As such, we are able to reconstruct a demographic snapshot of the population for each of the census years as well as track individuals between censuses which were taken in 1770, 1772, 1773, 1774, 1778, 1779, 1781, 1784, 1787, 1788, 1790, 1791, 1792, 1794, 1795, 1796, 1805, 1822, 1823, 1824, 1826, and 1834. Following the demographic study of two enslaved populations from Tidewater Virginia conducted by Lorena Walsh (Walsh 1997:109-170), we can identify a similar pattern in the Wye House censuses. Because of the economic considerations of plantation-based slavery and the kinds of labor required to grow agricultural commodities for export, many slave owners deliberately purchased younger, working-age males. Furthermore, during much of the seventeenth 102 and eighteenth centuries, the enslaved population in colonized North America as well as in the British colonies, was largely not self-sustaining. That is, there were more deaths than births and in order maintain the size of the enslaved workforce, enslaved workers had to be imported into the colonies. While this trend was less pronounced in the Chesapeake when compared to South America or the West Indies, it is still present. To work in the agricultural fields, plantation owners would need productive, young, male laborers. Given the requirements of the tobacco monocrop agricultural system throughout the Chesapeake and the availability of new laborers through the transatlantic slave trade, Walsh describes this as an expected demographic pattern as slave owners such as the Carters or the Lloyds are shaping their decisions regarding enslavement for this purpose (Walsh 1997:139). In her analysis, Walsh sees this pattern in the pre-Revolution Carter censuses at Carter’s Grove and when examining the 1770 Lloyd censuses of Wye House and its outlying plantations with this demographic pattern of age cohorts in mind, we can identify several cohorts or clusters of similarly aged individuals in the dataset. These clusters of between ages 15-25, 35-40, 45-55, and 70-80 as seen in 1770 (Figure 4.13) may represent several of these cohorts, purchased to supply labor for the plantation as the cohort above became physically less productive and new labor was needed on the plantation. As historians have noted, at some point in the eighteenth century the enslaved population in American undergoes a fundamental demographic and cultural shift. Researchers at the University of Richmond’s Digital Scholarship Lab have used historic Census returns between 1810 and 1860 to track the spread and decline of slavery through in-migrations and out-migrations as the demographics and 103 topography of enslaved agricultural production changes (Nelson et al. 2017). They specifically note changes in demographics as the population in many areas of the upper South, especially the Chesapeake, becomes self-sustaining in the late eighteenth and early nineteenth century as the birth rate begins to overtake the death rate and the result is a general out-migration to areas further south and west as slave holders adjust their agricultural and labor practices to follow (Nelson et al. 2017). Coupled with a national ban on the importation of slaves in 1808, this demographic shift marks a transition from African slavery in the Americas to the creation of a new, uniquely African American slave community. Walsh identifies this trend at Carter’s Grove with the disappearance of specific cohorts which are replaced by a more natural population distribution. The censuses at Wye House reveal this historical trend as well. By 1834, not only are males and females roughly equally distributed, but the age distribution is much closer to what one would expect to find in a naturally-reproducing population. So while Douglass tells us in My Bondage, My Freedom that there are African-born slaves still present in 1855, writing, “At the time of which I am now writing, there were slaves there who had been brought from the coast of Africa” (Douglass 1855:76), the importation of new slaves onto the plantation from Africa had been illegal for more than a generation. This demographic pattern is not one that is created by purchasing cargos of enslaved individuals but by a self-sustaining population. While Douglass was enslaved at Wye House, he was enslaved by Aaron Anthony, the overseer hired by Edward Lloyd to manage and run Wye House and as such, does not appear on the Lloyd’s slave censuses. 104 The People of Wye House database is also useful in that it tracks many of the same people through time. Reproduced here and formatted by Pruitt for inclusion on the website are the population distributions recorded from the 1770, 1792, 1805, and 1826 censuses (Figure 4.13). Note that the cluster of 10-25-year-olds in 1770 appear again as a cluster of 25-to-40-year-olds in 1792, again as a cluster of 35-to-50-year- olds in 1805, and again as a cluster of 55-to-70-year-olds in 1826. From the censuses, we can track this group from childhood to adulthood and into old age. 105 This group is of particular note because when Douglass writes of those enslaved at Wye, these are the elders in that community. He writes, “Strange, and even ridiculous as it may seem, among a people so uncultivated, and with so many stern trials to look in the face, there is not to be found, among any people, a more rigid enforcement of the law of respect to elders, than they maintain” (Douglass 1855:69). This cohort represents Douglass’s elders. These individuals possess familial knowledge, community history, and represent a physical connection to the past. The consequence and importance of this transition in demographics in terms of the creation of a uniquely African American culture among the enslaved population between 1770 and 1834 cannot be overstated. In 1770, we see evidence of a group of people who were purchased explicitly to work on a plantation and were replaced as they became less efficient at producing agricultural crops. By 1826, this pattern is gone, and in its place we see one of a self-sustaining community; a community with its own history, memory, and culture. Having these kinds of datasets allows us to understand the demographic makeup of Talbot County just prior to the American Civil War. Transcribing and digitizing the federal Census as well as internal plantation censuses where and when available allows us to calculate these kinds of statistics that reveal some of these demographic trends. Future work with these kinds of datasets from elsewhere in the American South can compare and contrast the effects of slavery to those populations subjected to it. Frederick Douglass opens Chapter 4 of My Bondage and My Freedom by writing, “It is generally supposed that slavery, in the state of Maryland, exists in its 106 mildest form, and that it is totally divested of those harsh and terrible peculiarities, which mark and characterize the slave system, in the southern and south-western states of the American union” (Douglass 1855:61) He goes on to explain why this notion is a fiction. While many of the “peculiarities” Douglass refers to come from the “cruelty and barbarity of masters, overseers, and slave-drivers” and are either tempered or fostered as a result of “public sentiment”, components of this system and its severity may be visible on a demographic level, in the form of skewed age and sex distributions, noted disabilities, or other recorded attributes. As has been shown above, the 1860 US Census represents a remarkable dataset for understanding historical demographics at a particular moment in time. Unfortunately, while it and other similar datasets hold a great deal of information about these people who lived in the past, they generally do not contain spatial information that would enable the researcher to map these data. There are no addresses listed in 1860 nor do the returns appear to be ordered spatially. This dissertation is ultimately concerned with historic distribution, configuration, and use of space. This demographically rich dataset does not have the kinds of spatial information needed to understand the spatial distribution of that data but there is a way forward. To solve this problem, it is possible to use the tools of historical GIS to join these two historic datasets—the 1860 US Census and the 1858 Dilworth map— that is, link the spatial information contained within the historic map to the demographic data contained within the 1860 US Census. Others have already joined census data to spatial datasets. Notably, the National Historical Geographic Information System (NHGIS) run by the Minnesota 107 Population Center at the University of Minnesota (www.nhgis.org) has census data available aggregated at the state and county level since the first United States federal Census in 1790. While useful for many applications, aggregating this data at the county level does not allow researchers to go beyond county-level analyses of demographic trends and patters. Ian Gregory and Humphrey Southall accomplish a similar union between British census tracts and census returns through time (Gregory and Southall 2002), but again, this data is aggregated at the level of the census tract. With this data, one cannot locate individuals in the past. The scale at which this spatial information is recorded does not allow for the kinds of fine-grained analyses presented below. Joining the 1858 Dilworth map with the corresponding census data will create a new spatial and demographic dataset at an extremely high spatial resolution. Embedding census data into spatial data sources is not new; however, what is new here is the scale at which this process is carried out. From the work discussed above, we can use the digitized copies of both of these documents to create a demographically-rich spatial dataset and a spatially- embedded demographic dataset. On the census, each free individual in the county is listed by first name and last name. As discussed, enslaved individuals are not listed by name in the 1860 census; rather, they are listed by slave owner. On the historic map, property ownership is recorded with the first initial and last name of the property owner, which is generally enough to identify a specific individual as the head of household. Because it is shared between the two datasets and is mostly unique, this name of the property owner or head of household is the primary key by which these two datasets can be merged or joined. The names of the property owners listed on the 108 map can be matched to the names of heads of household on a census (Figure 4.14). This is ultimately a problem of juggling and merging spatial data with a database of associated data, a task GIS is well-positioned to handle. Once this link has been established between the 1860 US Census and the Dilworth map using the names of property owners as the common link, that is, once the individuals named on the map are located in and tied to the census, the names and data from the census will be attached to the boundaries and points representing properties and structures on the 109 map through the names on the map. Extending the link between these datasets one step further, we can use the unique dwelling house number assigned to each of these located property owners/heads of household to identify all of the other individuals living at the same locations. These people not listed on the Dilworth map, but through this attribute we can locate them and embed this additional demographic data within our map as well. This combination of data represents new geographic and historical knowledge. One of the motivations behind the creation of this spatial demographic dataset was to make it widely available for the public to use in order to find ancestors in time and space or to satisfy their curiosity about the history of the places where they live. To this end, part of the 2013 grant from the Future of Information Alliance described above that supported the creation of these digital datasets also paid for the creation of an online WebGIS component to this project. Taking advantage of the online mapping capabilities of the same GIS software that was used for the above analyses, this platform allows users to search for and locate specific names in the spatial census database, view county-wide distributions of people, overlay the historic maps and data layers with contemporary map layers, and otherwise explore the spatial dataset created above. It allows access to many of the same spatial tools used to create the above analyses without subjecting the end user to the steep learning curve or costs associated with producing them. The following work was funded by this grant and this author was assisted by University of Maryland undergraduates Clio Grillakis and Marcella Stranieri and University of Maryland graduate students Stefan Woehlke, Beth Pruitt, and Kathryn Deeley in digitizing these two datasets and publishing the 110 results in an online site titled Locating People in the Past at http://aia.umd.edu/locate/ along with an interactive map based on the ArcGIS for Flex platform. Of the 1,169 number of structures visible on the Dilworth map, 651 appear to be residential structures and are labeled with the naming convention described above (first initial followed by a period, sometimes second initial followed by a period, last name) (Figure 4.15). It should be noted again that all 1,169 structures on the map do not represent the total building stock of Talbot County in 1858 nor does it represent every residential structure in the county. Schedule 1 of the census lists 2,024 unique dwellings in the county and Schedule 2 lists 145 unique slave quarters. The “country village” described by Douglas at Wye House is represented with only two structures on the Dilworth map. Nonetheless, these 651 labeled structures, when coupled with the link between the two datasets, gives us the framework onto which to place people back onto this map. In order to join the owners of these residential structures to the same names in the census, they need to be in the same format so that the computer will be able to identify those names that are the same. For a human observer with a little bit of background information, it is relatively simple to note and understand that “Col. E. Lloyd” on the map is the same individual as “Edward Lloyd” in the census. For the GIS program performing this join, it has no way of making this connection unless the data is first processed in such a way that allows for these comparisons. When digitized, the names on the Dilworth map were placed into a single column as they are written on the map. The names from the census were transcribed as two separate 111 attributes as they were originally recorded in the census. First name and middle name/initial were transcribed in one column and last name was placed in a second. The next step is to structure both sets of names so that they will be in similar formats. In Excel, by using the expression “=LEFT([cell with first name],1)&"."” ] we can 112 create a new column of data that will extract the first letter from each of the first names in the census and add a period after it. This represents the first initial of each individual as found in the census but formatted like the Dilworth map. The expression “=[cell with first initial and period]&" "&[cell with last name]” will combine this newly created column of formatted first initials with the corresponding last names, giving us a new column with census names formatted to match the format of the Dilworth map. The next step is to join these two datasets in ESRI’s GIS program ArcMap using these formatted names as the primary key between the two datasets. A join between two datasets such as these will find corresponding records between the two datasets and make a connection between the two sets of data wherever a match exists. In this case, this process first finds all instances in which the name on the Dilworth map matches exactly the formatted name from the census, that is, those places where the two records are identical, and then adds all of the additional data associated with that name in the census such as age, sex, occupation, and, as will be extremely important below, dwelling number, to the corresponding record in the table. This join is the most crucial part of combining these two datasets because this is the step that embeds the demographic data from the census into the spatial dataset that is the Dilworth map. Before carrying out this join, it is important to understand both how the program carries out the join and how to optimize our datasets in order to maximize the number of matches. The join function of ArcMap will only join two datasets if the primary key is a perfect match, that is, both strings of characters are exactly the same. 113 An individual in the map will be matched with his or her associated information from the census if and only if the string of characters in the maps’ name cell matches exactly the string of characters in the census’ name cell. This is why it is so important to ensure that both datasets are formatted properly before attempting the join. As seen above, while Excel makes it relatively easy to automate this formatting in most instances, there are specific cases that require additional attention. Individuals whose names on the map contain a middle initial cause problems for several reasons. Generally, the census does not record this information or it is inconsistent when recording middle names. Having middle initials present when attempting to join these two datasets will cause the join to fail because one dataset will have an initial and the other will not and therefore they will not match. The solution is to remove all middle initials from the names from the Dilworth map prior to the join for the sake of consistency; however, it should be ascertained that this middle initial was not needed to distinguish between two or more individuals. For example, the Dilworth map records a property owner named W.T. Goldsborough. There are no W.T. Goldsboroughs in the census; however, there is a William Goldsborough and a Wm [William] Goldsborough, both of whom could be this W.T. Goldsborough from the map. Upon examining these two census records, Wm Goldsborough is listed in the census as a 58-year-old white male who is a farmer with almost $100,000 of real and personal property whereas William Goldsborough is a free 7 year old black male with no recorded property. Therefore, the W.T. Goldsborough on the map is almost certainly Wm Goldsborough from the census, and not William Goldsborough. However, before this link can be made, we must first remove the middle initial from 114 the map record to read “W. Goldsborough” and second, we must make sure that this record joins to Wm Goldsborough and not William Goldsborough when the join is made. Likewise, the names on the Dilworth map also sometimes include titles and honorifics. These include military ranks such as Gen. (General), Col. (Colonel), or Capt. (Captain), professions or trades such as Dr. (Doctor), Rev. (Reverend), or Capt. (non-military ship captain), political office such as Gov. (Governor), personal titles such as Mrs., Miss, and Mr., and a notation that indicates the specific property is owned by the heirs of a recently deceased individual. These are all placed before the name of the individual on the map and need to be removed from the dataset before attempting the join for the same reason middle initials must also be removed. As is the case with middle initials, it should be determined that these honorifics and titles are not used to distinguish between two or more individuals. For example, the owner of Wye House Edward Lloyd is listed both in the census and on the map. In the census, he is listed as Edward Lloyd. On the map, he is listed as Col. E. Lloyd. Here, we must remove “Col.” from the record so it reads, “E. Lloyd” before our join can be successful. Another problem that arises when attempting to join these two datasets is actually a pair of related issues. First, when transcribing thousands of records recorded in a loopy, sweeping cursive hand from hundreds of faded, torn, and damaged pages, there are bound to be transcription errors that creep into the digitized dataset. In the original census, some records are difficult to read and the transcriber has had to use their judgment in determining the contents of entries throughout the 115 dataset. Similarly, a second issue to consider when working with these mid- nineteenth-century datasets is the accuracy and consistency in spelling of the names not just by the transcriber as above, but by the census taker when initially creating this dataset as well as by those individuals whose names appear in these documents. It is possible that while literate, the census taker might not be familiar with the spellings of the names of the people he recorded. It is also possible that those whose names are recorded in the census who may not be literate themselves and the fidelity of the spelling of their own names may be more casual. In either case, inaccurate or inconsistent spellings of names will and do appear in the census, an error that will be perpetuated in the transcribed dataset. Both of these issues leave us with the problem that names in the census are not going to match names on the map not because they represent different people but because they are just spelled differently. The join performed by the ArcMap software will not recognize these records as belonging to the same individual because it makes a link only where there is a perfect match between the two datasets. Fortunately, while this problem is caused by our reliance on a computer to match records, we can use a computer to solve this issue. Rather than rely on rigid one-to-one matching, there are ways to apply a fuzzy match or fuzzy lookup to our digitized datasets in order to find those cases that mostly match each other and determine whether or not these records are legitimate matches. Microsoft provides a plugin for Excel called the Fuzzy Lookup Add-In for Excel which accomplishes this task. It is designed to match data records that are similar, but not identical, and can be used to smooth over variations resulting from common issues arising from working with transcribed 116 documents including spelling errors, inconsistent data entry, abbreviations or incomplete entries. All of these issues are present in our two digitized historical datasets. This add-in uses Jaccard similarity, which simply defined is a measurement or quantification of the similarities between two records—the number of common characters—divided by the total characters used—that is all of the characters used in both records. This results in a score for potential matches between 0.0 and 1.0 where the closer the two records, the closer to 1.0 the score will be. From this, we can identify and manually edit those records with high Jaccard similarity; those that fall into this category of misspelled or otherwise misidentified names. Now we can run our join a second time and catch these otherwise missed records. The result is an even better join between the names in the digitized census and the names on the digitized map. After accounting for these kinds of textual discrepancies, we can finalize the join between the 1858 historic map and the corresponding 1860 census and begin to analyze the results. Of the 1,169 structures on the map, 668 are listed as the residences of named individuals. The exactly 500 other structures are either not labeled clearly enough to indicate ownership to a specific individual, are non- residential structures such as churches, post offices, stores, schools, or mills, or are listed as residential structures but are labeled as “Negro Lots” with no other identifying information. Of these 668 identified structures, 18 are listed as belonging to the heirs of an individual or to an individual’s estate. In these cases, the named individual is deceased by the time Dilworth makes his map in 1858 but the property has not yet been inherited by another party. As such, the deceased will certainly not 117 appear in the 1860 census. This leaves 651 residential structures from the Dilworth map to locate in the 1860 United States Census. Of these 651 labeled residential structures on the map, our join is able to find matches with 396 names from the census, a success rate of 60.83% (Figure 4.16). 118 Through these 396 matched property owners between the map and the census, we are also able to embed those individuals who also live in these same households with the named property owner. Using the dwelling number assigned to each structure by the census and listed next to each individual therein, we are able to place 1,866 free individuals listed in the census to our map. Schedule 1 of the census lists 11,101 individuals living in the county. While we were able to locate the owner of 60.83% of the residential structures seen in the Dilworth map, this process is only able to place 16.81% of the total free residents of the county in 1860 (Figure 4.17). The average age of these located individuals is 23.60 years with a standard deviation of 18.04 years. 991 of them are male (53.11%) and 875 are female (46.89%). 1,493 are listed as white, 334 are listed as black, and 39 are listed as mulatto. The average amount of real estate property held by these individuals is valued at $1,748.89, however, the vast majority (1,634 or 87.57%) do not own any real property at all. Of those that own property, the average amount is valued at $14,066.49. Likewise, the average amount of personal property held by all located individuals is $1,398.26; however, the majority (1,509 or 80.87%) are not recorded as owning any personal property. Removing these individuals without personal property, the average amount of personal property held by those that hold personal property is $6,962.08. Because we are joining the census to the map through the landed, generally older, largely male, property owning individuals, we can expect this resulting dataset to be skewed toward this demographic. Because so much of this spatial research depends on how well the two datasets merge, it is important to understand the quality of the join. This entails 119 understanding which records from the map were not matched to records in the census and which records in the census were not matched to records in the map as well as determining whether or not those individuals we were able to place on the map are representative of the population at large. From this smaller pool of 255 unmatched 120 structures, we can try to understand where and why the join failed. The most common reason for failing to find a match is that there simply is not somebody with a similar name in the census. There are many reasons why this can occur. Frequently, the same surname will appear in the census but there will not be an individual with the correct first initial as indicated on the map. In 38 cases, it seems as if there is an individual with the same name living in Talbot County in the 1850s but that they have either died or moved between the time Dilworth created his map in 1858 and the census was conducted in 1860. These individuals can be located by going back to the 1850 Census where they appear in Talbot County. They are generally older individuals and seem to die before the 1860 census is taken. The two-year gap between Dilworth recording his map in 1858 and the census being conducted in 1860 certainly provides enough time for individuals to disappear from the record in this manner. In 23 cases, an individual could not be located in the Talbot County census returns; however, this is because they appear listed in neighboring counties. These include Queen Anne’s, Dorchester, Caroline, and Kent counties. In these cases, while they own property in Talbot County, because they do not reside in Talbot County in 1860, these individuals are excluded from this analysis because they are not recorded in the Census returns for Talbot County. In a larger regional study, it would be important to include these records; however, the nature and limits of these datasets forces us to focus exclusively on Talbot County. In nine cases, Dilworth only records the first and last initials or the first, middle, and last initials of the individual. This makes it incredibly difficult to identify the specific property owner. The initials “J.B.” could refer to 157 individuals from the 1860 121 Talbot County Census, “J.M.” could refer to 151 individuals, “J.N.H.” could refer to 107 individuals, “J.G.” could refer to 80 individuals, and “W.H.” could refer to 68 individuals. Including these nine cases where Dilworth only provides initials, there are a total of 62 cases where there are simply too many individuals that share the same name as is on the map. Without doing significant amounts of additional genealogical research, it is extremely difficult to determine which individual in the census is the individual on the map. With the 396 matched structures and these 123 understood failures caused by properties within our dataset, we are unable to match or explain why we could not match 132 labeled structures, or only 20.27%. Perhaps the individuals were residents in 1858 to appear on the map but not in 1850 or 1860 to appear in the census. Perhaps they own land in Talbot County but do not reside in one of the immediate surrounding counties. Perhaps their names are misspelled badly enough in one or both places to make them unrecognizable and unmatchable. Furthermore, there are several reasons to help explain why we were able to locate only 16.81% of those individuals listed on the free census. Primarily, the Dilworth Map does not show the totality of the building stock of Talbot County in 1858. Most properties only show one structure, owned by one individual. Even at a place such as Wye House, which Douglass describes as “[wearing] the appearance of a country village”, Dilworth only shows two structures. Furthermore, the urban centers of the county—Easton, St. Michaels, Trappe, Oxford, Royal Oak, and Wittman—only contain 140 structures between them, and this count includes civil and religious buildings such as post offices, stores, schools, court houses, and 122 churches. Those who rent property or their homes from somebody else will not have a link to the map and will not be located using this process. Several properties are marked “Negro Lots” and it will not be possible to identify the residents on these pieces of land through this process. While the process described above does a good job linking named property owners on the map to their corresponding records in the census, the map is missing key information about many of the households listed on the census that would allow us a more complete join between it and the census. This process can be repeated with the map and Schedule 2 of the 1860 Census using the names of the slave owners in the same manner the heads of household were used above to create the connection between the records in the census with properties on the map. While there are no names of the enslaved in this census, creating this join still embeds the other demographic information such as age, sex, and race. Of the 3,724 individuals listed on the enslaved census in 1860, the join between Schedule 2 of the census and the Dilworth map through the names of the slave owners is able to locate 1,985 individuals or 53.30% of those enslaved in the county in 1860 (Figure 4.18 and 4.19). The average age of these individuals is 19.96 years with a standard deviation of 16.20 years. 1,020 of these matched individuals are male and 965 are female. 1,739 are listed as Black and 246 are listed as Mulatto. These individuals are joined to the spatial dataset through a total of 138 individual slave holders. Between them is recorded 90 separate slave quarters or dwellings. Frequently, researchers can reconstruct the paths that census takers followed by going line by line through the census (Conzen 1969; Schnore and Bowden 1975:156). Unfortunately, using the spatial data produced by the join above, it does 123 not appear that the order recorded in the census conforms to a door to door survey of the county. By plotting the order of the matched structures from the map, the listings in the census jump haphazardly within each district. This makes it impossible to even roughly locate an unmatched structure by locating the closest matched structures 124 above and below it. It is unknown why the records to not conform to a geographic order like many other censuses. This kind of historical GIS research has the potential to offer a fresh new dimension to historical research. Spatial questions such as those addressed above 125 have the potential to reveal much we do not know about the past and can form the basis for studies of landscapes. While doing this kind of work is both time and labor intensive and requires training and expensive software and hardware, the potential benefits outweigh these hurdles. As shown in this chapter, being able to ask “where” of a historical dataset opens the door to new avenues of research. This spatial dataset embedded with demographic information will be especially important below as we analyze the spatial configurations of Frederick Douglass’s autobiographies. Future research can expand on the work above with Schedules 1 and 2 of the 1860 United States Census for Talbot County, Maryland. Of particular interest is Schedule 4, the agricultural census, which records improved and unimproved land, value of farming implements, numbers and value of livestock, and agricultural products. Also of note is Schedule 5, the industrial census, which records types of industries in the county, types and quantities of raw materials used, numbers of employees, and their wages. As with the population schedules used above, these demographic data can be tied to spatial data by using the names found on both the census and in the map. Being able to combine multiple datasets that at first seem incongruent or unrelated such as the United States Census and Dilworth’s map enables us to locate people in the past that have not left a heavy historical footprint. Through this work, we can combine what little information may exist regarding the lives of those enslaved in Talbot County and literally put these people back on the map. 126 Chapter 5: Mapping the Plantation: Using Airborne LiDAR and Historical GIS (hGIS) to Survey Extant Plantation Sites LiDAR (Light Detection and Ranging) Survey Underlying and antecedent to most of the questions asked by archaeologists is the question, “When?” Perhaps antecedent to that question, however, this dissertation considers the question, “Where?” It is incredibly difficult to answer this first question if you cannot find or do not understand the spatial patterns of a site. With the introduction and revolution of computer-based mapping and spatial technologies in the last decades of the twentieth century, archaeology is in the middle of a revolution in how we answer this second question, a revolution at least one archaeologist has argued is on the order of the invention of radiocarbon dating (Neustupný 1995:134). One of these new mapping technologies utilizes pulses of laser light to quickly and accurately map large areas of topography, even through vegetation. LiDAR, sometimes spelled lidar, LIDAR or LIDaR, is either an acronym for “Light Detection and Ranging” or is a combination of the words “Light” and “Radar” (Brooker 2009:12), as it is a mapping technology similar to RADAR or SONAR. It sends out pulses of energy from a sensor platform and converts the time it takes for that energy to reflect off of the surface or object to be mapped and return to the sensor (Bolstad 2008:249). The principles behind LiDAR mapping are relatively simple. Because the speed of light (referred to as c) is a known constant (c=299,792,458 meters per second in a vacuum and approximately 299,702,547 meters per second in Earth’s atmosphere), the time it takes for a pulse of light to be emitted, travel to the surface to 127 be mapped, reflect off of that surface, and travel back to the sensor platform, can be timed and converted into a distance by applying the following formula: distance from platform=time of flight*speed of light/2. This distance, combined with spatial coordinates of the sensor platform and both the horizontal and vertical angle of the sensor platform, can be used to calculate the coordinate of the point at which the pulse of light reflected off of the surface to be mapped. By repeating this process hundreds of thousands or even millions of times across the entire surface to be mapped, researchers can generate extremely detailed digital topographic maps of relatively large areas quickly and relatively cheaply. Furthermore, the timing, shape, and intensity of the return pulse can reveal additional information about the surface being mapped including material type and even the species of vegetation. Dong and Chen (2018) provide an overview for applications of LiDAR technology. For a technical introduction to LiDAR mapping, see Brooker (2009:215-243). Airborne LiDAR has become a popular archaeological survey tool since the mid-2000s. Archaeologists the world over have turned to LiDAR to identify and map historic resources. The strong tradition of landscape archaeology in the British Isles that traces its origin back to W.G. Hoskins in the 1950s and his The Making of the English Landscape (1955) has strongly contributed to LiDAR’s widespread acceptance in archaeological surveys and studies there. Remote sensing and LiDAR prospecting has largely been incorporated into the cultural resource management process, site reconnaissance, and research methodology in Great Britain (Crutchley 2006; Powlesland et al. 2006; Schindling and Gibbes 2014). Others working in this 128 context (Challis et al. 2011; Hesse 2010) have experimented with computer processing and visualization techniques to better interpret LiDAR returns. Further afield, LiDAR is also frequently used by archaeologists working in the jungles of Mesoamerica, often to spectacular effect (Chase et al. 2012; Fernandez- Diaz et al. 2014; Rosenswig et al. 2015). LiDAR’s ability to penetrate dense forest canopies and tree cover makes it possible to record the topography of the ground surface in areas where the ability to conduct pedestrian survey and visibility are extremely limited. These archaeologists have been able to use LiDAR to produce jaw-dropping images of pre-Columbian cities, temples, and pyramids, buried or lost in the dense tropical jungle as well as reveal the wide-ranging geographical extent to which these urban centers were surrounded by previously unidentified, dense, complex settlements. In Southwest Asia, there also exists long a tradition of archaeological studies that relies on remote sensing and aerial recordation of the landscape. These started with aerial observation and photography and have expanded to include modern multi- spectral satellite imagery (Hritz 2014). To this program of remotely-sensed landscape data, much potential exists for using the detailed topographic models provided by LiDAR mapping but the technology has not been widely used for a variety of economic and geopolitical reasons which have limited the availability and practicality of LiDAR data collection. Bjoern Menze and Jason Ur (2012) combine multi-spectral satellite data with coarse topographic surveys to quantify the intensity of human occupation across an entire region through a clever combination of remotely-sensed spatial data. They first identify sites of human occupation revealed by anthropogenic 129 soils and identified in Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) imagery and then calculate the volume of these anthropogenic soils by examining the vertical relief of these sites from Shuttle Radar Topography Mission (SRTM) global elevation data to approximate intensity of human occupation. Adrian Myers (2010) explains that the political realities on the ground in many of the countries in Southwest Asia (as well as in some American contexts like at the military detention facilities at Guantanamo Bay, Cuba) prevent the collection of the kinds of aerial LiDAR datasets discussed in this chapter; however, given the kinds of remotely sensed data and GIS spatial analysis already being conducted in the region, many researchers and research programs would greatly benefit from the inclusion of DEMs created by aerial LiDAR surveys in addition to other kinds of remotely-sensed data.. Recently, others have adapted the technology not for scanning the land but reaching under water from air and from space instead (Abdallah et al. 2013; Shih et al. 2014). LiDAR scanning for submerged resources is especially useful in shallow water where more traditional side-scan sonar or magnetometry is unable to operate. For airborne bathymetric LiDAR survey, the principles of laser range finding remain the same with only a few minor changes. The electromagnetic characteristics of the laser pulse are fine-tuned to better travel through the water column and the angle at which the sensor platform scans is much shallower to avoid reflection of the air/water boundary. LiDAR has also been used on North American historic period sites. Katharine Johnson and William Ouimet (2014) have used airborne LiDAR to locate, identify, and characterize abandoned agricultural landscapes in rural New England. They are 130 able to distinguish between several different kinds of above-ground archaeological landscape features including stone walls, building foundations, and abandoned roads and paths. They use this ability to lay out a methodology of archaeological field reconnaissance that incorporates LiDAR survey into the preliminary site identification process. They, as others have also done, argue that this LiDAR data can be combined with other kinds of historical data like maps, deeds, probate records, or censuses to make a more robust dataset. Lastly, Johnson and Ouimet remind their audience that these newly-available LiDAR datasets can be used within an anthropological framework to not only describe the features encountered in the scans but also to address the cultural processes that created them. Of most relevance to this dissertation, James Harmon, Mark Leone, Stephen Prince, and Marcia Snyder lay out in a 2006 article in American Antiquity titled “LiDAR for Archaeological Landscape Analysis: A Case Study of Two Eighteenth- Century Maryland Plantation Sites” a research methodology that maps plantation landscapes by incorporating new advances in airborne LiDAR with traditional archaeological survey and research. Harmon and his coauthors use two different LiDAR datasets to look at two extant plantation landscapes in the state of Maryland—Wye Hall in Queen Anne’s County (William Paca’s Eastern Shore home not to be confused with Wye House also discussed below) and Tulip Hill in Anne Arundel County. This article serves as a direct inspiration for the investigation conducted in this chapter and their suggestions for future research form a significant inspiration for this dissertation. 131 The authors contracted Airborne-1 Corporation to commission a private aerial survey in order to create their LiDAR dataset. From this data, the researchers were able to produce digital elevation models (DEMs) of these landscapes and use this data to begin characterizing these formal landscapes and their elements. At Wye Hall, the landscape components that were identified include the axes along which the main house and its gardens and outbuildings are arranged, terraces, garden beds, and other earthworks constructed throughout the plantation. While many of features are visible to the unaided eye, their vast scale, obfuscation by grass and other vegetation, and slight topographic relief make it difficult to characterize this landscape without a form of systematic, high resolution recordation. At Tulip Hill, researchers also characterized the extant landscape while also prospecting for potential archaeological features through subtle variations in topography. They note the descending terraces found on some Chesapeake plantations and a widening central path that creates converging and diverging lines of sight. Harmon et al. also experiment with different methods for visualizing their LiDAR-derived digital elevation models. There are many ways to visualize topographic data and in their article they provide three. Perhaps the most familiar to archaeologists is the contour map. The authors also experiment with color by utilizing a color ramp from blue to red with one end of this range representing low elevation values while the other end representing high elevation values. Lastly, they visualize the terrain through a hillshade visualization which simulates sunlight and shadow across the landscape and provides for a more natural-looking representation of the 132 mapped surface. While each of these visualizations has its strengths, they each have weaknesses which makes it important to consider how the dataset is being displayed. Building on these efforts, I would like to suggest a fourth method of displaying digital elevation data (Figure 5.1 shows these visualization methods at 133 Ratcliffe Manor, also discussed below). Utilizing color to display elevation data is a useful way for conveying depth or topographic relief across the landscape; however, it is less useful to discerning slight variations in topography that may be associated with human intervention or archaeological features. Likewise, the hillshade is useful in that it is able to display subtle variation in the landscape through the use of simulated shadows; however it can usually only display the shadow from a single simulated light source (the sun) at a time and it is difficult to compare elevations of different features and analyze overall topographic relief throughout the scene. As such, I have found that displaying both colorized DEM data as well as slope data simultaneously makes it possible to assess topographic relief while still picking up subtle topographic distinctions. By using the ability of GIS programs to display multiple overlapping layers simultaneously by adjusting the transparency of one of these layers so that the other will be partially visible beneath it, both the absolute topographic value of each pixel (elevation) and the local change in elevation (slope) can be displayed and considered together. Through Wye Hall and Tulip Hill, Harmon and others explore a new technology to investigate plantation landscapes in the Chesapeake and throughout the American South. They conclude: The implication is that one cannot analyze archaeological landscapes with LiDAR alone….The application of these methods will ultimately be most successful when the data are used in conjunction with information derived from the field, the laboratory, and from archival sources. This is in fact also the case with the majority of archaeological interpretations derived from the use of similar techniques, including aerial photography, and reinforces the need for integration of multiple techniques when attempting to understand archaeological landscapes. (Harmon et al: 2006:668) 134 They recognize some of the limitations beyond the technical ones of using LiDAR to examine historic landscapes; namely, that they are temporal palimpsests. The laser scanning platform will only record the ground surface at the moment in time the scanner is active. If a historic feature is not present at the surface of the ground in the form of topographic variation, it will not be detected by the laser scanner. The result is that the dataset produced by LiDAR scanning is the culmination of all activity on a landscape from the geologic creation of that landform through prehistoric and historic period occupation of that landscape through the present. For this reason, they recommend augmenting the LiDAR dataset with other kinds of datasets that fill in some of the gaps produced by the methodological limitations that result from the use of a temporal palimpsest. While they don’t explicitly state it, what the authors call for is a combination of LiDAR data into a historical GIS framework. Both Johnson and Ouimet (2018) and Grammar and others (2017) reiterate this call by stressing the importance of applying historical context and supplemental information to interpret LiDAR-derived elevation data on archaeological landscape remains. William Paca House and Garden, Annapolis, Maryland Before examining this call to integrate multiple techniques in historical landscape archaeology might look like, I would first like to turn to William Paca’s better-known Annapolis home—usually referred to simply as the William Paca House—before using the methods and ideas suggested here to detail several known, extant plantation landscapes. 135 The William Paca House (AA-657) is a five-part brick Palladian structure located at 186 Prince George Street in the heart of Annapolis, Maryland (Figure 5.2). Work began on this dwelling in 1763 (NRHP 1971:5). The center, five-bay-wide, two-and-a-half-story, gable-roofed block is connected to two one-story, brick, forward-facing, gable wings by one-story hyphens. The entrance on the south, street- facing façade is in the center bay and reached by climbing a stair to the raised first story that sits on a cellar. The other four bays on this side contain nine over nine hung sash windows capped with rubbed brick jackarches and flanked by shutters. Directly above the arches is a four-brick belt course. On the second story, directly above each opening in the first is a nine over six hung sash window, also topped with a rubbed brick jackarch; however, these windows do not possess shutters. This plan is repeated 136 on the north, garden-facing façade except for the presence of the dominating central pavilion and larger cellar windows. The entire lot is raised up several feet and surrounded by an all-header brick retaining wall. The building sits above the surrounding lots by several feet. The steep gable roof is pierced by five evenly spaced dormers on the south, road-facing side and two dormers that flank the large two-story, projecting pavilion that dominates the center of the north, garden-facing façade. The house was reworked and hidden by the addition of the Carvel Hall Hotel on the north side of the structure in 1907, which was subsequently removed in the late 1960s following intense architectural and archaeological study and restoration (NRHP 1971:4-5). Archaeologically, the original ground surface of William Paca’s garden was encountered under the hotel parking lot and nine feet of fill, as was evidence of the original perimeter wall, terrace falls, irrigation system, a pond, a springhouse, and a large garden pavilion (NRHP 1971:4-5). Many of these archaeological features can be seen in a 1772 portrait of William Paca painted by Charles Willson Peale (Leone 2005:71). Perhaps most famously, the Paca House in Annapolis is the subject of Mark Leone’s seminal 1984 chapter “Interpreting Ideology in Historical Archaeology: Using Rules of Perspective in the William Paca Garden in Annapolis, Maryland” published in Tilley and Miller’s Ideology, Representation and Power in Prehistory. Leone uses Marxian concepts of power, hierarchy, domination, and ideology to describe why Paca built his garden and how it functioned (Leone 1984). A significant inspiration for these studies has been the work of Michel Foucault, especially his work on discipline, surveillance, and observation. Leone’s argument contends that 137 gardens like William Paca’s were conceived of and built by white, land-owning males interested in either maintaining their own power and position in society by naturalizing and masking inequalities or in an attempt to acquire this power and position which they do not yet have. These gardens function to replicate the dominant ideology through the use of perspective. It has been established by others that underpinning the layouts of plantation landscapes are formal designs rooted in mathematical principles and underlying these designs are the same ideologies underlying the Enlightenment (Deetz 1977; Leone 1984). Archaeologists and landscape historians have noted the geometric relationships that can be found within formal gardens and landscapes designed throughout this period. That this mathematical basis exists should not come as a surprise to researchers. There exists an entire body of period literature that explains how best to design, create, and maintain a garden both in terms of as a living thing and as a mathematical exercise. These books, which were widely available in the eighteenth and nineteenth centuries, can be found listed in inventories of many of the elites. These include Leonard Meager’s The New Art of Gardening (1697), Jean de La Quintinie’s The Complete Gard’ner (1704), Batty Langley’s New Principles of Gardening (1728), and Philip Miller’s The Gardener’s Dictionary (1735). In his 1984 chapter, Leone argues that the adoption of these principles of garden and landscape design suggest an adoption of the ideologies of the Enlightenment, and more specifically, adoption of the ideologies of individualism as described by Deetz (1977). 