By Sarah Peterson

Future Projects:

During the semester there were many fantastic ideas proposed for final projects. I made attempts at two projects in particular. The projects went uncompleted for several reasons, but I believe they would provide valuable insight into the life and times of Tiwanaku. In sharing my collected work on Facial Reconstruction and Landscape modeling, my hope is that someday someone will continue the work I was unable to finish.

Facial Reconstruction:

In our very first class, Dr. Alexi Vranich mentioned a newly discovered collection of skulls from Tiwanaku as an example of the diverse research topics available to those interested in the site. This immediately caught the attention of both myself and Mollie Epstein as a potential project. By the end of September the two of us had declared our intention to work on a skeletal reconstruction project. Told to “dream big,” my original vision was that, using CT scans of the recovered, reconstructed (if necessary), skulls and computer imaging software, digital 3D facial compositions would be compiled. Ideally, touch-screen terminals would be set up in the exhibits at Tiwanaku each with a different face of an ancient inhabitant, visitor, or victim. By touching different symbols on the screen, a visitor would be able to ask the face for information about different aspects of the site, and the face would narrate their own life story through the places and things they encountered during their time at Tiwanaku. This project would be a direct fulfillment of our class goal to “People the Past” by literally establishing a dialogue between the past and the present.

The work looked promising during the beginning. Mollie and I met together after class to discuss the overall plan of our project. As a physical anthropologist, Mollie works closely with Dr. Janet Monge and agreed to discuss with her what would be the best course of action. I dove into research. I found Kathleen Reich and Emily Craig’s article “Facial Approximation: Procedures and Pitfalls” particularly useful and there was also a significant amount of material available online. On October 13th, Dr. P. Thomas Schoenemann came into our class to give presentation on his work with Penn’s Morton Collection, which was very encouraging. Dr. Schoenemann works with Dr. Monge on a project called The Penn Cranial CT Database Project, but is a professor at the University of Michigan.

The difficulties began in mid October. Though Dr. Monge and Dr. Schoenemann were excellent resources, we had difficulty arranging a meeting time because Dr. Schoenemann was in Michigan during the week. It was also becoming apparent that we would not actually have access to the new Tiwanaku skulls but would be able to do reconstructions from already owned Andean skulls. As a class, we read “Reanimating the Dead: Reconstruction of Expressive Faces from Skull Data” by Kolja Kähler, Jörg Haber and Hans-Peter Seidel and had a class discussion on the process and limitations of facial reconstruction. By the end of the discussion the class consensus seemed to be that facial reconstruction was not worthy of the effort since there was very little historical accuracy. However, the ultimate justification for putting off the facial reconstruction project was that we, as a class, were defining our overall project much more clearly and facial reconstruction was not essential to the class project. With the limited amount of time to complete the class project, priority had to be given to more essential tasks.

I continue to believe strongly that personal connection with the people of the past is the key to generating interest and spreading knowledge. For that reason, I hope that my vision will someday be completed, and I hope that the information I have shared here will be of use to any person who undertakes its completion.

http://grape.anthro.upenn.edu/%7Elab/pennct/

Taylor, Karen T. Forensic Art and Illustration. Boca Raton: CRC Press, 2001.

http://medialab.di.unipi.it/Project/Mummia/SIGGRAPH99/

http://www.buffalo.edu/news/fast-execute.cgi/article-page.html?article=52160009

Reichs, Kathleen J. and Emily Craig. “Facial Approximation: Procedures and Pitfalls.” Forensic Osteology: Advances in the Identification of Human Remains. Ed. Kathleen J. Reichs. Springfield: Charles C Thomas Publisher, Ltd. 1998. 491-513

Landscape Modeling:

Before I begin, I should apologize here for all the vocabulary and terminology. ArcGIS is a very complicated program, and learning the technical terminology was my first and biggest obstacle during this project. A lot of the terminology I still do not understand. ArcGIS is the program and any other “Arc_____” is an extension within the program which allows you to do different things with a dataset, like a map or a picture. MAYA and SketchUp are other modeling programs where most of the class did its architectural reconstructions. If you get confused with other terminology, know that I was and am just as confused.

In early November Dr. Erickson asked whether anyone in the class was interested in helping him with the landscape modeling. Mollie Epstein and I volunteered, and met with Dr. Erickson soon after to discuss the methods and goals of the project. He informed us that we would be working with a program called ArcGIS 9.1. The goal was to make a series of 3D maps: one of the local landscape- the temples and plazas; the second of the urban area- the temples, plazas and urban sprawl; the third of the surrounding region-the site, local hills and ridges; the final of the broad regional landscape- the site, valley, extending out to all the distant mountains visible in the Tiwanaku skyline. The 3D maps would be converted into VMRL format and opened in MAYA. Then all the class’ MAYA and SketchUp models would be imported and placed on the landscape. The original geographic models could be used to analyze the environment. We had plans to do population models, simulated pilgrimages through the mountains, flyovers and tours.

Dr. Erickson had already put a lot of work into this project. He had collected the materials and begun entering and manipulating them. DEMs from the University of Maryland Geology Department and LandSat ETMs, satellite pictures, were available for generating the large regional maps. Various topographic maps and aerial photographs had been obtained for the local maps. These pictures were entered through ArcScan to be adjusted in ArcMap and viewed in ArcScene and/or ArcGlobe. With all the work Dr. Erickson had already done, he was very close to completing the project. However, he had gotten stuck, and wanted us to go through the tutorials and help menu to see if a fresh perspective would help solve the problem.

The first thing I did was to copy the many tutorials and help manuals onto CD so I could review them on my home computer. Mollie had the only key to the office with the one computer that had ArcGIS installed, so I spent over five hours that week pouring over the tutorials and manuals. The manuals were in the form of PowerPoint presenations, often with 500+ slides, and there were so many to go through, I had to be selective about which ones I would read. 3D Analyst, Spatial Analyst and ArcScan were several that would be important, as well as ArcCatalog and ArcMap.

Then, just before Thanksgiving, I spent an entire day, over 8 hours, working on the program, going through some more tutorials and sorting through the material Dr. Erickson had left for us. The day was extraordinarily frustrating and entirely fruitless. What I learned was that I needed a teacher to walk me through ArcGIS. When I returned from Thanksgiving, it appeared Gina, one of our TAs, had made some progress, and the project, or at least the undergraduate participation in the project, had effectively stalled. With other project deadlines approaching, Mollie and I decided to focus on more attainable goals.

In the final days of the semester, I spoke with Dr. Erickson and he assured me that the landscape models will be completed. Yet, I would still encourage others to pursue this topic in future years because of the richness in analytical material. There are an almost infinite number of potential smaller projects when an accurate 3D model exists of the landscape, and I hope that these possibilities are fully explored in the future.