My name is Robert Z. Selden, Jr. (my colleagues call me Z or Zac), and I work at the Heritage Research Center at Stephen F. Austin State University (SFASU), where I am engaged in a wide variety of cultural heritage research projects using 3D meshes and models. Those efforts range from shape analyses enlisting ceramics from the American Southwest, American Southeast, Brazil, and Peru, to analyses of bronze cannons and French trade axes recovered from the La Belle shipwreck, as well as morphological analyses of human remains recovered from Sudan and Tanzania. Given my heavy investment in digital technologies, some have considered me a non-traditional archaeologist—I personally prefer 21st-Century archaeologist—however, I do still spend a month or two out of every year sinking square holes into sediment looking for previously-unrecorded archaeological sites. My investment in anthropological and historical research using 3D data has thus far yielded a new method of analyzing axial asymmetry in ceramics (Selden Jr., 2017), the discovery of a spatial boundary that correlates with a significant shift in Caddo material culture products (Selden Jr. et al., 2018a, Selden Jr., 2019a,b, Selden Jr., 2018a,b), discussions of precision and accuracy in 3D prints (Hammond et al., 2018, Selden Jr. et al., 2018b), differences in fossil reconstruction efforts that may have detrimental impacts on interpretation (Bergstrom et al., 2018), and the documentation and dissemination of 3D meshes that articulate with stone tools (preforms, points, blades, and blade cores) manufactured and used by some of the earliest immigrants to the Americas (Selden Jr et al., 2018, Selden Jr. and Crawford, 2018).
Using 3D for Analysis
One of the many benefits of collecting data in 3D is that it allows me to shift analytical directions quickly as new insights are discovered. This stands in stark contrast with the practice of collecting traditional linear and orthogonal measurements, where subsequent observations would require a return to each repository to revisit the same specimens to collect new data. Collecting data at a museum or repository is a very involved undertaking, and includes scheduling, submitting new paperwork with new research designs, requests for permission related to access and data collection, submitting a request for a new letter of support from the Tribes, and a new source of funding to cover time, travel, and access fees. For instance, if during an analysis of ceramic bottle shapes, I happen to realize that the angle of the attachment between the neck and body appear to be changing through time, I can immediately pursue that question upon completion of the current analysis, or run the two studies concurrently. The ability to pursue exploratory research designs (Figure 1) where multiple hypotheses can be formulated and tested, or hierarchically-nested hypotheses, which can be tested in subsequent confirmatory analyses has truly allowed me to capitalize on the morphology of a whole object, providing for a greater return on our investment in travel. While the production of 3D meshes does require a more substantial initial investment of time (compared with linear and orthogonal measures), 3D data have tremendous reuse potential across a wide variety of applications.
Tools and Workflow
For data collection, I use Creaform scanners or a ZScanner 700CX (depending on the needs of the project), and have a NextEngineHD that I use for training. I am able to scan directly into Geomagic Design X, where I process and model those data, and apply reference geometry in advance of analysis. I also use Geomagic Control X for inspections between CAD models and meshes, or between meshes that have been processed in different software or using different sequences associated with our post-processing protocols.
I was lucky to have realized the reuse potential of 3D models early in the development of my digital workflow, and collaborated with Dillon Wackerman at the Center for Digital Scholarship to find a way of incorporating those models into our Institutional Repository using Sketchfab (Wackerman and Thompson, 2014, Wackerman and Selden Jr., 2016). We had previously identified Zenodo @ CERN as an ideal option for digitally curating the 3D meshes, where each would receive a digital object identifier. Zenodo allows users to update the metadata associated with each entry as new works are published (linking the publication back to those data used in the study), which is a tool that I have continued to use as my efforts have expanded, allowing me to better track the various publications that each of my 3D meshes is used for in my own work, as well as the work of others.In addition to Zenodo, I use GitHub and the Open Science Framework throughout the development of my projects.
ScholarWorks (bepress) is the Institutional Repository at SFASU. Wackerman and I were able to work with bepress to identify a plugin (embedly) that allows me to include a 3D model from Sketchfab that is easily viewed and manipulated in ScholarWorks. The result is a suite of tools that leverages the many benefits of digital curation at Zenodo with the search and discovery tools of ScholarWorks paired with Sketchfab to produce a means of making my 3D meshes as visible as possible, reaching universities, museums, companies, classrooms, and anyone with an interest in cultural heritage and an internet connection.
Using 3D as a Training Tool
The bulk of those specimens used in my work articulate with the Native American Graves Protection and Repatriation Act (NAGPRA), and I regularly teach a section of the NAGPRA for Archeologists workshop facilitated by the National Center for Preservation Technology and Training (NCPTT). I was also able to work alongside NCPTT to develop the Advanced 3D Imaging and Morphometrics for Archeologists workshop that we launched in the Arizona State Museum at the University of Arizona last fall. Much of the content of these workshops, the sections that I teach in particular, regularly link back to my work with our Institutional Repository, making it easy for me to discuss and work through specific challenges in real-time using the meshes on Sketchfab that are embedded in ScholarWorks.