138 In an attempt to understand the underlying geometry governing the design and construction of William Paca’s Annapolis property, Barbara Paca-Steele with assistance from St Clair Wright identifies and imposes a geometric grid over the site which she believes to be able to explain the locations of features of the garden (Paca- Steele 1986). Paca-Steele is able to “describe the practical and theoretical basis” (Paca-Steele 1986:299) for the placement of many of the elements in the archaeologically-known and reconstructed garden through ratios commonly cited and used in the construction of similar eighteenth-century gardens, especially 3:4:5. Paca-Steele’s argument regarding the design and construction of the garden starts from both an archaeological and a historical understanding of the garden and cites the period literature with which William Paca and his gardeners would have been familiar and turned to in order to construct his garden. These include texts Paca would have been exposed to as part of the curriculum at the College of Philadelphia, including works with names like A Treatise of Practical Geometry; Elements of Plane Geometry; and Short, but yet Plain Elements of Geometry. Paca-Steele notes that William Paca was also exposed to Isaac Ware’s translation of Andrea Palladio’s I quattro libri dell'architettura, or The Four Books of Architecture, which stresses proportion, ratio, and balance, as well as Architecture with Fortification and The Theory and Construction of Fortification. Chandra Mukerji has explored the relationship between military engineering and garden design in her Territorial Ambitions and the Gardens of Versailles (1997). These similarities include not just the surveying techniques required to implement the designed plan and construction 139 techniques to move significant quantities of earth but also the social hierarchies and control that are present in both. Paca-Steele begins by describing the dimensions of Paca’s property in Annapolis, two adjacent lots, Lots 93 and 104, each of which measures 198 feet by 198 feet, which result in a 2:1 rectangle which is 396 feet by 198 feet. Paca-Steele notes that these lots can be further subdivided evenly by using a commonly used surveying tool, the Gunter’s Chain, a 66 foot long chain, into three by three 66 foot by 66 foot squares. Historic garden and landscape researchers have noted that many of these landscapes are built using a key measurement or ratio of the main house or main room within that house from which these spaces were meant to be viewed. In a way, this projects the symmetry, dimensions, and proportions of the house out onto the landscape. As Paca-Steele points out, the William Paca House is no exception. The main block of the William Paca house is built to the ratio of 3:4:5. The width of the house is 36 feet, the length is 48 feet, and the hypotenuse is 60 feet. The upper-most terrace on which the house sits is also proportioned as a 3:4:5 ratio with measurements of 99 feet:132 feet:165 feet and with the house sitting in its center. Paca-Steele is further able to divide the property into a grid of 3:4:5 rectangles whose measurements are based on the Gunter’s Chain (49.5 feet by 66 feet). Along these divisions in the garden are the locations of the first fall between terraces as one descends into the garden and the bottom of the second fall. Using these proportions of 3:4:5, Paca-Steele is able to explain the locations of some of the terrace divisions and other features of the garden that have been 140 recovered archaeologically as well as justify some of the interpretive decisions made in the reconstruction of the present garden. However, it is important to remember that the present archaeologically-based reconstruction of the Paca garden is an invention dating to the late 1960s. Likewise, it is important to understand which features are original, which are interpretive decisions based in historical and archaeological research, and which are informed inventions made to fill in where no historic or archaeological data exists. Onto this garden which Paca-Steele has identified the use of the 3:4:5 ratio, we can identify a second proportion at work. It needs to be stressed that the Paca garden as it exists in the present and as studied represents a reproduction based partially on archaeological evidence and care must be exercised before researchers conclude this second proportion existed in the eighteenth century. Incorporated within and alongside the geometry described by Paca-Steele’s 3:4:5 proportions is a use of the classical golden ratio or mean (Figure 5.3). Measuring approximately 1.61:1, the golden mean is a recurring ratio found throughout classical architecture as well as in nature. The supposed harmony within the golden mean comes from its self-referential nature, which is fractal-like. This proportion is achieved when length a is to length b as a+b is to a. Because the ratio references itself to derive its other part, it can be replicated from a single base measurement at either a larger or smaller scale ad infinitum. This makes it useful for extending from the central part of the plantation, the main house, out into the landscape around it. Another good example of an asymmetrical urban house and garden that has been largely unaltered is the Annapolis home of Charles Carroll of Carrollton, one of 141 Paca’s co-signers of the Declaration of Independence. Situated on a triangular lot on Spa Creek, archaeological research has shown the Carroll House garden consists of six terraces divided by five falls. Four walkways run perpendicular to the slope of the garden with the one furthest west aligning with the southern entrance of the house. 142 Like the Paca garden, this garden was built using a key measurement from the house. Four years of archaeological survey conducted between 1987 and 1990 was able to reveal a largely extant, asymmetrical garden that was contemporary with and a peer of Paca’s garden (Leone and Shackel 1990). Plantation Survey The above academic exercise in superimposing geometry onto a formal space begs the question, “Of what use is this knowledge?” In the case of archaeological landscapes, it allows us to make informed predictions regarding the layout of these spaces given limited information that typically comes with archaeological investigation. In many cases, they have used central axes to locate fences and gateways and the rules of symmetry to mirror known elements across yard spaces, predict openings or breaks in fence lines, or locate key intersections or structures that were previously unknown (Yentsch and Kratzer 1994). Likewise, as will be seen below, once armed with these proportions, it may be possible to link extant landscape features to the main house and establish their approximate age. Understanding the geometry embedded within a landscape has a certain practical usefulness in terms of archaeological survey. Furthermore, understanding the rules of landscape design allows for reconstructions based on limited archaeological and historical evidence (Heath 2012:12), like the kinds seen at the Paca House (South 1968), at Monticello (Kelso 1982), at Poplar Forest (Trussell 2012), or in the Upper Flower Garden at George Washington’s Mount Vernon (White 2016). Julie Ernstein (2004) conducted a survey of five surviving formal gardens and landscapes in Prince George’s County, 143 Maryland using traditional topographic mapping, documentary research, and archaeological excavation. She places these gardens within a context of social action and in her conclusion calls for more comparative landscape studies. Expanding this work, the Maryland-National Capital Park and Planning Commission has published Antebellum Plantations in Prince George’s County, Maryland: A Historic Context and Research Guide (2009), a survey and analysis of plantation resources in the county designed to provide an overview and assist researchers working on these sites. To these ends, we can use the new digital mapping tools used by Harmon and others at Wye Hall and Tulip Hill to conduct the kind of investigation described by Paca- Steele at the Annapolis Paca House and Garden. The remainder of this chapter is a survey of ten plantation landscapes in Talbot County, Maryland using GIS and a LiDAR dataset as well as modern and historic aerial imagery alongside several other kinds of historic plantation surveys to map extant physical features and understand the rules by which they were originally planned. These include National Register of Historic Places nomination forms, State of Maryland Inventory of Historic Properties forms, early architectural and archaeological research, and several period maps that depict elements of the plantation landscape as it was on the eve of the American Civil War. These maps include the 1858 William H. Dilworth Map used extensively in the previous chapter as well as an 1847 U.S. Coast Survey map titled, No 18. Map of St. Michaels River, Third Haven Creek, Easton &c. (Figure 5.4), an 1862 U.S. Coast Survey chart titled Chesapeake Bay from tis head to Potomac River: No. 35 From Magothy River to 144 Choptank River (Figure 5.5), and a historic aerial photograph index created in 1937 by the U.S. Agriculture Stabilization and Conservation Service (ASCS) (Figure 5.6). As suggested by Harmon and others, with detailed digital elevation models of existing plantation landscapes generated from a LiDAR dataset, we can begin to piece 145 together the mathematical parameters that govern their design and in doing so, we can begin to understand how they were put together and where we can expect to find archaeological remnants of these historic landscapes. Combining LiDAR survey data with a hGIS framework addresses two problems encountered when working with 146 these plantation landscapes; first, the large size of these pieces of cultural material, and second, access to them. Fortunately, utilizing LiDAR datasets solves both of these issues. 147 The research presented in this chapter was initially undertaken in 2010 and used the State of Maryland’s then-current LiDAR dataset. In the early 2000s, the State of Maryland contracted with Spatial System Associates, Inc. as well as several other companies and commissioned the collection of LiDAR data for most of the state for purposes of floodplain analysis and shoreline erosion studies. The initial data collection flights were conducted by the Sanborn Map Company, Inc. in April and May of 2004. Data processing such as feature and artifact removal was conducted by Computational Consulting Services, LLC. The quality of the collected and processed data was tested by Dewberry & Davis. Post-processing was performed by Spatial Systems Associates, Inc. The area covered by this dataset is bounded by the following coordinates (in latitude and longitude): West: -76.472538; East: -75.297853; North: 39.124828; South: 37.881527. From the initial LiDAR data-collection flights, a number of data products have been produced. These include first and last returns (first returns represent the earliest recorded signal return and generally represent the tops of trees and other vegetation while Last returns represent the last recorded signal return and generally represent the returns that penetrate canopy and vegetation cover), bare-earth mass points (these represent the ground surface after all vegetation and man-made structures like buildings have been removed), a resampled gridded digital elevation model or DEM (this data product is resampled from the bare-earth mass point data and is a cleaner representation of the topography at the expense of some horizontal accuracy), and a return intensity raster (this represents the strength of the returning signal at any given point and is generally indicative of reflectivity of the ground 148 surface). This LiDAR dataset and its derivatives were made available to researchers and analysts for purposes requiring detailed elevation data. This dataset was housed by the Maryland Department of Natural Resources in Annapolis, Maryland and was made available to the public. In the years between the following research was first conducted and this dissertation published, a second, aerial LiDAR data collection survey was conducted over the county. The Federal Emergency Management Agency (FEMA) ordered the creation of a detailed elevation dataset of coastal areas following Hurricane Sandy, which impacted the Caribbean and eastern United States in late October and early November 2012. Because this new dataset was more accurate, it was acquired, and the analyses below were reconducted using FEMA’s Post-Hurricane Sandy LiDAR instead of the State of Maryland’s described above. The subset of the Post-Hurricane Sandy LiDAR dataset that includes Talbot County was collected by Woolpert, Inc. at an elevation of approximately 2,000 meters (~6500 feet) over Talbot County, Maryland by a pair of Leica ALS70 scanners onboard either a Cessna 310 or Cessna 404 aircraft over 17 separate flights between December 7th 2014 and January 2nd, 2015. The average point spacing of this dataset is 0.7 meters (2.3 feet) and the absolute vertical resolution meets or exceeds 15.6 centimeters (6.14 inches) and the relative resolution is better than this standard. The State of Maryland’s current LiDAR dataset, including links to download a variety of LiDAR-derived data products can be found at: https://imap.maryland.gov/Pages/lidar.aspx. Metadata for this LiDAR product can be found online at: 149 https://lidar.geodata.md.gov/metadata/Talbot/2015/MD_PA_Sandy_Lidar_DEM_IM G.xml. Working from the list of plantations in three Maryland counties listed on the National Register of Historic Places compiled by University of Maryland undergraduate Ryan O’Connor and presented as Table 2.1, a sample of ten extant plantations in Talbot County has been chosen from the National Register of Historic Places. This subset is presented as Table 5.1 and mapped as Figure 5.7. Note, Ratcliffe Manor is not listed on the National Register of Historic Places; however, it is remarkably intact, historical documentation is readily available, and has been deemed eligible for inclusion on the Register. It should be noted that this is not a representative sample of plantations, surviving or otherwise. There are many biases that have influenced whether a plantation has survived into the present and whether or not it is a good candidate for this kind of case study in using this survey technique to characterize and map formal plantation landscapes in the Chesapeake. An analysis of Maryland Inventory of Historic Properties ID Plantation Name Construction Date T-52 The Anchorage 1720s T-53 Myrtle Grove 1734 T-54 Wye House 1781-84 T-90 Hope House 1740 T-146 Compton before 1794 1780-90, expanded T-149 The Wilderness 1815 1720-30, expanded T-164 Otwell 1800 T-189 Clay's Hope 1783-98 T-244 Sherwood Manor 2nd half of 1700s not listed on NRHP Ratcliffe Manor 1755 Table 5.1 150 plantation valuations by Aubrey Land (1965) suggests that plantations like these listed in Table 5.1 are certainly in the top percentages of all plantations in terms of wealth, an observation that is supported by the census analysis presented here in Chapter 4. Nevertheless, this sample was chosen because these plantations represent 151 well-documented, surviving, and mostly-intact plantation landscapes, three characteristics required for the below survey. The many similarities between these ten plantations perhaps should not come as a surprise. They are mostly found in the geographic area between the dendritic riverine peninsulas on the eastern half of the county and the upland areas to the west and were used for largely the same purpose on the eve of the American Civil War. Many of the plantations are directly connected through familial relations. Because of their shared historical trajectories, uses, and social relations, it is not surprising that there is a general model to describe these plantations. While these ten plantations differ in scale and execution, the underlying similarities are hard to dismiss. Before proceeding with this survey, it should be noted that these data sources have been chosen to maximize the breadth and coverage of the survey while still maintaining a quick and efficient research design. They have been chosen for being readily-available, publicly-accessible historic and physical documentation of these landscapes. This methodology is designed to provide broad geographic coverage of a large area with enough data to identify and target sites of interest without requiring physical access to these sites. It should be reiterated here before proceeding that embedded within the use of GIS as an analytical toolkit is a worldview or ideology that favors particular understandings of space over others. I repeat here the argument made by J.B. Harley which was presented in Chapter 3 because of the hazards of uncritically accepting GIS as an objected method of studying landscape. Paca- Steele’s and others’ use of a grid or net of squares is a literal imposition of an ideology onto the archaeological record. Likewise, the spatial sources used have their 152 own implicit biases and heavily influence the below analyses. Furthermore, by relying on National Register of Historic Places and Maryland Historic Inventory forms as well as the body of historical documentation behind them, we run the risk of holding up these well-documented sites as typical Maryland plantations and reinforcing the plantation narrative this dissertation seeks to address in the first place. The remainder of this chapter and Chapter 6 relies on these better-documented sites as a series of test cases in integrating historic GIS into historical and archaeological research design. Once shown to be successful, others can incorporate these tools and techniques elsewhere. Clay’s Hope (T-189) Clay’s Hope (Figures 5.8, 5.9, and 5.10) is a late eighteenth-century brick house located above a small neck of land, near Bellevue in southwest Talbot County. It is “a vernacular structure of excellent, although simple, workmanship” (NRHP 1976:3). Named for Henry Clay, who patented the land in 1662 and then sold it two years later in 1664 to the Coulson or Colston family, Clay’s Hope is significant not only for its use of vernacular architecture but also for its continued agricultural history (NRHP 1976:3). The main block of the plantation house is a three-bay-wide, two-and-a-half- story-tall brick structure. Tax records indicate the present structure was begun by James Coulson sometime between 1783 and 1798 (NRHP 1976:3). The main block is of brick construction laid in Flemish bond on the southern, land-facing façade and 153 common bond on the sides, all on top of a below-grade stone foundation. This wing measures 36 feet long by 26 feet wide (NRHP 1976:2). It has a beveled or champhered water table at its base and a three-brick tall belt course between the first and second stories. The moderately-pitched gable roof is pierced by a pair of symmetrical dormer windows and an interior end chimney on each gable end. The floor plan of this original wing includes an entry living room in the northeast corner, a stair hall behind that in the southeast corner, and a large den in the west half. This room was originally divided into two smaller rooms with shared corner fireplaces (NRHP 1976:2). Despite this alteration, much of the original woodwork on this floor, as well as the second floor and the attic, survives, including a dog-leg staircase with feather-edged paneling (MIHP 1991:8). 154 Adjoining the original wing of the house to the west is a slightly lower, three- bay-wide, two-story frame wing dating to the mid-to-late-nineteenth century that probably replaced an earlier 20 foot long by 20 foot wide brick kitchen (MIHP 1991:9). Records indicate it was built by Alexander Harrison shortly before 1860 155 (NRHP 1976:3). The fenestration of this wing has been changed since the mid- nineteenth century with the addition of a bay window and several others on the first floor (MIHP 1991:28). Like the older brick wing, this addition also has a pair of chimneys; the one on the adjoining wall interior and the opposite one exterior. This 156 wing is built with a center hall floorplan with one room on each side and retains much of its original woodwork on the second floor (MIHP 1991:15). Interestingly, the porch has been removed from the northern, water-side of the main block and is used as a gazebo elsewhere on the property (MIHP 1991:16). Immediately to the south of the house along the driveway is an eighteenth-or-early- nineteenth-century tobacco barn, unique in Talbot County. This frame structure is built on a low brick foundation and is fastened with wrought-nails, some of which still hold the original beaded weatherboards, attesting to its early construction. Shed additions have been added during the early twentieth century and the structure has been converted for use as a hay barn (MIHP 1991:17). Several other early twentieth century agricultural structures on the property include a stable, a corn crib, a chicken coop, a carriage house, and a tool shed built to resemble an early smokehouse (MIHP 1991:13). Adjacent to the nineteenth century frame wing to the west is a semi- detached garage (MIHP 1991:13). The National Register listing for Clay’s Hope includes 37.29 acres of farmland surrounding the house, which is notable for being under cultivation for over 300 years (NRHP 1976:3). The location of the main house at Clay’s Hope at the beginning of a small neck of land on Tar Creek means that the plantation core of Clay’s Hope is relatively small at about 11.5 acres. The 1862 U.S. Coastal Survey map shows only approximately 3 acres not in cultivation at the core of Clay’s Hope. The placement of the main house at the top of a small rise at the beginning of this small neck of land essentially cuts off access to this neck from the mainland, establishing a form of control of it by the house. In the 1862 Coastal Survey, an 157 orchard appears at the end of this peninsula, occupying a removed and protected space on this landscape. At present, at the end of this neck of land at the point and extending into Tar Creek, is a pier by which the plantation could communicate and travel with the larger Chesapeake and Atlantic. According to the depths indicated in the Coastal Survey, historically the water was very shallow here, perhaps less than three and a half feet. Placing the landing as far out into Tar Creek as possible would help to maintain access to deep water shipping. Neither rigid adherence to symmetry nor right angles are found in the landscape at Clay’s Hope. The orientation of the core and the most central components of landscape are largely defined by the orientation of the main house; however, the other agricultural outbuildings like the tobacco barn are oriented independently from the main house. Likewise, the current driveway from the main road to the house at Clay’s Hope is not straight and comes to an angle near the house. The tobacco barn and other outbuildings along the west side of the driveway and the cove on the east side of the driveway form a pair of diverging lines with the house as the focal point, a common arrangement creating an optical illusion designed to bring the object of focus closer to the viewer. Despite these multiple orientations in the landscape, imposing a grid of squares over the landscape reveals several additional alignments. The original block of the house measures 36 feet long by 26 feet wide. The house is situated 780 feet from the edge of the main road by the plantation. This distance is 30 times the width of the house. Likewise, the house is situated 1,440 feet from the edge of the water. This distance is 40 times the length of the house. Together, this creates a 3-4-5 right 158 triangle 30 boxes (or sometimes they are called squares, despite not being equal in length and width) wide by 40 boxes (or squares) long. In 1967, Henry Chandlee Forman writes that, “the box garden, of two concentric circles, has entirely disappeared” (Forman 1967:134). It is possible that the remains of this box garden are just barely visible in the LiDAR-elevation data just to the northwest of the house; however, without prompting by Forman, it is extremely difficult to make out this feature. Compton (T-146) Compton (Figures 5.11, 5.12, and 5.13) is located on Grubbin Neck on LaTrappe Creek, two miles southwest of Trappe in Talbot County (NRHP 1973a:2). The current dwelling is comprised of two main parts, an original early eighteenth- century, one-and-a-half-story block with a larger, early nineteenth-century two-and-a- half-story addition (Weeks 1984:184). Historically, Compton is notable for being the home of Samuel Stevens, the eighteenth Governor of the State of Maryland (NRHP 1973a:4). Stevens inherited Compton from his father in 1794 and was responsible for adding the one-and-a-half-story east wing after 1798 and raising the west wing to its present height of two and a half stories (NRHP 1973a:4). Architecturally, the house is notable for being an example of early eighteenth-century vernacular architecture that has been incorporated into major Federal period additions and renovations (NRHP 1973a:4). The original house at Compton probably dates to the second quarter of the eighteenth century and has since been significantly altered in several phases (NRHP 159 1973a:2-3). The main part of the current house dates to the eighteenth century. It rests on an English basement with arched louvered windows and a quarter-round molded water table. The southwest side of the house faces the water and is five bays wide with a three-brick belt course between the first and second floors. The gable roof is pierced by three dormers on each side and a pair of interior end chimneys, one on gable end. Additionally, on the gable ends are the remains of a two-brick belt course slightly lower than on the main façade. The eastern wing contains a kitchen and dining room and was built sometime after 1798 (NRHP 1973a:2). The basement does not continue under the one-and-a- 160 half-story wing. It is flush with the north east land-facing façade and is set back about twelve feet from the southwest water-facing façade (NRHP 1973a:2). The fenestration is on this wing is irregular and there are a pair of asymmetrically placed chimneys. 161 Much of the interior woodwork is consistent throughout the house and is Federal in style, consistent with the 1798 Stevens’ renovations; however, there are several decorative and architectural elements that predate these late eighteenth- century modifications. These include a pair of doors, a window, arches in the cellar, 162 remnant gable end belt course, and segmental arches on the basement windows. It is believed this earlier version of Compton had an “L” shaped plan (NRHP 1973a:2). To the east of the main house is a surviving dairy or milkhouse and to the north is an allée of mature maple and cedar trees (NRHP 1973a:2-3). The dairy is a two-story brick structure with the first story located behind an underground moat-like retaining wall (NRHP 1973a:4). One of the limitations of the methodology proposed here for rapidly surveying plantation landscapes remotely is that it is largely dependent on the availability of key information about the property. This includes date, architectural features, as well as measurements of the structure. Ideally, when laying out a grid of squares over one of these plantations in GIS, the measurements of the main house should ideally come from an actual measurement taken off the structure by hand, either by the individual who prepared the National Register of Historic Places nomination form (or another similar document) or as reported by one of these records from early tax records. In some cases, this piece of information is missing from this documentation or historic sources. In these cases, the dimensions of the main house can be estimated by measuring the outline of the structure as seen in aerial photography or directly from the LiDAR returns. In the case here at Compton, the dimensions of the structure do not seem to be recorded in its historical documentation. From the aerial imagery and LiDAR dataset, it is estimated that the main block of Compton measures approximately 40 feet long by 30 feet wide, another 3-4-5 ratio. One of the most prevalent landscape features at Compton is the long allée running north, away from the house. This strong, tree-lined central axis measures 163 approximately 120 feet wide, or 3 times the width of the main block of the house. It runs north to a small inlet that cuts into the landform approximately 1,200 feet away, which is 40 times the 30-foot-wide house. On the south side of the house, the lines of the allée continue and define a 120-foot-by-120-foot-square between the house and the water. So being equal in length and width, this area is 3 grid squares wide by 4 grid squares long, another 3-4-5- ratio. To the west of the house, 1,200 feet or 30 times the 40 feet long house is the end of the neck of land on which Compton is built. Further south across the narrow inlet, 1,200 feet from the house or 40 grid squares is the edge of the next landform. These measurements would seem to indicate that Compton was sited and scaled with these specific alignments in mind. In 1862, it is shown that the driveway to Compton cut through a large woodlot. This woodlot still exists, and the remnants of this driveway are visible in the LiDAR dataset as a faint trace through the woods. Many of the field divisions shown in the 1862 map are also recognizable in the present. Hope House (T-90) Hope House (Figures 5.14, 5.15, and 5.16), is a seven part dwelling located on a peninsula between two branches of Woodland Creek at the confluence of the Wye River and the Miles River where they flow into Eastern Bay, near Tunis Mills, northwest of Easton, Maryland. Hope House is located just south of Wye House and their histories are deeply interconnected. A 100-acre tract of land called “Hope” was patented in 1665 by Henry Hawkins, a close relative (brother-in-law or stepson) of Edward Lloyd I of Wye House (NRHP 1977:3). It was purchased by James Lloyd, 164 Edward Lloyd II’s younger brother in 1720 (NRHP 1977:3). Hope House is historically significant for its connection to the Tilghman and Lloyd families as well as to mid-twentieth-century artist Ruth Starr Rose (NRHP 1977:4; Paca et al. 2015:14). Architecturally, Hope House is similar to Baltimore’s Mount Clare in many respects and is notable for its Federal period central block and extensive Colonial Revival period renovations and additions to the building and landscape starting in 1907 (Weeks 1984:169-170). The structure as it currently stands is notable because it represents an early twentieth-century reinterpretation and adaptation of an original plantation house and its surrounding landscape. The first record of the current Hope House dates to 1808 when the property was inherited by Tench and Margaret Tilghman, although tradition holds it was built 165 in 1740 (MIHP 1977:10; NRHP 1977:6). The central block of the structure is said to have been built in 1740; however, it does not appear in the tax assessments of 1783 nor 1798 (NRHP 1977:6). This core of this Flemish bond brick structure is three bays wide and two-and-a-half stories tall. The north-facing land side of the house is 166 centered by a two-story pavilion with a twentieth-century Doric porch on the raised first story, a three-part Palladian window on the second story, and an elliptical window at the apex. This entrance has a single wide door flanked by fluted pilasters and sidelights. The windows are nine-over-six sash with stone sills and louvered 167 shutters. This block is lacking a water table or belt courses. The gable roof was raised in the early twentieth century and is pierced by a pair of dormers which appear to match those seen in nineteenth-century photographs of the house (NRHP 1977:2; Weeks 1984:83). There is an interior end chimney at each gable end of this main block. The pavilion is repeated on the south-facing water side of the house where it is less pronounced because of the flat-roofed columned porch that runs the width of the first story and the pair of nine-over-six sash windows that crowd both stories of the pavilion in addition to the pair that flank it on the main façade as on the other side. At the apex of the pavilion is a lunette. There is no door on this side of the main block; rather, two sets of jib doors below the two central windows on the first story open onto the porch. The pair of dormers on the north side is repeated on the south side. The floorplan of this central block consists of a stair hall across the entire north side of the block that opens into two parlors along the south side of the block. The woodwork in this part of the house dates to its original construction in the early nineteenth century (NRHP 1977:2). Each of the hyphens and wings consist of one room only each on the first floor. The flanking wings and hyphens of Hope House were added after the house was bought in 1907 by a wealthy lumberman named William J. Starr, whose wife is said to have complained that he “bought little more than a staircase” (NRHP 1977:6). These additions, executed in the Colonial Revival style by reportedly the best available local craftsmen (NRHP 1977:3), make Hope House a unique blend of early nineteenth-century plantation architecture and early twentieth-century Colonial 168 Revival changes. A pair of nineteenth-century photographs shows the house prior to these renovations. One shows the house with two one-story ogee hyphens connected to a pair of one-and-a-half-story, gable-roofed frame wings and the other shows just one of these ogee hyphens bricked up to a flat roof with the original ogee still visible in the brick work and with no adjoining wings (NRHP 1977:3). These were replaced in the 1907 renovations by a pair of unique ogee hyphens that start two stories but decrease to one and a half stories, clearly inspired by the ghost roofline seen in the earlier photograph. These are three bays wide with tall arched windows on the first floor and custom windows that follow the ogee curve on the second. They attach to a pair of two-and-a-half-story, forward-facing, gable-roofed wings. The wings are two bays wide and three deep with a single chimney each toward the outer side of the structure. Lastly, a pair of gable-roofed one-and-a-half-story wings with the gable clipped on the exterior side of the structure. In addition to the major changes to the house, the Starr renovations extended to the surrounding landscape to include the garden between the house and water. These include the porches on both sides of the house, plantings, walls, allées, formal gardens, and a new driveway designed by Ms. Starr (NRHP 1977:6). The current National Register listing contains approximately 76 acres of the original plantation and includes the family cemetery and a pre-twentieth-century brick icehouse. The National Register Nomination forms do not list the dimensions of Hope House and the aerial imagery is partially obscured by vegetation. The best visual approximation of the main block of the structure is 40 feet by 30 feet. 169 Because of the extensive modifications made here after 1907, any attempt to understand Hope House or its surrounding landscape needs to take into account the 1907 renovation and expansion of the house and gardens and should serve as a cautionary tale to the landscape researcher to always consider the temporal origins of the features being examined, as the majority of the current landscape at Hope House seems to date to the early twentieth century, not the eighteenth or nineteenth centuries. The current plantation driveway that stretches north from the house and loops around the landform to the east is one of these twentieth-century inventions. The mid- nineteenth-century maps of Hope House show that the land approach to the house is a straight road from the east. When approaching the house by car today, this is the road by which you approach the house. However, standing on the steps of Hope House looking down the long, straight, tree-lined allée, the intent of this newer driveway is unmistakable. Aside from its absence on the historic maps, the scale at which this feature is executed is wrong for the Colonial or Federal periods. My hypothesis is that this is a twentieth-century invention built to mimic an older, historic driveway. Indeed, much of the twentieth-century redesign of Hope House can be described as grandiose in scale, a characteristic which distinguishes it from the older, un- remodeled landscapes described in this chapter. In addition to extensive renovations to the house and driveway after 1907, the garden on the south side of the house was also extensively remodeled by Mrs. Starr after she and her husband purchased the house. These gardens are well documented and include extensive terracing. 170 Looking outward from the main core of the plantation, the orientation of the different components of the landscape are notable in that there are two different orientations on the landscape. In the vicinity of the main house, the orientation of the gardens, driveways, fields, and other structures take their orientation from the main house. Further out, the fields take their orientation from the main road into the plantation, except for one field to the north of the main house, which is on the house’s orientation and no that of the road. The existence of two slightly different orientations on the plantation as well as the manner in which the driveway approaches the main house from the side means that there is not a grand approach to Hope House the way there is with the other plantations in this chapter. In the 1862 Coastal Survey, the plantation core at Hope House is depicted as very dense. In addition to the agricultural fields near the house, there are also several orchards or garden and at least six other structures located within one quarter mile of the house. The LiDAR elevation data shows extensive alterations and disturbance on the peninsula to the west of Hope House and any original garden or landscape features here seem to be masked by later, twentieth-century landscape changes. Myrtle Grove (T-53) Myrtle Grove (Figures 5.17, 5.18, and 5.19) is located on Goldsborough Neck, a point of land between Goldsborough Creek and the Miles River, just northwest of Easton, Maryland, and has been the seat of the Goldsborough family and their descendants from the time it was built in 1734 until 1989 when it was bought by its current owners (MIHP 1992:3,23-25). Nicholas Goldsborough came to Maryland 171 with his wife in 1669 and settled on Kent Island (NRHP 1973b:3). He died shortly thereafter; however, his sons Robert and Nicholas both moved to Talbot County after 1678. Nicholas settled near Oxford and Robert settled on the Miles River and established an estate he called Ashby (NRHP 1973b:3). Robert’s son, Robert Goldsborough II, built Myrtle Grove on a part of Ashby in 1734 (NRHP 1973b:3). Like Compton, Myrtle Grove is an amalgamation of several distinct phases of construction. The current house is a three-part structure with its origins in the early eighteenth century. The first of these wings dates to 1734. This early frame section of the house is 40 feet by 26 feet and built on an English bond brick foundation on top of 172 a stone foundation and full basement (MIHP 1992:12). It is five bays wide, one-and- a-half stories tall and built in the post-medieval style with a steeply pitched roof. The entrances to this wing are in the center bay through a wide, ten-panel door. There are three dormers piercing the gable roof, one over the center entrance and the other two 173 between the first-floor windows, in addition to a massive chimney on the northwest end of this wing. On the northeast-facing land side of the wing is a one-story porch with a slightly less steep roofline, supported on square posts that spans its entire 174 length. The floor plan of this original section of the house includes an ell-shaped hall, dining room, sitting room, and stairway. Attached to this wing to east is a Federal, four-bay-wide, two-story-tall brick wing measuring 44 feet by 36 feet built in 1790 (MIHP 1992:12,15). This wing is laid in Flemish bond on a raised, full basement with a molded brick water table on the land-facing northeast and water-facing southwest facades. The entrances are in the northwestern most bay on both of these facades and consist of a door and large surround. To the southeast on the first floor are two additional windows. Under the two windows, at ground level, are louvered shuttered windows in the basement. On the second floor are three windows, centered over the openings on the first floor. On the southeast façade of the house are six windows. Two on the corners of the first floor, two on the second floor centered above the pair on the first floor, and a fifth also on the second floor centered between the two in the corners. All of the sash windows in this wing are capped by a rubbed brick jack arch. A sixth window is an elaborate elliptical window in in the triangular pediment formed by the low-pitched gable roof. A second elliptical window can be found in the pediment on the opposite gable. This is the only window on this façade as most of it is attached to the older wing of the house; however, there is a door leading to the basement on the northeast side of this face. Two pairs of interior end chimneys rise through the house and pierce the roofline near its apex. On the first floor of this wing, a hall connects the two entrances and runs the full depth of the house. Off this hall are two adjoining parlors. The second floor contains a pair of bedrooms as well as two bathrooms, which were added later. 175 Attached to the original wing to the west is a low, Colonial Revival frame kitchen wing built in 1927 (NRHP 1973b:5). Covered in beaded clapboard, it is one- and-a-half stories tall and two bays wide with dormers over the windows. This wing replaces an earlier wood kitchen attached to the main house by a covered way. A porch that is detachable and non-permanent was added off the southwest wall of this kitchen wing to avoid permanently altering any more of the structure. In addition to the main house, there are several surviving outbuildings at Myrtle Grove, including an 1805 pyramidal-roof brick dairy (MIHP 1992:21), a partial-brick and partial-frame smokehouse (NRHP 1973b:6), and a detached 1790 law office (NRHP 1973b:6) which has recently been converted into a cottage. The approach and driveway to Myrtle Grove retains a parklike setting with many old trees and garden features are still discernable on the northeast and southeast sides of the house. The LiDAR returns from Myrtle Grove are notable because they capture two significant changes to this landscape that occurred between the original collection of the 2004 LiDAR returns and the more recent 2015 returns. In the time between when these two scans were conducted, six rectangular garden beds on axis to the west of the main house and whose origins are at least as old as the early twentieth century and are probably much older were removed and a large, man-made pond was installed on axis to the south of the main house. These are two significant, intensive modifications to the landscape recorded in successive scans of the landscape just over a decade apart. Here at Myrtle Grove, even more so than at Compton, things are complicated by the multiple, overlapping elements of this landscape. If there is a grid of squares to 176 be found here, the first issue must be to determine which the landscape is built upon. The allée that runs northeast from the house is centered on the older 1734 block, which indicates that it was probably built prior to the construction of the later 1790 wing, which upset the balance of symmetry in front of the house. When it was built on the opposite side of the 1734 block in the 1930s, the newer kitchen wing returned balance to Myrtle Grove in respect to the central allée. The early 1734 block of Myrtle Grove measures 40 feet by 26 feet. This ratio of length to width is approximately 1.54:1, very close the 1.61:1 ratio of the classical golden mean. The width of the allée running northeast from the main house measures 120 feet at the house, or approximately three squares. At the far end of this allée, it has narrowed to only 40 feet, or one square. At this point, the driveway crosses a drainage ditch and bends toward the north and loses sight of the main house at Myrtle Grove. In addition to the main allée to the northeast of the house, the allée to the southeast of the house is also built using this key measurement from the earlier block of the house. Unlike the grand allée, the sides of this allée are parallel with each other and on grid with the main house. It is defined by a strong terrace on the northeast side and lined with trees in a park-like setting. The Goldsboroughs, like the Lloyds, owned vast tracks of land throughout Talbot County (see Chapter 4 and the 1858 William Dilworth map). Like many of the other plantations described in this chapter, the original tract of land on which Myrtle Grove stands has since been subdivided and developed. Given the footprints of the large homes and depths of the swimming pools many of them possess, the situation is 177 not ideal for the preservation of some of the outlying landscape features. In some cases, it is relatively clear which boundaries are oriented with Myrtle Grove and probably predate development. In others, the relationship between original landscape feature and development is less clear. Otwell (T-164) Otwell (Figures 5.20,. 5.21, and 5.22) is an early eighteenth-century, T-shaped house with three telescoping nineteenth-century wings located at the end of a neck of land between Goldsborough Creek and Trippe Creek near Oxford, Maryland. It is architecturally significant for its early T-shaped plan and string of nineteenth-century 178 additions built at the same time to replicate the appearance other plantation structures comprised of more organic growth (NRHP 1980:3). Originally surveyed in 1659, Otwell passed into the Goldsborough family by marriage between Sarah Turbutt and Nicholas Goldsborough, III around 1720 (NRHP 179 1980:3). Construction of the earliest part of Otwell began sometime between 1720 and 1730. It is a two-story-tall, gambrel-roofed, T-shaped dwelling, although architectural evidence suggests it may have originally been built as a one-and-a-half story, gable-roofed structure (NRHP 1980:3,5). The crossbar of the T is oriented 180 north/south, is three bays wide and two bays deep, and on the first floor is a pair of parlors (NRHP 1980:5). On the second floor are a pair of irregularly placed dormers that piece the steeply pitched portion of the gambrel roof and end just below the flatter portion. A large interior end chimney pierces the center of the roofline on each gable end of this wing. The wing that comprises the stem of the T intersects at the center of the crossbar wing and is approximately the same length east/west as the crossbar is north/south. Fenestration in the stem of the T is irregular but consists of the same twelve-over-twelve hung sash on the first floor and six-over-six hung sash in the dormers on the second floor. The entrance to the earliest portion of Otwell is through a stair hall that runs through the width of the wing. The two parlors in the crossbar of the T are accessible to the west through this hall (MIHP 1976:9). To the east in the remainder of this early T-shaped block is a dining room, also accessible through the stair hall. Running the full length of the stem and the next wing as well along the southern water-facing side is a modern enclosed porch (MIHP 1976:9). Attached in a line extending to the east from the base of the T are three additional blocks, each with varying rooflines. These telescoping wings date to the first decade of the nineteenth century and appear to have been constructed at roughly the same time (NRHP 1980:3). From the east gable end of the earlier T wing and moving further to the east, these include a two-story, one-bay-wide, gambrel-roofed stair hall, a two-story, three-bay-wide, gable-roofed kitchen with an entrance on the north side of the wing, and a two-bay-wide, one-and-a-half-story-tall, gable-roofed bedroom wing. The latter two wings each have an interior end chimney piercing the center of their roofline on their east sides. 181 A renovation in 1946 and then a fire in 1958 destroyed most of the original interior of the structure; however, the current floorplan remains largely the same (NRHP 1980:6). The post-fire renovations represent an attempt at authentically reproducing the original interior (MIHP 1976:13). Otwell is listed on the National Register of Historic Places along with 37 acres of its surrounding landscape. On this parcel are surviving boxwood and hedges east and northeast of the house on the land side, a brick wall, and the drive as well as active agricultural land. The shape, size, and orientation of Otwell pose a challenge for applying the same kind of grid of squares analysis used for the other plantations in this chapter. The original 1720s T-shape plan oriented 90 degrees to the land approach to the house throws out the rules of symmetry that are supposed to have guided the design and layout of these formal plantation houses and their core landscapes. The measurement of the driveway to the north of the house, terracing to the east and south suggest that 42 feet is a reoccurring measurement on this landscape. Neither the National Register form nor any other source define the dimensions of the irregularly- shaped main house at Otwell, however, measurements from aerial imagery suggest that the top of the T is approximately 42 feet long. Otwell seems is situated to take advantage of a V-shaped indentation in the coastline. The early date for the construction of Otwell may be responsible for the lack of landscape modification and impact here. Like other early houses, here in Talbot County, Otwell does not seem to have been built with the same kinds of rigid, formal symmetry that extend out into the surrounding landscape. 