(August 11, 2021)—Archaeological data allow us to address issues of contemporary relevance, such as identity, migration, and inequality. Unlike other social sciences, archaeological data are well-suited to examining changes in social networks, demographic trends, and wealth over the course of centuries and even millennia. The US Southwest is an ideal testing ground for such studies because of the vast amount of high quality data collected during the past century from academic and cultural resource management (CRM) projects.
But synthesizing this data? Well, that has proven to be a daunting task, due to the sheer volume of information and the numerous recording systems used by thousands of projects.
Over the past two decades, members of the cyberSW team—including a small army of graduate research assistants and community scientists—have been building regional databases of major pre-Hispanic settlements across much of the US Southwest with the generous support of the National Science Foundation (NSF). Much of this information was already available in published and unpublished reports, site records, and museum files. We also added information from other regional databases, especially from Northwest Mexico. And we conducted new analyses in areas where little data was available.
The primary goals of these efforts were to reconstruct population trends using room counts and settlement habitation dates, and to conduct social network analyses using similarities in painted ceramics as a measure of social connectivity.
In 2017, we were awarded a grant from NSF’s Resource Implementations for Data Intensive Research (RIDIR) program to standardize and merge these regional databases, and to make them available online. This was a collaborative effort by the University of Arizona, Archaeology Southwest, Arizona State University, and University of Colorado Boulder. The result is the cyberSW 1.0 science gateway, which launched in June 2020 after some COVID-related delays (link cyberSW launch blog).
The database behind cyberSW 1.0 merges all data from these previous projects using a graph database platform developed by Neo4j. This structure makes it much easier to add, organize, visualize, and revise complex archaeological data than traditional databases such as Microsoft Access. (Fun fact: Instagram, Twitter, Facebook, and Amazon all utilize graph databases.)
Currently, cyberSW 1.0 has data from more than 13.7 million ceramics representing thousands of types and type variants; nearly 11,000 obsidian artifacts from over fifty sources; and more than 1,500 public architecture features from 25,000 pre-Hispanic settlements across the US Southwest and Northwest Mexico. The science gateway is open access, allowing anyone with an internet connection to explore and analyze this massive database at the settlement level. (Site locations are obscured.) We were able to hire a full-time Database Manager (Joshua Watts) and Systems Developer (Andre Takagi) to develop and expand the functionality of cyberSW as a living research database.
The cyberSW 1.0 science gateway is ideal for exploring and analyzing the archaeological data from settlements across the Southwest/Northwest while maintaining social distance in the COVID era. It will also save considerable time and money on travel to museums and other site-record repositories—we have done much of that work already!
There are two basic groups of functions in cyberSW 1.0: “explore” and “analyze.” These are separate tabs in the science gateway. Some of the analyses are beyond my skill level, but I really enjoy exploring cyberSW by site, customized project area, and material type. One of my favorite pastimes is to generate heat maps of various combinations of ceramic and obsidian distributions. Queries such as “which settlements have both Roosevelt Red Ware and Mule Creek obsidian?” can be answered in a few seconds.
And that’s not all! Users can easily generate publication-ready graphics. Citations for each data source are displayed as individual records associated with a particular site. Data tables can be exported either alone or in user-defined batches in various file formats used by other software (Microsoft Excel, for example). The analysis tab has a built-in toolkit that includes various statistics used in network analyses, as well as tools for dating sites, reconstructing room counts through time, and reconstructing and visualizing networks. This toolkit is being expanded on a regular basis, based on user feedback.
Since the cyberSW launch, more than 320 people have registered. The browser has been visited more than 33,500 times with an average of 100 per day over the last two months.
The collaborative team for cyberSW 1.0 was recently awarded a new grant through NSF’s Human Networks and Data Science program to develop cyberSW 2.0. This version will greatly expand the capacity and structure of cyberSW 1.0 to add and explore archaeological data within settlements.
Using the US Census as a conceptual model, we will identify and conduct analyses on households and perhaps neighborhoods. Households defined by archaeologists and the Census are both based on spatial and circulation patterns. For example, the Census defines a household as all of the people who occupy a “housing unit” with inferred social and economic connections among these people. A housing unit is a house, an apartment or other group of rooms, or a single room that is inhabited as a separate living quarters with direct access from the outside or through a common hall. It’s easy to envision a similar definition for a Hohokam compound, Mimbres room suite, or Unit Pueblo. The Census records information on household size, cultural identity, and wealth—variables that can be measured directly or indirectly from archaeological data. Drilling down from settlements to households provides a deep-time perspective for other scientists who also use household census data.
To develop cyberSW 2.0, we will first expand the graph database infrastructure to include various features (for example, rooms, middens, and storage facilities) and contexts (such as floor, fill, roof/wall collapse) within settlements. We will also develop a chronology tool that dates and groups contemporaneous features and contexts. Architectural and artifact data from a sample of excavated settlements across the US Southwest will be standardized and added at this more refined level. Large-scale excavations with digital data will be emphasized, including CRM projects across the Hohokam area and long-term academic research projects such as those at Homol’ovi, Silver Creek, and Cibola in the Ancestral Pueblo region.
Next, we will reconstruct households from contemporaneous features and contexts using the regional expertise of the Principal Investigators—no computer program can do this yet! A variety of analyses will be developed to reconstruct household networks, identify cultural background, and measure wealth through time. These analyses will be added to cyberSW’s analytical toolkit.
A Tribal Working Group and Research Assistant will facilitate communication between archaeologists and Indigenous groups in building out the infrastructure. The goal is to develop cross-cultural vocabularies that will be incorporated in cyberSW 2.0. This is an essential part of the project.
cyberSW 2.0 will move us toward a critical mass of data volume and standardization to be adopted by future projects, especially those in the CRM sector that are generating much of the new data in the US Southwest. The science gateway and research tools developed in this grant will extend access to data within settlements. cyberSW fills a critical need for a living research database that provides broad access to information about regional archaeology in a format that is amenable to casual exploration and intensive research. In order to realize the enormous potential of cyberSW, Archaeology Southwest is committed to its long-term maintenance and expansion.