3D for Outreach and Education
I was introduced to Dr. Bernard K. Means through Dr. John P. Hart early in my career, which spun up a productive professional relationship. This led to our bringing Means on board as a Research Affiliate at the Heritage Research Center, and my post as a Research Fellow in the Virtual Curation Laboratory at Virginia Commonwealth University. His experience with 3D printing is far superior to my own, and he has been instrumental in educating me with regard to public archaeology and outreach possibilities related to 3D meshes and models within the realm of cultural heritage. Means also taught me how to animate my scans, and early on he even took the time to animate a few for me himself. It was during this period that I began experimenting with building 3D puzzles for some of my own models, and my wife—Lauren Selden, Professor of Metalworking and Jewelry at SFASU—introduced me to the laser cutter in the School of Art at SFASU. That first 3D puzzle was cut out of an old moving box that had been stowed away in our attic for some time (Figure 2), and I would eventually revise the plans for these models so that they could be printed on a standard 8.5” x 11” piece of paper, glued to cardboard (any old cardboard will do), and cut out, making them perfect rainy day activities for elementary and middle school students. The models are based upon real Caddo burial vessels, and were designed with the intent of providing a means of physically interacting with an otherwise intangible piece of human history. All of these models are open access, and continue to be some of the more highly downloaded designs that I have uploaded to ScholarWorks. Links to some of these are included in my data papers, but you can find and download a selection of those here, here, and here. Since they are based upon real 3D models of Caddo burial vessels, I consulted with the Caddo, who suggested that they be constructed out of cardboard or wood and not plastic. More recently, I have been collaborating with SFASU faculty in the College of Education to find new ways of integrating some of our 3D scans into middle and high school curricula.
In addition to my duties as a research associate, I am also a member of the graduate faculty at SFASU, and regularly serve on graduate committees. Two of my current graduate (MFA, ceramics) students, Jack Shanklin and Erik Ordaz, have scanned some of their recent works that are included below. My wife and I recently worked on an academic poster with Erik, where he used the 3D scan to create a mold that he subsequently cast in metal as a means of exploring “[t]he idea of subtle changes in everyday life, the fast moment in which we make a decision and how it affects the course of our time in this world.”
Data Exploitation and Management
Citation networks are among the more novel components of my 3D-related research designs, and I regularly leverage them to aid in the identification and exploitation of research domains where others are not currently working, or where previous investigations have shifted to a dormant cycle (Figure 3). The utility of this approach helps me to ensure that my work has substantive impact across the field, and is a requirement of my project designs, since the time commitment of a 3D project is significant. Leveraging the results of these networks has helped me to stay abreast of new developments, and to forecast shifts across various research domains (those in which I am most interested or otherwise engaged) in real time. Should a new analytical approach, tool, or method of analysis become available, I spend the time needed to learn how to use it properly, integrate it into my workflow, and continue to push forward.
While my processed data are uploaded to Zenodo, ScholarWorks, and Sketchfab, each project that I work on is also included on the Open Science Framework (OSF). This is where my unprocessed 3D files are stored and made available following an embargo period (usually five years from data collection), and are organized in such a way that they are directly linked to each project in which they are used. Preprints for each of my recent works are also posted to SocArXiv (and more recently also to SSRN) in advance of publication as a means of democratizing my work by making it publicly accessible outside of the publisher’s paywall. The reason that I use OSF instead of uploading everything to Zenodo is due to the size of the unprocessed data (Zenodo currently has a 2GB limit per file), and ease of access when it comes to identifying which scan was used for what project. I realize that this may seem like overkill; however, from a personal and ethical perspective, it is absolutely necessary. Those data that I produce were collected with public funds (I work at a state [Texas] university, and leverage federal [US] grants); therefore, they should be made available for the public to use with no restrictions, and cited accordingly.
Additionally, I have found that when 3D analysts repurpose 3D data in their projects, which is happening at an increasingly regular rate across the sciences, it is best for them to do so by including the unprocessed scan data and integrating those data using a post-processing workflow identical to that of the rest of their project. While this does extend the time needed to locate, process, integrate, and analyze new datasets, it is absolutely necessary, since the many post-processing algorithms and workflows do alter 3D data.
Current and Future Projects
My research is dynamic, and—due to the many benefits of enlisting 3D data—evolves quickly. Current projects include the analysis of Chimú bottles and jars from Peru, where I have found a previously unrecognized difference in the shape of jars and bottles adorned with anthropomorphic and zoomorphic sculptural elements. I was recently awarded permission and funding from the Caddo Nation of Oklahoma and the National Forests and Grasslands in Texas (NFGT) to begin work on a study of differential incision morphology on Caddo ceramic sherds found on NFGT sites, as a means of better characterizing the morphology of those tools (the business end of those tools, at least) used to apply Caddo decorative motifs. I am grateful to my colleagues at NCPTT for providing access to their optical profilometer (Dr. Mary Striegel, Dr. Catherine Cooper, and Tad Britt in particular) that I used to scan a series of sherds that I had previously incised and fired while taking a ceramics workshop at Arrowmont with a suite of standard sculptural implements, as well as the Caddo sherds from NFGT (Figure 4). I am also working on two new papers that deal with analyses of the impacts that differential post-processing methods can have on 3D data, and while these papers are a bit more technical, preliminary results indicate that the digital tools that 3D analysts use (and the order in which we apply them) are worthy of further consideration and scrutiny, particularly within systematic studies of geometric shape.