182 An interesting feature at Otwell identified by this landscape survey and of interest to follow up with an on the ground survey is a circular mound with sunken center, southeast of house on the water. It appears to be very similar to a feature identified at The Wilderness, both of which are reminiscent of the remains of a known ice house at Myrtle Grove. Ratcliffe Manor (T-42) Ratcliffe Manor (Figures 5.23, 5.24, and 5.25) is situated on a point of land at the confluence of the Tred Avon River and Dixon Creek, just a mile west of Easton, Maryland. Unlike the other properties in this chapter, Ratcliffe Manor is not on the National Register of Historic Places, although it has been deemed eligible for 183 inclusion on the Register and certainly does not lack integrity or significance (MIHP 2003:2). Work began on this two-and-a-half-story, brick house by Henry Hollyday in 1755, was completed by 1762 and stayed in the Hollyday family until 1902 (MIHP 184 2003:62-63). The main block of the house is a Flemish bond, five bays wide with a distinctive clipped-gable or jerkinhead roof. It is built on a raised English bond brick foundation with a beveled watertable at the base of the first floor and a five-course molded belt course between the first and second floors. The northeast land-facing and 185 southwest water-facing facades are almost identical to each other and are each five bays wide with a raised entrance on the first floor in the center bay and two twelve- over-twelve hung sash windows on either side. Above each of these openings are rubbed brick jack arches. On the second floor are five slightly smaller twelve-over- eight hung sash windows directly above the openings on the first floor. Directly above each of these openings is a modillion block cornice that extends up to the eave line. Piercing the jerkinhead roof are three dormers with broken triangular pediments and six-over-six hung sash windows with a semicircular light at the top. One is centered above the entrance and the other two are placed between the outer two windows on the lower floors. Two sets of paired interior end chimneys pierce the northwest and southeast facing clipped gable roof faces. The most visible difference between the two facades is the presence of a forward-facing gabled roof portico over the northeast entrance of the house that may date to the early nineteenth century; however, the outline of a matching porch can be seen in the brickwork on the southwest façade. On this side of the house, the belt course is unmolded and only three bricks instead of five. The southeast gable façade is three bays wide with a side entrance to the study in the northeast corner and a six over six sash window with a bulkhead entrance to the cellar below it in the southeast corner. The northwest gable façade is partially obscured by the one-and-a-half-story, gable-roofed kitchen wing. This wing is three bays wide with an entrance placed closest to the main block of the house and two windows equally spaced to the northwest. Above and piercing the gabled roof is a pair of dormers with six-over-six hung sash windows. These dormers are joined by a 186 shed addition on the southwest side of the wing. A large interior end chimney pierces the apex of the gable roof on the northwest façade of this wing. A large, enclosed porch was added on the southwest, water-facing front of this wing in the second quarter of the twentieth century around the same time the interior of this wing was remodeled and reoriented (MIHP 2003:67). The floorplan of the first floor includes a central stair entrance hall which opens onto a study and a library along the landside of the house and leads to a parlor and a dining room which run the entirety of the waterside of the house. Approximately 75% of the woodwork in the house survives from the Georgian period, including the original mid-eighteenth-century stair in the hall, arched doorways, chair rails, baseboards, and wainscoting (MIHP 2003:70). Particularly of note is the woodwork in the parlor, located in the southeast corner of the house, and represents “the best in Georgian style craftsmanship to survive in Talbot County” (MIHP 2003:68). The living room in the southwest corner of the house was refinished in the early nineteenth century and contains Federal style woodwork (MIHP 2003:68). The floorplan of the second floor of the main block is like the first in that a central hall opens onto four rooms, originally bedrooms (MIHP 2003:68). The attic was finished sometime around 1800 and contains three bedrooms off a central hallway (MIHP 2003:68-69). The dormers that provide light to this space were probably added around the same time as they possess several Federal period stylistic elements like semicircular lights above the six-over-six hung sash, broken triangular pediments, and fluted pilasters surrounding the window frames. 187 The Maryland Historical Trust Inventory of Ratcliffe Manor has been amended since it was originally prepared in 1976 to include a collection of twelve early twentieth-century agricultural and dairy buildings. These include a tenant house, barn, silo, animal houses, a smokehouse, and sheds, generally located northeast of the main house along the entry driveway (MIHP 2003:1). Like both the exterior and the interior of the house, the surviving landscape at Ratcliffe Manor is especially notable. Originally, the Hollydays possessed over 1,000 acres of land at Ratcliffe Manor (MIHP 2003:63). Between the house and the water there is a large, terraced garden comprised of three descending terraces to produce a four-level garden, a lunette at the top of the garden, the remnants of a boxwood alley, and earthen ramps between these levels. In addition, there is a smokehouse, an icehouse, and two other surviving outbuildings to the northwest of the main house. The garden located to the south of Ratcliffe Manor is a remarkable example of a descending terrace garden, similar to the one at the William Paca House in Annapolis and the one at Tulip Hill in Anne Arundel County. The terraces and falls of this garden show up remarkably well on both LiDAR datasets, as does the lunette at the top of the first fall, and the ramps connecting each terrace. In addition, there is a rectangular garden feature adjacent to the steps of the garden on the northwest side that is relatively flat and runs its full length and there is a kind of ramp that leaves the garden to the southeast from the second terrace from the top and includes a path to the pier or wharf. There are additional topographic features that may be garden terraces, beds, or pathways along the southeast coast here. The siting of Ratcliffe Manor here 188 on this peninsula is significant in that ship traffic to Easton by way of the Tred Avon River must pass below this house and garden. One tradition of plantation garden design evident here is that of the transition from formal to wilderness in the garden. Philip Miller (1735) describes the ways to plan and build wildernesses in a garden. Of wildernesses, he writes, “if rightly situated, artfully contrived, and judiciously planted, [they] are the greatest Ornaments to a Fine Garden” (Miller 1735 Vol.2:513) and, “there is nothing so agreeable to terminate the Prospect, as a beautiful Scene of the various Kinds of Trees judiciously planted” (Miller 1735 Vol. 2:514). The main house at Ratcliffe Manor and its supporting outbuildings on top of the hill together with the straight, rigid lines of the descending terraces garden form the formal landscape here. As one descends the garden and reaches the bottom, the object of view is the peninsula or “island” at the end of the neck of land. Accessible only by walking through the garden, this area is offset from the rest of the plantation and stands in contrast to the rigidity of the garden, and as described by Miller, the sightlines of the garden terminate at this wooded peninsula beyond the garden. From the tax assessment of 1798, the main block of the house at Ratcliffe Manor measures 44 feet long by 34 feet wide (MIHP 2003:63). Extending this grid out from the house and into the landscape is a productive exercise and is able to explain the placement and dimensions of this unique descending terrace garden here. Each of the top three terraces measure 220 feet by 102 feet (5 squares by 3 squares) and the bottom terrace measures 220 feet by 68 feet (5 squares by 2 squares). 189 One way to measure the impact a plantation has on its surrounding landscape is to look at how far its grid or orientation extends across that landscape. At Ratcliffe Manor, this grid extends up the peninsula all the way to the main road. The grid extends to either side of the driveway as well onto the adjacent peninsula to the northwest and represents more than 1,000 acres of coverage. While the axis defined by the house runs up the driveway to the main road, the driveway is not perfectly straight. At one point approximately two-thirds of the way up the driveway, the driveway bends to the east to avoid a small drainage gulley. Here, the LiDAR survey indicates the presence of a surviving portion of the straight driveway between the extant driveway and the centerline of the main axis. Sherwood Manor (T-244) Sherwood Manor (Figures 5.26, 5.27, and 5.28), not to be confused with nearby Sherwood Forrest (T-334) another Talbot plantation, is a two-story, five-bay- wide, brick house located four miles north of St. Michaels, Maryland. Located on a narrow point of land on Hemmersley Creek, this structure dates to the late eighteenth century, sometime after the American Revolution (NRHP 1975:2). The primary north and south facades both face water as does the eastern gable end of the structure. It is distinctive for its early use of brick panels inset into the brickwork on the west, land- facing gable façade of the structure, reminiscent of bricked-in windows (NRHP 1975:2). Sherwood Manor sits on a tract of land originally called “Sherwood’s Neck”, patented by Philip Sherwood from Lord Baltimore in 1713 (NRHP 1975:3). It was 190 purchased in 1771 by Matthew Tilghman, future member of the First Continental Congress, Speaker of the Maryland House of Delegates, and the largest landholder in Talbot County at the time (NRHP 1975:3,7). At his death in 1790, he left it to his wife Ann, to be given to his youngest son Lloyd Tilghman upon her death, which occurred in 1794 (NRHP 1975:7). In 1851, the house was purchased by John Covey (NRHP 1975:9). The present brick structure first appears in the tax assessment of 1798 (NRHP 1975:8). The five-bay wide south facades of the house is arranged symmetrically with an entrance in the center bay and a pair of twelve-over-twelve hung sash windows spaced evenly to either side on the first floor and eight-over-twelve hung sash windows over each of these openings on the second floor. Each of these windows is 191 topped by a brick jack arch. The house is set on a raised brick foundation at the base of which are short windows with vertical bars over them and segmental brick arches above. Separating the basement and the first floor is a water table and separating the first floor and the second floor is a four-brick belt course. Above the watertable, the 192 south facade and west gable end are laid in Flemish bond and below the watertable they are laid in English bond. The fenestration of the north façade is irregular. This entrance is slightly offset to the east of the south entrance on the other side of the central stair hall in 193 order to account for the stair that runs along the western half of the entrance hall. A later addition, a corridor enters into this central passage from an entrance on the east side of the structure. A single twelve-over-twelve hung sash window mirrors the outer or eastern-most window on the south facade. A pair of twelve-over-twelve hung sash windows on the west end of this façade mirror their south façade counterparts across a large living room that runs the full depth of the block. Fenestration on the second floor is repeated on the second floor except the center window in the center stair hall is offset to the west over the stair landing instead of to the east over the entrance. A pair of interior end chimneys pierce the roof at its apex, one on each gable end of the house. Flanking each chimney stack as they pass through the attic are a pair of small, four-pane casement windows that provide the only sunlight into the attic. Evidence of former additions can be seen in the brickwork of the east and west gable ends of the structure. To the east of the main house is a stuccoed brick smokehouse (NRHP 1975:16). Like Clay’s Hope, the landscape at Sherwood Manor is small, modest, and largely contained on a small neck of land which juts out from a larger neck of land. The main block of the house is 50 feet long by 22 feet wide (NRHP 1975:8) and the grid defined by its orientation extends the length of this larger neck of land. Sherwood Manor’s driveway intersects the main road up this neck 1,100 feet grid west from the house. 1,100 feet is 22 times the length of the house, which is 55 feet. 1,100 feet grid east of the house is the tip of the next neck of land. Around the main house, there is very little evidence of a surviving garden or other landscape modifications. The house appears to be built on a very slight rise and is situated to 194 take advantage of a rectangular cut in the hillside down to the water that may or may not be a natural feature. The 1862 U.S. Coastal Survey map indicates the presence of several structures along the driveway to Sherwood Manor between the house and the end of the main road. This small work area is surrounded by several plots that appear to be orchards and agricultural fields. Further inland was a woodlot, some of which still exists in the present. There is a small landing on the south side of the small neck of land on which Sherwood Manor is situated. Here, the Coastal Survey indicates a water depth of 11 feet. On the north side of the neck, the water depth is only two and three quarters feet. The Anchorage (T-52) The Anchorage (Figures 5.29, 5.30, and 5.31) is located on the Miles River approximately three miles northwest of Easton, Maryland and is not to be confused with The Anchorage (T-316), a nineteenth-century plantation house located on the Choptank River near Trappe, Maryland and also in Talbot County. Across the Miles River from The Anchorage is the Rest (T-339), home of Franklin Buchanan, an Admiral in the Confederate Navy who is buried in the cemetery at Wye House (Weeks 1984:231-232), and just up the river from The Anchorage is Myrtle Grove. The Anchorage is a five-part brick and frame house comprised of a center block, and a pair of wings connected to the center block by hyphens. The Anchorage was bought by the Lloyd family of Wye House in 1831 (NRHP 1973c:3) and the southeast façade is not unlike Wye House. 195 The land on which The Anchorage stands on was originally patented as Daniels Rest in 1677 by Daniel Walker, who operated a ferry over the Miles River (NRHP 1973c:3). The earliest parts of the present structure were built in 1810 and perhaps incorporated an earlier 1720s brick building into its frame construction (NRHP 1973c:2,6). The main block is seven bays wide and two-and-a-half stories tall and is dominated by a projecting tower fronted by a two-story-tall Greek Revival portico supported by four Doric columns which covers the center three bays of the southeast or water-facing façade and was probably built around 1830 along with the similarly styled wings (NRHP 1973c:2). The entrance is located in the center of the main wing and is comprised of double doors surrounded by sidelights and a large rectangular fanlight above. Still under the forward-facing gable of the porch, to either 196 side of the entrance is a single six-over-six hung sash window with hinged jib doors at the bottom. On the second floor, centered above the entrance is a six-over-six hung sash window flanked by a pair of two-over-two vertical lights. There are two additional bays on either side of the central portico, each with a window; six-over- 197 nine sash on the first floor and four-over-four sash on the second floor. A pair of dormers pierces the southeast facing gable roof between the two outer bays on the lower floors and each dormer has a six-over-six hung sash window topped by a semicircular light, a broken pediment, and pilasters. 198 On either side of the protruding center frame section of the house are one- story, enclosed, flat-roofed porches. Here, the brickwork is painted white while elsewhere it is finished with stucco (NRHP 1973c:2). A single interior end chimney is located on either gable end of the main block. The hyphens and wings were added around 1830 at the same time as the central portico (NRHP 1973c:2). Built in the same Greek Revival style, the one-story, flat-roofed hyphen is one bay wide and the forward-facing gable-roofed wings are two bays wide. Each wing contains an additional interior end chimney at their rear and a pair of six-over-six hung sash windows. The northeast side of the house has two additional six-over-six hung sash windows while the southwest side has a Palladian window. Behind the main block of the house to the northwest is an additional two- story wing. Attached to the northeast of this wing is a modern kitchen and to the southwest an additional porch (NRHP 1973c:2). To the west of the structure is a log smokehouse, windmill, and slave quarters (NRHP 1973c:5,7). The Anchorage is built around an earlier, much smaller structure. It was not expanded until it was bought by the Lloyd family in the 1830s. Because the house started as a much more modest dwelling, it is not surprising that the landscape at The Anchorage lacks some of the expansive and rigid definition that other of the plantations in this chapter possess. Examining the Post-Sandy LiDAR dataset, there do appear to be some cultural topographic relief. Between the house and the water to the southeast are some subtle terrace features defined by a line of tree plantings. To the southwest is a garden bounded and separated by slight topographic relief. The house is not situated on the highest point of ground, which is located to the northwest, 199 but it is situated at the top of the first rise from the river. This puts the house in an extremely visible position. The earlier building believed to be incorporated into The Anchorage measured 28 feet long by 16 feet wide (NRHP 1973c:7). The east wing of the current structure is approximately this size and may be its survival. If the current landscape features adhere to these measurements, they will probably date to this earlier period. If the landscape was modified or rebuilt after 1831, it will probably not adhere to these earlier measurements. However, the layout of the current structure is relatively irregular which makes determining its dimensions difficult. The central pavilion measures approximately 22 feet square. The total length of the house is approximately 120 feet. From the LiDAR dataset, there does not appear to be many topographic features associated with The Anchorage. Despite this, there is at least one feature which appears to be built according to a measurement derived from the main house. A small garden and walkway to the southwest of the house appears to be built using 16 feet as its base measurement. While there does not appear to be much surviving topographic evidence of a geometric built landscape here at The Anchorage, the 1847 U.S. Coastal Survey indicates that it may have been built using the rules of perspective. The road leading to the ferry and the fence line that defines the field behind the house form a pair of lines that diverge as they move uphill, away from the house. As one crosses the Miles River and looks up at The Anchorage, these lines would appear to be straight and parallel and would not appear to converge like true parallel lines do. 200 The Wilderness (T-149) The Wilderness (Figures 5.32, 5.33, and 5.35) is a multipart brick house laid in Flemish bond situated on the Choptank River in southern Talbot County, approximately five miles south of Trappe, Maryland. In 1659, this tract of land was part of a 3,040-acre land patent to Edward Lloyd, the first of his name, and known as Hier Dier Lloyd (NRHP 1973d:3). The Wilderness was divided from this tract in 1683 but work on the current structure did not begin until the last quarter of the eighteenth century (NRHP 1973d:3). The Wilderness is historically significant for being the home of Daniel Martin, Maryland’s 20th governor and architecturally 201 significant for being an example of Maryland vernacular architecture (NRHP 1973d:3). The oldest portion of the house is a two-and-a-half-story-tall, four-bay-wide Georgian brick dwelling dating to 1780-1790 (NRHP 1973d:2). There are two 202 chimneys in this section of the house, an exterior end chimney on the east end and a slightly taller interior end chimney on the west end. The floorplan of this wing consists of two rooms but originally may have been a center hall with a room on each 203 side (NRHP 1973d:2). Some of the original woodwork appears to remain in this section of the house (NRHP 1973d:2). Attached to the east is a taller two-and-a-half-story, three-bay-wide Georgian wing dating to about 1815 (NRHP 1973d:2). The fenestration in this wing is more regular with an entrance located in the bay closest to the original wing of the house and covered by a small porch and two evenly spaced windows to the west. Located above these on the second story are three identical windows and located above the outer two bays are small dormers that pierce the gable roof with small six-over-six sash windows. Inside the western façade of the structure is a pair of interior end chimneys, taller than the other two in the earlier wing of the house. The floorplan of this wing consists of a side hall that runs the depth of the block and two rooms that lead off it. In the early twentieth century, a partially enclosed one-story porch was added to the southwest, water-facing side of the house and a small kitchen wing was added to the southeast (NRHP 1973d:2). The entire structure was painted white around this time as well (NRHP 1973d:2). In line with the main house to the east are a smokehouse, a dairy, and a now-renovated slave quarter (NRHP 1973d:2). At The Wilderness, the area between the south façade of the house and the Choptank River appears to have been initially laid out as a formal garden space. There are extant standing lines of trees and a grassy allée between the house and the river; however, no discernable topographic variation can be seen in the LiDAR dataset. 204 Unfortunately, the National Register nomination form for The Wilderness does not list its dimensions, nor does any of the published literature about the structure. From the aerial imagery, the 1815 wing appears to measure approximately 35 feet long by 38 feet deep. The older wing is more difficult to measure because it is partially blocked by tree cover. The best approximation that can be come by from this aerial imagery is 36 feet long by 32 feet wide. This would make the entire structure 71 feet long and 38 feet wide at its deepest point. The current driveway to the Wilderness is centered on the 1815 west wing of the house. The Wilderness, like Myrtle Grove and Compton, is a multi-part structure built across several generations. As such, determining the layout and orientation of the landscape surrounding the house can help to identify to which period it dates. Because the current approach to the house is centered on the later wing, it should date to this post-1815 time period. Interestingly, it appears as if the earlier driveway may still exist along the edge of the field just to the east of the post-1815 driveway. Additionally, the LiDAR dataset reveal an intricate system of mainly on-grid drainage ditches crisscrossing the agricultural fields at The Wilderness. The core of the plantation adheres to the grid defined by the main house. However, toward the periphery of the plantation, a regular gird can still be identified but it is several degrees off this alignment. The LiDAR survey at The Wilderness has revealed an interesting feature worth further investigation. At the end of a small ridge that ends at the Choptank River, is a small, circular rise with a smaller, circular depression in the center of it. Similar features can be seen at just southeast of the house at Otwell and just east of 205 the house at Myrtle Grove. At Myrtle Grove, this feature is believed to be the remnants of an ice house. Given that ice houses are common features on some of these larger plantations, there is a strong possibility that these two other pits are also ice houses. Wye House (T-54) Wye House (Figures 5.35, 5.36, and 5.37) is a symmetrical seven-part frame Federal house with early Greek Revival characteristics with a central, forward-facing gable two-story central block flanked by one-story connecting hyphens and shorter, attached, forward facing two-story dependencies attached to one-story pavilions. The plantation is located on the Wye River in northwest Talbot County, near Easton, 206 Maryland. The present structure was built sometime between 1781 and 1784, was completed in its current configuration by 1799 (NRHP 2009:7) and is perhaps the second or third to stand on the property (Forman 1967:61-65). It is currently owned by Richard and Beverly Tilghman, 11th generation descendants of Edward Lloyd I, 207 who settled along the Wye River in 1658 (NRHP 2009:7). Wye House is designated as a National Historic Landmark and was the boyhood home of Frederick Douglass, who writes about the plantation in each of his autobiographies. 208 The two-and-a-half-story-tall, five-bay-wide central block is 47 feet wide by 42 feet deep (NRHP 2009:4). The south, land-facing façade is dominated by the large, forward-facing gable capped with a triangular pediment or tympanum with an arch window in its center. Large pilasters rise to the base of the tympanum and give the appearance of a Greek temple. The entrance to the main block is in the central bay with a Palladian portico that protrudes from the façade. Above the door is a fanlight and to the side are two sidelights. To either side are two additional bays with nine- over-six hung sash windows slightly offset to the outside to account for the wider central bay. Fenestration on the second story is similar to the first with six-over-six hung sash windows hung over the first-floor openings and the central window over the entrance with a pair of two-over-two side lights. The north façade, which does not have water-facing views because of the placement of a late eighteenth-century Orangery, is similar to the south façade except for the addition of a one-story, flat- roofed enclosed porch that runs the full width of the block and the subtraction of the sidelights around the central window. Four interior chimneys rise through the lower, outside edges of the roofline toward the corners of the block. The floorplan of the house consists of an entry hall with a T-shaped hall junction that leads to either wing of the house. Two additional rooms are in the north side of the main block and access to the enclosed porch is through a pair of jib windows, one in each of these rooms. The center block of this frame structure is built on a brick foundation and its walls are nogged with brick (MIHP 1989:2). This central block is connected to a pair of shorter two-story-tall, two-bay- wide pavilions by one-story hyphens, each with a single window in each bay. The 209 hyphens contain a single room each, in addition to the passage that leads into the outer wings and are covered by a pent roof that is not visible from the south approach to the house (NRHP 2009:4). Both hyphens and wings have six-over-six hung sash windows on the first floor. The wings have short, three-over-three hung sash in each bay on the second story, truncated by the triangular, forward facing tympanum that echoes the one in the main block, including a pair of pilasters that rise to their bases, although without the arched semicircular window in its center (NRHP 2009:4). Each wing has an additional interior chimney on their outside face, just south of the centerline of the house and the western wing has a second interior chimney at its rear. Capping the ends of the structure are one-story-tall, two-bay-wide end units. These have half-hipped gable roofs that start as a traditional gable where they intersect the taller two-story wings and terminate at the outside edges of the structure as a hipped gable. The one to the west is the Lloyd’s library containing over 1,000 volumes (Wolf 1969) and the one to the east is a recently renovated kitchen that also used to contain quarters for the enslaved (NRHP 2009:4). In total, the seven parts of Wye House measure to 151 feet long (NRHP 2009:4). Extending the measurement of the center block of the current house (47 feet long by 42 feet wide) across the landscape as a grid of squares reveals that much of the extensive agricultural fields at Wye House fall along these grid lines. One hundred yards north of the main house is a late eighteenth-century greenhouse, frequently referred to as the Orangerie. Built of stuccoed brick, some of which has been made to look like quoins, the Orangerie is the result of several phases of construction, the last of which dates to the 1790s (MIHP 1989:3). The structure is 210 85 ½ feet long and consists of a central, two-story-tall, four-bay-wide main block that protrudes slightly from the two symmetrical, flanking, one-story-tall, three-bay-wide wings (MIHP 1989:3). Because it was built as a greenhouse, the southern facing windows are extremely large; thirty-over-thirty hung sash on the first story of the central block and thirty-over-twenty-four topped with a semicircular fanlight on the wings. The second story of the main block was at one point a billiard room (MIHP 1989:3) and has four eight-over-eight hung sash windows directly above the four windows on the first floor. The roof on the main block is hipped and the roofline on the two wings has the same half-hipped profile as the two end wings on the main house. A pair of chimneys rise through either end of the back of the Orangerie and are a part of the hypocaust or heating system that enabled the Lloyds and their workers and slaves to grow tropical fruits such as oranges and lemons year-round (MIHP 1989:3). Elsewhere on the property are a dairy, smoke house, Captain’s House, carriage house, tenant house, and barn complex (MIHP 1989:15-20). The plantation has undergone several transformations, most notably in the 1770s when Edward Lloyd IV inherited the plantation and continuing until just after the American Revolution. This transformation included the construction of the current plantation house and a switch from tobacco to grain and livestock. Wye House has been studied extensively by architects, landscape architects, and archaeologists. The history, architecture, landscape, and archaeological investigations of Wye House are discussed at length in the following chapter. Among these is Henry Chandlee Forman, whose research has formed an important cornerstone of the archaeological investigations at Wye House conducted by Archaeology in Annapolis 211 (most notably reported in Forman 1967:51-81). Working at mid-century, much of Forman’s work is historic now in its own right. Like the early aerial photograph of Wye House, Forman’s work records the landscape at Wye House from more than half of a century ago. This is especially helpful for researchers when it comes to studying dynamic, living parts of the landscape like formal gardens, boxwoods, and hedges. Forman’s map of the garden surrounding Wye House and its Orangery is incredibly detailed and reproduced as the endpapers to his 1967 Old Buildings, Gardens, and Furniture in Tidewater Maryland. The LiDAR dataset from Wye House confirms many elements of Forman’s garden map, but others are harder to see or nonexistent. Comparing the LiDAR to this map is illustrative of some of the limitations of using LiDAR and other kinds of remote sensing to record and identify plantation landscape features (Figure 5.38). It is tempting to consider LiDAR scanning as all- encompassing and infallible, but this example reminds us that many landscape features are too small to be captured by current LiDAR technologies, do not leave a topographic footprint, or have been completely removed from the landscape at some point in the past. Largely for this reason, the survey methodology proposed and explained in this chapter cannot be considered a replacement for more detailed, in- depth studies. Rather, it is proposed as a quick and wide-ranging survey technique designed to cover a large amount of area at little cost. Across the landscape at Wye House are numerous small ditches and trench- like features. These are usually found along field boundaries and act to divide the vast agricultural holdings of the Lloyds. These boundary ditches form a rectilinear network of fields with two distinct orientations; one located to the south of the house 212 and garden and the other to the north. From this LiDAR dataset, we can extract the sides of these ditches and examine their profiles, including attributes like depth and slope. Forman also notes the presence of these two distinct angles or orientations at Wye House and attributes them to the late seventeenth-century house and the present 213 late eighteenth-century house (Forman 1967:57-61). The configuration of the landscape at Wye House prior to the construction of the current structure is somewhat of a mystery to researchers. The following chapter, Chapter 6, will contain a synthesis of the previous archaeology conducted at Wye House as well as an attempt to use this technique to conduct a more focused survey to locate archaeological sites related to plantation slavery. Discussion From this analysis of ten plantations and their landscapes in Maryland’s Talbot County, we have accomplished several goals. First, we have examined and explored ten extant plantation landscapes in Talbot County, Maryland and attempted to combine their histories, architecture, and surviving landscapes features. We have identified a number of mathematical or geometric principles that seem to underpin the design and layout of these spaces, namely symmetry, perspective, the use of ratios and proportions, and the use of a grid of squares. In so doing, we have also begun to develop a useful methodology for rapidly surveying plantation structures and landscapes without needing physical access to the plantation. Together, these ten plantations are not intended to represent a representative sample of all plantations or even plantations in Talbot County, Maryland. They were chosen for appearing relatively intact and for possessing a modicum of previously- conducted background historical research. From this readily-available background information as well as the State of Maryland and Post-Sandy LiDAR datasets, we 214 have been able to document and explore these landscapes remotely, quickly, and cheaply. From the available material on these plantations, the dichotomy between the periphery and core is clear. The main plantation houses themselves have received the most attention followed by the outbuildings that surround the main house. The gardens and surrounding landscapes have received little to no attention in the common forms of plantation documentation. The fields, forests, driveways, roadways, waterways, and, as will be shown in the following chapter, the quarters of the enslaved can be identified and located if there are any extant remains that affect the ground surface. From this survey, it is clear that most, if not all, of these plantations show evidence of being laid out according to a rigid grid. In most cases, the grid takes its orientation and dimensions from the house. Embedded within most of these grids is a proportion derived from the dimensions of the house. While this largely confirms what is already know about the construction of eighteenth-and-nineteenth-century plantations and their landscapes, it is useful to see this in application across so many sites. The analysis presented in this chapter also lays out a way to establish a chronology for the creation of these gardens and landscapes. From the ten plantations presented here, many of the earliest ones do not possess the rigid geometry and extensive terracing that the later ones do. Likewise, those earlier houses that later saw extensive modifications and additions frequently have landscapes and gardens oriented and scaled to these later additions. 215 One notable finding from the survey of Talbot County plantations here, as well as in the larger survey conducted by Ryan O’Connor, is a lack of two features found elsewhere in the literature on eighteenth-and-nineteenth-century plantation landscapes in the Chesapeake. As documented by Harmon and others at both Tulip Hill and Wye Hall as well as by Leone at the Paca House in Annapolis, large, built, descending terraces are one feature that was expected on many of these Talbot County landscapes. However, with the exception of Ratcliffe Manor, none of the other nine plantations had any similar kind of ground surface modification. While several have limited amounts of landscaping such as boundary or drainage ditches, garden beds, or slight terracing, none approach the level of modification seen at the plantations described by these authors elsewhere in Maryland. This kind of feature is probably much rarer than is generally understood, which is probably not surprising given the amount of labor required to produce them and the biases in favor of their survival and documentation. Likewise, another landscape element commonly discussed in Chesapeake plantation literature has been identified on only some of these plantations: the use of perspective or of converging and diverging lines. At Clay’s Hope, Otwell, Myrtle Grove, The Anchorage, and Wye House, there is evidence of the use of the rules of perspective. The possibility remains that converging and diverging lines could have existed on these other plantations in the form of vegetation or since-filled-in-ditches; however, the datasets used here are unable to identify them as they may not have left a topographic or cartographic trace. Of the plantations here that do manipulate the rules of perspective in the built landscape, four of them are older plantations with 216 origins in the seventeenth or early eighteenth century with connections to two of the wealthiest families in Talbot County. hGIS and LiDAR survey of ten Talbot County plantations and their surrounding landscapes should complicate the ways in which archaeologists and historians think about these spaces. Furthermore, it needs to be stressed again that these ten examined here are not a representative sample of plantations from across the county but were chosen because of their historic integrity and availability of documentation. Further work could perhaps use the structures identified in 1858 by William Dilworth on his map in conjunction with the LiDAR survey data, historic tax lists and census returns, deeds, plats, wills, and other legal documents to document and categorize these historic features in a systematic manner. In addition to teasing out a model of how these plantations were constructed and laid out, the two LiDAR surveys used in this chapter (2004 and 2015) represent a record of this landscape at two moments in time. As long as this data is kept and maintained, an imprint of it will continue to exist, just as the earlier US Coast Survey maps continue to record the landscape from the mid-nineteenth century. At Myrtle Grove, where two significant changes to the landscape were recorded in the 11 years between LiDAR surveys, going forward into the future more and more changes will begin to accumulate across the landscape of Talbot County. In 50 years, as landscape features are erased, or built over, or fade away, these two datasets may prove to be as useful as Henry Chandlee Forman’s research has been to Archaeology in Annapolis at Wye House or the US Coast Survey maps have been to the research presented in this chapter. 217 This methodology provides the researcher with a way to access a historic resource that is largely inaccessible. As will be discussed in the following chapter, for plantation landscapes of the Chesapeake or elsewhere in the American South, it represents a way to access an African American heritage that is locked away, forgotten and buried on private lands. But it should be emphasized that the line of inquiry presented here only represents the first step in a larger research project. From here, the next step is to start with this overview and survey and combine it with further historical and archaeological investigations. Guided by these results, a deeper dive into the historical and archaeological records can lead to important archaeological discoveries. The following chapter provides one such example. 218 Chapter 6: Locating That Which Has Been Lost: Using Historical Geographic Information Systems (hGIS) to Locate Quarters at Wye House hGIS and LiDAR It has been shown by Harmon and others in their 2006 article in American Antiquity, by Johnson and Ouimet in a pair of articles (2014 and 2018), and by this author in the previous chapter that aerial LiDAR datasets can be used to map existing landscape features and that in their current configurations can be found earlier eighteenth (and in some cases perhaps seventeenth) century underlying geometries and geographies. From this knowledge, we can begin to make predictions as to the location of structures no longer present on the contemporary landscape and understand the rules by which they were originally built by incorporating existing historic cartographic representations of that landscape. While this methodology can be useful for identifying the design principles behind these landscapes as well as for establishing the age of particular components, it should be noted that data collected in the present such as the LiDAR-derived elevation data used here represents the modern configuration of that space, of which we can only attempt to tease out the underlying geometry. That being said, the above authors all argue for the integration of modern geospatial data with the historic record and especially historic maps and other kinds of archaeological datasets in order to arrive at a more robust historic geospatial dataset that can be used to locate and understand archaeological resources. To assist with the task of establishing the historical antecedents embedded within the modern landscape as recorded by modern LiDAR datasets and aerial 219 imagery, some form of archaeological interrogation is required. Here I use archaeology in its broader, literal, etymological sense, that is, a study of beginnings or of the old, rather than as it is more commonly used to refer to a specific course of study involving excavating, dirt, and artifacts. This is not to say that there will not be excavation involved in examining the historic character of the modern landscape. Physical archaeological excavation is ultimately required in order assess the validity of our claims regarding human occupation and manipulation of space and place that come from this proposed research methodology. One of the premises underlying archaeology as a field as it is practiced today is that we can arrive at some approximation of past human activity in a specific place through an examination of the cultural material left behind at that place. It is through archaeological excavation that material culture can be directly identified and assessed. Unfortunately, one of the greatest challenges in doing landscape archaeology is that often times very little physical material evidence is left behind to recover and the areas covered are often very large. This means that an archaeological inquiry of landscape must frequently rely on alternative forms of evidence to supplement traditional archaeological survey and excavation when seeking to understand past configurations of landscape. This approach is not novel or revolutionary by any means. The necessity to combine multiple lines of historic evidence with on-the-ground investigation of material remains has been embedded within the study of historic landscapes from the beginning of modern landscape studies. W.G. Hoskins relied heavily on maps of the English countryside along with a peripatetic, boots-on-the-ground approach to understand centuries of historic land use and change in his influential and discipline- 220 defining The Making of the English Landscape (1955). Hoskins relied on the palimpsestual nature of landscapes as embedded in the arrangement of space in the form of roads, forests, fields, town plans, and place names as recorded in contemporary maps to tease out vanished landscapes of the past. Similarly, archaeologists can and have turned to similar datasets to investigate past spatial configurations. Hoskins relied on the vestiges of previous configurations and uses of land surviving in some form into the present. While he was generally interested in a much older past than landscape archaeologists working in the Chesapeake, there is no reason this theoretical outlook and methodology cannot be applied here. Through his work with Ordinance Survey maps, Hoskins was able to examine features of the Medieval, Roman, and Neolithic landscapes in England that have survived into the present. Furthermore, given his datasets and intellectual focus, as one approaches the present, the eighteenth-and-nineteenth-century configurations of landscape seem to be right there at the surface, self-evident and visible to all, as they are largely unobscured by subsequent landscape changes. These more recent landscapes sometimes seem like an afterthought to Hoskins before he launches into a tirade against the mid-twentieth-century highways and American strategic air bases that he sees as blight on an otherwise picturesque landscape. That is not to say that he does not engage with the railways, canals, parks, and estates that cover and crisscross eighteenth-and-nineteenth-century England, but rather, by the eighteenth and nineteenth centuries, there are other, and perhaps in some ways better, kinds of documents and records with which one can investigate these more recent landscapes along with the kinds of contemporary maps that form the basis for Hoskins’ study. So 221 rather than be left to tease out the historic depth behind landscape features seen on a modern map, one can examine directly a historic map from this more recent past. This is largely the same methodology that underlies historical archaeology and the study of the more recent past; our investigations of material culture can be greatly augmented by incorporating other kinds of textual and cartographic evidence where it exists. With a broadened definition of what constitutes archaeology to include these kinds of landscape studies, we are well-situated to investigate previous and “lost” configurations of landscape. It should be emphasized that these landscapes are frequently not truly lost. As Hoskins has shown, as we have seen in the previous chapter, and as will be discussed below, these landscapes are not completely gone, they still survive in various forms and fragments and in various places. It is up to the landscape archaeologist to find them and understand the significance of their remains. One notable study to attempt to characterize surviving cultural landscapes in the present relevant here is the study of two Maryland plantations using airborne LiDAR-derived elevation data by Harmon and others (Harmon et al. 2006). Whether the authors made it explicit or not, their survey follows in the same tradition as Hoskins in that it uses present configurations of landscape as recorded in a contemporary spatial dataset to understand survivals of historic landscapes into the present. Here, they examine the built terraces and other landscape features found in the formal gardens and landscapes of Wye Hall in Queen Anne’s County and Tulip Hill in Anne Arundel County, Maryland as well as prospect for additional archaeological features. In their conclusion, Harmon and others call for the integration of their LiDAR dataset with additional historical datasets to create a more 222 robust historical GIS database. They show that in addition to mapping the physical terrain of plantation landscapes, historic maps can be used to identify structures and layout on plantations. When it comes to investigating plantation landscapes at the scale of individual features and structures, we can follow this recommendation to incorporate digital elevation data “in conjunction with information derived from the field, the laboratory, and from archival sources” (Harmon et al. 2006:668). They conclude by suggesting the use of LiDAR be combined with more traditional archaeological datasets, certainly including historic maps, early aerial photography, and archaeological excavation. Following this recommendation, the methodology presented in this chapter combines LiDAR-derived elevation models, historic maps, historic aerial photographs, and modern aerial photographs in order to aid ongoing archaeological excavations by Archaeology in Annapolis at Wye House plantation outside Easton, Maryland (see Figure 2.2), including revisiting a previous archaeological interpretation, identifying the locations of a pair of slave quarters believed to have been on the Long Green at Wye House (the area on the plantation where many of the enslaved on the plantation lived and worked), and ground-truthing, or testing, to show that this methodology can be successful in locating archaeological features. One of the primary benefits of working within a GIS framework is that it allows for the combination and comparison of multiple spatial datasets simultaneously. This includes spatial datasets from multiple time periods. Whereas Hoskins was largely working with a single contemporary map, historic GIS allows for different maps and data to be overlaid with each other and connections to be made 223 between them where before no such connections existed. It helps establish time-depth across the landscape being studied, fills in gaps in the cartographic record, and assists in understanding change and continuity across time (Knowles 2002; Knowles and Hillier 2008). As shown in the previous chapter through combining the William H. Dilworth map and the United States Census, doing so allows us to marshal the strengths of these datasets while offsetting some of their weaknesses, resulting in new geographic, spatial, and historic knowledge. Likewise, spatial datasets that are drawn at different scales, with different thematic focuses, or from different time periods can be combined to create new geographic knowledge. Following this call by Harmon and others, this methodology can be employed at Wye House to locate and excavate archaeological features which have disappeared from the landscape. Taken as individual components, none of this proposed methodology is radically new to the field of archaeology. What is noteworthy about this use of GIS to conduct archaeological site reconnaissance is the integration of all of these data sources into a single digital geographic framework, including especially non-planimetric oblique aerial photography, which offers a powerful tool for unifying these sources in order to locate previously unidentified archaeological sites to a high degree of spatial accuracy. This deeper dive into one specific plantation site will show the potential of robust hGIS databases to archaeological research projects. Once historic maps and other historic datasets have been georeferenced, they can be used to begin understanding past configurations of space as well as the continuity of historic features into the present. On a Chesapeake plantation, it is easy enough to locate a main plantation house. Most are built on topographic high spots. 