Format: Paper presentations with discussion
Convenors:
James Flexner, Archaeology, University of Sydney, Australia, james.flexner@sydney.edu.au
Katharine Watson, Christchurch Archaeology Project, New Zealand, katharine.watson@christchurcharchaeology.org
Sara Gonzalez, Dept of Anthropology, University of Washington, USA, gonzalsa@uw.edu
Archaeology is increasingly experienced as a digital pursuit, with everything from fieldwork data to publications mediated through electronic devices and screens. This session seeks to explore the intersections between the digital and the modern worlds in two directions. First, by defining and critically examining the ways that digital archaeologies contribute to what in many parts of the world are still called “historical archaeologies.” How are we increasingly using the digital tools available to archaeologists, from photogrammetry to geographical information systems to online databases, to understand the material transformations that defined the emergence of modernity during the last 500 years? Second, to experiment with the methodologies and theories of historical archaeology for understanding the emergence of the digital world that increasingly dominates contemporary experience. How does archaeology help us to understand the industrial processes, systems thinking, environmental impacts, and socio-cultural implications of recent and contemporary digital life?
Papers:
3-d for What? Scanning Mangareva’s 19th Century Buildings
James L. Flexner, University of Sydney, Australia
Recent technological developments have made 3-d modelling easy, cheap and fast. Almost every mobile phone on earth now sports a camera that can be used for photogrammetry, and many devices also feature a lidar scanner. Apps abound to take advantage of this hardware to produce models quickly and intuitively. The problem with these kinds of technical developments is they can develop without a commensurate advance in research design, methodology, or theory. If almost everyone carries a device in their pocket that can model any archaeological site, immediately and practically for free, we might all start doing so, but why? As part of a current project I have scanned dozens of 19th century Catholic buildings in the Mangareva Islands of French Polynesia using an iPad Pro and the software Modelar, in addition to producing more ‘traditional’ hand-drawn plans and elevations. This presentation reflects on what the 3-d models add to the visualisation of these culturally significant heritage sites, as well as limitations of the method, arguing that 3-d is a supplement, not a replacement, for traditional recording methods.
Vegetation Patterns as Low-investment Remote Sensing
Tristan Newberry-Cushman, University of Denver, USA
Patterns in vegetation have long been used by archaeologists to make inferences about subsurface phenomena. Usually, visual identification of vegetation patterns is used as an informal indicator of potential data. Advanced remote sensing techniques are both time- and resource-intensive, making them impractical for many projects designed by students or heritage management firms. The methods presented here require no specialised equipment, relying solely on visual identification of vegetation patterns and widely available consumer technology.
At the historical Dallas townsite, GIS and handheld GPS units were used to develop a sampling design for shovel test pits. Statistical tests were designed to determine if the assemblages associated with different vegetation zones could show differential utilization of space during the period of occupation. The distribution of vegetation proved a reliable indicator of subsurface artefact deposits. This enabled inferences to be made about the spatial distribution of the site, despite major disturbances. The results show promise for future use of this method as an empirical form of sensing.
The Hamtramck Explorer: Deep Mapping for Community Resilience and Heritage Sustainability in Metro Detroit
Krysta Ryzewski, Professor and Chair, Dept of Anthropology, Wayne State University, Detroit, Michigan, USA
Dan Trepal, Senior Geospatial Research Scientist and Visiting Assistant Professor, Social Sciences. Great Lakes Research Center, Michigan Technological University, Houghton, Michigan, USA
Don Lafreniere, Professor of Geography and GIS, Director, Geospatial Research Facility, Michigan Technological University, Houghton, Michigan, USA
Hamtramck, a small Muslim-majority city surrounded by Detroit, is defined by its century-long reputation as a welcoming home for foreign-born immigrants, who comprise over 40% of the city’s present-day population. This presentation details the creation of the Hamtramck Explorer prototype by a collaborative team that includes the authors, the Hamtramck Historical Museum, and community stakeholders.
The Hamtramck Explorer is the first successful deep map to integrate historical, archaeological, and geospatial data. The free, web-based deep map consists of an interactive digital atlas on the front end, powered by a robust historic spatial data infrastructure (HSDI) on the back end. The HSDI connects millions of data points between myriad bodies of historical evidence, including archival, archaeological, and spatial data related to Hamtramck across a 150-year time span.