The Heritage Research Center is also affiliated with the JMD DYCLAM program in France, and I hold a research post as an associate professor in the Cultural Heritage Department at Jean Monnet University (UJM). Our universities have brokered a fantastic exchange program for our graduate students, and each spring two talented graduate students from UJM come to the Center to work with me while taking classes, and two of our students attend UJM. During the course of a previous project funded by the Texas Historical Commission to reverse-engineer one of the cannons found with the La Belle shipwreck that is curated at the Bullock Texas State History Museum, I found some rather dramatic differences in the morphology of the cannon handles. This provided my graduate students and I with an opportunity to test whether all of the cannon handles may have been produced with different molds (or finishing techniques), or whether there were two distinct left and right molds used in their manufacture. The Office of Research and Graduate Studies at SFASU provided funding for Julie M. Villing (pictured below) and Emmanuel Gobez (both from UJM) to travel with me to Bay City and Corpus Christi, Texas in 2018 to scan the handles of the other two cannons for further comparisons, which supported the hypothesis that each handle was, in fact, unique. This spring, I am working on an analysis of French trade axe morphology with two new graduate students from UJM, Julia Cid and Indye Zennaf.
It has been very exciting for me to watch my students gain familiarity with these tools, and enlist them in new and creative analyses of material culture. While my own adventure has proven to be quite enjoyable and full of many exciting prospects for current and future research, nothing makes me happier than to be present to witness those moments of pure joy when one of my students develops or conceives of an exciting new analytical approach or concept that opens up a new way of thinking about a complex problem. I just need to find a way to keep pace with them!
Many thanks to the Heritage Research Center at Stephen F. Austin State University, to my current and former students, and to my many very talented (and patient) colleagues for their support of my previous, current, and future research.
BERGSTROM, K., BUTARIC, L. N., GIDNA, A., MUSIBA, C., SELDEN JR., R. Z. & MAGORI, C. 2018. Taphonomic and Fossil Reconstructive Analyses of the Ngaloba (LH 18) Skull. Austin: Paleoanthropology Society.
HAMMOND, L., BERGSTROM, K., SELDEN JR., R. Z. & DE RUITER, D. 2018. Precision in 3D Prints: The Case of MH1 (Australopithecus sediba). American Association of Physical Anthropologists. Austin.
SELDEN JR, R. Z., WILLIAMS, T. J., VELCHOFF, N. & COLLINS, M. B. 2018. 3D Scan Data for Selected Clovis-Age Artifacts from the Gault Site (41BL323). CRHR Research Reporta, 4, Data Paper 1.
SELDEN JR., R. Z. 2017. Asymmetry of Caddo Ceramics from the Washington Square Mound Site: An Exploratory Analysis. Digital Applications in Archaeology and Cultural Heritage, 5, 21-28.
SELDEN JR., R. Z. 2018a. Ceramic Morphological Organisation in the Southern Caddo Area: Quiddity of Shape for Hickory Engraved Bottles. Journal of Archaeological Science: Reports, 21, 884-896.
SELDEN JR., R. Z. 2018b. A Preliminary Study of Smithport Plain Bottle Morphology in the Southern Caddo Area. Bulletin of the Texas Archeological Society, 89, 63-89.
SELDEN JR., R. Z. 2019a. Ceramic Morphological Organisation in the Southern Caddo Area: The Clarence H. Webb Collections. Journal of Cultural Heritage, 35, 41-55.
SELDEN JR., R. Z. 2019. Louisiana Limitrophe: An Iterative Morphological Exegesis of Caddo Bottle and Biface Production. In Ancestral Caddo Ceramic Traditions, edited by Duncan P. McKinnon, Jeffrey S. Girard and Timothy K. Perttula, in press. Baton Rouge: LSU Press.
SELDEN JR., R. Z. & CRAWFORD, G. T. 2018. 3D Scan Data for Selected Artifacts from Blackwater Draw National Historic Landmark. CRHR Research Reporta, 4, (Data Paper 2).
SELDEN JR., R. Z., DOCKALL, J. E. & SHAFER, H. J. 2018a. Lithic Morphological Organisation: Gahagan Bifaces from the Southern Caddo Area. Digital Applications in Archaeology and Cultural Heritage, 10, e00080.
SELDEN JR., R. Z., MEANS, B. K., IGLESIAS, E. G. & MOSIER, K. 2018b. Morphological Variation in Three-Dimensional Printed Replicas. CRHR Research Reporta, 4.
WACKERMAN, D. M. & SELDEN JR., R. Z. 2016. Committing to the Non-Traditional: The Path to the Incorporation of 3D Models in an Online Journal. Dallas: Southern Methodist University.
WACKERMAN, D. M. & THOMPSON, A. 2014. Content Management Systems and 3D Models: Creation, Interaction and Display. Nacogdoches: Stephen F. Austin State University.