224 Many are built overlooking water. In the event it is not still standing, good maps exist starting at least around the time of the Civil War and aerial photographs from the early twentieth century that show at least the general location of the main house. More difficult to find are the outlying buildings and structures that were required for the agricultural work that occurred on a planation. Frequently located away from the main house are the work areas and agricultural fields which provide the economic base of the plantation and are of interest to archaeologists studying African American life on plantations (Vlach 1993). Unfortunately for researchers, the locations of many of these quarters are unknown, not just at Wye House, but across the entire Chesapeake. In some cases, their location is explicitly noted on historic maps, identified on a National Register of Historic Places nomination form, or remembered in oral tradition; however, these cases where the locations of these resources is actively known seems to be relatively rare. Some historic spatial datasets, like those maps produced by the United States Coast and Geodetic Survey, go further and show farms and plantations in more detail, including fields, forests, and outlying structures like quarters or barns, but these are less common and coverage is far from universal. Data Sources Fortunately, building a historic Geographic Information System at a place like Wye House is not terribly difficult, largely because it has been documented as well as it has been, both in the present by researchers and in the past by map makers, photographers, and others. As seen in the previous chapter, the kinds of spatial evidence used here at Wye House are not unique to Wye House. As such, similar 225 hGIS databases can be created for other locations or this one can be expanded to include neighboring plantations and used to the same effect. hGIS at Wye House is presented here as a fully-developed case study with potential to be copied as a model elsewhere. With the most recent phase of excavations at Wye House beginning in 2005, Archaeology in Annapolis has benefited from the existence of historical documentation of the plantation in the form of both primary documents and accounts produced by the Lloyd family and by Frederick Douglass and others who were at the plantation as well as from secondary sources produced by historians and other scholars over the past century. Many of these datasets are either explicitly spatial in nature, making them suitable for the kind of research program suggested by Harmon and others or by Johnson and Ouimet, or implicitly spatial in nature, making them suitable for a more qualitative spatial analysis as will be seen in the following two chapters. These first set of explicitly spatial documents that will be discussed in this chapter include the historic maps, historic aerial photographs (both vertical and oblique), maps derived from archaeological and historical research, modern LiDAR elevation data, and modern aerial imagery. To these, we can also bring in other contemporary spatial datasets including infrared photography, modern as well as historic shoreline maps, wetlands inventories, or any other spatial dataset detailing present or past land use. Combining these historical documents with the digital elevation model provided by a LiDAR elevation dataset creates a robust spatial dataset embedded with historical depth that enables us to locate archaeological resources which have disappeared from the landscape but were either recorded in the 226 past or have left some physical trace on the ground surface in the present. The hGIS is not limited to the kinds of explicitly spatial datasets described above, for example, historic textual descriptions of the landscape, like those of Wye House found in Frederick Douglass’s autobiographies, can be incorporated into the hGIS as will be seen in the next chapter. The core of the hGIS database is comprised of a series of historic maps. These historic maps show a series of static snapshots of the landscape at several points in time and help locate historic features and allow us to begin to understand the time- depth of the specific landscape depicted. The maps used will vary based on geographic location, time period being studied, and the specific features that one wishes to be able to capture. At Wye House, the specific maps identified for including in this hGIS database include an 1847 US Coast Survey map of the St. Michaels River (US Coast Survey 1847; Figure 6.1), a map also from the US Coast Survey that appears to be at least partially based on this earlier 1847 survey titled, Chesapeake Bay from its Head to Potomac River (US Coast Survey 1862; Figure 6.2), the 1858 William H. Dilworth map discussed at length in Chapter 4 (Dilworth 1847; Figure 6.3), and the 1877 Lake, Stevenson, and Griffing map of Talbot County, Maryland (Lake et al. 1877; Figures 6.4 and 6.5). As a warning, each of these historic spatial datasets are the unique product of specific circumstances and contain their own biases, blind spots, and weaknesses. While working with the datasets above, it is important to understand the reasons behind their creation in order to understand why their creators did or did not include particular elements. It is easy to forget that these historic maps depict only a 227 specifically-chosen subset of the built and natural environment and that even those parts they do depict should not be assumed to be accurate or complete. As they were in the past, contemporary landscapes are palimpsests and evidence of the past can be 228 found in the physical shape of the landscape alongside and under modern configurations of the same place. As with its use with the Census described in Chapter 4, the Dilworth map is a useful starting place in that it gives us a regional view of the historic plantation 229 landscape. It can be useful for situating Wye House into a larger neighborhood, establishing the historic depth of the contemporary regional landscape (especially in terms of roads and property boundaries because the map clearly shows these features), and when combined with the Census can serve as a spatial directory of 230 social relationships in the county. From an archival survey of Talbot County maps done by this author for the identification of potential spatial data sources for this chapter, these US Coast Survey maps probably represent the first systematic recordation of the built and natural landscape in Talbot County. At Wye House, on 231 the larger scale 1847 map we clearly see the main plantation house, the greenhouse, rows of ornamental trees lining the mail driveway, work buildings, barns, quarters, fence lines, orchards, and a windmill. On the lower scale, later survey, the details are 232 harder to discern, but they generally agree with the earlier map in the overall configuration of the plantation landscape. In addition to these maps which are explicitly spatial datasets, we can also utilize early aerial photography. With advances in both photography and aircraft, the first wide-spread use of aerial photography occurred during the early twentieth century. Archaeologists have been using aerial photography as early as the 1910s and 1920s (Renfrew and Bahn 2004:83). Aerial photography can be acquired from either directly above the subject in a vertical or top-down view or from the side at an oblique angle. Vertical aerial photography is perhaps easier to work with because its perspective is most similar to a map, distortions related to the position of the observer and subject are minimized, it is relatively easy to georeference these images, and distance, dimensions, and orientation can easily be obtained. The angled perspective introduced by oblique aerial photography provides for a view of the subject more similar to natural human vision but at the cost of its usefulness in a mapping environment. Distances and bearings between objects are not uniform across the image; therefore, oblique photographs are useful to look at but not analyze as a map or vertical aerial photograph without some kind of processing that will be discussed below. On the wall of the library located in the west wing of Wye House, an aerial photograph taken by the Dallin Aerial Survey Company can be found hanging, showing the house and its surroundings as they existed in the early twentieth century (Figure 6.6). J. Victor Dallin, a pilot in the Royal Air Force during World War I, came to the United States and started the Dallin Aerial Survey Company. Based out 233 of Philadelphia, the company operated by acquiring aerial imagery of events, cities and town, infrastructure, and estates like Wye House and then selling the imagery to interested parties, as in this case, the owners of Wye House (Hagley 2017; Martin 2008). While the exact date of this photograph is not currently known, the Dallin Aerial Survey Company archives held by the Hagley Library has imagery acquired over Wye Island, directly across the Wye River to the north dated June 23, 1927 and September 5, 1927 and over Easton, Maryland August 16, 1929. As these three batches are geographically the closest known aerial photography in this collection with dates, the best estimation is that this imagery of Wye House was also acquired on one of these three dates in the late 1920s. 234 Because this photograph of Wye House is shot from the side and not directly overhead, it is difficult to immediately determine either the relative or absolute locations of structures visible in the photograph. One way to address this problem is to turn our oblique photograph into a map, that is, a spatial document with internally- consistent scale and orientation. This is accomplished by warping and distorting the image in such a manner that it corresponds to the same map projection as our GIS database. We are interested in being able to determine the real-world coordinates of any given pixel in the image and this is also accomplished through the process of georectification. By georectifying this image, it will be possible to determine the coordinates of any pixel in the photograph which will allow us to identify the locations of structures and features that are present in the historic photograph but no longer exist. This can be accomplished by following a digital version of the methodology described by Scollar (1975) and Estes and others (1977). The distortion of the oblique photograph that they accomplish with a rubber sheet (which is why this process is sometimes also known as rubber sheeting) can also be accomplished using GIS (Figure 6.7). Just as with georectifying maps, this is accomplished here by creating a database of ground control points between known locations in the coordinate system and the same locations in the aerial photograph. These points can be used to compute the transformation needed to warp or distort the oblique aerial photograph into a top- down or vertical map. The resulting warped image can be used as any vertical aerial photograph would and the locations of structures or objects in this photograph correspond to real-world coordinates and therefore can be used to pinpoint on the 235 ground where these structures stood. However, this process comes with a few caveats. Because the image needs to be heavily distorted in order to convert it from an oblique image to a flat, top-down projection, many ground control points need to be identified and well-distributed across all areas of the image in order to account for the low angle 236 at which the photograph was taken as well as any distortion introduced by the camera lens. Areas without ground control points or those near the edges of the image should be used with caution until it can be shown that they are accurate. The transformation will need to make more assumptions about areas without GCPs, especially around the edges of the image, and the resulting accuracy will suffer. Likewise, because this process tries to apply an extreme correction to the oblique image, the resulting image will be heavily warped to make it fit and therefore will be difficult to interpret visually without referring to the original image as reference. Historic vertical aerial photography also exists for Wye House and the surrounding area. The earliest of these systematic surveys was carried out over Talbot County in 1937 (Figure 6.8). Georectifiying these aerial images is similar to rectifying a historic map; GCPs need to be established between known points in the aerial photograph and on the modern landscape. One advantage of historic aerial photography is that, unlike a map, it captures image data for every pixel in the scene, resulting in coverage for the entire survey area. Assuming conditions were good and cloud cover was limited, this means there should be no blank spots in the survey area whereas historic maps only record data where the cartographer chose to record a building or road or forested area. Aerial imagery can frequently reveal detail about areas overlooked by historic map makers for one reason or another. In terms of historical research and landscape analysis, these oblique and vertical aerial photographs represent an important dataset. They provide detailed coverage of large landscapes. While the spatial coverage of these oblique datasets are not uniform in the same way the systematic vertical surveys are, it would seem that 237 those large, surviving plantation houses like Wye House are more likely to have been photographed in the past and it is more likely for that photography to have found its way into an archive (or framed on a wall) to be preserved. Vertical aerial photography generally does not have the same kind of problem. Starting in the 1930s, much of this 238 kind of imagery was acquired in a systematic manner by governmental entities and was frequently designed to be used stereographically, which necessitates many overlapping frames to produce the illusion of 3D topography from two flat pieces of paper (Bolstad 2008:216, 230-231). Therefore, the spatial coverage of these surveys is generally universal or near-universal within a given region, especially as one approaches the present and satellite-based imaging platforms come into service. This makes both kinds of aerial photography useful datasets for archaeologists working in these kinds of plantation contexts. Because of the costs of aerial photography, the temporal coverage of these datasets can also be irregular, but any coverage from the early twentieth century represents a snapshot in time positioned roughly halfway between slavery and the present and can provide important stepping stone for recovering those past landscapes. To these spatial datasets that were originally produced in the nineteenth-or- early-twentieth century that were created by observing the then-current configuration of this landscape, we can add secondary spatial datasets that were produced through archaeological and historical research. The Henry Chandlee Forman map of Wye House, dated 1956-1965 and reproduced as the endpapers of Forman’s 1967 Old Buildings, Gardens, and Furniture in Tidewater Maryland (Figure 6.9), is an interesting hybrid of historic map that shows the configuration of space in the past, now more than 50 years old, and an archaeological site map, partially informed by Foreman’s investigations at Wye House. For this reason, Forman’s map holds significant potential for researchers not only because it depicts both the seventeenth- and-eighteenth-century layouts of the plantation (at least as Foreman understood 239 them), but also because the structures on it are labeled with descriptive names indicative of their use (which has since been lost from public memory) and other data generated by Foreman’s early archaeological and architectural studies. Of particular note here are the structures Forman identifies as the “Br. Row Quart.” and the “2- 240 Story Quarter”, located in an inset of the Long Green in the bottom right of his map (Figure 6.10). Identifying the location of these structures is the first step in a larger archaeological research project involving their excavation, analysis, and interpretation. Because Foreman’s document already appears to be drawn to scale, all 241 that needs to be done in order to incorporate it into a GIS framework is for it to be translated, rotated, and scaled to an existing coordinate system. If Forman did a good job, the map should be an accurate representation of what once existed on the ground. Once the map is georeferenced, structures on the map that still exist should correspond to their real-world locations and we will be able to use the map to determine the locations of non-extant features. If this is not the case, warping and stretching based on known GCPs across the map may be required to georectify the map appropriately. Because the map contains an inset of the Long Green, this process should be repeated twice, once for the main map and once for the inset as they require separate transformations to georectify accurately. Additionally, whereas this map depicts a palimpsest consisting of four centuries of construction, erasure, and redevelopment at Wye House, Forman also provides in his book a copy of a plat of Wye House dating to 1784 (Figure 6.11). This earlier map is notable in that it shows that the main block of the current house is standing by 1784 along with two unattached (now attached) dependencies. It also shows a series of buildings along the east side of the secondary plantation driveway in the area known as the Long Green. Unfortunately, it also omits several buildings known to be standing in 1784 such as the Captain’s Cottage and simplifies the area around Lloyds Cove in such a manner that makes it difficult to accept that this is anything but a highly stylized depiction of these structures, meant to convey only their general details and property boundaries. To these historic maps and aerials can be added a digital elevation model (DEM) of the terrain generated from the 2004 Maryland Department of Natural 242 Resources’ LiDAR dataset, the 2014 MD/PA Post-Hurricane Sandy Supplemental Lidar Data Acquisition, or any subsequent LiDAR data acquisition (Figures 6.12 and 6.13). Like these modern datasets, other contemporary spatial datasets that cover Wye House can be obtained and mapped, including street maps, parcel maps, utility maps, 243 and shoreline maps which, when combined with the historic spatial datasets above can be used to help tease out the historic depth on the landscape in a manner not unlike the way W.G. Hoskins uses British Ordnance Survey maps to survey the English landscape 244 While these maps and spatial datasets are excellent sources to map both specific and regional plantation landscapes, they are not without their shortcomings. The primary shortcoming to remember when working with these historic maps is that they only show that which the cartographer chose to represent, or, in the case of aerial 245 photography, are entirely dependent on the researcher’s ability to identify, analyze, and interpret the cultural significance of features on the photograph. Likewise, the representation of the landscape on the final map can only be as good as the cartographer’s knowledge of the landscape they were mapping. Together, this means there are biases and distortions all over these historic maps, not just in terms of geographic representation and accuracy (these errors can frequently be identified by comparing multiple maps to each other and to the present conditions on the ground, and by having a familiarity with the other historic records that describe the same area), but also in terms of the deeper cultural geographic knowledge embedded in local geographies such as place names, residents, exploitable resources, trails and pathways, etc. While historic maps can provide an important base on which to begin exploring past configurations of physical and cultural space, they should not be the only source to consult. In order to best take advantage of a robust historic GIS database, the researcher should be well-versed in the geography and history of the area being investigated. Working with these cartographic and textual descriptions of the Wye House plantation landscape, we can guide archaeological investigations. Furthermore, once excavations have been completed, this new archaeological data can be added to the hGIS database as an additional means to ground-truth historic documentation and as a research tool. GIS can be used to interpret archaeologically-excavated features as well as guide and shape archaeological survey and excavation. The survey and excavation generated as a result of the creation of this hGIS database and described below are not the first archaeological, architectural, or historical investigations at Wye House; 246 however, they can be seen as a part of an evolving research design spanning the entire twentieth century and into the present that tracks along with the development of the field. Previous Historical and Archaeological Research at Wye House The earliest historical engagement with Wye House took the form of a series of articles, chapters, and entries in larger volumes that are largely descriptive accounts. Notably, these include John Martin Hammond’s Colonial Mansions of Maryland and Delaware (1914), a pair of articles by McHenry Howard titled “Lloyd Graveyard at Wye House, Talbot County, Maryland” (1922) and “Wye House, Talbot County, Maryland” (1923), both appearing in Maryland Historical Magazine, Katherine Scarborough’s Homes of the Cavaliers (1930), and Alice Lockwood’s Gardens of Colony and State (1934). Some of these accounts, especially the early ones, include flowery and poetic descriptions of the house and surrounding landscape. Others, mainly those produced toward mid-century, begin to include technical architectural descriptions of the structure. Also included in these early works are brief biographical sketches of prominent figures to live at Wye House and chronologies of the major events and building episodes to occur at Wye House. Two similar traditions or motivations can be seen underlying and guiding these early efforts at documentation. The first is a romantic engagement with the past as a simpler, bygone time worthy of admiration, emulation, and remembrance. The second is recognition at the time that these historic houses, structures, and places were rapidly being lost to modernity. These two sentiments were probably at least 247 partially responsible for the wave the early historic preservation starting in the 1930s, including the Historic American Buildings Survey as well as the formation of some of the earliest historic districts in the country. Forman’s extensive fieldwork and documentary research at Wye House is most prominently featured in his 1967 Old Buildings, Gardens, and Furniture in Tidewater Maryland, but can also be found in some of his other publications, including Tidewater Maryland: Architecture and Gardens (1956) and Maryland Architecture: A Short History from 1634 through the Civil War (1968). Forman’s body of work and the article published by J. Donnell Tilghman in Maryland Historical Magazine in 1953, simply titled “Wye House” are good examples of the nature of architectural and archaeological inquiry at mid-century. Unfortunately, from reading Forman’s work at Wye House, it is not entirely clear what methodology or data went into the creation of his plan of Wye House produced as the endpapers of Old Buildings. However, it is clear that he spent time at Wye House to produce measured drawings of the Greenhouse for the Historic American Buildings Survey (HABS) and other architectural observations that formed the basis for his publications. The map could not have been produced without being out in the formal landscape surrounding the house and on the Long Green without surveying equipment (or at the very least a compass and a good sense of his own pacing). A note in the lower right corner indicates Forman is using the 1784 plat at least as a cross-reference to indicate which buildings were present by that date. By providing the locations and names of structures indicated to be “gone” by the time of his fieldwork in the 1950s and 1960s, we can deduce that he has arrived at this 248 knowledge by either walking the plantation to find structural ruins like those seen in the Dallin aerial photograph and still protrude from the ground in one or two spots today, or by talking to people then at Wye House who remember the locations and names of these structures. While primarily interested in the main house and its formal garden, Forman did pay brief attention to those areas of the plantation occupied by those enslaved there such as quarters, work buildings, fields, and cemeteries. From reading his books, it is clear that these early efforts drew on a mix of historical research, architectural investigation, limited archaeological excavation, and oral histories from local people who had seen many of these buildings before they were removed from the landscape, which makes his early work at Wye House useful to consult. In the 1990s. relatively limited excavations were conducted by ACS Consultants and the Lost Towns Project in conjunction with planned renovations and an expansion of the Captain’s Cottage or Captain’s House (Kerns and Gibb 1998). This survey revealed significant ground disturbance around the structure, including a large depression located immediately to the south. They agree with Forman’s assessment that a large brick building once stood here; however, based on a lack of seventeenth-or-early-eighteenth-century artifacts amongst the material recovered, could not concur with the circa 1660 date Forman provides for the missing structure. The most recent phase of archaeology at Wye House began by Archaeology in Annapolis in 2005. For more information regarding these excavations, please consult Phase II Archaeological Testing on Wye Greenhouse (18TA314), Talbot County, Maryland, 2008 (Blair, Cochran, and Duensing 2009); Phase II Archaeological 249 Testing on the Interior of the Wye Greenhouse (18TA314), Talbot County, Maryland, 2009 (Blair and Duensing 2009); Archaeological Excavations on the Long Green (18TA314), 2005-2008, Talbot County, Maryland, 2009 (Blair, Duensing, Cochran, Kraus, and Gubisch 2009); Archaeological Mitigation of the Great House Front Steps (18TA314), Talbot County, Maryland, 2010 (Tang and Knauf 2010); and Archaeological Excavations at the Middle Building (Locus 2) on the Long Green (18TA314), 2006-2010, Talbot County, Maryland, 2010 (Tang 2010). The work on the South Long Green and the East Cove described below can be found reported in Phase I and II Archaeological Testing of Two Structures at Wye House (18TA314), Talbot County, Maryland, 2-Story Quarter and Brick Row Quarter, 2011-2014 (Skolnik and Pruitt 2019) with the initial Phase I shovel test survey already published as Shovel Test Survey at Wye House (18TA314), East Cove / South Long Green, April 2011 (Skolnik 2011). Most of the archaeological work conducted by Archaeology in Annapolis between 2005 and 2014 was carried out in areas of the plantation in which the enslaved would have lived and worked; however, limited testing has occurred on the main house itself. In July 2010, three units were placed around the front south steps from the main house’s center block after they were removed for renovation and replacement (Tang and Knauf 2010). The goals of this work were to determine the age of the earlier brick stairs revealed beneath the current stone steps, to determine what impact to the archaeological record this restoration work would have, and to mitigate this impact. The brick stairs revealed under the stone steps were dated to the 250 1780s, which corresponds to the known date of construction of this portion of the main house. Archaeological work in and around the Greenhouse was conducted in 2008 and 2009, with limited additional testing in 2012. In October and November 2008, seven units were excavated outside of the Greenhouse and inside the rear shed additions in anticipation of foundation stabilization work (Blair, Cochran, and Duensing 2009). This work continued in July 2009 when three additional units were placed within the main room of the Greenhouse (Blair and Duensing 2009). The goals of these excavations were to assess the archaeological integrity of the Greenhouse and surrounding area, to evaluate the potential effects of planned construction work, and to investigate the architectural development of the Greenhouse and its use by the Lloyds and the enslaved who lived and worked in the structure. Historical evidence in the form of an inventory taken of the estate of Edward Lloyd III at his death in 1770 and a 1783 tax list suggests the main block of the Greenhouse was constructed sometime between those two dates and archaeological evidence from these excavations agrees that the Greenhouse was built in the 1770s while the wings and hypocaust were not added until the mid-1780s. Archaeology in Annapolis excavators encountered stratified deposits in these units, showing the Greenhouse retains much of its archaeological integrity, including evidence of the building’s construction, modification, and use as a greenhouse. These include scaffolding postholes, construction surfaces, builder’s trenches, and buried walls. They also uncovered evidence that may point to an older, previous garden structure standing where the present greenhouse stands now, perhaps dating to the 1760s or 251 even earlier. Forman also notes that the brick that comprises the rear wall of the present Greenhouse is larger than brick found in late eighteenth-and-nineteenth- century construction elsewhere at Wye House and is of the same size as that used in the earlier Captain’s Cottage, suggesting that portions of the Greenhouse may be contemporaneous with that earlier structure. Furthermore, artifacts recovered from the shed addition on the Greenhouse’s north side reveal it was occupied as a domestic space for the enslaved between approximately the 1790s and the 1840s. During these 2008-2009 Greenhouse excavations, soil samples were collected from inside the main room of the Greenhouse as well as from the quarters located in the shed additions on the north side of the structure and sent to the Andrew Fiske Memorial Center for Archaeological Research at the University of Massachusetts Boston for specialized pollen analysis. Of interest to researchers were the identification of plants grown in the Greenhouse, the identification of plants used by the enslaved in the attached quarters on the rear of the structure, and regional historical environmental reconstruction. Susan Jacobucci and Heather Trigg identified pollen from 128 different botanical taxa, including from naturally occurring vegetation, plants likely grown in the Greenhouse, and plants probably brought into the quarters by the enslaved for their own use (Jacobucci and Trigg 2010). They note a marked difference between pollen levels recovered in the main block of the greenhouse and pollen levels recovered from the quarter behind the structure. In contrast to the main room of the greenhouse, the quarter was a relatively small, enclosed area without the large openings that would have been left open during parts of the year. For this reason, they conclude pollen recovered from contexts from the 252 main block of the greenhouse are probably more representative of regional, windblown, background pollen assemblies than pollen recovered from contexts from the quarter, which were probably introduced to the quarter and used on site. From the Greenhouse, Jacobucci and Trig identify pollen from the following taxa known to be non-native to the area or known to have been cultivated historically in greenhouses: Caryophyllaceae (Pink family, including carnations, pinks, and sweet William), Citrus (oranges, lemons, limes, citrons, and shaddocks), Crocus (commonly used in gardens), Daphne (indigenous to Europe and Asia), Geranium (indigenous to the region but its pollen is not wind-borne), Iris, Liliacae (Lily family, used commonly as ornamentals), Musaceae (tropical plants from African and Asia, including bananas, plantains, heliconia, and bird of paradise), Nymphaea (pond lilies), Phlox (indigenous to the region and used for both decorative and medicinal purposes), Rosaceae (Rose family, made up of approximately 1200 species, including burnet, strawberry, peaches, and roses), and Saxifraga (flowering plants found locally and known to be used prehistorically for both food and medicine). Musaceae and Citrus have no known naturally-occurring wild species and their presence in the Greenhouse assemblages reflect active human cultivation. From the quarter, they also identified pollen from the following taxa, notable for their known uses or non-local origin: Asarum (wild ginger), Brassicaceae (Mustard family, including pepperweed, broccoli, and mustard), Lilaceae, Musaceae, Physalis (ground cherry is a member of the Nightshade family, which also includes tobacco, tomato, eggplant, and potato), Poaceae (European cereals), Rosaceae, Sagittaria (arrowhead), and Vaccinium (blueberries and cranberries). 253 Together, the pollen assemblages recovered from the Greenhouse and the quarter represent agricultural, dietary, and medicinal uses of plants in and around these space as well as background regional botanical activity from the late eighteenth century into the present. Because of the difference in both use and architectural structure of these two spaces, the different levels of pollen recovered from each can highlight these different uses of plants. In conjunction with this work in and around the Greenhouse, a geophysical survey was conducted in the fall of 2009. Four locations, including one around the Greenhouse, were surveyed with ground penetrating radar (GPR), one of which was also surveyed with a magnetic gradiometer. This survey revealed numerous subsurface anomalies, including suspected structures, plantation driveways, natural features, and even prior archaeological investigations (Haley 2009). The areas surveyed include in and around the Greenhouse, in the area of the Middle Building on the north Long Green, across the plantation driveway north of the Captain’s Cottage, and along the western edge of an agricultural field east of the cove. In the remotely-sensed data from these survey areas, modern and historic roads are clearly visible, as are other anomalies possibly representing modern utilities and buried historic architecture. However, Haley warns that without excavation, it is difficult to discern these types of features from prehistoric cultural features and natural targets and these results should be used as a guide for future archaeological survey and not as a final conclusion in these areas (Haley 2009:10-11). Results of note from this survey include a small rectangular anomaly near the southeast corner of the Greenhouse that may represent architectural remains and the lack of any brick 254 foundation where Forman’s map suggests there should be one in the field east of the cove. In addition to the surviving Greenhouse, several additional garden structures are known to have been at Wye House. As a follow-up to this geophysical survey and its results, limited Phase II testing was conducted at one of the anomalies identified just southwest of the Greenhouse, believed to be a 16 foot long by 16 foot wide hothouse. Two units were excavated in this area, in which a brick foundation was discovered and artifacts including flower pot fragments, window glass, and disarticulated brick support this interpretation (Pruitt 2013). Archaeology in Annapolis also conducted a series of excavations on the area of the plantation known as the Long Green. This name is used by Frederick Douglass to describe the work area on the plantation in which the enslaved lived and worked. While still built as part of an overarching layout or plan within the plantation, this area can be contrasted to the formal layout of the main house, Greenhouse, gardens, and other dependencies that make up the plantation core. This semi-structured space stretches along the plantation driveway to the east of the house, greenhouse, and gardens, on both sides of a shallow cove in Lloyd’s Creek. Between 2005 and 2010, three structures were identified and excavated on the west side of the cove along the plantation driveway and one standing structure was investigated on the east side of the cove. Prior to the work conducted on the 2-Story Quarter and Brick Row Quarter described below, at least 180 shovel test pits (STPs) were excavated across the Long Green (in both the north area where the North Building, Middle Building, and Tulip 255 Poplar Building were identified), in the south area (mainly to the north of the split in the driveway to the Red Overseer’s House and east of the split in the driveway to the main house, but also in an extremely limited capacity across from the barn near the 2- Story Quarter), and on the east side of the cove near the Red Overseer’s House. These Shovel Test Pit surveys were performed as a part of seasonal fieldwork with the Archaeology in Annapolis summer field school and the results guided and informed excavation strategy between 2005 and 2010. In 2005 and 2006, fourteen test units were excavated around a large Tulip Poplar tree at the southern edge of a small hill on the northern part of the Long Green in order to investigate a structure identified there (Blair, Duensing, Cochran, Kraus, and Gubisch 2009). Locus 1, or the Tulip Poplar building is the southern-most area investigated on the northern part of the Long Green, and is characterized by a park- like setting of mature trees and grassy lawn on the east side of the plantation driveway above the cove. The Tulip Poplar building’s foundations measure 16.5 feet long by 16.5 feet wide and there is a brick chimney fall associated with these remains. The artifacts recovered suggest an occupation date of between the mid-eighteenth century and the early nineteenth century. Prehistoric stone tools and ceramics were also encountered in the lower stratigraphic levels of these excavation units and a Native American human burial, probably from the Early or Middle Woodland period (1000 B.C.E. to 500 C.E.) was partially uncovered in Test Unit 4, several feet north of the north wall of the structure. The natural and cultural stratigraphy within and around the Tulip Poplar building is largely disturbed by its namesake tree, which took root sometime in the 256 past 100 years. While the brick foundation is intact, associated stratigraphy is full of root burrows, mixing many contexts. Archaeologists estimate this structure was used primarily sometime after the first quarter of the nineteenth century until the Civil War, but earlier mid-and-late-eighteenth-century material was also recovered from these contexts, leading them to suggest a possible second, earlier structure. Between 2006 and 2010, seventeen units were placed several feet to the north, in and around the Middle Building in Locus 2 of the Long Green (Tang 2012). This structure appears to have been a wood frame building measuring approximately 40 feet long by 30 feet wide. It is believed that this building was built sometime around the end of the seventeenth century or the beginning of the eighteenth century. This area was initially chosen for testing because several bricks were observed protruding through the ground surface in this area. Under this structure, two cellar features were excavated, including a root cellar and a cold cellar. While not in use at the same time, both could have been used for food storage. The foundations of this structure are somewhat unusual in that portions of it are continuously-mortared brick, dry-laid brick, and post-in-ground construction. In conjunction with architectural renovation of the structure in 2006, archaeological monitoring work was carried out under the Red Overseer’s House (Locus 4) on the east side of the Long Green. Douglass describes this extant-early nineteenth-century residence in his autobiographies, writing “There was the little red house up the road, occupied by Mr. Sevier, the overseer” (Douglass 1855:66). Two units were excavated; one in front of the chimney on the north side of the house and one near the chimney on the south side of the house. Dendrochronological analysis of 257 wood from the structure support a construction date of 1815 (Worthington and Miles 2007) and artifacts recovered support this early nineteenth-century date (Kraus 2006), which is further supported by Douglass’ descriptions of it in his autobiographies. Of note is that this structure dates to a period after the plantation reorganization believed to have occurred in the early 1780s and represents another phase of change and growth on the plantation. During the 2006, 2007 and 2008 field seasons, thirteen units were placed in the northern section of the northern Long Green to evaluate exposed brick foundations identified during a 2005 walkover survey (Blair, Duensing, Cochran, Kraus, and Gubisch 2009). This building in Locus 3, referred to as the North Building, is thought to measure 30 feet wide by 40 feet long (Figure 6.14). In the site report, excavators conclude that the structure had undergone multiple phases of use and that these are the remains of a multiple-story brick building, likely the one described by Douglass as being, “perched upon a hill…a very tall, dilapidated, old brick building”. It was considered that this foundation belongs to the Corn Crib known to have stood in the area and documented by as a part of the Historic American Buildings Survey (HABS) in 1936 (Figure 6.15) and captured in the 1930s Dallin aerial photograph (Figure 6.16), but this hypothesis was ultimately rejected by Archaeology in Annapolis archaeologists at the time. Being able to combine multiple spatial datasets in our hGIS allows us to revisit and re-inform previous archaeological interpretations as well as to guide future excavations. 258 Revisiting Previous Archaeological Interpretations: the Wye House Corn Crib The potential interpretation that the North Building excavated between 2006 and 2008 is the Corn Crib documented for the Historic American Building Survey by E. H. Pickering in 1936 was initially questioned based on an inability at the time to determine the location of the Corn Crib on the landscape and ultimately rejected based on an estimation of the width of the Corn Crib from the HABS photograph to between 16.5 feet, or one perch, and 22 feet, far short of the estimated 30 foot wide foundation encountered archaeologically. To reach this measurement, archaeologists estimated the width of the central door at between three and four feet and used that to 259 measure the remainder of the structure in the photograph (Blair, Duensing, Cochran, Kraus, and Gubisch 2009:45). Not discussed in the site report, but HABS documentation on a card prepared in 1973 generally agrees with this estimation, and describes this building as being “approx. 20’ (three-bay front) x 30’, one-and-a-half stories, gabled roof, brick foundation; central passage with flanking storage bins.” While it agrees with the assessment made by archaeologists, this measurement being approximate would seem to indicate that it was not measured in the field and instead is an estimation produced after the fact, possibly in much the same manner as the estimate made by Archaeology in Annapolis archaeologists, along with the notation that the structure was by then demolished. 260 Exploring the idea that the North Building, excavated between 2006 and 2008, may be the Corn Crib documented in 1936, Archaeology in Annapolis archaeologists write: From the [historic] images it is difficult to determine with any accuracy where this building falls in relation to the current structures. The topography of the area has changed over the last 100 years between the changes in the coastline…and rapid growth of brush. These make a basic assessment of the images in relation to the landscape difficult. (Blair et al. 2009:39) This geographic uncertainty is precisely the kind of spatial problem that can be overcome by loading the archaeological site map of the North Building into our hGIS geodatabase along with modern aerial imagery, the historic Dallin aerial photograph, and Forman’s map of the Long Green. In doing so, it is clear that the 261 location of the corn crib as seen in the early twentieth-century aerial photograph, the building labeled “Corn H.[ouse]” seen on Foreman’s map, and the archaeological footprints of the recovered and documented structure all overlay each other in the same location. Furthermore, once the sources are georectified, the features seen in them can be measured, including the Corn Crib in the Dallin aerial photograph. Using the hGIS, the Corn Crib in the aerial photograph measures approximately 40 feet by 30 feet, the same dimensions as the North Building. Together, these two facts (location and size of the Corn Crib) are both fairly strong evidence that the North Building is the Corn Crib that was initially rejected without this GIS evidence to support it; however, with this strengthened hypothesis, we can revisit the reasons it was rejected initially. Given that this corn crib is an agricultural building and not a domestic structure, interpreting the opening in the central bay as a standard three-foot-wide door is problematic. Given that hogsheads are frequently more than three feet wide when turned on their side and rolled, this opening should be scaled to at least this measurement. By scaling this doorway to be larger, the two smaller openings on either side scale up to be closer to the standard width doors used more commonly on domestic structures. Furthermore, the angle at which the 1936 HABS photograph is taken complicates efforts to estimate length and width of the structure from the dimensions of the central opening. Because the southwest corner of the structure is closer to the position of the camera and the northwest and southeast corners are farther away, distance is not uniform or linear across these two facades in the photograph, and 262 objects in the foreground are larger than objects in the background. This makes estimating sizes difficult without taking this distortion of distance into account. While this effect is not as pronounced in the oblique Dallin aerial photograph, the effect is similar and still present. As such, any estimation that tries to extrapolate overall dimensions using this methodology should be used cautiously. Instead of using door width as a means to estimate building size, given their fairly uniform size, using the bricks visible in the photograph of the corn crib’s foundation as a means to estimate overall building dimensions may produce a more reliable measurement. While the brickwork in the Corn Crib’s foundation is irregular in places (a fact also noted by archaeologists in their excavation of the North Building, the foundation of which appears to be a combination of piers and brick walls with some irregular brickwork), in the 1936 photograph it appears to be somewhere between 52 and 54 brick lengths long and approximately 40 brick lengths wide, a length to width ratio of between 1:1.3 and 1.35, depending on which approximation is used. Before continuing, these numbers should be considered approximations because the brickwork is irregular in places and in others cannot be counted because the brickwork is either missing or obstructed by debris. In the case of both the length and width (south and west facades, respectively), a single course of bricks could not be counted, so an estimate was made by counting multiple courses or creating placeholders where the brick was obstructed. A building of dimensions 30 feet by 20 feet has a length to width ratio of 1:1.5 whereas a building of dimensions of 40 feet by 30 feet has a length to width ratio of 1:1.33, almost exactly the ratio observed in the 1936 photograph of the corn crib. 263 Furthermore, while brick sizes are not standardized historically (Crews 2005), we can again use an approximate brick measurement to determine the size of the structure seen in the 1936 HABS photograph. Modern modular bricks have a length of 7 5/8”, and allowing for a 3/8” mortar joint, 54 bricks by 40 bricks of this size produces a building 36 feet long by 26.67 feet wide. Modern standard bricks are 8” long, and allowing for a 3/8” mortar joint, 54 bricks by 40 bricks of this size results in a building 37.69 feet long by 27.92 feet wide. As frequently noted, modern bricks are smaller than many bricks found in historic buildings, a fact supported by the size of the bricks found archaeologically at the North Building. While an exact measurement of the brick excavated as part of the North Building is hard to determine from field photographs, these bricks appear to measure around 9 long each when compared to the ruler visible in the frames used as a scale bar. Therefore, a building 54 bricks long by 40 bricks wide, each 9 inches long and allowing for a 3/8” mortar joint would measure 42.19 feet long by 31.25 feet wide. Using bricks as a way to measure overall building size, we arrive at the conclusion that the building in the 1936 HABS photograph of the Corn Crib at Wye House has the same length to width ratio and approximately the same dimensions as the North Building excavated by Archaeology in Annapolis between 2006 and 2008. Furthermore, in their rejection of the Corn Crib as the North Building, Archaeology in Annapolis archaeologists argue that “the area surrounding the building in the photo is perfectly level with effectively no variation in ground elevation.” However, examining the HABS photograph reveals this analysis to be incorrect. Comparing the ground surface against the courses of brick in the 264 foundation shows at least approximately a foot of topographic relief between the northwest corner and center of the south wall of the building. Like with the estimation of the building from counting the number of bricks that make up its length and width, the irregular brickwork and missing and obstructed sections complicate this analysis. However, there are approximately five courses of brick between the lowest topographic point along the foundation (the northwest corner of the structure) and the highest topographic point along the foundation (at the second wall post from the southwest corner on the south façade). Modern standard bricks have a height of 2.25 inches. Adding a 3/8” mortar joint between each brick and multiplying by five brick courses gives a height difference of 13.125 inches between these two points on the foundation. Allowing for slumping after the building was destroyed and for the accumulation of soil since then, this seems to be close to the amount of topographic relief as indicated on the stratigraphic profiles drawn of this structure. Because of this hGIS analysis of the Dallin aerial photograph and the HABS photograph, we can revisit the interpretation that the North Building is not the Corn Crib. First, the analysis shows that the Corn Crib was standing in the same location as the North Building. No other overlapping building of this description was encountered during the excavations of the North Building and no other structures appear near here in any of the historic aerials and maps. Second, once georectified, features seen in the Dallin aerial photograph become measurable and the Corn Crib seen there measures approximately 40 feet long by 30 feet wide. Third, an understanding of the function of the corn crib as an agricultural work building would lead one to reject the interpretation of the opening in the central bay as only three feet 265 wide in favor of a width of at least four feet wide or wider. Fourth, door size is probably not a reliable way to estimate the size of a building because they are not standardized and the angle was not accounted for. Fifth, an estimate of the size of the building in the 1936 HABS photograph of the Corn Crib using the number of bricks in its foundation produces a figure that generally matches the dimensions of the Corn Crib as seen in the Dallin aerial photograph of the same time and the North Building as excavated by Archaeology in Annapolis between 2006 and 2008. Lastly, using bricks again in order to estimate the topographic relief seen in the HABS photograph, we can reject the claim that the terrain here is perfectly flat and instead roughly matches the terrain documented during the excavations of the North Building. Taken singly, none of these arguments is enough evidence to conclude whether or not the Corn Crib as seen in the HABS photograph and the Dallin aerial photograph is the same building as the North Building excavated by Archaeology in Annapolis between 2006 and 2008. However, these six points together, largely derived from hGIS analysis, strongly suggest the North Building is the Corn Crib. Locating Lost Structures: the 2-Story Quarter and the Brick Row Quarter Using historical GIS to locate structures which have disappeared from the landscape but were not still standing to be captured by early aerial photography can be a little more difficult than documenting buildings still standing and observable in this dataset. The following case study at Wye House will highlight some of the strengths and weaknesses of using hGIS as a part of an archaeological research program. 266 The two areas in which Archaeology in Annapolis has concentrated its excavations at Wye House are the Greenhouse (Orangery) and the Long Green. While the Greenhouse still stands, there are no extant structures on the Long Green. Previous work by Archaeology in Annapolis has examined the remains of several of these vanished structures on the Long Green; however, a pair of quarters as identified on the Henry Chandlee Forman map as the 2-Story Quarter and the Br[ick] Row Quarter had eluded researchers. While the North Building (shown above to probably be the Corn Crib), Middle Building, and Tulip Poplar Building on the northern half of the Long Green have been explored through archaeological testing, these two other structures, the 2-Story Quarter on the South Long Green and the Brick Row Quarter east of the cove have not. Because of the research potential held by these structures purported to be quarters for the enslaved, locating these structures is of importance to Archaeology in Annapolis and its ongoing research project at Wye House. Unfortunately, the size, scale, and line weight of Forman’s map, as well as the Long Green’s confinement to an inset in the corner of the map, mean that it is not accurate enough on its own to locate these structures without additional work. By combining this map with a historic 1847 United States Coast Survey map, the LiDAR-derived digital elevation model (DEM), historic aerial photographs, and other spatial datasets, the probable locations of these two structures can be identified and an archaeological research plan can be designed to evaluate these two sites. This hGIS analysis and subsequent Phase I shovel test survey and four seasons of Phase II test unit excavations have confirmed the locations of these two structures and their characterization as quarters. The results of these investigations at Wye 267 House between 2011 and 2014 are reported in Phase I and II Archaeological Testing of Two Structures at Wye House (18TA314), Talbot County, Maryland, 2-Story Quarter and Br[ick] Row Quarter, 2011-2014 (Skolnik and Pruitt 2019) and Shovel Test Survey at Wye House (18TA314), East Cove / South Long Green, April 2011 (Skolnik 2011). In total, 44 Phase I shovel Test Pits were excavated by University of Maryland, College Park graduate and undergraduate Anthropology students on April 8th and 9th, 2011 in an attempt to ground-truth the results of the initial historical GIS survey of the 2-Story Quarter and the Brick Row Quarter (Figure 6.17). Twelve Phase I shovel test pits were excavated around the suspected location of the 2-Story Quarter on the South Long Green and 32 Phase I shovel test pits were excavated around the suspected location of the Brick Row Quarter on the East Cove. In situ architectural elements were encountered on the South Long Green and in both loci architectural rubble and large concentrations of historic material strongly suggested the presence of additional archaeological features. Of particular note during these Phase I excavations were a section of mortared brick, lying on edge, in Shovel Test RQ/R3, initially interpreted as a fallen wall or chimney, brick path, or intact foundation wall (Figure 6.18), a significant quantity of historic artifacts in Shovel Test R3.5, initially interpreted as an episode of refuse dumping within the remains of the structure, probably dating to after its destruction, and bricks encountered in multiple shovel tests along Transect Y (Figure 6.19), parallel to the plantation road and perhaps related to the eastern wall of that suspected structure. From the distribution of historic ceramics, glass, architectural materials, and 268 other artifacts, as well as the apparent presence of in situ or mostly in situ architectural remains, it was determined that these two locations on the South Long Green and the East Cove were likely the locations of the two structures identified by Forman on his map as the 2-Story Quarter and Brick Row Quarter. Based on these 269 results, additional excavation was recommended with the stated goals of delineating the precise location and boundaries of these structures and, once located, to assess the archaeological integrity of these structures. Following the shovel test pit survey conducted in April 2011, four three-week seasons of archaeological excavation were conducted as a part of the Archaeology in Annapolis field school at Wye House starting in June 2011 and continuing until July 2014. In total, 25 test units were excavated across these two areas, 14 in and around the 2-Story Quarter on the South Long Green and 11 in and around the Brick Row Quarter on the East Cove (Figure 6.20). As expected, these excavations revealed the remains of two structures. At the 2-Story Quarter, some sections of intact foundation walls were encountered and there was evidence of multiple phases of construction and demolition. At the Brick Row Quarter, the architectural material encountered 270 suggests that this structure was initially more ephemeral and more lightly built than the 2-Story Quarter and is currently more poorly preserved than the 2-Story Quarter. Both of these structures appear to be the ones identified by Forman in his map of Wye House and both of them appear to have been identified on the 1847 US Coast Survey map that includes Wye House plantation. At the 2-Story Quarter on the South Long Green, there is evidence for two distinct building episodes; one earlier building with a foundation comprised of mortared brick and stone and one later building, probably frame structure, set on brick piers dug into the rubble of the earlier structure. The artifact assemblage from this early structure consists of primarily whiteware and ironstone, with lesser amounts of earlier pearlware and creamware, but very little that dates to before the plantation 271 was reorganized in the early 1780s. Some black basalt and white salt-glaze stoneware was recovered; however, most of the material dates to the early nineteenth century. Intact brick foundations of this structure were not encountered during these excavations, however, on the west and south sides archaeologists excavated deep, 272 brick-rubble-filled trenches which appear to be robbed foundation trenches (Figure 6.21). From the layer of material resting directly on top of the rubble of this structure, it appears this building was demolished sometime between the end of the Civil War and the late 1870s. This date fits with changing labor practices at Wye House following Emancipation. It is this earlier (c. 1780s-1870s) structure that is depicted on the 1847 map. It was also probably standing on the landscape when Frederick Douglass was enslaved here as a child. While intact brick foundations were not encountered, the location of the suspected collapsed brick wall or chimney on the east side of the structure and the robbed brick trench on the west side provides a rough estimate of the width of this structure at somewhere under 18 or 20 feet. Similarly, measuring from the southern edge of the robbed trench northward to where there is no longer archaeological evidence of a robbed trench (in roughly this location is a ditch 273 running in approximately the same orientation as this anticipated northern wall) produces a building length of approximately just under 20 feet. Cut into this brick rubble layer that dates to the destruction of this first structure is evidence of a second structure, probably a wood frame building set on brick piers. At least two brick piers survive at least mostly intact, and there is strong evidence for at least one more pier that has since been robbed, and at least two more possible pads on which former piers might have been set were encountered (Figure 6.22). From the incomplete set encountered, it would appear that these piers were initially set 6 feet apart and the second building would have measured approximately 20 feet long by 16 feet wide. While complete dimensions could not be found for the earlier structure, it was probably close to this size as well. 274 Furthermore, a stone foundation was also encountered at this location that may date to one or both of these buildings (Figure 6.23). The wall is even on its interior face, suggesting a stone-lined cellar under the building; however, from the units excavated, it is unclear how the wall articulates with either the earlier brick wall foundation or later brick pier foundation. The brick rubble around this location is of two distinct types, corresponding to the two different buildings here. The lower, earlier brick rubble is comprised of larger brick fragments, bats, and even whole bricks. The significant quantity of this lower, larger brick rubble as well as the section of mostly-intact wall fall near the east side of this structure and the deep robbed trenches filled with this rubble are evidence of this mortared foundation. Resting directly on top of this rubble layer is a deposit of material dating to between the 1860s and 1880s (see below for more on this deposit). 275 This gives a terminus ante quem (date before which) for the destruction and robbing of this earlier foundation and building as well as a rough terminus post quem (date after which) for the construction of the second building on top of this rubble. The second, upper layer of brick rubble resting on top of the first is much smaller and less frequent than the larger rubble below. Because of the change in size and frequency, as well as the presence of brick piers and robbed brick piers, this smaller rubble most likely dates to the second, wood frame structure on this location. Largely outside the scope of this chapter detailing the use of hGIS to locate historic resources on a plantation are at least three collections of material recovered along the western and southern edges of the 2-Story Quarter that may have spiritual or religious origins and connotations. These have been reported elsewhere, notably by Leone and others (2018) in Historical Archaeology and by Pruitt (2015:159-165) in her dissertation. The most recent of these deposits was placed near the center of the western wall, directly on top of the architectural remains of the later structure, probably in the early twentieth century. It consisted of a clothes iron, glass bottle base and body, other glass fragments, and an oyster shell. The size, clustering, and location of these objects compared to a lack of other similarly sized, or clustered objects at this depth made them stand out as anomalous. A second deposit was encountered beneath the destruction and fill layers related to the second structure and directly on top of the rubble and fill layers related to the destruction of the first structure sometime between 1860 and 1880 (Figures 6.24 and Figure 6.25). These materials were mostly circular in shape, such as an iron 276 cart wheel, several glass bottle bases, canning jar lids, and several pieces of round stone with flattened tops and bottoms, or non-circular objects arranged in a circle, all arranged in a rough line on top of the robbed foundation trench and all along the western edge of this structure. This deposit would have sat on top of the ground surface underneath the floorboards of the second, wood frame structure that stood here. Given the topography of this location that slopes down to the east, the structure’s location along the plantation driveway, and using standing wood-frame- on-brick-pier architecture at Wye House as a guide, the floor of the structure along the western edge would have probably been flush with or only slightly raised from the ground surface while the eastern edge would have been raised up higher off the ground on piers. This would have caused this deposit to have been largely 277 inaccessible from the outside of the structure and only accessible from above through the floorboards, if at all. A similar horizon of flat or round artifacts was not encountered on top of the earlier destruction layer where it was present elsewhere in this locus on the South Long Green. Both of these facts suggest this deposit was not randomly accumulated refuse that collected under the outside edges of this structure. Furthermore, its presence only under the suspected entrance to this structure reinforces the argument it was deliberately placed here with these intentions. A potential third, and earlier, deposit of material was found contained with the brick and mortar rubble fill dating to the destruction of the first structure. It consisted of many iron tools, mostly complete and unbroken. These included a plow, shovel, 278 hoe, ax, wheel hub, and several horseshoes. Given that the date of the destruction of this earlier structure corresponds roughly with the end of the American Civil War and Emancipation, it may be that these iron tools were simply discarded as a result of changing labor patterns on the plantation and an increase in the use of mechanization for agricultural work and not placed with spiritual intent. The intent of this dissertation is not to analyze the spiritual significance of these objects nor to determine whether they were placed deliberately. Regardless of whether these objects and deposits were the result of deliberate placement or the result of the random accumulation of refuse from across the plantation, the symbolic connection of these objects to West African religious and spiritual beliefs and modern African American Christianity is intriguing and warrants additional attention. For more on these deposits, see Leone and others, where it is argued these last two deposits contain symbolism embedded in both many eighteenth-and-nineteenth- century Western African religious and spiritual beliefs as well as modern forms of African American Christianity such as the use of iron and the wheel. Any future excavation planned at the 2-Story Quarter should attempt to better define the architectural dimensions of the first or earlier building here, determine the original location and function of the collapsed brick wall or chimney on the west side of the structure, better understand how the stone foundation or cellar articulates with either or both of the earlier robbed brick foundation or the later brick pier foundation, and address the articulation of this building with the larger landscape. It should also seek to identify any other potential deposits like the ones encountered here already as well as consider whether to explore the unexcavated portions of the same. 279 While the remains of the 2-Story Quarter on the South Long Green (probably two successive structures on the same location) are evident in the archaeological record, the remains of the Brick Row Quarter less than 100 yards to the east in the East Cove are much more ephemeral. As revealed by these excavations, the Brick Row Quarter appears to be more lightly built than the 2-Story Quarter and the archaeological evidence of these less-robust foundations does not appear to have survived as well as the that of the 2-Story Quarter (Figures 6.26 and 6.27). A twentieth-century agricultural silage pit or trench has been cut across the western portion of the suspected location of this structure, possibly disturbing much of the archaeological footprint of this building. Furthermore, no subfloor pits or cellars were discovered under or around the Brick Row Quarter. Given its location on the plantation on the other side of the cove from the main core, adjacent to an agricultural 280 field, and its elongated three-part plan as drawn on Foreman’s map, it is a reasonable assumption that this structure was intended to house enslaved agricultural laberors working in the adjacent fields whereas those quarters closer to the core of the planation were for those enslaved workers working in the main house and its surrounding outbuildings and their industries. Given the ephemeral nature of the archaeological remains of this structure, it is difficult to determine its dimensions. Archaeologically, its dimensions can be estimated to be between 36 and 54 feet long and approximately 16 feet wide. As with the 2-Story Quarter, the cultural material recovered from the Brick Row Quarter seems to be primarily early-to-mid-nineteenth-century ceramics like whiteware and ironstone; however, there is some late eighteenth-century material such as creamware and pearlware but very little material that dates to earlier. This supports a construction 281 date in line with the reorientation of the plantation in the early 1780s and a destruction date of after the Civil War and certainly by the late 1920s when the Dallin aerial photograph reveals no structure in this location. As shown by this work at Wye House, it is possible to combine historic maps, modern elevation data, and various types of archaeological data to locate archaeological remains no longer visible or evident surface of the ground. With the ability to locate lost structures and other historic resources remotely with a high degree of accuracy, one might ask, “Why do archaeologists need to dig anymore?”. There are several reasons why a hGIS-based site identification phase should be incorporated into a larger program of archaeological research and the results ground- truthed. Of fundamental consideration is that an hGIS can only be as good as the kinds of data that are loaded into it. If the initial historic cartographic data that one is relying on to determine the locations of historic structures is wildly inaccurate, warped or distorted, or misinterpreted by the analyst, the resulting location of that structure will be incorrect. Likewise, if a historic structure is not depicted on any historic cartographic source nor is it detectible in the kinds of remotely sensed aerial datasets frequently used in hGIS databases, there is no way to locate that structure archaeologically. Archaeological excavation based on the results of an hGIS site survey serve to ground-truth the historic maps and datasets that produced the initial prospective locations. Over time, by ground-truthing these results against archaeological survey data as well as against still-standing structures also visible on maps or aerial photographs, one can better assess the accuracy and reliability of those datasets for future use. 282 Whereas a blank spot on a map is unable to tell a researcher what, if anything, once stood at that location, archaeological excavations in that location prompted by other pieces of spatial evidence help to fill in that cartographic blank spot. In the case of plantations throughout the American South, these blank spots extend outward from the main plantation house into the agricultural fields and forests and rivers that make up the larger plantation and inter-plantation landscape. As it has been shown to be successful with this case study at Wye House, this technique can be used elsewhere with similar datasets, hopefully to the same effect. Historic maps, early aerial photography, period textual descriptions, and archaeological data can all be combined into a single spatial dataset in GIS in order to create new geographic and historical knowledge. 283 Chapter 7: Mapping the Enslaved Landscape Through Frederick Douglass: Integrating Qualitative Data into Historical Geographic Information Systems (hGIS) Frederick Douglass and Wye House This chapter will use the quantitative tools of GIS to map qualitative data extracted from the autobiographies of Frederick Douglass. This will take the format of a kind of textually annotated map or spatially annotated text wherein qualitative descriptions of the plantation landscape are pinned to quantitative descriptions based in GIS and mapping techniques. This will embed Douglass’s descriptions of Talbot County and Wye House from the 1820s and 1830s within the hGIS framework developed in Chapters 4, 5, and 6. Thus far, this dissertation has tended toward the quantitative, coordinate-based version of landscapes and landscape analysis. There are several reasons for this methodological focus. It is partially because the kinds of datasets used above such as maps, aerial photographs, and historic censuses more readily lend themselves to these kinds of analysis. It is also largely because of the specific tools used to analyze them, namely Geographic Information Systems, are thoroughly grounded in quantitative computer science and spatial data analysis. It is important to highlight here that far from being neutral, value-free, impartial records of spatial knowledge, maps and GIS are culturally-specific forms of knowledge, embedded within a cultural framework that prioritizes a specific spatial worldview and in turn reinforces that same spatial worldview (Elwood 2008; Harley 1988; Kwan 2002; Pavlovskaya 2018; Schuurman 1999; Wilson 2017). 284 Given the multivocality of landscape, that is, the ability for it to simultaneously communicate multiple messages, we should be open to the idea that different groups of people can look at, move through, and understand the same space differently. This is the thesis advanced by Dell Upton in two pieces on racial plantation landscapes (Upton 1984, 1990) described in this dissertation. In the following two chapters, we can build on the historical GIS work presented in previous chapters and turn to the writings of Frederick Douglass to incorporate his descriptive accounts of place and geography as well as his relationship to this landscape into a historical GIS dataset and test Upton’s hypothesis. With the kinds of historical and spatial datasets described in previous chapters already loaded into a GIS, one is able to spatially annotate Douglass’s autobiographies in order to ground them in specific places in the present as well as conduct specific spatial inquiries into the historic landscape as inspired by Douglass. This has many applications for researchers working on historic sites or with historic documents, in preservation and management contexts, or in the realm of historic tourism with the creation of historic trails or highway tours. Douglass’s writings can be used to help us navigate the physical and social landscapes of early nineteenth-century Talbot County and Wye House and are an invaluable resource to guide scholars as they follow in his footsteps. Before proceeding, it should be noted that when Douglass was describing Wye House in his first autobiography, Narrative of the Life of Frederick Douglass which was written in 1845, he was already 27 years old, writing about a place at which he had lived for only a few years as a child and had not seen in over close to 20 years. His 285 recollections of his brief childhood home are certainly shaped by the physical and temporal distance from which he was writing. Furthermore, Douglass writes with a specific anti-slavery message in mind, certainly coloring his account. When attempting to pin these locations to a map, one should keep in mind the age and brief time spent there by the author as well as his self-admitted biases. That being said, with only a few exceptions, Douglass describes Wye House in a manner that is readily recognizable to a visitor to the plantation in the present. Keeping these warnings in mind, we can still turn to Douglass to help locate ourselves on the landscape at Wye House and fill in our map with specific textual details. This textual layer of cultural geographic knowledge can be added to our spatial understanding of the plantation landscape. This is not a novel project. Others have reoriented plantation landscapes from the perspective of the enslaved as well as other African American landscapes, notably Whitney Battle-Baptiste at the Hermitage (2007, 2010), Alemy and others in Setauket, New York (2017). Many organized tours at heritage sites such as Mount Vernon or Colonial Williamsburg seek to reorient the visitor in order to discuss slavery or African American heritage at the site. Perhaps Douglass’s best description of Wye House and its environs can be found in Chapter 4, “A General Survey of the Slave Plantation” in My Bondage, My Freedom (1855); however, Douglass recycles material between publications and relevant passages can be found throughout the opening chapters of all three of his autobiographies. With the hGIS database created in the previous chapter, we can pin Douglass’s childhood recollections to the physical landscape and reinforce the accuracy of his account as well as add a layer of textual description to our 286 understanding of the historic landscape. Therefore, the following is a spatial annotation of Chapter 4, “A General Survey of the Slave Plantation” from Douglass’s 1855 My Bondage, My Freedom (Figure 7.1). 287 In his account, Douglass provides the names and functions of the buildings that comprise the plantation landscape. At Wye House, he describes the “great many houses, human habitations full of the mysteries of life at every stage of it”. These include the still-standing Red Overseer’s House, “the little red house up the road, occupied by Mr. Seveir, the overseer” and “the long quarter”, a building he places “a little nearer to my old master’s” and describes as “a long, low, rough building literally alive with slaves of all ages, sexes, conditions, sizes, and colors” (all from Douglass 1855:66). On Henry Chandlee Forman’s map of Wye House, Forman depicts a long building as three rectangles laid end to end and is labeled as a “Br.[ick] Row Quart.[er]”. Archaeology in Annapolis, guided by these descriptions, excavated in this location and was able to identify the ephemeral remains of a roughly-built long brick structure (see Chapter 6, Skolnik and Pruitt 2019, and Skolnik 2011). Cryptically, Douglass also describes “a tall, dilapidated old brick building, the architectural dimensions of which proclaimed its creation for a different purpose, now occupied by slaves, in a similar manner to the long quarters” (Douglass 1855:43). He locates this building “perched upon a hill east of our house” (Douglass 1855:43). Here, in a way, Douglass is trying his hand at being an archaeologist, deducing past cultural activity from then-present configurations of material. His reader, however, is left guessing the meaning of this passage on two fronts. First, while he describes this building as being used as a quarter, its original function seems known to Douglass (or at least he suspects it) but does not bother to share this with his reader. Given the context of the statement and the emphasis on the building’s size, the possibilities include a large agricultural building like a barn or stable, an antecedent to the current 288 greenhouse, or the former plantation house that would have been replaced by the current Wye House when it was built in the last quarter of the eighteenth century. Forman believed the Captain’s Cottage to be the remnant kitchen wing of an earlier Wye House that he locates adjacent to the south of the Captain’s Cottage and just west of the 2-Story Quarter also described in Chapter 6. The second area of ambiguity here is the location of this building. While at Wye House, Douglass writes that he was living in the Captain’s House or Cottage, located about 350 feet northeast of the main house. Immediately to the east of the Captain’s House is a low-lying area which would have been a shallow cove prior to silting in and beyond this the terrain rises again and opens to an agricultural field. However, located at the top of this slight rise is the suspected location of the long quarter described by Douglass above and which Archaeology in Annapolis has confirmed the presence of a structure. The dimensions and ephemeral nature of these remains suggest they are of the long quarter, a “long, low, rough building” described by Douglass and the Brick Row Quarter described by Forman. From this location, the ground continues to rise very slightly to the east and at the highest point on this peninsula is a small, wooded area marking the plantation’s graveyard for the enslaved. No building is known to have stood in this location, making Douglass’s description of “a tall, dilapidated old brick building” as being “perched upon a hill east of our house” unclear. The highest point here only rises approximately 13 feet above sea level, and only approximately 6 feet above this landform once it rises off the water (Figure 7.2). Perhaps the answer to this question is that this slight topographic rise in front of the Brick Row Quarter simply looked much larger to young Douglass. 289 In addition to these buildings, Douglass casually refers to “numerous other slave houses and huts scattered around in the neighborhood, every nook and corner of which were completely occupied” (Douglass 1855:66). The historic 1847 Coast Survey map of Wye House (Figure 7.3) as well as the 1784 plat traced by Forman and 290 Forman’s map of Wye House all support Douglass’s description of numerous other buildings scattered around the plantation landscape. The 1847 Coast Survey map shows at least 17 other buildings comprising the core of Wye House plantation in addition to the main house, greenhouse, Captain’s House, Red Overseer’s house, long 291 quarter, and the tall, dilapidated brick building. Douglass’s list of these include most of the buildings one might expect on a working agricultural plantation, including “barns, stables, storehouses, tobacco-houses, blacksmith shops, wheelwright shops, [and] cooper shops” (Douglass 1855:66). In his initial plantation survey published in 1845 (and only slightly reworked in subsequent editions of his autobiography), Douglass describes his master’s house on the plantation, the house Douglass lives in (known today as the Captain’s Cottage), in just one sentence, writing “Old master’s house, a long brick building, plain but substantial, was centrally located, and was an independent establishment” (Douglass 1881:28), before moving onto Wye House itself. Douglass also describes the plantation windmill, “with its wide-sweeping white wings” (Douglass 1881:28), away from the core of the plantation on a point of land between the Miles River and the Wye River. This windmill is also depicted on the 1847 Coastal Survey map as well as on the 1784 plat of Wye House traced by Henry Chandlee Forman (Figure 7.4). Douglass recalls that the mill was “a commanding object to a child’s eye” and that he “spent many hours here watching the wings of this wondrous mill” (Douglass 1881:28). Like many other plantations in Talbot County, Wye House is surrounded on many sides by water. The Wye River flows by to the north and west to where it meets with the Miles River on its way into the Chesapeake Bay. To the east, Wye House is bounded by Lloyd Creek (which was known in the nineteenth century as Lloyd’s 292 Creek), a small tributary of which divides the core of the Wye House plantation with some of its outlying fields and quarters to the east. This silted-in cove still provides a natural boundary on the modern landscape at Wye House, dividing the residential and historic core of the plantation from the agricultural farm. Douglass provides a 293 colloquial name for Lloyd Creek and the tributaries off Wye House—the “Swash”. He also places in the “Swash” the Sally Lloyd, the plantation sloop responsible for taking the plantation’s and its neighbors’ agricultural products to market in Baltimore and bringing back anything required by the plantation. It is interesting to note that Douglass ascribes a fixed geographic location to an object whose purpose for existing is to be capable of motion. The sloop Sally Lloyd is also noteworthy to mention here in the context of the hGIS database described above because in the Dallin Aerial Survey photograph of Wye House (Figure 7.5), the hull of a sloop-like ship can be seen laying quietly, just as Douglass described, stripped of its tackle, fittings, rigging, and mast, in a cove at the edge of Lloyd’s Creek, exactly where Douglass says he saw 294 it a century earlier. It is not clear that this vessel is the Sally Lloyd; however, from the limited evidence available to us, it is probably a good hypothesis. Turning from the agricultural and industrial work areas of the plantation to the main house, Douglass provides his readers with a textual description of the approach to Wye House in much the same way Dell Upton describes the grand procession to Mount Airy and other plantations, writing: The carriage entrance to the house was a large gate, more than a quarter of a mile distant from it; the intermediate space was a beautiful lawn, very neatly trimmed, and watched with the greatest care. It was dotted thickly over with delightful trees, shrubbery, and flowers. The road, or lane, from the gate to the great house, was richly paved with white pebbles from the beach, and, in its course, formed a complete circle around the beautiful lawn. Carriages going in and retiring from the great house, made the circuit of the lawn, and their passengers were permitted to behold a scene of almost Eden-like beauty. Outside this select inclosure, were parks, where--as about the residences of the English nobility--rabbits, deer, and other wild game, might be seen, peering and playing about, with none to molest them or make them afraid. The tops of the stately poplars were often covered with the red-winged black-birds, making all nature vocal with the joyous life and beauty of their wild, warbling notes. (Douglass 1858:67-68) These landscape features (although minus the wildlife), can be identified between the main house and the road on the spatial datasets that form the hGIS database of Wye House. These are also mostly still present on the contemporary landscape. In 2009, archaeologists excavating in and around the Greenhouse at Wye House identified a grouping of objects placed under the entryway just outside the door of the quarter attached to the rear of the Greenhouse. Located between the domestic quarter and the Lloyd family cemetery immediately adjacent to the north, this grouping of objects consisted of a colorless quartzite projectile point, chert projectile point, and a coin (or possibly a button), placed deliberately in order to direct spirits away from the entrance to the domestic space, which faces the Lloyd 295 family cemetery (Blair, Cochran, and Duensing 2009). Douglass describes this cemetery and the accompanying unease and aversion to it by the enslaved at Wye House, writing: A short distance from the great house, were the stately mansions of the dead, a place of somber aspect. Vast tombs, embowered beneath the weeping willow and the fir tree, told of the antiquities of the Lloyd family, as well as of their wealth. Superstition was rife among the slaves about this family burying ground. Strange sights had been seen there by some of the older slaves. Shrouded ghosts, riding on great black horses, had been seen to enter; balls of fire had been seen to fly there at midnight, and horrid sounds had been repeatedly heard. Slaves know enough of the rudiments of. theology to believe that those go to hell who die slaveholders; and they often fancy such persons wishing themselves back again, to wield the lash. Tales of sights and sounds, strange and terrible, connected with the huge black tombs, were a very great security to the grounds about them, for few of the slaves felt like approaching them even in the day time. It was a dark, gloomy and forbidding place, and it was difficult to feel that the spirits of the sleeping dust there deposited, reigned with the blest in the realms of eternal peace. (Douglass 1855:68) While conducting oral history interviews on present-day African-American gardens, Grey Gundaker and Judith McWillie took note of found objects that serve to denote the entrances to yards. Separation of spaces between exterior and interior is important, and in this interpretation the cache found at the Wye Greenhouse guards a transition between two worlds, one of which requires protection from the other (Gundaker and McWillie 2005:121). A similar interpretation can be made of the cache found under the front wall of the 2-Story Quarter at Wye House excavated between 2011 and 2014, and briefly described in the previous chapter and by Leone and others (2018) and Pruitt (2015). As argued by these authors, this horizon of circular, reflective, crossed, white, iron, or glass objects arraigned in a line under the edge of this structure can be charged with the symbolisms of Western African 296 spirituality as well as early African American Christianity and interpreted as an attempt to connect to or control the elements of the spirt or spiritual world. After describing the built and natural environment of the plantation landscape, Douglass concludes his tour by writing, “These all belonged to me, as well as to Col. Edward Lloyd, and for a time I greatly enjoyed them” (Douglass 1855:68). Clearly, Douglass is not suggesting these belong to himself in the legal sense the way they did Colonel Lloyd; rather, Douglass is positioning himself socially on this landscape. In the context of his narrative, Douglass makes clear that his social position as a child at first exempted him from many of the restrictive elements of control and discipline built into the plantation landscape. This should not be taken to mean that as an enslaved child there was no regime of control or discipline extended around Frederick Douglass. In fact, he explicitly writes about such a regime only pages later. Rather, Douglass is suggesting here that the coded messages and ideologies that were built into, transmitted, and reinforced through plantation and landscape architecture were not intended for him and as such, as a child had identified one such “hole in the net” as described by Dell Upton and discussed in the following chapter. At the same time Douglass is exploiting one such hole in the net of ideology, he is still entrapped in other parts of it. Despite being able to pass through some of the restrictive controls of the plantation because of his position on this landscape as a child, he is still trapped within other parts of its messaging and is enmeshed in its discourse. Douglass prefaces his description of Wye House itself, writing: …above all, there stood the grandest building my eyes had then ever beheld, called, by every one on the plantation, the "Great House." This was occupied by Col. Lloyd and his family. They occupied it; I enjoyed it. The great house was surrounded by numerous and variously shaped out-buildings. There were 297 kitchens, wash-houses, dairies, summer-house, green-houses, hen-houses, turkey-houses, pigeon-houses, and arbors, of many sizes and devices, all neatly painted, and altogether interspersed with grand old trees, ornamental and primitive, which afforded delightful shade in summer, and imparted to the scene a high degree of stately beauty. The great house itself was a large, white, wooden building, with wings on three sides of it. In front, a large portico, extending the entire length of the building, and supported by a long range of columns, gave to the whole establishment an air of solemn grandeur. It was a treat to my young and gradually opening mind, to behold this elaborate exhibition of wealth, power, and vanity. (Douglass 1855:66-67) Interestingly, Douglass includes a similar passage in his 1881 edition of his autobiography but changes the final word from “vanity” to “beauty”. Despite his lived experience on this plantation, it is clear from this description that Douglass is influenced by and takes cues from the ideological messaging intended of the built environment, even if he is able to recognize and articulate what that message is. Physically, Wye House is one of, if not the largest, residential dwelling within miles. Its architectural cues and details are meant to signify the sense of importance and position on the landscape its owners wish to convey. It is clear to Douglass what the intended effect of this structure and landscape was, and this is reflected in his description of it in his autobiographies. Despite his opposition to it, the language of the Lloyds becomes the language of Douglass. In many places, this dissertation certainly falls into the same linguistic trap. The power of ideology at Wye House is such that it becomes the default language to talk about this place. Another interesting glimpse into the way Douglass internalizes this landscape is the order in which he introduces his reader to it. It is important to note that his “General Survey of the Slave Plantation” (Douglass 1855:61-78) begins with the total isolation of the slave on the plantation from the larger, outside world, moves to the objects of most interest to a small child, then to the buildings lived and worked in by 298 the enslaved on the plantation, and only then moves onto the main plantation house and its immediate outbuildings, the core of the plantation. It should be emphasized here that almost all other popular and scholarly documentation of Wye House and other plantations like it (like the kinds discussed in Chapter 5) are organized in the opposite direction. Usually, the main house comes first, an arrangement of space more in line with the slaveholder than the enslaved. Nearly every other description of Wye House and plantations like it, from National Register of Historic Places or National Historic Landmark nomination forms and the Historic American Building Survey to architectural and historical volumes like those of Scarborough, Lockwood, Forman, or Weeks to even their entries on Wikipedia begin with the main house and the immediate formal landscape, including carefully placed flankers and wings and outbuildings and gardens before moving onto the outlying agricultural buildings and quarters for the enslaved and the fields and forests beyond. Even the planned procession of a visitor through this landscape as described by Upton (1984) follows this hierarchy with the plantation house as the center of ones’ focus and everything else secondary and ancillary to that. The plantation landscape is purposely ordered and built in order to force this kind of hierarchical understanding of the landscape, its buildings and spaces, and the human interaction that is supposed to occur and be reinforced on it. The description of Wye House provided by Douglass is on the one hand reversed and on the other, uses the same ideological language to describe it. He begins with the outlying agricultural and enslaved landscape before describing the main plantation house and the core of the plantation. Where other authors have spent pages, chapters, or even volumes 299 describing the main house and the “Monumental Layout of the Grounds” (to borrow a phrase from Forman [1967:61]), Douglass only gets around to the main house after first describing windmills, barns, anchored sloops, a lack of school houses, and “a mysterious tall, dilapidated old brick building.” So while Douglass inverts his tour of the plantation and makes it clear that he understands and is fluent in the message and intent of this landscape, he still internalized this discourse on an ideological level, seemingly running into some of the same problems as modern tour guides of these plantations as described by Butler, Eichstedt and Small, and Modlin in Chapter 2. Frederick Douglass and Talbot County, Maryland Through GIS, we can use Douglass’s descriptions of Wye House and its landscape to produce a textually annotated map of the plantation landscape and add it alongside other historic sources of cartographic and spatial information. These Wye House-specific references assist Archaeology in Annapolis researchers as they excavate and interpret archaeological features at Wye House. Furthermore, Douglass’s descriptions are not limited to Wye House. His autobiographical accounts are full of other rich textural descriptions of space and place throughout Talbot County, Maryland. As such, we can expand our hGIS dataset to include these other locations as well. Douglass begins the narration of his autobiographies with a summary of his early life growing up in his grandmother’s cabin near Tuckahoe, Maryland in the northeast corner of Talbot County (Figure 7.6). He begins Chapter I of My Bondage, My Freedom by writing: 300 In Talbot county, Eastern Shore, Maryland, near Easton, the county town of that county, there is a small district of country, thinly populated, and remarkable for nothing that I know of more than for the worn-out, sandy, desert-like appearance of its soil, the general dilapidation of its farms and fences, the indigent and spiritless character of its inhabitants, and the prevalence of ague and fever. The name of this singularly unpromising and truly famine stricken district is Tuckahoe…. it is seldom mentioned but with contempt and derision, on 301 account of the barrenness of its soil, and the ignorance, indolence, and poverty of its people. Decay and ruin are everywhere visible, and the thin population of the place would have quitted it long ago, but for the Choptank river, which runs through it, from which they take abundance of shad and herring, and plenty of ague and fever. (Douglass 1855:33-34) In order to make sense of Douglass’s experiences as a slave in Talbot County, it is important to understand the geography of the county and the way in which it affects the economic system of slavery that was imported and developed there. Tuckahoe is quite literally a back-water region in Talbot County. The dendritic nature of the coastline and rivers in the county means that almost no point of the entire western half of the county (up to approximately the location of present day Route 50) is more than a mile from navigable water (see the following chapter for more on this). In this part of the county, one is almost always surrounded by water. For an agricultural economy that relies on access to transportation to get crops to local and international markets, this proximity is of paramount importance. Plantation owners and small farmers living in this part of the county have ready access to the shipping routes that connect their plantations and farms to the larger Atlantic world and can reap the economic and social benefits it entails. In marked contrast to this, Tuckahoe is in the far northeast corner of the county and the only water access to this part of the county is via the winding and shallow Tuckahoe River, itself a tributary of the Choptank River. Because of this geography, a large section of northeast Talbot County, southeast Queen Anne’s County, and western Caroline County is essentially without deep-water access to the Chesapeake Bay, a significant economic disadvantage to development and investment in the region. The “decay and ruin” and “general dilapidation” in this “singulary unpromising and truly famine stricken district” 302 described by Douglass should not be surprising given its lack of access to this transportation network and its relative position in the regional economy. While Douglass writes mostly about Talbot County and the historic work described in this dissertation is regarding Talbot County, Tuckahoe is in the northeast corner of Talbot County, adjacent to both Queen Anne’s and Caroline Counties. He describes the widespread reputation his grandmother had for making fishing nets, “not only in Tuckahoe, but at Denton and Hillsboro, neighboring villages”. These towns are both across the county line to the west and northwest in Caroline County, and Hillsboro extends across Tuckahoe Creek into the town of Queen Anne, in Queen Anne’s County. It should be noted that the neighborhood that Douglass describes includes parts of these other towns and counties that are outside the boundaries of Talbot County (Figure 7.7). To emphasize this point that immediate political boundaries are permeable, from a young age, Douglass reports that he is aware of people and places outside of his immediate neighborhood and lived experience. He writes, “…as I grew larger and older, I learned by degrees the sad fact, that the "little hut," and the lot on which it stood, belonged not to my dear old grandparents, but to some person who lived a great distance off, and who was called, by grandmother, "OLD MASTER" (Douglass 1855:38). The old master here to which Douglass refers is a man called Aaron Anthony. Anthony was a Maryland plantation owner born around 1760 and at the time of Douglass’s childhood, was employed by Edward Lloyd IV as his plantation manager and “lived a great distance off” (about 12 miles) at Lloyd’s estate, Wye House (Preston 1980:22-30). 303 In his writings, Douglass recalls some of the diversions near his grandmother’s cabin in Tuckahoe. He writes: Down in a little valley, not far from grandmammy's cabin, stood Mr. Lee's mill, where the people came often in large numbers to get their corn ground. It was a water-mill; and I never shall be able to tell the many things thought and felt, while I sat on the bank and watched that mill, and the turning of that 304 ponderous wheel. The mill-pond, too, had its charms; and with my pinhook, and thread line, I could get nibbles, if I could catch no fish. (Douglass 1855:45) Turning to the 1858 Dilworth map of Talbot County, we can identify these landscape features recalled from Douglass’s youth. To the north and west of Anthony’s property in Tuckahoe are a pair of mill ponds. Created by damming Highfield Creek, these ponds would have powered a series of mills. The 1858 Dilworth Map shows at least two mills here, one labeled “Whites Mill” and the other simply labeled “Mill”, and at least one more unlabeled building that could also be a mill or mill-related structure. The 1877 Lake, Griffing, and Stevenson Atlas also shows at least two mills here, one a grist mill and the other a saw mill (Figure 7.8). The remains of the mill dam can even be seen here on the LiDAR elevation dataset just west of MD-303 (Lewistown Road). It is relatively easy to mine Douglass’s narratives for spatial landmarks and identify their locations. Another way to examine the larger planation landscape of antebellum Talbot County, including the inter-plantation spaces is to recreate the journey across the county between Tuckahoe and Wye House as described by Douglass. As an adult, this journey is vividly remembered and recalled by Douglass, who writes: The distance from Tuckahoe to Wye river—where my old master lived—was full twelve miles, and the walk was quite a severe test of the endurance of my young legs. The journey would have proved too severe for me, but that my dear old grandmother—blessings on her memory!—afforded occasional relief by "toting" me (as Marylanders have it) on her shoulder. My grandmother, though advanced in years—as was evident from more than one gray hair, which peeped from between the ample and graceful folds of her newly-ironed bandana turban—was yet a woman of power and spirit. She was marvelously straight in figure, elastic, and muscular. I seemed hardly to be a burden to her. She would have "toted" me farther, but that I felt myself too much of a man to 305 allow it, and insisted on walking. Releasing dear grandmamma from carrying me, did not make me altogether independent of her, when we happened to pass through portions of the somber woods which lay between Tuckahoe and Wye river. She often found me increasing the energy of my grip, and holding her clothing, lest something should come out of the woods and eat me up. Several old logs and stumps imposed upon me, and got themselves taken for wild beasts. I could see their legs, eyes, and ears, or I could see something like eyes, legs, and ears, till I got close enough to them to see that the eyes were 306 knots, washed white with rain, and the legs were broken limbs, and the ears, only ears owing to the point from which they were seen. Thus early I learned that the point from which a thing is viewed is of some importance. (Douglass 1855:46-47) From Douglass’s account of his trip from Aaron Anthony’s plantation in the northeast corner of the county to Wye House, we can take the digitized road network created from the Dilworth map in Chapter 4, most likely a close proxy for the road network traversed by Douglass approximately three decades earlier, in order to determine the most likely route or routes that Douglass and his grandmother took by calculating the shortest and most efficient possible routes from the start point in Tuckahoe to the end point at Wye House (Figure 7.9). Using the 1858 Dilworth map, we can reconstruct the journey undertaken by young Frederick Douglass and his grandmother. He tells us this distance between Anthony’s property in Tuckahoe is 12 miles from Wye House. A direct line between Tapper’s Corner and Wye House measures 11.87 miles. Using the road network as it existed in 1858 as depicted on the Dilworth Map, the shortest distance between Tapper’s Corner and Wye House is 15.64 miles. Using modern road names, this shortest route is as follows: • Head west from Tapper’s Corner (at the intersection of Lewistown Road and Tappers Corner Road) along Tapper’s Corner Road for one mile. • Veer left onto Dukes Bridge Road and continue for 0.8 miles. • Turn right onto Cordova Road and make a quick left onto Asches Acres Road for 1.1 miles. • Continue straight onto Skipton Cordova Road. • Turn left onto Blades Road and continue straight for 1.9 miles where it turns into Old Cordova Road. • Continue for 2.9 miles and turn right onto Rabbit Hill Road. • Continue for 2.2 miles and turn left onto Longwoods Road. • Make a quick right onto Sharp Road and continue for 1.5 miles. • Continue straight onto Little Park Road for 1.5 miles. • Turn left onto Todds Corner Road and continue for 1.2 miles until it turns 307 into Bruffs Island Road. • Continue straight on Bruffs Island Road for 1.9 miles. • Turn right onto the Wye House plantation driveway and continue straight to your destination at the end of the driveway. 308 While this route is the shortest, there are several other possible routes that are of roughly similar lengths. A southern route along Chapel Road is approximately 15.9 miles and a northern route along Old Skipton Road is approximately 16.5 miles. In the present, using the modern road network, Google Maps (seemingly the standard for navigation and route-based distance calculation at the time of this dissertation) calculates the shortest distance to be 15.1 miles. The major difference between this route and the calculated historic one is the use of the straight Cordova Road (MD-309) which follows the route of the Maryland & Delaware Railroad instead of the more circuitous route of Old Cordova Road, Blades Road, Skipton Cordova Road, Asches Acres Road, and Dukes Bridge Road. The rail-line and corresponding route did not exist when Douglass walked across the county from Tuckahoe to Wye House in the early 1820s, which would certainly explain why it was not used by Douglass and his Grandmother to shorten their route to Wye House. Douglass does not describe much of his route except for one element: “the somber woods which lay between Tuckahoe and Wye river”. Looking at Douglass’s possible routes, he would have had to walk through the Lloyd Woodland that separated Wye House from the rest of Talbot County. Even today, much of this woodlot is intact and one must still drive through it to get to Wye House from just about anywhere else. In this author’s estimation, it still retains its somber character. Douglass returns to a different set of somber woods later as an adolescent, this time using them to escape and hide from Edward Covey, the man to whom his labor was rented, writing: I, however, darted back into the woods, before the ferocious hound could get hold of me, and buried myself in a thicket, where he lost sight of me. The 309 corn-field afforded me cover, in getting to the woods. But for the tall corn, Covey would have overtaken me, and made me his captive. He seemed very much chagrined that he did not catch me, and gave up the chase, very reluctantly; for I could see his angry movements, toward the house from which he had sallied, on his foray. Well, now I am clear of Covey, and of his wrathful lash, for the present. I am in the wood, buried in its somber gloom, and hushed in its solemn silence; hid from all human eyes; shut in with nature and nature's God, and absent from all human contrivances. Here was a good place to pray; to pray for help for deliverance--a prayer I had often made before. (Douglass 1855:234) While these woods are not the same as he passed through with his grandmother on his way to Wye House (they are probably to the west, on the other side of St. Michaels), it is interesting to note that in both cases visibility or the lack thereof are the relevant elements of the woods for inclusion in the narrative. Douglass’s engagement with the woods as a means to evade surveillance and capture and as loci of resistance will be addressed in the following chapter. As has been shown, Douglass’s qualitative descriptions of the plantation landscape can be annotated, located, and mapped using the quantitative geographic tools described above and some of Douglass’s spatial worldviews can be interrogated. This provides a powerful tool for documenting surviving historical features and landscapes with a connection to Douglass with potential uses for cultural resource management and heritage tourism. It also leads us to a richer, deeper textual understanding of cartographic features seen on historic maps. This adds a rich descriptive layer to our historical GIS database that otherwise would be missing and makes for a more robust historical dataset. Furthermore, we can also do the reverse; that is, we can take the quantitative digital mapping tools described above and use them to explore subjective, qualitative meaning on the landscape. 310 Remapping an Eighteenth-Century Garden As has been discussed in Chapter 5, LiDAR-derived elevation data is a useful tool for quickly and remotely surveying extant and surviving plantation landscapes. When combined with the appropriate historic maps, it can be used to tentatively identify and extract those portions of the landscape that appear to possess historic continuity and depth. In addition to the uses developed in Chapter 5, the LiDAR data can be used in a more speculative manner, pushed up to and even beyond the limits of its resolution as a spatial dataset with the intent not to positively identify historic features of the landscape but to serve as inspiration for a multivalent interpretation of historic garden features. As also has been discussed in Chapter 5, the landscape at Ratcliffe Manor is remarkable for its scale and in its survival into the present (Figure 7.10). Built in 1747 for James Hollyday and his wife Anna Marie Robins, Ratcliffe Manor consists of a two-and-a-half-story central block and an adjoining wing. Located just outside Easton in Talbot County, Maryland, it is described in the 1930 Homes of the Cavaliers, as having “been fortunate in many things, but in none more so than the disinclination of the three families who have owned it to ‘make improvements.’” The author continues, “[it is] virtually the same, that is to say, but with the addition now of…prodigal growth of boxwood which has claimed an entire garden for its own instead of girdling the flowerbeds primly with its green.” (Scarborough 1930:302) In the years since this passage was written, the boxwood at Ratcliffe Manor has all but died, the flowerbeds have been allowed to revert to grassy lawn, and no trace of the 311 original garden remains to be seen by the casual observer except the formal garden terraces. Because it appears that no substantial and purposeful human modification to the garden at Ratcliffe has occurred since it was originally designed, the palimpsest that is this landscape appears easier to read in that there are far fewer layers to disentangle. The gardens here appear well-suited to test the limits of the resolution of the LiDAR dataset in an attempt to reveal micro-topographic variation related to long-vanished landscape and garden features. The garden at Ratcliffe is comprised of a series of falling terraces. The location of these terraces is clear in the present and the placement of the paths and garden beds or outdoor rooms can be gleaned quickly from observation and a basic understanding of eighteenth-century formal garden design. What is not readily apparent on the surface is the original configuration within these planting beds. Other 312 garden archaeology projects have been able to at least partially recover designs and patterns of plantings within garden beds, but only through extensive remote sensing, excavation, laboratory analysis, and historical research (Bevan 1994; Brown and Samford 1990; Luccketti 1990; Weber et al. 1990; White 2016; Yentsch and Kratzer 1994). As an alternative or supplemental methodology, as demonstrated in Chapter 5, one can use elevation data generated by LiDAR survey to highlight and reveal topographic features within these spaces, some of which may not even be readily visible to the naked eye (Figure 7.11). Before proceeding, it should be emphasized that the use of the State of Maryland LiDAR dataset here stresses and even exceeds the upper limit of the dataset’s vertical resolution. As such, extreme care needs to be exercised when interpreting these results as random noise residing within this dataset can be mistaken as a landscape feature with cultural significance. As argued by Harmon and others (2006), turning to a dataset with higher spatial resolution may help alleviate this problem; however, the initial topographic variation between the landscape features being teased out here is so small that current airborne LiDAR technology is stressed to resolve them. Its use in this chapter is not to positively identify extant garden features, although it certainly does this up to the limits of the resolution of the dataset; rather, in attempting to resolve these features, an interesting observation can be made and pursued. Displaying the original State of Maryland LiDAR dataset collected beginning in 2004 and supplanted by the more recently collected FEMA dataset used in Chapter 313 5, we can examine this specific portion of the landscape in much more depth. Of interest are the following features identified through the use of LiDAR (Figures 7.12 and 7.13): 1) rectilinear bed feature; 2) use of converging lines; 3) potential walkway and planting bed; 4) bed outline and potential garden feature; 5) potential garden 314 feature; and 6) potential garden feature. These features are consistent with eighteenth- and-early-nineteenth-century garden features identified and documented elsewhere by both period gardeners and observers and by contemporary researchers. While no excavations are planned for the garden at Ratcliffe Manor, the identification of these features of interest could serve as the starting point for an archaeological research program in garden archaeology at Ratcliffe Manor. Other kinds of remotely-sensed 315 data surveys such as ground penetrating radar or electrical conductivity could have the potential to identify archaeological deposits relating to these features. If shown to be successful, this methodology could be applied at other plantation sites and serves as a departure point for any projects investigating surviving gardens elsewhere. Of this desktop survey of possible landscape features at Ratcliffe Manor, I would like to put aside for a moment this issue of ground-truthing these findings. From their appearance on the LiDAR elevation map, Features 1 through 3 seem likely to exist on and/or in the ground. We have already shown in Chapter 5 that LiDAR can be used to identify topographic relief in these plantation landscapes that potentially has historic origins and shown in Chapter 6 that it is possible to combine historic maps and LiDAR to identify archaeological features in the ground. The question of the limits of this methodology for detecting smaller and smaller topographic variation from an airborne platform is probably more appropriately considered in a dissertation in the physics of applied lasers and remote sensing, not in one using them in conjunction with historic documentation to determine the origins of features observed in the present. Certainly, as LiDAR technology develops and as processing and analysis techniques improve, these limits will recede. For the time being, it is probably sufficient to just have a handle on what exists, what might exist, and what is an artifact of data collection or just random noise. As a thought experiment in this historic landscape, for a moment I would like to assume that Features 4, 5, and 6, the cruciform features seen in the LiDAR dataset at Ratcliffe Manor, exist on the ground at this plantation garden and is not a random artifact generated by noise in the data. It is not the goal here in this part of this chapter 316 to argue whether specific configurations of space are or are not present at Ratcliffe Manor or anywhere else; rather, the goal here is to use this specific form, not out of the ordinary for formal gardens or landscapes of this kind, and interpret it in a new light, using a set of cultural rules that are not the rules by which these spaces were designed or intended to be interpreted by those that owned them. Instead, I wish to interpret these features and this garden by a set of cultural rules that might have been possessed by those whose labor made them possible and for whom these gardens were arguably not intended. Regardless of whether or not the features identified exist at Ratcliffe Manor or are introduced by noise in the dataset, we can take the geometry of the garden implied by the LiDAR survey. The following discussion was originally presented as a conference paper at the 2013 Society for Historical Archaeology conference in Leicester, England co-authored by this Beth Pruitt and this author (Pruitt and Skolnik 2013). Our goal here was to go beyond the traditional interpretations of these gardens as places of symmetry, hierarchy, order, pleasure and relaxation, nature and culture. While the colonial elite that owned these gardens were almost exclusively white, of European descent, and male, we wanted to go beyond this perspective and the interpretations it has created. While they were the ones who owned these spaces, they were frequently not the ones that designed them and certainly not the ones who toiled to build, plant, and maintain them. It became apparent to Pruitt and myself that the symbolism, imagery, and meaning embedded within these colonial gardens owned by the plantation elite did not have to be interpreted in the manner intended by their owners. This insight can 317 hardly be considered new and comes directly from a critique frequently leveled against Leone and his interpretation of the William Paca garden in Annapolis, Maryland (Leone 1984). Leone’s chapter on the Paca garden has been influential within the field of archaeology and has spawned much discussion not just of colonial gardens and their owners but of the way culture and ideology function. This critique, perhaps best attributed to Ian Hodder (Hodder and Hudson 2003:218-222; Hodder 1986), would have us question whether all visitors to the garden would have interpreted it as Paca intended. Hodder argues that Leone’s treatment of ideology in his influential 1984 chapter on the Paca Garden in Annapolis is overly focused on the top-down imposition of ideology from the economic and social elites like William Paca and his peers to everyone else. Furthermore, under the model of ideology used by Leone, which he adopts from Michael Shanks and Christopher Tilley (1982) and their critique of Louis Althusser (1971), if worldview is dictated by the cultural or economic elite, then there is little or no room for individual agency. Hodder questions the degree to which the ideologies identified by Leone are accepted at all levels of society. Rather than a model in which all members of society are taken in by or tricked by the ideologies of the ruling class, Hodder suggests that meaning is not fixed and the same material culture or landscapes may hold different meanings for different social groups. Read more broadly, this critique questions whether the archaeology of plantations and slavery is best understood as a top-down distribution of power, ideology, and capitalism, or if archaeologists should be interested in how individuals resisted these impositions and negotiated space for 318 themselves. This line of thinking encourages us to see the material culture of slavery from a perspective other than the slave owner. Influenced by this critique by Hodder, Pruitt and this author have suggested that enslaved Africans and African Americans would have brought their own botanical expertise and ideologies to the plantation landscape, at times independent of, parallel to, and/or in contestation with those of individuals like William Paca or Edward Lloyd. They would have brought worldviews shaped by their cultural backgrounds and experiences to the gardens, their quarters, and elsewhere, just as would individuals like William Paca or Edward Lloyd. We are certainly not the first to suggest as much. Neil Norman and Kenneth Kelly describe a similar situation encountered during their fieldwork at Savi in Bénin wherein researchers were surprised to learn landscape features “embodied differing cultural significance for specific individuals” of different cultural backgrounds (Norman and Kelly 2004:98). Western and African interpretations of the same space existed simultaneously and coterminously here. Garden historian Barbara Sarudy devotes an entire chapter in her 1998 book Gardens and Gardening in the Chesapeake, 1700-1805 to the laborers that made these gardens possible. Sarudy (1998) suggests that some enslaved gardeners were highly valued for their botanical knowledge, which was applicable to the warmer climates of the Chesapeake and American South. Douglass tells of another reason the enslaved might be in a garden, writing: This garden was not the least source of trouble on the plantation. Its excellent fruit was quite a temptation to the hungry swarms of boys, as well as the older slaves, belonging to the colonel, few of whom had the virtue or the vice to resist it. Scarcely a day passed, during the summer, but that some slave had to 319 take the lash for stealing fruit. The colonel had to resort to all kinds of stratagems to keep his slaves out of the garden. The last most successful one was that of tarring his fence all around; after which, if a slave was caught with any tar upon his person, it was deemed sufficient proof that he had either been into the garden, or had tried to get in. In either case, he was severely whipped by the chief gardener. This plan worked well; the slaves became as fearful of tar as of the lash. (Douglass 1845:15-16) If African bodies and African botanical knowledge were in these gardens in order to create and maintain them, then why cannot also African symbolism have been brought into these spaces in order to understand them? These gardens were certainly a place on the landscape occupied and used by the enslaved, who, as both Hodder and Upton would urge us to remember, would have also applied their own cultural logic to interpret and understand the space that comprised the garden. Returning to the garden at Ratcliffe Manor outside Easton, Maryland, or to the William Paca Garden in Annapolis, Maryland, or to any of thousands of other plantations throughout Maryland or the American South, an important point can be made. There is a limit at which LiDAR can no longer reliably detect landscape features and while the garden at the Paca House is a reconstruction, they both stand as visible proxies for the kinds of spaces that would have been present around plantation houses and which both the plantation owners and those enslaved on the plantation filtered through their own worldviews. These two modern gardens are very similar in design, even if their locations and settings are very different. Both modern gardens, as were their historic antecedents, consist of a series of falling or descending terraces and garden beds, connected by walkways. The question posed by Pruitt and this author was how could the symbolism and meaning and function of these garden spaces be interpreted from an African perspective? If cultural meaning is not solely 320 imposed by dominant ideology from above and these gardens need not be about power and hierarchy, what new meanings might these spaces take on to those enslaved Africans and African Americans who through their cultural backgrounds and lived experiences possessed a different worldview and set of ideological assumptions than the owners of these spaces did and intended to convey? Others investigating Western African cosmological systems have noted that crosses, cruciform shapes, or crossroads can be symbolic of the intersection of the physical and spiritual worlds (Brown 2004; Fennell 2007; Ferguson 1992; Leone et al. 2018). The garden at Ratcliffe Manor, like most of these formal landscapes consist of multiple, intersecting paths and at each intersection is a crossing. Bodies of water, especially rivers that are flowing, and especially their surfaces, are symbolic of the boundary between the living and the dead (Ferguson 1992). Aside from the botanical requirements these gardens had for access to water, eighteenth-and-nineteenth- century plantations in the Chesapeake like the ones discussed in this dissertation are generally found along or within sight of the at least one of the rivers or tributaries that feed into the Bay. Bounded on three sides by the Tred Avon River and two of its tributaries, the garden is surrounded by water. These spaces are surrounded by, situated above, or otherwise in association with this water, and as such, there exists the potential to make this symbolic connection between these gardens and their crossroads and this boundary. Furthermore, in some instances, certain colors take on symbolic meanings. The color white can be associated with spirits, especially of the dead (Fennell 2007). Crushed oyster-shells were frequently used to pave paths and whole shells were 321 sometimes used in garden beds to improve drainage. Furthermore, iron tools, which are necessary for the creation and maintenance of these gardens are associated with Ogun, who is not only a deity of blacksmithing, iron, and fire, but also of transition, including birth, death, and rebirth (Barnes 1989). Examined in this light, these garden spaces like the one at Ratcliffe Manor, or the ones at Wye House or the Paca House become powerful spaces that can mediate between the physical and the spiritual worlds. They are at the intersection of life and death and suggest the possibility of moving between them. In this reworked interpretation of formal garden design, these gardens become not the places where the ideology of hierarchy, domination and power play out, but rather, they become a symbolic representation of a West African spiritual cosmology. When approached in this manner, the plantation can lose its Western sense of hierarchy and power and becomes something else. To conclude, Pruitt and this author suggest that in addition to being places of power, hierarchy, and dominance, formal gardens like the ones that exist and would have existed at Ratcliffe Manor and the Paca House represent a setting upon which the physical symbols of West African spiritual beliefs and practices can be identified and can be given a new meaning through this symbolism. While there is no archaeological proof in the scientific sense to support this assertion, this claim is not intended to be treated as such. Perhaps future historical or archaeological work will be able to suggest that some elements within these formally-designed landscapes were actively and deliberately shaped for such purposes by enslaved gardeners. However, when viewed with the understanding that physical configurations of space 322 and materials can simultaneously embody multiple symbolic associations according to different cultural traditions or to different viewers or in different circumstances, this new garden landscape represents a subversion of the dominant narrative of landscape as power and hierarchy, and recast these familiar spaces in a fundamentally different and new light. In a broad sense, this chapter, and this dissertation, represent a spatial annotation of mid-nineteenth-century Talbot County Maryland, and its larger plantation landscape. Compiled from a range of sources including archaeology, Douglass’s autobiographies, and other historical documentation and research, this chapter adds a textual layer of description on top of the historical spatial documentation produced in the previous chapters. By using GIS, we have located and annotated places identified by Douglass, which in the following chapter will be used to work toward a conceptual model of racialized space as previously introduced by Dell Upton. 323 Chapter 8: Mapping the Enslaved Landscape Through Dell Upton: Integrating Qualitative Data into Historical Geographic Information Systems (hGIS) Dell Upton and Racialized Plantation Landscapes The theoretical underpinning of this dissertation on plantation landscapes has largely come from Dell Upton’s two pieces regarding plantation landscapes discussed at the beginning of Chapter 3. His 1984 article “White and Black Landscapes in Eighteenth-Century Virginia” in Places and his 1990 chapter “Imagining the Early Virginia Landscape” in William Kelso and Rachel Most’s Earth Patterns: Essays in Landscape Archaeology provide a theoretical framework with which we can combine spatial GIS analyses of space on plantations with the archaeological and historical record. My understanding of the racialized nature of plantation landscapes largely starts with Dell Upton’s treatment of the subject of his 1984 article titled “White and Black Landscapes in Eighteenth-Century Virginia”, which has been influential in framing discussions of race and plantation landscapes elsewhere. In this article, he argues that there was not one eighteenth-century landscape in colonial America, but at least two, operating simultaneously and coterminously. Across both of these works, his central thesis is that racially-based slavery, a social and economic reality with profound social and spatial implications, caused those Virginians of European (and especially Anglo) descent and those enslaved Virginians of African descent to have fundamentally different ways of moving through, inhabiting, and ultimately viewing the landscape. While his area of focus is explicitly Virginia, neighboring Maryland is 324 similar enough in terms of geography, economy, and settlement to consider both together as a part of the larger Chesapeake region. This plays out on two different axes, both race and class, but it is his attention to racialized plantation landscapes on which this dissertation focuses. The consequences of this difference manifest themselves not just as profoundly different experiences and cultural understandings of specific spaces encountered on a daily basis, but also as profoundly different cultural understandings of space in the abstract sense. From his analysis, Upton concludes that the plantation landscape was designed to be met as “a powerful and intense ideological statement” (Upton 1984:71), but with the added caveat that the Africans and African Americans enslaved on them were not a part of this intended audience. Upton explains, “the elements of raw power replaced those of ideological persuasion” (Upton 1984:71). As such, despite his forceful statement regarding the ideological nature of the plantation and its landscape, the plantation was not intended to function as ideology for the enslaved; in its absence, power and violence were the coercive forces that were used to maintain slavery. He uses the phrase “raw power” here, and with it, he invokes the kinds of violence most familiar in the public discourse on plantations, including whips, chains, and beatings as a form of discipline. In the years since Upton wrote this, others have since theorized the many forms that violence can take, notably Slavoj Žižek (2008) and Saidiya Hartman (1997). These authors call on us to broaden our understanding of what constitutes violence beyond that which was envisioned by Upton to include forms of systemic, symbolic, and subjective violence and other kinds of threats and terrors that do not necessarily involve immediate bodily harm. 325 Upton’s analysis begins with the observation that, “The great planter intended that his landscape would be hierarchical, leading to himself at the center. His house was raised above the other buildings and was often set off from the surrounding countryside by a series of barriers or boundaries—fences and terraces” (Upton 1984:654). While there is certainly ideology at work on these plantation sites and landscapes, Upton argues that most of its messaging is not intended for the enslaved audience. This is a significant reworking of the central thesis regarding ideology and plantation landscapes presented throughout Leone’s work in Annapolis, including in the Paca garden (Leone 1984). What Upton does here is provide the theoretical framework for integrating a top-down study of ideology and power on these sites with one that is also aware of the ways those who were subjected to those forces may or may not have been targeted, received, or accepted them. Upton’s work has two major aims which are relevant here. First, he addresses how plantation landscapes would have been encountered and negotiated by social actors. Upton, like others at the time, understood the plantation landscape through the binary racialization of a dichotomous black/white orientation of eighteenth-century plantation contexts. While Upton acknowledges some of the simplicities of this model (especially as related to class differences between white farmers and great planters) and begins to lay out the complexities of the reality in terms of conflicting intersections of class and race, he posits that the most important cultural divide as pertains to orientation, configuration, and movement across the landscape is that of race and slavery. On one hand, plantations were spaces where the plantation elite could both figuratively and literally elevate themselves above their white social and 326 economic subordinates as well as the enslaved individuals they owned, reinforcing this difference through physical arrangements of space as well as controlling access to these spaces. On the other hand, the enslaved Africans and African Americans who negotiated these spaces were able to access the same spaces used for social posturing through side doors and passageways and enter even the most formal spaces unnoticed and unobstructed. From this, Upton’s second aim is to argue that the formality of eighteenth- century plantations was not meant for or directed at the enslaved community. His most forceful point is that ideology was not necessary to maintain the black/white racial and social hierarchies that slavery required to exist in early Virginia. He concludes: When it was experienced as intended, [the landscape] could be a powerful and intense ideological statement. But the duties and personal experiences of slaves circumvented this experience. Blacks were not drawn into the social posturing of gentry society, and whites did not expect them to be. The elements of raw power replaced those of ideological persuasion” (Upton 1984:71). With this significant modification, Upton shows us where Leone’s model of ideology and landscape falls short. Partially freed of its ideological function, Upton’s treatment of racialized landscapes adds an important dimension of multivocality to landscape studies, concluding his chapter by writing, “[The landscape] must be read as a whole; it was neither uniform nor entirely dominated by the gentry. The meaning of the landscape could be read in more than one way” (Upton 1984:71). It is on this metaphorically broken terrain that we can begin to imagine a more nuanced and complicated picture of the plantation landscape. 327 As Upton argues, largely picking up on the work of sociologist Erving Goffman, specifically his 1956 The Presentation of Self in Everyday Life, that negotiating the social landscape can be thought of as a theatrical performance. Others have also applied Goffman’s work to the historical record, namely Rhys Isaac in his Pulitzer Prize-winning The Transformation of Virginia, 1740-1790 (1982), upon which Upton also leans heavily. According to Goffman, social individuals can be thought of as actors in a play, each of which is playing one or a multitude of roles at any given time using a script, props, costuming, and stage dressing to accomplish this task. There is some value to seeing social relations on the planation as a kind of theatrical production and the plantation landscape as a kind of stage on which this action takes place. Goffman explains, “A social establishment is any place surrounded by fixed barriers to perception in which a particular kind of activity regularly takes place” (Goffman 1959:238). Douglass’s characterization of the plantation landscape at Wye House as one of those “secluded and out-of-the-way places” certainly meets this definition. Of the purpose of writing his book, Goffman explains: I assume that when an individual appears before others he will have many motives for trying to control the impression they receive of the situation. This report is concerned with some of the common techniques that persons employ to sustain such impressions and with some of the common contingencies associated with the employment of these techniques…I shall be concerned only with the participant’s dramaturgical problems of presenting the activity before others. The issues dealt with by stagecraft and stage management are sometimes trivial but they are quite general; they seem to occur everywhere in social life, providing a clear-cut dimension for formal sociological analysis. (Goffman 1959:15) He also writes: 328 Within the walls of a social establishment we find a team of performers who cooperate to present to an audience a given definition of the situation. This will include the conception of own team and of audience and assumptions concerning the ethos that is to be maintained by rules of politeness and decorum. We often find a division into back region, where the performance of a routine is prepared, and front region, where the performance is presented. Access to these regions is controlled in order to prevent the audience from seeing backstage and to prevent outsiders from coming into a performance that is not addressed to them. (Goffman 1959:238-9). Importantly, if we seize upon the metaphor of actors and roles, stages and sets to approximate social interaction, it opens an outlet for resistance to the social theatrical production. Isaac notes that “relationships developed during interaction supply the threads of which social fabric is woven” (Goffman 1959:350). It is out of this relatively mundane and routine social interaction that culture and history unfold, especially in the aggregate sense that archaeologists typically recover material remains from the past. In this light, one way to view resistance and defiance is as any act that challenges or threatens this stagecraft. This is ultimately a kind of power. As Isaac also notes, “The social power of a person is commensurate with the capacity to make others’ actions subservient to his or her own ends” (Isaac 1982:350). This power can come in many forms. Ignoring the script, improvising ones’ carefully proscribed lines, or breaking any part of the set, or even the fourth wall are all possible avenues to subvert the narrative being acted out. The premise underlying Upton’s argument is that social position affects the ways in which one moves through and experiences the landscape. He writes: As an audience slaves shared in some respects the position of the white common planter, but their status as slaves worked in other ways to alter and even to undercut the intended effects of the processional landscape….The slaves were not intentionally a part of the audience. Few white planters imagined that slaves were susceptible to the legitimating functions of white 329 society; they recognized that the slave’s lack of standing made force the only sure legitimizer. (Upton 1984:66). Upton continues this argument in his 1990 chapter, titled “Imagining the Early Virginia Landscape” (1990). He posits that from the perspective of the enslaved, the plantation landscape was ultimately about control. There were areas like the main house where the plantation owners were able to exercise the most control, areas where the enslaved were allowed a larger degree of autonomy such as yards and gardens, and areas largely outside the control of the plantation owner such as the woods surrounding the plantation (Heath and Bennett 2000; Vlatch 1993). Upton argues that because of the nature of slavery and the means used to enforce it, the enslaved came to see their landscapes as a series of loci where they are either completely controlled by somebody else or possessed varying degrees of freedom. The data Upton draws upon here is rather limited; however, some is derived from accounts provided by from travelers to the area complaining how difficult it was to get directions from the enslaved. Rather than describing the travel route in relation to a traveler traversing an open network of space, directions included specific landmarks and these landmarks’ owners. Rather than relaying the path the traveler was to take, the direction-giving enslaved navigator highlighted points along that path and their owners. James Robertson describes a similar phenomenon in traveler’s accounts asking for directions in Spanish Town, Jamaica (Robertson 2009). The representational mental maps of European visitors were not congruent with the way in which locals understood their built environment. Like the accounts Upton draws upon from Virginia, the mismatch between these two spatial templates or worldviews resulted in more than one exasperated traveler to record this frustration for posterity. 330 This insight should act as a gateway to illustrate the multitude of cultural logics that operate on and help create racialized plantation landscapes in the Chesapeake. Rather than being viewed as messy, this multivocality should be embraced and explored. Here are two different-yet-overlapping cultural frameworks for conveying directions from one location to another. Upton writes, “If the master’s landscape was a network that implied connection and movement, the landscape of the slave was a static one of discrete places” (Upton 1984:69). One relies upon traversing existing fixed routes while the other relies upon the participant’s dynamic location to fixed points on that landscape. The first exists draped across the landscape as a kind of network upon and in which social action takes place. The second exists as a reaction to navigating a social landscape imposed by others. The two cultural understandings of the landscape presented by Upton and his suggestions for how one is to move through it are fundamentally different. Upton explains, “The slaves’ landscape was described from the point of view of someone surrounded by other people’s power, and its landmarks were plantation houses and fields differentiated by ownership” (Upton 1984:70). From this, he concludes that the level of frustration caused by these directions signify they represent a culturally foreign, “alien and, to them, irrational mode of thought” (Upton 1990:75). These travelers shared a common Western and/or white conception of space and navigation, reinforced through their experiences in this social position. He traces its origin to the Enlightenment of the seventeenth and eighteenth centuries and places it within a specific cultural trajectory and framework. Upton calls this mental template dynamic, flexible, continuous, and articulated, describing it as a network of flexible relations 331 that bound both the natural and social worlds which could be manipulated from many points within it (Upton 1990:80). Here, a specific idea—a way of organizing, looking at, and moving through space—is given a specific cultural origin whose path and adoption can be traced through time and space. In his two works, Upton attempts such a project, producing what Foucault might recognize as one of his genealogies or as a kind of archaeology, that is, a project grounded in creating a history of the present and tracing the origins of institutions and configurations back through time. Upton concludes by discussing the recourse left to those unable to posture as the gentry, that is, those unable to design and command the larger landscape as their personal stage and theater. These include those enslaved on the plantation as well as free African Americans, poor white farmers, women, children, and anybody else not included at the top of the social hierarchy. Upton differentiates these groups from the planter elite in that the former have a fundamentally different relation to landscape than the latter, but maintains that for the enslaved, the added dimension of race separates these landscapes in a fundamental way. He writes, “Subject to the overbearing demeanor and irascible conduct of their social betters, barely scratching a living from the soil or allowed only a fraction of the fruits of their labors, they had little reason to imagine such a useful world. They searched instead for the holes in the net.” (Upton 1990:85) It is these holes in the net for which we will be searching. Anna Agbe-Davies (2018), emphasizes that the seventeenth-century Chesapeake (and it is true of the eighteenth and nineteenth century, also) is a pluralistic historical context, that is, one comprised of many kinds of people. One of the underlying ideologies of this place is what Agbe-Davies calls an “ideology of 332 difference”. She explains the increase in spatial segregation in the region at the end of the seventeenth century, marked by separate quarters for the enslaved and more formal delineations of space, as an attempt to further define social distinctions between different people. She reminds us that social distinction and ethnic group identity is produced through interaction, not through isolation. Having looked backward, from this position we can continue Upton’s project into the present. The spatial worldview that he describes as belonging to the gentry- class of eighteenth-century Virginia is the dominant spatial worldview of the twenty- first century. J.B. Harley prefaces his 1988 chapter, “Maps, Knowledge, and Power” by writing: My aim here is to explore the discourse of maps in the context of political power, and my approach is broadly iconological. Maps will be regarded as part of the broader family of value-laden images. Maps cease to be understood primarily as inert records of morphological landscapes or passive reflections of the world of objects, but are regarded as refracted images contributing to dialogue in a socially constructed world.” (Harley 1988:278) He continues, “Maps are never value-free images; except in the narrowest Euclidean sense they are not in themselves either true or false” (Harley 1988:278). He also follows in the footsteps of Foucault, quoting the French philosopher, “Knowledge was thus a form of power, a way of presenting one’s own values in the guise of scientific disinterestedness” (Poster 1982:118-19; cited in Harley 1988:279). Harley’s premise is that if maps as a socially encoded document are a reflection of the society that created and produced them, then they will also be instilled with the values of that society. Maps are a distillation of the values of the social actors that produced them; chiefly, the state (and especially the military) the propertied, landowning class, and 333 commercial and business interests, and in turn maps reinforce the ideologies and values of these groups. In his chapter, Harley traces the history of the map as a technological development up through the present and inextricably ties it with the rise of the Western nation state. He emphasizes that, “Cartography remains a teleological discourse, reifying power, reinforcing the status quo, and freezing social interaction within charted lines….The iconology of the map in the symbolic treatment of power is a neglected aspect of cartographic history.” Here in Talbot County, the map must be seen as an expression of either the state or of the landed, white, gentry class (of which there is considerable overlap), mirroring its worldview and reflecting its interests. Its inclusions and silences must be considered in light of this fact and cultural meaning must be filtered through the reality that the marginalized are not the ones describing their own reality through these documents. At the time Harley wrote these words, the use of digital mapping technologies and Geographic Information Systems were just beginning to become widespread. In the years since, there has been an explosion in the use of digital maps. Lagging, but still present, has been a critical examination of how we use these digital cartographic representations of reality. There is little sign of a groundswell of critical human geographers wholeheartedly embracing GIS as a tool of their trade even as it becomes commonplace in commerce and government. Equally, it would be wrong to imagine that students of GIS are much troubled by the finer points of actor- network theory, or Foucault and Lefebvre. It seems even less likely that the numerous GIS users in commerce and government are paying very much attention to any of this.” (O’Sullivan 2006:783) 334 O’Sullivan characterizes some of the most successful examples of critical GIS as cases in which socially-informed researchers turn to the tools and technologies of GIS to answer their research questions. These can be broadly conceived of as many of the same directions and goals of more anthropologically or sociologically-minded scholars, including participatory or community GIS, gendered or feminist GIS, and issues of privacy, access, and ethics (Elwood 2008; Kwan 2002; Pavlovskaya 2018; Schuurman 1999; Wilson 2017). Denis Byrne presents a related study in an article on “nervous landscapes” in the context of Aboriginal negotiation of a different racialized landscape in Australia (Byrne 2003). He documents the introduction of the English cadastral system of property division, developed gradually through thousands of years of land use in England, to the continent of Australia almost immediately with little or no consideration for the physical terrain or cultural landscapes already present. He also examines the ways those subject to its control found to exploit that system, including fence-jumping, fishing, orchard raiding, and other forms of containment-breaking. Byrne, like both Upton and Douglass, is interested in the “gaps in the grid” where this imposed order breaks down and a different social space can be created, even if temporarily. Byrne’s treatment of “nervous landscapes” has already been introduced to plantation landscapes in Maryland by Megan Bailey (2018), where she uses the concept to examine power relations at and contextualize a plantation site called L’Hermitage managed by the National Park Service at the Monocacy National Battlefield. 335 While outside the scope of this dissertation, Upton’s hypotheses regarding the construction and maintenance of multiple, overlapping spatial frameworks can be further tested not in the archive or in the archaeological record, but by the cultural anthropologist. The social conditions required to produce a bifurcated, non- mechanized agricultural social landscape based on slavery no longer exist in the Chesapeake; however, there are modern-day analogs that combine ideology, power, social groupings, and space that researchers can turn to in order to examine the merit of Upton’s spatial and social hypothesis. Leaning on Foucault, these potential case studies include jails, prisons, hospitals, and schools, as well as modern-day plantations that revolve around a monocrop cash crop agriculture and/or undocumented workers. We can predict there the same kinds of spatial consequences as Upton categorizes in the Chesapeake and perhaps future work can test this hypothesis. In returning to the idea that landscapes come into being and take meaning through enactment and lived experience, one can turn to Christopher Tilley’s A Phenomenology of Landscape (1994). Tilley understands one’s movement through space as an expression of a kind of language, calling this movement a pedestrian speech act. Tilley writes, “Pedestrian ‘speech acts’ may be likened to the speech acts of language. Walking is a process of appropriation of the topographical system, as speaking is an appropriation of language. It is a spatial acting out of place, as the speech act is an acoustic acting out of language” (Tilley 1994:28). The phenomenological approach, he argues, would understand landscape as a fundamentally experiential phenomenon that must be lived-in if they are to be 336 understood. Therefore, it is through being in these landscapes that they come to take on their significance and meaning. By moving through them along certain paths, lingering at certain points, participating in some activities at some times and in some places and other activities at other times and in other places, a kind of narrative is formed and a “biography of place” begins to develops.” Bringing Tilley’s work to the plantation, movement of the enslaved through the landscape was carefully regulated (Bankole 1999; Berlin 1998:68; Upton 1984). As such, these biographies of place are inscribed with these power relations that define the cultural topography of the intra-plantation landscape as well as the inter- plantation landscape. These landscapes become traced with the impressions of slavery and in turn, further shapes the biographies of place into biographies of slavery and power. Through Tilley and Upton, we can build a connection between the navigation of space through routine everyday activities, personal biography, and geography. This connection is central to make the transition between plantation landscapes as abstract ideology and plantation landscapes as dynamic and creative places that can be used to mitigate or even subvert those ideologies. Unfortunately, these past landscapes no longer exist because the cultural forces that shaped them have been gone from them for more than 150 years and have since been overwritten with new ones. Therefore, the present challenge is for the archaeologist to recover their traces in order to identify Upton’s “holes in the net”. 337 Modeling Control on the Landscape One method of using the tools of Geographic Information Systems (GIS) to examine the antebellum plantation landscape of Talbot County, Maryland has been addressed here in this dissertation in Chapter 5 by using modern LiDAR data and historic maps to identify extant landscape features with historic origins. In this chapter, we can do the same considering Upton’s hypothesis and, more importantly, we can use them to find his “holes in the net”, that is, where this ideology of control and hierarchy breaks down spatially. Upton’s data is largely qualitative. It comes from traveler’s accounts, journals and diaries, and reflections on the configurations of space. To this body of data, we can, using Douglass as a guide, add a quantitative analysis of the plantation landscape drawn from historic maps and modern LiDAR- derived elevation data in an attempt to further refine Upton’s spatial hypothesis for plantation landscapes. As one of the nineteenth century’s most prolific writers and public speakers and first-hand witness to American slavery, Douglass and his body of work are situated to act as a conduit through which we can apply descriptive data to Upton’s hypothesis. In My Bondage, My Freedom, after describing the houses, workspaces, and landmarks of Wye House Plantation with a sense of child-like wonder and amazement, Douglass writes, “These all belonged to me, as well as to Col. Edward Lloyd, and for a time I greatly enjoyed them” (Douglass 1855:68). Later, in the 1881 edition of his autobiography, Douglass rewrites this sentence to read, “These all belonged to me as well as to Col. Edward Lloyd, and, whether they did or not, I greatly enjoyed them” (Douglass 1881:29). Inside this sentence is the crux of Dell 338 Upton’s hypothesis on plantation landscapes. Different people (or more accurately, groups of people) have different ways of seeing, moving through, and interacting with space depending on who they are as well as the social rules enforced in and through that space. Douglass certainly does not mean to imply that his and Edward Lloyd’s names are both on the deed to Wye House; their ownerships are of two different kinds, and yet he uses the same word to describe each case. Douglass describes the plantation as a “baronial domain” from Europe’s middle ages with Edward Lloyd as the lord of the estate (Douglass 1855:64). While the Lloyds own the plantation and those on it in a legal and economic sense, Douglass, as a child and a slave, still finds the social space to move and interact in this physical space, enough at least to consider it to belong to him and his curiosity. Douglass’s social position, in this instance foremost that of a child, but ultimately as an enslaved African American, shapes the way he moves through and is affected by this landscape. To this point, in the previous chapter we were able to use geospatial tools to locate and identify specific locations, people, or features described by Douglass in his autobiographies. To make the connection between Douglass and his description of the plantation landscape on the one hand and Dell Upton and his description of the plantation landscape on the other, we can turn to Foucault, particularly his work on observation, surveillance, dominance, and even governmentality, that is, on how to create governable, docile bodies. The issues that Foucault raises in Discipline & Punish (1977) ultimately revolve around the control of human bodies, a central concern on plantations whose purpose is to coordinate and facilitate an exploitative system of labor and are worth discussing here. 339 At the center of Upton’s argument about space on plantations is the issue of control. Whereas Upton can help us frame our discussion of power and control and surveillance on the plantation landscape, Foucault grounds our discussion of power and control. Through Foucault, we can identify an important characteristic to model across the landscape—that of visibility and seeing. In Discipline & Punish, his data, insofar as it can be said that Foucault uses data, largely comes from eighteenth-and- nineteenth-century France; however, his work is still relevant to the American plantation. Applied directly to the problem of historical archaeology by Mark Leone in his 1995 “A Historical Archaeology of Capitalism”, a body of archaeological studies has dealt with the concepts of power, control, resistance, and surveillance on the planation (see Delle 1998; Epperson 1990, 1999, 2001; McKee 1992; Singleton 2001; Singleton and Bograd 1995). In a 2001 article, Theresa Singleton discusses the spatial strategies behind the arrangement of Cuban coffee plantations as well as the enslaved responses to this imposed spatial order. She writes, “Nearly every aspect of plantation space from the location and arrangement of slave quarters…to the details of slave-house construction…resulted from conscious decision making on the part of planters to maximize profits, exercise surveillance and reinforce the subordinate status of enslaved people.” She notes the presence of fortified quarter complexes, enclosed by high walls which could be locked at night and watched by guards, a characteristic of Cuban plantation slavery and of interest to Singleton. She explains these features as being a response to the particular configuration of plantation slavery in this Cuban context, namely the need for panoptic surveillance, the need to control people coming 340 to and going from the quarter, and the need for protection from slave raiders. Conversely, the enslaved community had their own responses to these spatial arrangements ranging from mass-runaways, feigning illness, breaking or misplacing tools, maintaining underground storage pits, and the continuation of a host of African cultural practices (Craton 1982). Singleton concludes her article: In an effort to curb the rebellious activities of a growing slave population, slaveholders adopted management practices that resembled prison life. This jail-like character was particular evident in the built environment of slave quarters and the spatial order of plantations....What slaveholders failed to recognize was that enslaved laborers had their own notions of an ideal existence….Consequently, they sought various ways to undermine slaveholder authority. (Singleton 2001:110-111) Because one of the ways this authority was enforced was through landscape, it is through landscape that we can try to find some of the ways in which that authority was either incomplete or subverted. My own introduction to the profound impact of visibility, observation, and surveillance across the plantation landscape came to me while conducting fieldwork. Through Archaeology in Annapolis’ engagement with the excavations at Wye House, the site of Frederick Douglass’s boyhood enslavement, the opportunity arose to attempt to locate Douglass’s birthplace, located on the eastern side of Talbot County, far removed from the plantations on which he spent his childhood and adolescence before being sent to Baltimore and discussed briefly in the previous chapter. As recorded in his autobiographies, Douglass was born “in Talbot county, Eastern Shore, Maryland, near Easton, the county town of that county”. He continues: “there is a small district of country, thinly populated, and remarkable for nothing that I know of more than the worn-out, sandy, desert-like appearance 341 of tis soil, the general dilapidation of its farms and fences, the indigent and spiritless character of its inhabitants, and the prevalence of ague and fever. The name of this singularly uncompromising and truly famine stricken district is Tuckahoe….it is seldom mentioned but with contempt and derision, on account of the barrenness of its soil, and the ignorance, indolence, and poverty of its people. Decay and ruin are everywhere visible, and the thin population of the place would have quitted it long ago, but for the Choptank river, which runs through it, from which they take abundance of shad and herring, and plenty of ague and fever. (Douglass 1855:33-34) The Douglass birthplace survey project was guided by a combination of the kind of GIS modeling presented in this dissertation with a more traditional use of historical records familiar to historical archaeologists and is detailed in by this author (Skolnik 2015) and Grillakis and this author (2012). Using the 1858 farm boundaries from the William H. Dilworth map described above as the project area boundaries, contact was made with the present owner of the largest tract formerly owned by J.P. Anthony, Douglass’s former owner Aaron Anthony’s son, and an arrangement was made for Archaeology in Annapolis to conduct a limited Phase I survey of portions of the property in an attempt to locate the remains of Douglass’s birthplace. Much of this survey was conducted along the edge of a topographic feature known locally as the Kentucky Ravine. Running through much of the southern part of Aaron Anthony’s property, the Kentucky Ravine is a long, deep, branching, forested stream valley. The passage from My Bondage, My Freedom from which the title of this dissertation was taken (“The real distance was great enough”) also includes another relevant observation from Douglass in the sentence immediately before it. He writes, “But to understand, some one has said a man must stand under” (Douglass 1855:281, see page 6 of this dissertation for this full passage). It is relevant here because it was during this field survey at one of Douglass’s former homes that I literally stood under 342 this landscape of observation and control and it became clear to me one of the ways around it. One day during this Phase I survey along the top of the ravine, I took a few minutes to walk down to the bottom of the ravine, just to see what might be down there. It was here at the bottom of the Kentucky Ravine along an unnamed tributary of the Tuckahoe Creek, once home to a child named Frederick who would soon learn that he was a slave, that I was struck by the overwhelming solitude of this ravine and this small patch of forest. I was cut off from the surrounding agricultural fields. I was cut off from the road that brought me here. I was cut off from the field crew working above me. Nobody could see me down here, nobody knew where I had gone, nobody knew when I would be back. If I had wanted to, I could have walked from the top of the ravine to the bottom unnoticed. With a small boat and a moonless night as cover, I could have silently reached the Choptank and the ay beyond it. Here, at the bottom of the Kentucky Ravine, the possibility of visibility and isolation, secrecy and detection on the plantation landscape became real to me. In Douglass’s autobiographies, he discusses these elements of control and we can use the digital representations of the nineteenth-century plantation landscape produced in the previous chapters to model this across the regional county-wide landscape. One way to model this kind of control across the landscape is to understand it as Upton does. Upton sees the plantation landscape as a series of focal points, centered on the main plantation houses, where the gentry exercise their control over the land and people. Places closest to the plantation houses or along roadways 343 are those places under the most intensive kinds of control whereas those farther away will not be under as close scrutiny. In a GIS framework, it is a fairly simple process to calculate the distance between any point in space and any number of identified features. To calculate the distance between any two points or between a point and a specific feature, the GIS program uses the X and Y coordinates of each and solves for distance through the application of the Pythagorean Theorem. The difference between the X coordinates and the difference between the Y coordinates become the two legs of a right triangle. Squaring each of these values, adding them, and taking the square root of the result produces the length of the hypotenuse of this triangle, which is the distance between the two points or a point and a feature. Using the power of computer automation, it is a simple task to repeat this calculation for every pixel or cell in a map. The result is a map that displays the distance between each location and the nearest feature of interest. Here, we can recall Figures 4.3 and 4.9, maps highlighting the structures and roads as featured on the 1858 Dilworth map. These features include the main plantation houses that dot the landscape of Talbot County and the land-based transportation network that connected them, represented by the structures and roads depicted on Dilworth’s map. From the digitized structures layer from the Dilworth map, we can calculate the distance to the nearest structure for any location on the map. The result is a map that shows us those spaces that are closest and farthest from the structures mapped by Dilworth (Figure 8.1). If social power and control are exercised and maintained from these locations as a result of the social networks that emanate from them, then we can 344 begin to look for the “holes in our net” by moving as far away from these structures as possible. Similarly, instead of mapping the distance of each point to the nearest structure, we can map the distance of each point to the nearest road (Figure 8.2). 345 Roads are highly-trafficked, highly-visible pathways through the county. Not only does Upton discuss their importance in terms of control over landscape, but in a way so does Douglass. He recounts a meeting that took place on a country road between Edward Lloyd and one of his slaves: 346 It is reported of him, that, while riding along the road one day, he met a colored man, and addressed him in the usual way of speaking to colored people on the public highways of the south: "Well, boy, who do you belong to?" "To Col. Lloyd," replied the slave. "Well, does the colonel treat you well?" "No, sir," was the ready reply. "What? does he work you too hard?" "Yes, sir." "Well, don't he give enough to eat?" "Yes, sir, he gives me enough, such as it is." The colonel, after ascertaining where the slave belonged, rode on; the slave also went on about his business, not dreaming that he had been conversing with his master. He thought, said and heard nothing more of the matter, until two or three weeks afterwards. The poor man was then informed by his overseer, that, for having found fault with his master, he was now to be sold to a Georgia trader. (Douglass 1855:116-117) Roads are corridors along which one can travel from one place to another. Because they function to funnel and concentrate human movement into an area with an extremely small footprint, observation and interaction are much more likely to occur, making them places of elevated danger for the enslaved. In this above instance, a chance encounter on the road led to an unfortunate outcome for the enslaved individual described here. We can also combine these two distance maps to display the distance from each point to either the closest structure OR the closest road, whichever happens to be nearer. Given the distribution and prominence of these two classes of features on the landscape, this model allows us to approximate those locations where such an encounter might be most likely to occur. This map shows those locations where control represented by these features is closest and farthest away (Figure 8.3). It is important to remember that there are other ways of moving around and through Talbot County that do not involve roadways. As the title of Christopher Weeks’s 1984 volume on the historic architecture of the county reminds us, Talbot County is a place “Where Land and Water Intertwine”. As such, any consideration of distance, travel, observation, and movement must also consider travel by water. 347 We can use GIS to model the distance of any point on land from the coast or river bank (Figure 8.4). For the western half of Talbot County, up to approximately the location of present-day Route 50, the water is never very far away. In fact, no point west of Easton in Talbot County, is more than a mile from the water’s edge. The 348 water is not only a place for seeing and being seen, it is a place where movement, both sanctioned and proscribed, can happen. Not only does Douglass occasionally describe the sights along the coastline of Talbot County, he also contemplates 349 escaping the plantation by water in a small craft. Here he compares the benefits and hazards of a water route to a land route while plotting his escape: On the water, we had a chance of being regarded as fishermen, in the service of a master. On the other hand, by taking the land route, through the counties adjoining Delaware, we should be subjected to all manner of interruptions, and many very disagreeable questions, which might give us serious trouble. Any white man is authorized to stop a man of color, on any road, and examine him, and arrest him, if he so desires. (Douglass 1855:286) He continues this line of thinking several pages later: The feasibility of the plan, too, could the first steps have been taken, was marvelously plain. Besides, this was a new idea, this use of the bay. Slaves escaping, until now, had taken to the woods; they had never dreamed of profaning and abusing the waters of the noble Chesapeake, by making them the highway from slavery to freedom. Here was a broad road of destruction to slavery, which, before, had been looked upon as a wall of security by slaveholders. (Douglass 1855:306) For Douglass the water is simultaneously a corridor for movement as well as a place where one can see for great distances around them as well as been seen by others from far away; simultaneously a liability and an asset. Remarkably, Douglass tells his reader that the water was not regarded as an avenue of escape for the enslaved in Talbot County despite the proximity and access to it as well as the distances once could cover while on the water. Perhaps Douglass, writing in the year 1855, eight years prior to Emancipation, is trying to deflect attention from an obvious and useful escape route? Later, after the death of his owner and being sent away from Wye House, Douglass recalls his new home and considers his situation: Our house stood within a few rods of the Chesapeake bay, whose broad bosom was ever white with sails from every quarter of the habitable globe. Those beautiful vessels, robed in purest white, so delightful to the eye of freemen, were to me so many shrouded ghosts, to terrify and torment me with thoughts of my wretched condition. I have often, in the deep stillness of a 350 summer's Sabbath, stood all alone upon the banks of that noble bay, and traced, with saddened heart and tearful eye, the countless number of sails moving off to the mighty ocean. The sight of these always affected me powerfully. My thoughts would compel utterance; and there, with no audience but the Almighty, I would pour out my soul's complaint in my rude way, with an apostrophe to the moving multitude of ships: "'You are loosed from your moorings, and free; I am fast in my chains, and am a slave! You move merrily before the gentle gale, and I sadly before the bloody whip! You are freedom's swift-winged angels, that fly around the world; I am confined in bands of iron! O, that I were free! O, that I were on one of your gallant decks, and under your protecting wing! Alas! betwixt me and you the turbid waters roll. Go on, go on. O that I could also go! Could I but swim! If I could fly! O, why was I born a man, of whom to make a brute! The glad ship is gone; she hides in the dim distance. I am left in the hottest hell of unending slavery. O God, save me! God, deliver me! Let me be free! Is there any God? Why am I a slave? I will run away. I will not stand it. Get caught, or get clear, I'll try it. I had as well die with ague as with fever. I have only one life to lose. I had as well be killed running as die standing. Only think of it; one hundred miles straight north, and I am free! Try it? Yes! God helping me, I will. It cannot be that I shall live and die a slave. I will take to the water. This very bay shall yet bear me into freedom. (Douglass 1855:220) Surely, Douglass was not the only slave to take in these sights and have these thoughts. In earlier editions of his autobiography, Douglass obscures his method of escape in order to prevent action being taken to prevent other enslaved individuals from taking advantage of the same holes in the net as he. Likewise, he may be trying to do the same here, planting disinformation so as to make an obvious method of escape more practicable by writing it off and discounting it, in which case, this is a valuable insight into this landscape of slavery. Furthermore, embedded within these two paragraphs is perhaps Douglass’s own version of Upton’s spatial hypothesis. He contrasts the dynamic movement and freedom of the ships passing by his new home with the stagnant confinement of the terms of his enslavement. This highway on the water that connected places like Wye House and other plantations in Talbot County to 351 economic centers like Easton and Baltimore served as a constant reminder to Douglass that there was an alternative. These distance maps are an easy and simple way to model control across the landscape. The problem with these models is that they understand control as a function of distance and nothing else, whereas Foucault and others would caution us that the exercise of power or freedom from it is more nuanced. So, while distance models are useful as a starting place for thinking about how to model control across the plantation landscape, they have room for improvement. Continuing with Foucault in terms of determining levels of control, perhaps more important than distance to a feature is visibility to and/or from it. Borrowing from Foucault who in turn borrows from Jeremy Bentham, we can turn to the Panopticon, both a physical and metaphorical structure concerned with discipline and surveillance. As with our models based on distance, we can use a GIS program to model visibility from the structures, roads, and waterways identified in the Dilworth map. To do so, we require the same digitized layers of structures and roads that were used above as well as a digital elevation model like the LiDAR-derived one used throughout this dissertation. With these kinds of visibility models, the implication is that viewshed and visibility work in both directions. If you can see a point from your location, somebody at that point can see your location. If you can see the water from a place in Talbot County, then that location is visible from the water. Rather than display distance between features on the plantation landscape and the surrounding county, each of these previous maps can be reconfigured to display visibility between the same. Here, it is important to consider the ways visibility and 352 sight can affect the way people move through and interact in space. The center of the plantation model, both architecturally and logistically, is the manor house. Loosely borrowing from Quetzil Castañeda’s purposeful entanglement of the homophonic words “sights” and “sites” (Castañeda 1996), these plantation houses function as the plantation core largely because of their visibility. The structures depicted on the 1858 Dilworth map can act as an approximate representation of the most important plantation structures in the county. In his “New Principles of Gardening” (1728), Batty Langley describes to his reader where best to place a new house and garden, “The most noble and pleasant Situation of all others, is that on the Top of a Hill…where the Air is fine and clear, with noble views….And, on the contrary, where Situations are low, they often abound in too great a Plenty of Water, which, when stagnant, or of a slow motion, is very unhealthy” (Langley 1728:191-192). He concludes his section on the situation of gardens by quoting French Jesuit and poet René Rapin: If on thy Native Soil thou dost prepare T’ erect a Villa, you must place it there Where a free Prospect does itself extend Into a Garden; whence the Sun may lend His Infl’ence from the East; his radiant Heat Should on your House through various Windows beat: But on that Side which chiefly open lies To the North Wind, whence Storms and Show’rs arise, There plant a Wood; for, without that Defense, Nothing resists the Northern Violence, While with destructive Blasts o’er Cliffs and Hills Rough Boreas moves, and all with Murmurs fills; The Oak with shaken Boughs on Mountains rends, The Valleys roar, and great Olympus bends. Trees therefore to the Winds you must expose, Whose Branches best their pow’rful Rage oppose. (Rapin, cited in Langley 1728:192) 353 The period garden design literature suggests these parameters for placing villas or plantation houses for the practical reasons of health and pleasant disposition and offer suggestions on how to mitigate some of the problems associated with settling on an exposed position. Turning to Chandra Mukerji’s analysis of garden design as an outgrowth of European military technology, engineering, and strategy, a secondary set of practical reasons emerges for the placement of villas or plantation houses that instead revolve around observation, surveillance, and defense. These two sets of reasons for placing these features on the landscape are not mutually exclusive and are probably self-reinforcing. Turning to GIS, we can model the visibility of these structures across the landscape using the structures from the Dilworth map and the DEM of Talbot County as our inputs (county overview presented as Figure 8.5 and detail of Wye House region as 8.6). Despite a preference for placing primary plantation structures at these points of surveillance, the map of visibility from these locations reveals not just that there are many areas that cannot be seen from these vantage points, but that these areas are long, dendritic, and interconnected, just like the Kentucky Ravine discussed above. This would allow travel and movement across large geographic distances without being seen. Douglass discusses the importance of visibility from roads in an incident in which he is trying to return to St. Michaels to protest his treatment by a slave owner named Covey: Seeing no signs of my stopping, Covey caused his horse to be brought out and saddled, as if he intended to pursue me. The race was now to be an unequal one; and, thinking I might be overhauled by him, if I kept the main road, I walked nearly the whole distance in the woods, keeping far enough from the road to avoid detection and pursuit. But, I had not gone far, before my little strength again failed me, and I laid down. The blood was still oozing from the 354 wound in my head; and, for a time, I suffered more than I can describe. There I was, in the deep woods, sick and emaciated, pursued by a wretch whose character for revolting cruelty beggars all opprobrious speech--bleeding, and almost bloodless….The way was through bogs and briers, and I tore my feet often during the journey. I was full five hours in going the seven or eight miles; partly, because of the difficulties of the way, and partly, because of the feebleness induced by my illness, bruises and loss of blood. (Douglass 1855:227-228) 355 As with structures on the Dilworth map, we can model visibility from the roads on the same. (county overview presented as Figure 8.7 and detail of Wye House region as 8.8) as well as combine both into a single visibility map that highlights visibility from either of these features (county overview presented as Figure 8.9 and detail of Wye 356 House region as 8.10). Like Figures 8.5, 8.6, 8.7, and 8.8, these reveal pockets within the county that are not visible from either these residential structures or the roads; however, the inclusion of the road network significantly reduces the continuity of many of these areas. 357 Douglass frequently makes reference to the things he can see from the shore and while on the water. As with the distance to the coastline, we can also model visibility to/from the water. Of leaving Wye House for Baltimore, Douglass tells us: On setting sail, I walked aft, and gave to Col. Lloyd's plantation what I hoped would be the last look I should ever give to it, or to any place like it….After 358 taking this last view, I quitted the quarter deck, made my way to the bow of the sloop, and spent the remainder of the day in looking ahead; interesting myself in what was in the distance (Douglass 1855:136) We can also model visibility from the water (Figure 8.11). Not surprisingly, large portions of the western half of the county are visible from the water while large 359 portions of the eastern half are not. The pixels representing the westernmost edge of the county are categorized as “not visible" in this analysis not because they are not visible from the Chesapeake Bay, but because the raster analysis was unable to account for the presence of that water outside of the limits of the DEM. 360 A significant limitation of this kind of model and a key point for this discussion on landscape archaeology is a result of the data used to produce it. The digital elevation models used here and in similar kinds of historic view analyses are contemporary datasets. They depict the landscape as it exists in the present, compiled 361 from modern data sources, and frequently, from new technologies like LiDAR which did not exist until very recently. As such, these digital elevation models depict conditions in the present and our viewshed models will reveal viewsheds in the present. The results from these models should be considered with this caveat— cultural practices in the past and present actively and passively shape what is and is not visible across a given landscape through the processes of not just movement, but also construction, earth moving, planting, trimming, and neglect. These all come into the present by means of the palimpsest that is the present landscape. Furthermore, the input data here has been generated straight from the Dilworth map. These points and lines represent features with real-world locations, many of which can be knowable in the present with a little bit of background research. Given the 1,169 structures and 458.18 miles of road, verifying or correcting the positions of these features as mapped by Dilworth in 1858 to correspond to their real-world locations is a monumental undertaking. As such, for the purposes of this hGIS survey into Talbot County, the visibility model above was conducted using the location of these features as represented by Dilworth and not their locations as corrected in the real world. Therefore, our results can only be as good as the model that produced them and the model that produced them can only be as good as the data used to generate that model. This is in addition to the mental asterisk we must remember to assign to these data to remind us of their palimpsestual nature. In this case, the data used is contemporary digital LiDAR data, a contemporary invention that is only capable of capturing and recording the surface of the contemporary landscape. The problems of doing landscape archaeology in the presence of a modern palimpsest such as this have 362 already been addressed elsewhere in this dissertation; however, this point is important enough to reiterate here. This visibility/viewshed model can only be a visibility/viewshed model of the present. Buildings, landscaping, and probably most importantly, vegetation that has been added or removed means that this model can only approximate the visibility/viewshed of the antebellum plantation landscape. Another limitation to this methodology is that the digital elevation models used in this kind of process generally do not include representations of structures, vegetation, or other kinds of visual obstructions that can radically alter lines of sight across the landscape. These bare-earth representations of the ground have been processed to remove these built and natural additions to the landscape so as to reveal the underlying topography. This being said, it is possible to digitally account for visual obstructions such as forests or structures; however, one must first possess information as to the historic locations of these features, and second, add these parameters to the digital elevation models. Such historical data does exist for part of Talbot County in the form of the 1847 and 1862 United States Coastal Surveys introduced in Chapter 5; however, the area of coverage is limited to those coastal areas and as a result, is not included here as a result. Further historic landscape reconstruction in this area would be well-served to attempt such a recreation. Remapping the Plantation Landscape We can revisit the autobiographies of Frederick Douglass to take this project one step further. In light of the spatial thesis proposed by Dell Upton, we can identify a narrative of space that comes from the experience of slavery as described by 363 Douglass. Inside his narrative of slavery and freedom is a narrative of space that privileges movement. To be enslaved means a lot of things; however, in this chapter on space and landscape, one of the most immediate consequences of enslavement is an anxious relationship with movement (also see Byrne 2003 on “nervous landscapes”). One point emphasized by Upton and that can be picked up on in Douglass is the importance of control on the landscape. This can be seen in Douglass’s writings through many of the place names he uses to describe his surroundings. In Douglass’s autobiographies, we can find several examples where the enslaved, through movement, find what Upton calls “holes in the net”. Upton suggests we can see the plantation landscape as loci of control and Douglass explains how and where these loci function. He writes of the monthly distribution of food for the enslaved on all of the Lloyd plantations which occurred at Wye House: I have already referred to the business-like aspect of Col. Lloyd's plantation. This business-like appearance was much increased on the two days at the end of each month, when the slaves from the different farms came to get their monthly allowance of meal and meat. These were gala days for the slaves, and there was much rivalry among them as to who should be elected to go up to the great house farm for the allowance, and, indeed, to attend to any business at this, (for them,) the capital….Being selected, too, for this office, was deemed a high honor. It was taken as a proof of confidence and favor; but, probably, the chief motive of the competitors for the place, was, a desire to break the dull monotony of the field, and to get beyond the overseer's eye and lash. Once on the road with an ox team, and seated on the tongue of his cart, with no overseer to look after him, the slave was comparatively free; and, if thoughtful, he had time to think. (Douglass 1855:96-97) Here, Douglass illustrates a monthly occurrence in which the enslaved are afforded a degree of freedom in their movement. On their way to and from the epicenter of Lloyd’s plantation enterprise, they are temporarily above the system of local control 364 manifested in the plantation landscape. This monthly occurrence is an opportunity where the usual mode of place-based surveillance and control breaks down. On their way from one place to another, the enslaved are temporarily beyond the reach of the overseer and the plantation. Likewise, there were those enslaved who were required to move beyond the plantation as a part of their daily tasks, such as is the case of those on the Lloyd plantation sloop, the Sally Lloyd. In this case, the possibility for movement goes beyond the mere implication of being assigned to crew a vessel. It is a symbol of freedom, whose presence and subsequent absence from the plantation wharves cannot help but to invite thoughts of escape from that place. Douglass writes, “In the river, a short distance from the shore, lying quietly at anchor, with her small boat dancing at her stern, was a large sloop--the Sally Lloyd; called by that name in honor of a favorite daughter of the colonel. The sloop and the mill were wondrous things, full of thoughts and ideas. A child cannot well look at such objects without thinking” (Douglass 1855:65-66). Surrounded by this landscape of control, Douglass has identified several possible escapes, both in the literal sense and in the metaphorical sense. This theme continues throughout much of Douglass’s childhood recollections. For an enslaved child, this ship represents a world outside the plantation. It represents a way to move beyond the control of the slave owner, even if temporarily. And perhaps most importantly, it represents a way to move between these two worlds. Douglass’s wonder is not confined to just the Sally Lloyd; it extends to all of the ships sailing within sight of his plantation. He also writes, “From the mill we could see 365 other objects of deep interest. These were, the vessels from St. Michael's, on their way to Baltimore. It was a source of much amusement to view the flowing sails and complicated rigging, as the little crafts dashed by, and to speculate upon Baltimore, as to the kind and quality of the place” (Douglass 1855:72-73). While the plantation is a site that requires isolation and control of those enslaved there, it, by economic necessity, must also be enmeshed within a larger system of transportation and commerce, which requires movement and access to a larger world. This fascination extends not just to the ships themselves that facilitate movement between the plantation and the outside world but to the people doing that movement. Douglass writes about one of the enslaved employed aboard the Sally Lloyd: Tom was, sometimes, Capt. Auld's cabin boy; and when he came from Baltimore, he was always a sort of hero amongst us, at least till his Baltimore trip was forgotten. I could never tell him of anything, or point out anything that struck me as beautiful or powerful, but that he had seen something in Baltimore far surpassing it. Even the great house itself, with all its pictures within, and pillars without, he had the hardihood to say "was nothing to Baltimore." He bought a trumpet, (worth six pence,) and brought it home; told what he had seen in the windows of stores; that he had heard shooting crackers, and seen soldiers; that he had seen a steamboat; that there were ships in Baltimore that could carry four such sloops as the "Sally Lloyd." He said a great deal about the market-house; he spoke of the bells ringing; and of many other things which roused my curiosity very much; and, indeed, which heightened my hopes of happiness in my new home. (Douglass 1855:136) Being able to travel to and from the plantation aboard the Sally Lloyd and to see the City of Baltimore seems to be a privilege highly esteemed among the enslaved at Wye House. Douglass further tells us, “The hands belonging to Col. Lloyd's sloop, esteemed it a great privilege to be the bearers of parcels or messages to my new 366 mistress; for whenever they came, they were sure of a most kind and pleasant reception” (Douglass 1855:143). Douglass also recalls an incident wherein an overseer tried to whip an enslaved woman, Nelly, for “impudence”: There is no doubt that Nelly felt herself superior, in some respects, to the slaves around her. She was a wife and a mother; her husband was a valued and favorite slave. Besides, he was one of the first hands on board of the sloop, and the sloop hands--since they had to represent the plantation abroad--were generally treated tenderly. The overseer never was allowed to whip Harry; why then should he be allowed to whip Harry's wife? (Douglass 1855:93) In these cases, those enslaved individuals with connections to the Sally Lloyd are afforded greater privileges and esteem through their ability to move beyond the local planation landscape. As noted by Upton, the planation order is maintained through violence: as noted by Douglass, it is maintained through shocking episodes of physical violence, but also other kinds of violence perpetrated against the enslaved laborers—isolation, ignorance, and the control of movement. On this plantation landscape, movement, and especially freedom of movement, is one key to temporarily overcoming the system of surveillance, observation, power, and control that enforces the plantation’s threat of violence. These vignettes from Douglass’s autobiographies are illustrative of examples where movement and the permission of movement reveal one of the holes in slavery’s net, a hole to be discovered and exploited. In the context of plantation slavery, unlimited movement and the freedom of movement on the part of the enslaved is antithetical to the imposed conditions of their enslavement and the structures required to maintain that system. In Douglass’s texts, ships, and especially their unrestrained movement on the Chesapeake and its waterways, are a recurring theme and are 367 frequently used to juxtapose the status of freedom with Douglass’s enslavement, as can be seen in Douglass’s extended quote regarding “the Chesapeake bay, whose broad bosom was ever white with sails from every quarter of the habitable globe” cited above on pages 351-352 of this dissertation. Here, Douglass make it explicit to his reader the opposition between confinement and movement and links these to the conditions of slavery and freedom. Upton identifies control and dominance as the aspect that permeates throughout the landscape and we certainly see these points manifest themselves in Douglass’s accounts. However, we can use the specifics provided by Douglass to highlight this other key point we can take from Upton’s two pieces; the power of movement through space and its ability to circumvent at least some of the control embedded and exercised across the landscape. In his article and in his chapter, Upton only tangentially handles movement across the landscape in respect to the enslaved; however, I believe it to be a central piece to his hypothesis. Upton identifies the processional and articulated nature of the white landscape as enacted by the planter elite and their social inferiors. He describes the ways in which the built environment was designed to control the movement of the latter as they interact with the former, allowing that this manipulation and configuration of spaces was not intended to be directed at the enslaved. “If the master’s landscape was a network that implied connection and movement, the landscape of the slave was a static one of discrete places” (Upton 1984:69). In this distinction, Upton shows that the built landscape of formality and hierarchy on the plantation is not designed for the slave or as a part of the enslaved landscape. He 368 argues that there are other forms of control—violence and the threat of violence—to constrain social action on and to the plantation. In this, the lack of physical barriers in the landscape designed to limit the movement of the slave and impress upon them the importance of the social hierarchy of the landed gentry opens up the possibility for transcending these social constraints through movement. In the white landscape described by Upton, movement is permitted but constrained by the built landscape and social convention. The physical configuration of architecture, gardens, roads, and other built parts of the landscape act to constrain and limit this movement as required by the social elites who build them. The black landscape described by Upton is maintained not through the same use of architecture and social posturing, but through the surveillance, discipline, and violence more familiar to theorists like Foucault and Zizek. In those moments where the enslaved are temporarily beyond the control of this system such as the one Douglass recalls in his oxcart, as they would be in the act of movement from one place to another, we can find the limits of this control and the places where we might find the agency of the enslaved asserting itself against this system. Upton hints at the possibility of this when he describes the black landscape as extending beyond the quarter to encompass the woods and fields and waterways that surround, intersect with, and divide the plantation landscape (Upton 1984:70), an observation echoed by Isaac (1982:52-54). It extends from the quarters on one plantation through these avenues of movement that exist beyond the control of the planter who owns the plantation, to the quarters on neighboring plantations where ties of kinship and friendship connect the enslaved on the black landscape beyond the 369 confines of property lines. For the enslaved, this landscape implies movement between and beyond, both sanctioned and unsanctioned, not confined or controlled by the formal plantation landscape built for the purposes of social posturing by the gentry. Once reoriented thusly, the center of the plantation can be reconfigured to those quarters, connected not to the main houses on their plantations, but to each other, connected by the fields, forests, and rivers that would otherwise serve as boundaries. If these gaps between the net of ideology do exist, it is important to understand the ways in which the net is kept in place. Douglass writes: there are certain secluded and out-of-the way places, even in the state of Maryland, seldom visited by a single ray of healthy public sentiment--where slavery, wrapt in its own congenial, midnight darkness, can, and does, develop all its malign and shocking characteristics; where it can be indecent without shame, cruel without shuddering, and murderous without apprehension or fear of exposure. (Douglass 1855:61-62) He continues, “Just such a secluded, dark, and out-of-the-way place, is the "home plantation" of Col. Edward Lloyd, on the Eastern Shore, Maryland. It is far away from all the great thoroughfares, and is proximate to no town or village” (Douglass 1855:61-62). Douglass may not have intended to, but he describes for his reader one of the means by which this system of slavery is maintained: a lack of movement. For the enslaved, the possibility of movement is the possibility of freedom. If slavery is maintained in part through surveillance and control and limited movement within and beyond the plantation, the ultimate expression of resistance to this control is to take for oneself the right of movement and to run away, to physically move oneself off of this plantation context. In running away, not only does the enslaved exercise free 370 movement on the plantation landscape, a right typically denied to them, but they do it to permanently remove themselves from the system of slavery. In earlier editions of his autobiography like in his 1855 My Bondage, My Freedom, Douglass is reluctant to tell his readers how he ran away and escaped slavery. He explains: There are reasons for this suppression, which I trust the reader will deem altogether valid. It may be easily conceived, that a full and complete statement of all the facts pertaining to the flight of a bondman, might implicate and embarrass some who may have, wittingly or unwittingly, assisted him; and no one can wish me to involve any man or woman who has befriended me, even in the liability of embarrassment or trouble….Were I to give but a shadowy outline of the process adopted, with characteristic aptitude, the crafty and malicious among the slaveholders might, possibly, hit upon the track I pursued, and involve some one in suspicion, which, in a slave state, is about as bad as positive evidence….By stringing together a train of events and circumstances, even if I were not very explicit, the means of escape might be ascertained, and, possibly, those means be rendered, thereafter, no longer available to the liberty-seeking children of bondage I have left behind me. No anti-slavery man can wish me to do anything favoring such results, and no slaveholding reader has any right to expect the impartment of such information. (Douglass 1855:322) He concludes, “How I got away--in what direction I traveled--whether by land or by water; whether with or without assistance--must, for reasons already mentioned, remain unexplained” (Douglass 1855:334). Later, in his 1881 Life and Times of Frederick Douglass, published almost two decades after the abolition of slavery, he writes: In the first narrative of my experience in slavery, written nearly forty years ago, and in various writings since, I have given the public what I considered very good reasons for withholding the manner of my escape. In substance these reasons were, first, that such publication at any time during the existence of slavery might be used by the master against the slave, and prevent the future escape of any who might adopt the same means that I did. The second reason was, if possible, still more binding to silence--for publication of details would certainly have put in peril the persons and property of those who assisted. Murder itself was not more sternly and certainly punished in the 371 State of Maryland than was the aiding and abetting the escape of a slave. Many colored men, for no other crime than that of giving aid to a fugitive slave, have, like Charles T. Torrey, perished in prison. The abolition of slavery in my native State and throughout the country, and the lapse of time, render the caution hitherto observed no longer necessary. But, even since the abolition of slavery, I have sometimes thought it well enough to baffle curiosity by saying that while slavery existed there were good reasons for not telling the manner of my escape, and since slavery had ceased to exist there was no reason for telling it. I shall now, however, cease to avail myself of this formula, and, as far as I can, endeavor to satisfy this very natural curiosity. I should perhaps have yielded to that feeling sooner, had there been anything very heroic or thrilling in the incidents connected with my escape, for I am sorry to say I have nothing of that sort to tell; and yet the courage that could risk betrayal and the bravery which was ready to encounter death if need be, in pursuit of freedom, were essential features in the undertaking. (Douglass 1881:196) Douglass recognizes the importance of movement on the plantation landscape. In sharing with his readers his escape from slavery, one of his major motivations in withholding the details of his escape is to ensure that others wishing to take advantage of the same illicit movement of which Douglass took advantage can do so. In this later edition of his autobiography, published in 1881—close to two decades after the abolition of slavery—Douglass feels free to declassify and share the path he took to freedom. In the time that had elapsed between the 1855 and 1881 editions of his autobiography, the cultural landscapes of Talbot County had shifted dramatically and the power contained within such knowledge no longer had the same charge or danger it once did, illustrating the shifting temporal and cultural context of such spatial knowledge. In this chapter, we have used historical GIS and the writings of Frederick Douglass to work through the spatial hypothesis of racialized plantation landscapes as described by Dell Upton and used the spatial hypothesis of Dell Upton to help unlock Frederick Douglass as a social theorist on plantations and space. 372 Chapter 9: Epilogue This dissertation in historical Geographic Information Systems (hGIS) and mid-nineteenth-century plantation landscapes in Talbot County, Maryland has approached these racialized landscapes from several different directions. Using a historic map, we have been able to spatialize demographic data from the 1860 United States Census (Chapter 4). By combining LiDAR-derived elevation data with a series of historic maps and existing historical documentation, we have been able to create a methodology to quickly map and characterize the historic character of individual plantations and plantation features quickly without needing physical access to these spaces (Chapter 5). Diving deeper into one specific plantation, we have been able to incorporate this hGIS survey methodology into an ongoing archaeological research project to refine previous archaeological interpretations and to locate and excavate the archaeological remains of two lost quarters (Chapter 6). Turning to Frederick Douglass, who not only witnessed these landscapes but is a social theorist in his own right, as a guide, we have mapped the landscape as recorded by Douglass (Chapter 7). Lastly, we have considered ways to model kinds of control across the landscape by bringing together Dell Upton and Frederick Douglass into a spatial dialogue with each other (Chapter 8). This has largely been a methodologically-focused dissertation, one that works through how to use new technologies and techniques to answer questions and test theory already posed by others. Given that GIS is a spatial technology, most of the questions addressed within are some variation on the question, “where?” But it has 373 also addressed a few theoretical questions about racialized landscapes and cultural frameworks for understanding them. From their inceptions, plantations and their surrounding landscapes have been built on elements of control; that is, control of the enslaved Africans and African Americans whose coerced labor was used to build and maintain them and to enrich their owners. After Emancipation, the physical traces of slavery gradually disappeared from these landscapes as quarters fell into disrepair and were removed and as the formerly-enslaved moved away. Despite holding the physical remains of African American heritage, these properties largely remain privately owned by individuals who are not African American. This dissertation provides a way to use the panoptic tools of airborne LiDAR scanning, aerial imagery, and other cartographic resources to relocate these forgotten and lost sites and to digitally repopulate these landscapes. Situating the methodology described above in this context, it attempts to provide one answer to the question posed by Maria Franklin, “Why do historical archaeology?” (1997). Using this specialized toolkit, one can begin to address the erasures and silences that define plantation slavery in the present and highlight the spatial dimension of enslavement, not just in an American context, but anywhere this methodology might be used. One key point to take away from the title of this dissertation, “The real distance was great enough”, is that not all distances can be measured in miles or feet. GIS is a research tool with great potential for archaeology and other spatially-based research, but it remains difficult to work with non-quantitative data within a GIS framework. As such, creative uses of historic data and mapping technologies can 374 attempt to bridge this divide between quantitative and qualitative research methods and digital mapping technologies. In the end, through the several case studies in historic mapping, spatial research, and archaeology presented here, I hope that I have created some new piece of knowledge about either Talbot County, about Maryland, about slavery, about plantations, about Frederick Douglass, about the use of space, or about racial landscapes. I hope that I have given others the ability to pick up these tools and do the same, either here or in their own fields of study. And lastly, I hope that what is contained in these pages represents, as Douglass described it, “knowledge quite worth possessing.” 375 Bibliography Abdallah, Hani, Jean-Stephane Bailly, Nicholas N. Baghdadi, Nathalie Saint-Geours, and Frederic Fabre. 2013 Potential of Space-Borne Lidar Sensors for Global Bathymetry in Coastal and Inland Waters. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 6(1):202-216. Agbe-Davies, Anna S. 2018 Laboring Under an Illusion: Aligning Method and Theory in the Archaeology of Plantation Slavery. Historical Archaeology 52(1):125- 139. Alemy, Alexis, Sophia Hudzik, and Christopher N. Matthews 2017 Creating a User-Friendly Interactive Interpretive Resource with ESRI’s ArcGIS Story Map Program. Historical Archaeology 51(2):288-297. Althusser, Louis 1971 Ideology and Ideological State Apparatuses (Notes toward an Investigation). In Lenin and Philosophy and Other Essays. Translated by Ben Brewster, pp. 127-186. Monthly Review Press. New York. A Talbot Farmer 1858 For the Easton Star. Map of Talbot County. The Easton Star 20 April: Page 1, Column 4. Easton, Maryland. Babiarz, Jennifer J., Amelia Chisholm, and Mark P. Leone 2008 Phase I, II, and III Investigations of Wye Hall (18QU977), Queen Anne’s County, Maryland, for the Maryland Department of Natural Resources. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Bailey, Geoff 2007 Time perspectives, palimpsests and the archaeology of time. Journal of Anthropological Archaeology 26(2):198-223. Bailey, Megan May 2018 Landscapes of Tension: Exploring Nervousness and Anxiety on a Maryland Plantation. PhD dissertation, Department of Anthropology, University of Maryland, College Park. Bankole, Katherine 1999 Plantations without Slaves: The Legacy of Louisiana Plantation 376 Culture. In Plantation Society and Race Relations: The Origins of Inequality, edited by Thomas J Durant and J. David Knottnerus, pp. 193-204. Praeger, Westport, Connecticut. Barnes, Sandra T. 1989 Africa's Ogun: Old World and New. Indiana University Press, Bloomington. Battle-Baptiste, Whitney L. 2007 ‘In This Here Place’: Interpreting Enslaved Homeplaces. In Archaeology of Atlantic Africa and the African Diaspora, edited by Akinwumi Ogundiran and Toyin Falola, pp. 233-248. Indiana University Press, Bloomington. 2010 Sweeping Spirits: Power and Transformation on the Plantation Landscape. In Archaeology and Preservation of Gendered Landscapes, edited by Sherene Baugher and Suzanne M. Spencer-Wood, pp. 81-94. Springer, New York. Beaudry, Mary C., Lauren J. Cook, and Stephen A. Mrozowski 1991 Artifacts and Active Voices: Material Culture as Social Discourse. In The Archaeology of Inequality, edited by Randall H. McGuire and Robert Paynter, pp. 150-180. Basil Blackwell, Oxford. Berlin, Ira 1998 Many Thousands Gone: The First Two Centuries of Slavery in North America. Harvard University Press, Cambridge, Massachusetts. Bevan, Bruce 1994 Remote Sensing of Gardens and Fields. In The Archaeology of Garden and Field, edited by Naomi F. Miller and Kathryn L. Gleason, pp. 70- 90. University of Pennsylvania Press, Philadelphia. Blair, John E., Matthew David Cochran, Stephanie N. Duensing 2009 Phase II Archaeological Testing on Wye Greenhouse (18TA314), Talbot County, Maryland, 2008, for the Tilghman Family. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Blair, John E. and Stephanie N. Duensing 2009 Phase II Archaeological Testing on the Interior of the Wye Greenhouse (18TA314), Talbot County, Maryland, 2009, for the Tilghman Family. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Blair, John E., Stephanie N. Duensing, Matthew David Cochran, Lisa Kraus, Michael Gubisch 377 2009 Archaeological Excavations on the Long Green (18TA314), 2005- 2008, Talbot County, Maryland, 2009, for the Tilghman Family. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Bolstad, Paul 2008 GIS Fundamentals: A First Text on Geographic Information Systems. 3rd edition. Eider Press, White Bear Lake, Minnesota. Brandt, Lydia Mattice 2009 Re‐Creating Mount Vernon: The Virginia Building at the 1893 Chicago World’s Columbian Exposition. Winterthur Portfolio 43(1):79–114. Brooker, Graham 2009 Introduction to Sensors for Ranging and Imaging. SciTech, Raleigh, North Carolina. Brown, Kenneth L. 2004 Ethnographic Analogy, Archaeology, and the African Diaspora: Perspectives from a Tenant Community. Historical Archaeology 38(1):79–89. Brown, Marley R. and Patricia M. Samford 1990 Recent Evidence of Eighteenth-Century Gardening in Williamsburg, Virginia. In Earth Patterns: Essays in Landscape Archaeology, William M. Kelso and Rachel Most, editors, pp. 104-121. University Press of Virginia. Charlottesville. Butler, David L. 2001 Whitewashing Plantations: The Commodification of a Slave-Free Antebellum South. International Journal of Hospitality and Tourism Administration 2(3-4):163–175. Bryne, Denis R. 2003 Nervous Landscapes: Race and Space in Australia. Journal of Social Archaeology 3(2): 169–193. Campbell, Colin 2016 As Confederate Symbols Come down, 'Talbot Boys' Endures. Baltimore Sun 12 July. Baltimore, Maryland. www.baltimoresun.com/news/maryland/eastern-shore/bs-md-talbot- boys-memorial-20160516-story.html, accessed March 11, 2019. Cangelosi, Robert J. 378 2013 Colonial Revival Architecture. In New Encyclopedia of Southern Culture: Volume 21: Art and Architecture, edited by Judith H. Bonner and Estill Curtis Pennington. University of North Carolina Press, Chapel Hill. Castañeda Quetzil E. 1996 In the Museum of Maya Culture: Touring Chichén Itzá. University of Minnesota Press, Minneapolis. Challis, Keith, Chris Carey, Mark Kincey, and Andy J Howard 2011 Airborne Lidar Intensity and Geoarchaeological Prospection in River Valley Floors. Archaeological Prospection 18(1):1-13. Chappell, Edward A. 1999 Museums and American Slavery. In “I, Too, Am America": Archaeological Studies of African-American Life, edited by Theresa A. Singleton, pp. 240–258. University Press of Virginia, Charlottesville. Chase, Arlen F,, Diane Z. Chase, Christopher T. Fisher, Stephen J. Leisz, and John F. Weishampel 2012 Geospatial Revolution and Remote Sensing Lidar in Mesoamerican Archaeology. Proceedings of the National Academy of Sciences of the United States of America 109(32):12916–12921. Chowdhury, Amitava 2014 Maroon Archaeological Research in Mauritius and Its Possible Implications in a Global Context. In The Archaeology of Slavery: A Comparative Approach to Captivity and Coercion, edited by Lydia Wilson Marshall, pp. 299-323. Southern Illinois University Press. Carbondale. Conard, Rebecca 2001 Applied Environmentalism, or Reconciliation Among ‘The Bios’ and ‘The Culturals’. The Public Historian 23(2):9-18. Conolly, James, and Mark Lake 2006 Geographical Information Systems in Archaeology. Cambridge Manuals in Archaeology. Cambridge University Press, United Kingdom. Conzen, Michael P. 1969 Spatial Data from Nineteenth Century Manuscript Censuses: A Technique for Rural Settlement and Land Use Analysis. The Professional Geographer 21(5):337–543. Craton, Michael 379 1982 Testing the Chains: Resistance to Slavery in the British West Indies. Cornell University Press, Ithaca, New York. Crews, Ed 2005 Making, Baking, and Laying Bricks. CW Journal Winter 2005-2006. Crutchley, Simon 2006 Light Detection and Ranging (Lidar) in the Witham Valley, Lincolnshire: An Assessment of New Remote Sensing Techniques. Archaeological Prospection 13(4):251–257. Cuddy, Thomas W. and Mark P. Leone 2008 New Africa: Understanding the Americanization of African Descent Groups through Archaeology. In Collaboration in Archaeological Practice: Engaging Descendant Communities, edited by Chip Colwell- Chanthaphonh and T.J. Ferguson, pp. 203-224. Alta Mira Press, New York. Dann, Graham M.S. and Robert B. Potter 2001 Supplanting the Planters: Hawking Heritage in Barbados. In Slavery, Contested Heritage and Thanatourism, edited by Graham M. S. Dann and A. V. Seaton, pp. 51-84. Haworth Press. New York. Dann, Graham M. S. and A. V. Seaton (editors) 2001 Slavery, Contested Heritage and Thanatourism. Haworth Press. New York. DeBats, Don 2015 Voting Viva Voce: Unlocking the Social Logic of Past Politics. Electronic document, sociallogic.iath.virginia.edu, accessed March 12, 2019. Deetz, James 1996 In Small Things Forgotten: The Archaeology of Early American Life. Originally published 1977. Doubleday, New York. 1990 Landscapes as Cultural Statements. In Earth Patterns: Essays in Landscape Archaeology, edited by William M. Kelso and Rachel Most, pp. 1-4. University Press of Virginia. Charlottesville. Delle, James A. 1998 An Archaeology of Social Space: Analyzing Coffee Plantations in Jamaica's Blue Mountains. Contributions to Global Historical Archaeology. Plenum Press, New York. Dilworth, William H. 380 1858 Map of Talbot County. The Easton Star 20 April: Page 3, Column 1, Easton, Maryland. Dilworth, William H. 1858 Map of Talbot County; with farm limits. Map. Rae Smith Engraver and Printer, New York. Electronic document, https://www.loc.gov/item/2002624021, accessed March 12, 2019. Dincauze, Dena Ferran 2000 Environmental Archaeology: Principles and Practice. Cambridge University Press, United Kingdom. Dong, Pinliang, and Qi Chen 2018 Lidar Remote Sensing and Applications. Remote Sensing Applications. CRC Press, Taylor & Francis Group, Boca Raton, Florida. Douglass, Frederick 1845 Narrative of the Life of Frederick Douglass, an American Slave. Anti-Slavery Office, Boston. 1855 My Bondage and My Freedom. Miller, Orton & Mulligan, New York. 1881 Life and Times of Frederick Douglass. Park, Hartford, Connecticut. Durant, Thomas J, Jr. and J. David Knottnerus (editors) 1999 Plantation Society and Race Relations: The Origins of Inequality. Praeger, Westport, Connecticut. Eichstedt, Jennifer L. and Stephen Small 2002 Representations of Slavery: Race and Ideology in Southern Plantation Museums. Smithsonian Institution Press. Washington, DC. Elwood, Sarah 2008 Volunteered geographic information: future research directions motivated by critical, participatory, and feminist GIS. GeoJournal 72(3-4):173-183. Epperson, Terrence W. 1990 Race and the Disciplines of the Plantation. Historical Archaeology 24(4): 29-36. 1999 Constructing Difference: The Social and Spatial Order of the Chesapeake Plantation. In "I, Too, Am America": Archaeological Studies of African-American Life, edited by Theresa A. Singleton, pp. 159-172. University of Virginia Press, Charlottesville. 381 2001 ‘A Separate House for the Christian Slaves, One for the Negro Slaves’: The Archaeology of Race and Identity in Late Seventeenth-Century Virginia. In Race and the Archaeology of Identity, edited by Charles E. Orser Jr., pp. 54-70. University of Utah Press, Salt Lake City. Ernstein, Julie Hevener 2004 Constructing context: Historical archaeology and the pleasure garden in Prince George’s County, Maryland, 1740-1790. PhD dissertation, Graduate School of Arts and Sciences, Boston University, Boston. Estes, John E., John R. Jensen, and Larry R. Tinney 1977 The Use of Historical Photography for Mapping Archaeological Sites. Journal of Field Archaeology 4(4):441–447. Fairbanks, Charles H. 1984 The Plantation Archaeology of the Southeastern Coast. Historical Archaeology 18(1):1-14. Feder, Kenneth L. 1994 A Village of Outcasts: Historical Archaeology and Documentary Research at the Lighthouse Site. Mayfield, Mountain View, California. Fellows, Kristen R., and James A. Delle 2015 Marronage and the dialectics of spatial sovereignty in colonial Jamaica. In Current Perspectives on the Archaeology of African Slavery in Latin America, edited by Pedro Paulo A. Funari, pp. 117- 132. Springer, New York. Fennell, Christopher 2007 Crossroads and Cosmologies: Diasporas and Ethnogenesis in the New World. Cultural Heritage Studies. University Press of Florida, Gainesville. Ferguson, Leland G. 1992 Uncommon Ground: Archaeology and Early African America, 1650- 1800. Smithsonian Institution Press, Washington, DC. Fernandez-Diaz, Juan Carlos, William E. Carter, Ramesh L. Shrestha, and Craig L. Glennie 2014 Now You See It… Now You Don’t: Understanding Airborne Mapping Lidar Collection and Data Product Generation for Archaeological Research in Mesoamerica. Remote Sensing 6(10):9951–10001. Foner, Eric 382 1999 The Strange Career of the Reconstruction Amendments. Yale Law Journal 108(8)2003-2010. Forman, Henry Chandlee 1934 Early Manor and Plantation Houses of Maryland: An Architectural and Historical Compendium, 1634-1800, with Three Hundred Twenty Photographs and One Hundred Forty-Five Plans, Details, and Sketches. Printed privately for the author, Easton, Maryland. 1956 Tidewater Maryland Architecture and Gardens. Bonanza Books, New York. 1967 Old Buildings, Gardens, and Furniture in Tidewater Maryland. Tidewater, Cambridge, Maryland. 1968 Maryland Architecture: A Short History from 1634 through the Civil War. Tidewater, Cambridge, Maryland. Fortin, Marcel, and Jennifer Bonnell (editors) 2014 Historical GIS Research in Canada. Canadian History and Environment Series, 2. University of Calgary Press, Alberta. Foucault, Michel 1977 Discipline and Punish: The Birth of the Prison. Pantheon Books, New York. Fowles, Severin 2010 The Southwest School of Landscape Archaeology. Annual Review of Anthropology 39:453-468. Franklin, Maria 1997 “Power to the People”: Sociopolitics and the Archaeology of Black Americans. Historical Archaeology 31(3):36-50. Gaffney, Vincent L, Zoran Stančič, and Helen Watson 1995 The Impact of GIS on Archaeology: A Personal Perspective. In Archaeology and Geographical Information Systems: A European Perspective, edited by Gary Lock and Zoran Stančič, pp. 211-229. Taylor and Francis, London. Gillings, Mark, David J. Mattingly, and Jan van Dalen 1999 Geographical Information Systems and Landscape Archaeology. Oxbow Books, Oxford, United Kingdom. Glassie, Henry 1975 Folk Housing in Middle Virginia. University of Tennessee Press. Knoxville. Goffman, Erving 383 1959 Presentation of Self in Everyday Life. Doubleday, New York. 1967 Interaction Ritual; Essays in Face-to-face Behavior. Aldine, Chicago. Gottschalk, L. 1945 Effects of Soil Erosion on Navigation in Upper Chesapeake Bay. Geographical Review 35(2):219-238. Goucher, Candice and Kofi Agorsah 2014 Excavating the Roots of Resistance: The Significance of Maroons in Jamaican Archaeology. In Out of Many, One People: The Historical Archaeology of Colonial Jamaica, edited by James A. Delle, Mark W. Hauser, and Douglas W. Armstrong, pp. 144-160. University of Alabama Press, Tuscaloosa. Grammer, Benedikt, Erich Draganits, Martin Gretscher, and Ulrike Muss 2017 Lidar-Guided Archaeological Survey of a Mediterranean Landscape: Lessons from the Ancient Greek Polis of Kolophon (Ionia, Western Anatolia). Archaeological Prospection 24(4):311–333. Gregory, Ian, and Humanities Data Service 2003 A Place in History: A Guide to Using GIS in Historical Research. History Data Service. Oxbow Books, Oxford, United Kingdom. Gregory, Ian, and Paul S Ell 2007 Historical GIS: Technologies, Methodologies and Scholarship. Cambridge Studies in Historical Geography, 39. Cambridge University Press, United Kingdom. Gregory, Ian N, and A Geddes 2014 Toward Spatial Humanities: Historical GIS and Spatial History. The Spatial Humanities. Indiana University Press, Bloomington. Gregory, Ian N, and Humphrey R Southall 2002 Mapping British Population History. In Past Time, Past Place: GIS for History, edited by Anne Kelly Knowles, pp. 117-130. ESRI Press, Redlands, California. Grillakis, Clio and Benjamin A. Skolnik 2012 Archaeological Applications of Remote Sensing and GIS: The Search for the Frederick Douglass Birthplace. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Gundaker, Grey and Judith McWillie 2005 No Space Hidden: The Spirit of African American Yard Work. University of Tennessee Press, Knoxville. 384 Hagley Museum and Library 2017 A Bird's Eye View of the Delaware Valley: The Dallin Aerial Survey Company. Electronic document, www.hagley.org/research/digital- exhibits/birds-eye-view-delaware-valley-dallin-aerial-survey- company, accessed March 11, 2019. Haley, Bryan S. 2009 A Geophysical Survey of Portions of the Wye House Grounds, Talbot County, Maryland. University of Mississippi Center for Archaeological Research. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Hammond, John Martin 1914 Colonial Mansions of Maryland and Delaware. J.B. Lippincott, Philadelphia, Pennsylvania. Handler, Richard and Eric Gable 1997 The New History in an Old Museum: Creating the Past at Colonial Williamsburg. Duke University Press, Durham, North Carolina. Hardwick, M. Jeff 1997 Homesteads and Bungalows: African-American Architecture in Langston, Oklahoma. Perspectives in Vernacular Architecture 6:21– 32. Harley, J.B. 1988 Maps, knowledge, and power. In Iconography of Landscape: Essays on the symbolic representation, design and use of past environments, edited by Denis Cosgrove and Stephen Daniels, pp. 129-148. Cambridge University Press, New York. Harmon, James M., Mark P. Leone, Stephen D. Prince, and Marcia Snyder 2006 LiDAR for Archaeological Landscape Analysis: A Case Study of Two Eighteenth-Century Maryland Plantation Sites. American Antiquity 71(4):649-70. Hartman, Saidiya V. 1997 Scenes of Subjection: Terror, Slavery, and Self-Making in Nineteenth- Century America. Oxford University Press, New York. Heath, Barbara 2012 A Brief History of Plantation Archaeology in Virginia. In Jefferson's Poplar Forest: Unearthing a Virginia Plantation, edited by Barbara J. Heath and Jack Gary, pp. 31-51. University Press of Florida, Gainesville. 385 Heath, Barbara J., and Amber Bennett 2000 ‘The Little Spots Allow’d Them’: The Archaeological Study of African-American Yards. Historical Archaeology 34(2):38–55. Heath, Barbara J, and Jack Gary (editors) 2012 Jefferson's Poplar Forest: Unearthing a Virginia Plantation. University Press of Florida, Gainesville. Hesse, Ralf 2010 Lidar-Derived Local Relief Models - a New Tool for Archaeological Prospection. Archaeological Prospection 17(2):67–72. Hicks, Dan 2005 ‘Places for Thinking’ from Annapolis to Bristol: Situations and Symmetries in ‘World Historical Archaeologies’. World Archaeology 37(3):373-391. Hodder, Ian 1986 Reading the Past: Current Approaches to Interpretation in Archaeology. Cambridge University Press, Cambridge. Hodder, Ian and Scott Hutson 2003 Reading the Past: Current Approaches to Interpretation in Archaeology. Third Edition. Cambridge University Press, New York. Hoelscher, Steven 2003 Making Place, Making Race: Performances of Whiteness in the Jim Crow South. Annals of the Association of American Geographers 93(3):657–686. Hoskins, W.G. 1955 The Making of the English Landscape. Hodder & Stoughton. London. Howard, McHenry 1922 Lloyd Graveyard at Wye House, Talbot County, Maryland. Maryland Historical Magazine 17(1):20–33. 1923 Wye House, Talbot County Maryland. Maryland Historical Magazine 18(4):293–299. Hritz, Carrie 2014 Contributions of GIS and Satellite-Based Remote Sensing to Landscape Archaeology in the Middle East. Journal of Archaeological Research 22(3):229–276. Isaac, Rhys 1988 The Transformation of Virginia 1740-1790. W. W. Norton, New York. 386 Jackson, Antoinette T 2008 Imagining Jehossee Island Rice Plantation Today. International Journal of Heritage Studies 14(2):131-155. Jacobucci, Susan A. and Heather B. Trigg 2010 An Analysis of Pollen Recovered From the Greenhouse at the Wye House Plantation, Easton, Maryland. Andrew Fiske Memorial Center for Archaeological Research at University of Massachusetts Boston. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Jenkins, Tracy 2015 Phase I and II Archaeological Testing at the Talbot County Women’s Club, 18 Talbot Lane, Easton, Maryland, 18TA439. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Jenkins, Tracy H. and Benjamin A. Skolnik. 2013 Phase I and II Archaeological Testing at 321 and 323 South Street, Easton Maryland, Home of the Family of the Buffalo Soldier, 2012. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. John, Beverly M. 1999 The Construction of Racial Meaning by Blacks and Whites in Plantation Society. In Plantation Society and Race Relations: The Origins of Inequality, edited by Thomas J. Durant and J. David Knottnerus, pp. 41-52. Praeger, Westport, CT. Johnson, Katharine M, and William B Ouimet 2014 Rediscovering the Lost Archaeological Landscape of Southern New England Using Airborne Light Detection and Ranging (LiDAR). Journal of Archaeological Science 43(1):9–20. 2018 An Observational and Theoretical Framework for Interpreting the Landscape Palimpsest through Airborne Lidar. Applied Geography 91:32–44. Johnson, Matthew H. 2007 Ideas of Landscape: an Introduction. Blackwell, Oxford, United Kingdom. 2012 Phenomenological Approaches in Landscape Archaeology. Annual Review of Anthropology 41:269-284. Jones, Steven L. 387 1985 The African-American Tradition in Vernacular Architecture. In The Archaeology of Slavery and Plantation Life, edited by Theresa A. Singleton, pp. 195-215. Academic Press, New York. Joseph, J. W 2004 Columns and Rows, Allees and Quarters: The Archaeology of Plantation Landscapes and the Landscape of Plantation Ideology in the Lowcountry of South Carolina. South Carolina Antiquities 36(1-2):1- 12. Kelso, William M. 1982 Jefferson's Garden: Landscape Archaeology at Monticello. Archaeology 35(4):38-45. Kelso, William M., and Rachel Most (editors) 1990 Earth Patterns: Essays in Landscape Archaeology. University Press of Virginia, Charlottesville. Kerns, Michelle and James G. Gibb 1998 Wye House. 18TA314. Survey and Excavation Report. The Lost Towns of Anne Arundel Project. Anne Arundel County Department of Planning and Zoning. Kinney, William A. 1954 Roving Maryland’s Cavalier Country. National Geographic 105(4):431-470. King, Thomas F. 2004 Cultural Resource Laws and Practice: An Introductory Guide. 2nd edition. Heritage Resources Management Series. Altamira Press, Walnut Creek, California. Knowles, Anne Kelly (editor) 2002 Past Time, Past Place. ESRI Press. Redlands, California. Knowles, Anne Kelly, and Amy Hillier (editors) 2008 Placing history: how maps, spatial data, and GIS are changing historical scholarship. ESRI Press. Redlands, California. Kraus, Lisa 2006 Archaeology of the Red Overseer’s House. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Kwan, Mei-Po 388 2002 Feminist Visualization: Re-Envisioning GIS as a Method in Feminist Geographic Research. Annals of the Association of American Geographers 92(4):645-661. Land, Aubrey C. 1965 Economic Base and Social Structure: The Northern Chesapeake in the Eighteenth Century. Journal of Economic History 25(4):639-654. Langley, Batty 1728 New principles of gardening: or, The laying out and planting parterres, groves, wildernesses, labyrinths, avenues, parks, & c. Printed for A. Bettesworth and J. Batley, London. La Quintinie, Jean de 1704 The complete gard'ner: Or, directions for cultivating and right ordering of fruit-gardens and kitchen gardens. By Monsienr [sic] De la Quintinye. Now compendiously abridg'd, and made of more use, with very considerable improvements. By George London, and Henry Wise. Printed for Andrew Bell. London. LaRoche, Cheryl Janifer 2007 Resistance to Slavery in Maryland: Strategies for Freedom. Special History Study. National Underground Railroad Network to Freedom. Philadelphia, Pennsylvania. 2013 Free Black Communities and the Underground Railroad: the Geography of Resistance. University of Illinois Press. Urbana. Leone, Mark P. 1984 Interpreting Ideology in Historical Archaeology: Using Rules of Perspective in the William Paca Garden in Annapolis, Maryland. In Ideology, Representation and Power in Prehistory, edited by Christopher Tilley and Daniel Miller, pp. 25-35. Cambridge University Press, United Kingdom. 1988 The Georgian Order as the Order of Merchant Capitalism in Annapolis, Maryland. In The Recovery of Meaning: Historical Archaeology in the United States, edited by Mark P. Leone and Parker B. Potter, Jr, pp. 235-261. Smithsonian Institution Press. Washington, DC. 2005 The Archaeology of Liberty in an American Capital: Excavations in Annapolis. University of California Press, Berkeley. 2010 Critical Historical Archaeology. Left Coast Press. Walnut Creek, California. Leone, Mark P. and Silas D. Hurry 1998 Seeing: The Power of Town Planning in the Chesapeake. Historical Archaeology 32(4):34-62. 389 Leone, Mark P. and Paul A. Shackel 1987 Forks, Clocks, and Power. In Mirror and Metaphor, edited by Daniel Ingersoll and Gordon Bronitsky, pp. 45-61. University Press of America, Lanhamn, Maryland. 1990 Plane and Solid Geometry in Colonial Gardens in Annapolis, Maryland. In Earth Patterns: Essays in Landscape Archaeology, edited by William M. Kelso and Rachel Most, pp. 153-167. University Press of Virginia. Charlottesville. Leone, Mark P, Elizabeth Pruitt, Benjamin A Skolnik, Stefan Woehlke, and Tracy Jenkins 2018 The Archaeology of Early African American Communities in Talbot County, Eastern Shore, Maryland, U.S.A., and Their Relationship to Slavery. Historical Archaeology 52(4):753–772. Lewis, Kenneth E. 1985 Plantation Layout and Function in the South Carolina Lowcountry. In The Archaeology of Slavery and Plantation Life, edited by Theresa A. Singleton, pp. 35-66. Academic Press, New York. Little, Barbara, J. 2007 Historical Archaeology: Why the Past Matters. Left Coast Press. Walnut Creek, California. Llobera, Marcos 2003 Extending GIS-Based Visual Analysis: The Concept of Visualscapes. International Journal of Geographical Information Science 17(1):25–48. Lock, Gary 2003 Using Computers in Archaeology: Towards Virtual Pasts. Routledge, London. Lockwood, Alice G.B. 1934 Gardens of Colony and State: Gardens and Gardeners of the American Colonies and of the Republic before 1840. Charles Scribner’s Sons, New York. Luccketti, Nicholas 1990 Archaeological Excavations at Bacon’s Castle, Surry County, Virginia. In Earth Patterns: Essays in Landscape Archaeology, edited by William M. Kelso and Rachel Most, pp. 23-42. University Press of Virginia. Charlottesville. Malinsky-Buller, Ariel, Erella Hovers, and Ofer Marder 390 2011 Making Time: 'Living Floors', 'Palimpsests' and Site Formation Processes - a Perspective from the Open-Air Lower Paleolithic Site of Revadim Quarry, Israel. Journal of Anthropological Archaeology 30(2):89-101. Little, Margaret Ruth 2012 Getting the American Dream for Themselves: Postwar Modern Subdivisions for African Americans in Raleigh, North Carolina. Buildings & Landscapes: Journal of the Vernacular Architecture Forum 19(1):73-86. Martí-Henneberg, Jordi 2011 Geographical Information Systems and the Study of History. Journal of Interdisciplinary History 42(1):1-13. Martin, Kevin 2008 The Dallin Aerial Survey Company Collection at the Hagley Museum and Library. Microform & Imaging Review 37(3):130–33. Maryland Historical Trust State Historic Sites Inventory (MIHP) 1973a Compton, Easton, Trappe, Maryland, T-146 1973b The Anchorage, Easton, Talbot, Maryland, T-52 1973c The Wilderness, Trappe, Talbot, Maryland, T-149 1975 Sherwood Manor, St. Michaels, Talbot, Maryland, T-244 1976 Otwell, Easton, Oxford, Maryland, T-164 1977 Hope House, Easton, Talbot, Maryland, T-90 1989 Wye House, Easton, Talbot, Maryland, T-54 1991a William Paca House, Annapolis, Annapolis, Maryland, AA-657 1991b Clay’s Hope, Bellevue, Talbot, Maryland, T-189 1992 Myrtle Grove, Easton, Talbot, Maryland, T-53 2003 Ratcliffe Manor, Easton, Talbot, Maryland, T-42 Maryland-National Capital Park and Planning Commission 2009 Antebellum Plantations in Prince George’s County, Maryland. Prince George’s County Planning Department. Upper Marlboro, Maryland. Matthews, Christopher and Matthew Palus 2007 A Landscape of Ruins: Building Historic Annapolis. In Envisioning Landscape: Situations and Standpoints in Archaeology and Heritage, edited by Dan Hicks, Laura McAtackney and Graham Fairclough, pp. 226-250. Berg, Oxford, United Kingdom. McCoy, Mark D. and Thegn N. Ladefoged 2009 New Developments in the Use of Spatial Technology in Archaeology. Journal of Archaeological Research 17(3):263-295. 391 McDavid, Carol 1997 Descendants, Decisions, and Power: The Public Interpretation of the Archaeology of the Levi Jordan Plantation. Historical Archaeology 31(3):114–131. McEwan, Dorothy Graves 2012 Qualitative Landscape Theories and Archaeological Predictive Modelling—a Journey through No Man's Land? Journal of Archaeological Method and Theory 19(4):526–547. McKee, Larry 1992 The Ideals and Realities Behind the Design and Use of 19th Century Virginia Slave Cabins. The Art and Mystery of Historical Archaeology: Essays in Honor of James Deetz, edited by Anne E. Yentsch and Mary C. Beaudry, pp. 195-213. CRC Press. McNulty, Anne 2012 A Place Called Wye House. What's Up? Eastern Shore July, pp. 48– 51. Annapolis, Maryland. Meager, Leonard 1697 The new art of gardening: With the gardener' almanack : containing the true art of gardening in all its particulars ... : to each head is added an almanack, shewing what is to be done every month in the year. Printed for Henry Nelme, London. Melnick, Robert Z. 2000 Considering Nature and Culture in Historic Landscape Preservation. In Preserving Cultural Landscapes in America, edited by Arnold R. Alanen and Robert Z. Melnick. Johns Hopkins University Press, Baltimore, Maryland. Menze, Bjoern H., and Jason A. Ur 2012 Mapping Patterns of Long-Term Settlement in Northern Mesopotamia at a Large Scale. Proceedings of the National Academy of Sciences of the United States of America 109(14):778–787. Miles, Tiya 2010 The House on Diamond Hill: A Cherokee Plantation Story. University of North Carolina Press, Chapel Hill. Miller, Naomi Frances, and Kathryn L Gleason (editors) 1994 The Archaeology of Garden and Field. University of Pennsylvania Press, Philadelphia. Miller, Philip 392 1735 The gardeners dictionary: Containing the methods of cultivating and improving all sorts of trees, plants, and flowers, for the kitchen, fruit, and pleasure gardens; as also those which are used in medicine. With directions for the culture of vineyards, and making of wine in England. In which likewise are included the practical parts of husbandry. Printed for the author, London. Modlin, E. Arnold, Jr. 2008 Tales Told on the Tour: Mythic Representations of Slavery by Docents at North Carolina Plantation Museums. Southeastern Geographer 48(3):265-287. Moyer, Teresa S. 2015 Ancestors of Worthy Life: Plantation Slavery and Black Heritage at Mount Clare. Cultural Heritage Studies. University Press of Florida, Gainesville. Mukerji, Chandra 1997 Territorial Ambitions and the Gardens of Versailles. Cambridge University Press. Cambridge. Myers, Adrian 2010 Camp Delta, Google Earth and the Ethics of Remote Sensing in Archaeology. World Archaeology 42(3):455–67. National Parks Service 1999 Section 1: How to Prepare National Historic Landmark, National Register of Historic Places Bulletin. National Register Publications, National Park Service, U.S. Department of the Interior. Electronic document, www.nps.gov/nr/publications/bulletins/nhl/nhlpt1.htm, accessed 11 March 2019. 2018 National Register of Historic Places (U.S. National Park Service), U.S. Department of the Interior. Electronic Document, www.nps.gov/subjects/nationalregister/index.htm, accessed December 3, 2018. National Register of Historic Places (NRHP) 1971 Paca House and Garden, Annapolis, Annapolis, Maryland, NR ID 71000364. 1973a Compton, Trappe, Talbot, Maryland, NR ID 74000970. 1973b Myrtle Grove, Easton, Talbot, Maryland, NR ID 74000967. 1973c The Anchorage, Easton, Talbot, Maryland, NR ID 74000968. 1973d The Wilderness, Trappe, Talbot, Maryland, NR ID 74000971. 1975 Sherwood Manor, St. Michaels, Talbot, Maryland, NR ID 77000701. 1976 Clay’s Hope, Bellevue, Talbot, Maryland, NR ID 79001142. 1977 Hope House, Easton, Talbot, Maryland, NR ID 79001143. 393 1980 Otwell, Oxford, Talbot, Maryland, NR ID 82002819. 2009 Wye House, Easton, Talbot, Maryland, NR ID 70000264. Nelson, Robert K, Edward L. Ayers, Justin Madron, Nathaniel Ayers, and Scott Nesbit 2017 The Forced Migration of Enslaved People in the United States: 1810- 1860, Visualizing Emancipation. Electronic document, http://dsl.richmond.edu/forcedmigration.html, accessed March 12, 2019. Neustupný, Evžen 1995 Beyond GIS. In Archaeology and Geographical Information Systems: A European Perspective, edited by Gary Lock and Zoran Stančič, pp. 133-140. Taylor and Francis, London. Nicholas, George P. 2006 Decolonizing the Archaeological Landscape: The Practice and Politics of Archaeology in British Columbia. American Indian Quarterly 30(3- 4):350-380. Nicholas, Wendy 2010 Collaborating to Save Whole Places. Forum Journal 25(1):7-10. Norman, Neil L. and Kenneth G. Kelly 2004 Landscape Politics: The Serpent Ditch and the Rainbow in West Africa. American Anthropologist 106(1):98-110. Orser, Charles E. 1996 A Historical Archaeology of the Modern World. Contributions to Global Historical Archaeology. Plenum Press, New York. Orser, Charles, and Pedro Funari 2001 Archaeology and Slave Resistance and Rebellion. World Archaeology 33(1):61–72. O’Sullivan, David 2006 Geographical Information Science: Critical GIS. Progress in Human Geography, 30(6):783–791. Paca, Barbara, Leslie King-Hammond, and Nina Khrushcheva 2015 Ruth Star Rose (1887-1965): Revelations of African American Life in Maryland and the World. Reginald F. Lewis Museum of Maryland African American History & Culture. Paca-Steele, Barbara 1986 The Mathematics of an Eighteenth-Century Wilderness Garden. The 394 Journal of Garden History 6(4):299–320. Palmer, Catherine A. 1994 Tourism and Colonialism: The Experience of the Bahamas. Annals of Tourism Research 21(4):792–811. Pavlovskaya, Marianna 2018 Critical GIS As a Tool for Social Transformation. Canadian Geographer 62(1):40-54. Pickering, E.H. 1936 Wye House, Corn Crib, Bruffs Island Road, Tunis Mills, Talbot County, MD, photograph for the Historic American Buildings Survey (HABS). Library of Congress Prints and Photographs Division, Washington DC. Poster, Mark 1982 Foucault and History. Social Research 49(1):116-142. Powlesland, Dominic, James Lyall, Guy Hopkinson, Danny Donoghue, Maria Beck, Aidan Harte, and David Stott 2006 Beneath the Sand—Remote Sensing, Archaeology, Aggregates and Sustainability: A Case Study from Heslerton, the Vale of Pickering, North Yorkshire, UK. Archaeological Prospection 13(4):291–299. Preston, Dickson J. 2018 Young Frederick Douglass. Johns Hopkins University Press, Baltimore, Maryland. Preston, Dickson J, and Norman Harrington 1983 Talbot County: A History. Tidewater, Centerville, Maryland. Pruitt, Beth 2013 Phase II Archaeological Testing of the Hothouse Structure (18TA314), Talbot County, Maryland, May 2012. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. 2015 Reordering the Landscape: Science, Nature, and Spirituality at Wye House. PhD dissertation, Department of Anthropology, University of Maryland, College Park. Pruitt, Beth and Benjamin Skolnik 2013 Challenging Landscapes: Alternate Perspectives of Chesapeake Plantation Gardens. Paper presented at the Society for Historical Archaeology conference, Leicester, England. Rabinow, Paul 395 1984 The Foucault Reader. Pantheon Books. New York. Rahier, Jean Muteba, and Michael Hawkins 1999 ‘Gone with the Wind’ versus the Holocaust Metaphor: Louisiana Plantation Narratives in Black and White. In Plantation Society and Race Relations: The Origins of Inequality, edited by Thomas J. Durant and J. David Knottnerus, pp. 205-233. Praeger, Westport, Connecticut. Reich, Jerome R. 1998 Colonial America. Prentice Hall. Upper Saddle River, New Jersey. Renfrew, Colin and Paul Bahn 2004 Archaeology: Theories, Methods and Practice. 4th edition. Thames & Hudson, London. Robertson, James 2009 Giving Directions in Spanish Town, Jamaica: Comprehending a Tropical Townscape. Journal of Urban History 35(5):718–742. Robson, T K. 1858 Map of Talbot County. The Easton Star 20 April: Page 2, Column 7, Easton, Maryland. Rosenswig, Robert M, López-Torrijos Ricardo, and Caroline E Antonelli 2015 Lidar Data and the Izapa Polity: New Results and Methodological Issues from Tropical Mesoamerica. Archaeological and Anthropological Sciences 7(4):487–504. Royce, Edward Cary 1993 The Origins of Southern Sharecropping. Labor and Social Change. Temple University Press, Philadelphia, Pennsylvania. Rua, Helena, and Pedro Alvito 2011 Living the Past: 3d Models, Virtual Reality and Game Engines As Tools for Supporting Archaeology and the Reconstruction of Cultural Heritage - the Case-Study of the Roman Villa of Casal De Freiria. Journal of Archaeological Science 38(12):3296–3308. Ruffins, Faith Davis 2006 Revisiting the Old Plantation: Reparations, Reconciliation, and Museumizing American Slavery. In Museum Frictions: Public Cultures/Global Transformations, edited by Ivan Karp. Duke University Press. Durham, North Carolina. Russo, Jean Burrell, and J. Elliott Russo 396 2012 Planting an Empire: The Early Chesapeake in British North America. Regional Perspectives on Early America. Johns Hopkins University Press, Baltimore, Maryland. Samford, Patricia 2007 Subfloor Pits and the Archaeology of Slavery in Colonial Virginia. University of Alabama Press, Tuscaloosa. Sarudy, Barbara Wells 1998 Gardens and Gardening in the Chesapeake, 1700-1805. Johns Hopkins University Press, Baltimore, Maryland. Sayers, Daniel O. 2014 A Desolate Place for a Defiant People: The Archaeology of Maroons, Indigenous Americans, and Enslaved Laborers in the Great Dismal Swamp. University Press of Florida, Gainesville. Scarborough, Katherine. 1930 Homes of the Cavaliers. Macmillan, New York. Schindling, James, and Cerian Gibbes 2014 Lidar As a Tool for Archaeological Research: A Case Study. Archaeological and Anthropological Sciences 6(4):411–23. Schnore, Leo Francis, and Martyn J Bowden 1975 The New Urban History: Quantitative Explorations by American Historians. Quantitative Studies in History. Princeton University Press, Princeton, New Jersey. Schuurman, Nadine 1999 Critical GIS: Theorizing an Emerging Science. Cartographica 36(4):7-99. Scollar, Irwin 1975 Some Pitfalls in the Interpretation of Air Photographs. Aerial Reconnaissance for Archaeology, edited by David R. Wilson, pp. 59- 69. Council for British Archaeology, London. Seaton, A.V. 2001 Sources of Slavery—Destinations of Slavery: The Silences and Disclosures of Slavery Heritage in the UK and US. In Slavery, Contested Heritage and Thanatourism, Graham M. S. Dann and A. V. Seaton (editors), pp. 107-130. Haworth Press, New York. Shackel, Paul A. 397 1993 Personal Discipline and Material Culture: An Archaeology of Annapolis, Maryland, 1695-1870. University of Tennessee Press, Knoxville. Shanks, Michael, and Christopher Tilley 1987 Re-Constructing Archaeology: Theory and Practice. New Studies in Archaeology. Cambridge University Press, London. Shanks, Michael and Tilley, Christopher 1982 Ideology, Symbolic Power and Ritual Communication: A Reinterpretation of Neolithic Mortuary Practices. In Symbolic and Structural Archaeology, edited by Ian Hodder, pp. 129-154. Cambridge University Press, London. Shannahan, John Henry Kelley 1907 Tales of Old Maryland: History and Romance on the Eastern Shore of Maryland. Press of Meyer & Thalheimer, Baltimore, Maryland. Shih, Peter Tian-Yuan, Ya-Hsing Chen, and Jie-Chung Chen 2014 Historic Shipwreck Study in Dongsha Atoll with Bathymetric Lidar. Archaeological Prospection 21(2):139–146. Singleton, Theresa A. 1985 The Archaeology of Slavery and Plantation Life. Studies in Historical Archaeology. Academic Press, Orlando, Florida. 2001 Slavery and Spatial Dialectics on a Cuban Coffee Plantation. World Archaeology, 33(1):98-114. Singleton, Theresa A. and Mark D. Bograd 1995 The Archaeology of the African Diaspora in the Americas. Guides to the Archaeological Literature of the Immigrant Experience in America, No. 2. Society for Historical Archaeology, Ann Arbor, Michigan. Skolnik, Benjamin A. 2011 Shovel Test Survey at Wye House (18TA314), East Cove / South Long Green, April 2011. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. 2015 Phase I Archaeological Testing at the Bailey Homestead, Tuckahoe, Talbot County, Maryland, 2013. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Skolnik, Benjamin and Beth Pruitt 2019 Phase I and II Archaeological Testing of Two Structures at Wye House (18TA314), Talbot County, Maryland. 2-Story Quarter and 398 Brick Row Quarter. 2011-2014. Manuscript in preparation, Archaeology in Annapolis, College Park, Maryland. Smith, Monica L. 2014 The Archaeology of Urban Landscapes. Annual Review of Anthropology 43:307-323. South, Stanley 1968 Restoration Archaeology at the Paca House, Annapolis, Maryland. The Conference on Historic Site Archaeology Papers 1967 2(1):23-32. Stanton, Lucia C. 2012 "Those Who Labor for My Happiness": Slavery at Thomas Jefferson's Monticello. Jeffersonian America. University of Virginia Press, Charlottesville. Talmage, Valerie 2010 Lessons for Land Conservation. Forum Journal 25(1):11-17. Tang, Amanda 2012 Archaeological Excavations at the Middle Building (Locus 2) on the Long Green, (18TA314), 2006-2010, Talbot County, Maryland. With contributions by Jocelyn Knauf and Benjamin Skolnik. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. 2014 ‘Fried Chicken Belongs to All of Us’: The Zooarchaeology of Enslaved Foodways on the Long Green, Wye House (18TA314), Talbot County, Maryland. PhD dissertation, Department of Anthropology, University of Maryland, College Park. Tang, Amanda and Jocelyn Knauf 2010 Archaeological Mitigation of the Great House Front Steps (18TA314), Talbot County, Maryland, 2010. University of Maryland, College Park. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Taylor, Alan 2001 American Colonies. The Penguin History of the United States. Viking, New York. Thompson, Mary V. 2019 "The Only Unavoidable Subject of Regret": George Washington, Slavery, and the Enslaved Community at Mount Vernon. University of Virginia Press, Charlottesville. Tilghman, J. Donnell 1953 Wye House. Maryland Historical Magazine 48(2):89–108. 399 Tilghman, Oswald, and Samuel Alexander Harrison 1915 History of Talbot County, Maryland, 1661-1861. Williams & Wilkins, Baltimore, Maryland. Tilley, Christopher Y. 1994 A Phenomenology of Landscape: places, paths, and monuments. Berg, Providence, Rhode Island. Tischauser, Leslie Vincent 2012 Jim Crow Laws. Landmarks of the American Mosaic. Greenwood, Santa Barbara, California. Trussell, Timothy 2012 A Landscape for Mr. Jefferson’s Retreat. Jefferson's Poplar Forest: Unearthing a Virginia Plantation, edited by Barbara J. Heath and Jack Gary, pp. 72-85. University Press of Florida, Gainesville. United States Agriculture Stabilization and Conservation Service 1937 Aerial Photography Index. Scale ca. 1:20,000. https://jscholarship.library.jhu.edu/handle/1774.2/36481, accessed March 13, 2019. United States Bureau of the Census 1860 Eighth Census, U.S.; Instructions &c. Instructions to U.S. Marshals. Instructions to Assistants. Census Office, Department of the Interior. 1864 Population of the United States in 1860; Compiled from the Original Returns of the Eighth Census under the Direction of the Secretary of the Interior. Census Office, Department of the Interior. United States Census Bureau 2017 U.S. Census Bureau QuickFacts: Talbot County, Maryland. Census Bureau QuickFacts. Electronic document, www.census.gov/quickfacts/talbotcountymaryland, accessed March 11, 2019. 2018 1860 Census: Population of the United States. United States Census Bureau, U.S. Department of Commerce. Electronic document, https://www.census.gov/library/publications/1864/dec/1860a.html, accessed 11 March 2019. United States Coast Survey USCS 1847 No 18. Map of St. Michaels River, Third Haven Creek, Easton &c., T- Sheet. U.S. Coast Survey, Washington, DC. Electronic document, https://historicalcharts.noaa.gov/historicals/preview/image/T00224-00- 1847, accessed March 3, 2019. 1862 Chesapeake Bay from Its Head to Potomac River: No. 35 From 400 Magothy River to Choptank River, Nautical Chart. U.S. Coast Survey, Washington, DC. Electronic document, https://historicalcharts.noaa.gov/historicals/preview/image/LC00135 _00_1862, accessed March 3, 2019. Upton, Dell. 1984 White and Black Landscapes in Eighteenth-Century Virginia. Places 2(2):59-72. 1990 Imagining the Early Virginia Landscape. In Earth Patterns: Essays in Landscape Archaeology, edited by William M. Kelso and Rachel Most, pp. 71-86. University Press of Virginia. Charlottesville. Vlach, John Michael 1993 Back of the Big House: The Architecture of Plantation Slavery. The Fred W. Morrison Series in Southern Studies. University of North Carolina Press, Chapel Hill. Walsh, Lorena Seebach 1997 From Calabar to Carter's Grove: The History of a Virginia Slave Community. Colonial Williamsburg Studies in Chesapeake History and Culture. University Press of Virginia, Charlottesville. Weeks, Christopher 1984 Where Land and Water Intertwine: An Architectural History of Talbot County, Maryland. Johns Hopkins University Press, Baltimore, Maryland. Weik, Terry 1997 The Archaeology of Maroon Societies in the Americas: Resistance, Cultural Continuity, and Transformation in the African Diaspora. Historical Archaeology 31(2):81-92. Wescott, Konnie, and R. Joe Brandon (editors) 2000 Practical Applications of GIS for Archaeologists: A Predictive Modeling Toolkit. Taylor and Francis, London. Wheatley, David, and Mark Gillings 2002 Spatial Technology and Archaeology: The Archaeological Applications of GIS. Taylor & Francis, New York. White, Esther C. 2016 ‘As It Was Originally Laid Out by the General’: George Washington and His Upper Garden. Historical Archaeology 50(1):45–60. Wilkie, Laurie A. 401 2004 Considering the Future of African American Archaeology. Historical Archaeology 38(1):109–23. Wilson, Matthew W. 2017 New Lines: Critical GIS and the Trouble of the Map. University of Minnesota Press, Minneapolis. Wiseman, James, and Farouk El-Baz 2007 Remote Sensing in Archaeology. Interdisciplinary Contributions to Archaeology. Springer, New York. Witte, Brian 2011 Statue of Frederick Douglass Arrives after Years of Debate. NBCUniversal News Group, June 16, 2011. Electronic document, www.nbcnews.com/id/43430504/ns/travel-destination_travel/t/statue- frederick-douglass-arrives-after-years-debate, accessed March 13, 2019. Wolf, Edwin 1969 The Library of Edward Lloyd IV of Wye House. Winterthur Portfolio 5:87-121. Worthington, M.J. and D.W.H Miles 2007 The Tree-Ring Dating of the Red House, Wye Plantation, Easton, Talbot County, Maryland. Oxford Dendrochronology Laboratory. Manuscript on file, Archaeology in Annapolis, College Park, Maryland. Yamin, Rebecca, and Karen Bescherer Metheny (editors) 1996 Landscape Archaeology: Reading and Interpreting the American Historical Landscape. University of Tennessee Press, Knoxville. Yentsch, Anne E. and Judson M. Kratzer 1994 Techniques for Excavating and Analyzing Buried Eighteenth-Century Garden Landscapes. In The Archaeology of Garden and Field, edited by Naomi F. Miller and Kathryn L. Gleason, pp. 168-201. University of Pennsylvania. Philadelphia. Žižek, Slavoj 2008 Violence: Six Sideways Reflections. Picador, New York. 402