The three-fold purposes of the Hamtramck Explorer are to preserve historical information, support research, and facilitate participatory storytelling about places and people associated with the city and its residents, past and present. Its open-access design is aimed at maximising inclusivity, accessibility, and sustainability. In doing so, it promotes increased participation in local history, archaeology, and historic preservation among a city population that continues to struggle with the decades-long effects of disinvestment and deindustrialisation.
LiDAR and GIS Based Analysis: Reanimating Activity on a Peripheral Goldfield
Richard MacNeill, Dept of Archaeology and History, La Trobe University, Melbourne, Australia
The isolated Mount Misery diggings, peripheral to better-known and more comprehensively recorded goldfields in central Victoria, spans the life of the gold industry across the last half of the 19th century. Lying within land unsuitable for activities that erased other better-known goldfields, features not scoured by time or destroyed by bushfire remained intact but largely invisible.
GIS based spatial analysis of a detailed LiDAR based elevation model revealed a distribution of domestic, industrial and communal features largely concealed by understorey and leaf-litter. This paper describes two complementary approaches to the analysis and interpretation of these features and the landscape across which they are distributed. The first distinguishes features. The second models movement across and between these features. These two approaches combine to provide insights into industrial and communal space within the digging and the character and dynamics of a distinct goldfields society. Together, they inform methods and principles of non-invasive archaeological survey.
Historical Archaeology in Comics as Digital Data Visualisations
Paulina F. Przystupa,
Archaeology as a digital pursuit relies on data literacy to make sense of and effectively use the digital tools available to archaeologists. However, data literacy begins with the ability to read and write. These skills are regularly mediated through digital frameworks in Western contexts because much of what we read and write, archaeological and not, appears through screens and other devices as text, audio, and images. This paper explores the need for archaeological data literacy within digital historical archaeologies using the example of the data story Tome Reader. This data story, part of the larger work of the Alexandria Archive Institute’s Data Literacy program, aims to cultivate archaeological data literacy by having archaeology-interested learners read, engage with, and discuss comics with archaeological themes. This paper will focus on the comics recommended in this data story that explore digital, historical, and contemporary archaeologies in their content and accessibility to see how these works created for non-archaeology publics, and legible through digital means, help us better teach and learn about archaeologically relevant concepts and the way that the digital consumption of those works allow for the unique interactions with digital historical archaeological knowledge.
ARCHAEOVISION: Optimising Classification of Microvertebrate Remains with AI, Optics and Neural Networks
Juan C. Rofes, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines; National Museum of the Philippines; Archéozoologie, Archéobotanique Sociétés, Pratiques et Environnements, CNRS/MNHN, France; National Museum of the Philippines
Chara Deanna F. Punzal, Data Science Program, College of Science, University of the Philippines Diliman, Quezon City, Philippines
Khristian G. Kikuchi, Data Science Program, College of Science, University of the Philippines Diliman, Quezon City, Philippines; College of Computer and Information Science, Mapua Malayan Colleges, Laguna, Philippines
Patricia S. Cabrera, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines
Gabrielle Anne B. Gascon, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines
Timothy J. Vitales, National Museum of the Philippines
Francisco N. Claravall, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines
Ranzivelle Marianne Roxas-Villanueva, University of the Philippines Los Baños, Los Baños, Laguna, Philippines
Hermine Xhauflair, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines; National Museum of the Philippines; UMR 7149 Histoire Naturelle de l’Homme Préhistorique, Institut de Paléontologie Humaine, France
Armand Salvador B. Mijares, School of Archaeology, University of the Philippines Diliman, Quezon City, Philippines
Giovanni A. Tapang, Data Science Program, College of Science, University of the Philippines Diliman, Quezon City, Philippines; National Institute of Physics, University of the Philippines Diliman, Quezon City, Philippines
Microvertebrate remains are usually recovered in high amounts from archaeological and paleontological sites. Depending on the factor of accumulation, they can be excellent proxies for taphonomic processes, palaeoecological inferences and/or human utilisation. To optimise the labor-intensive and time-consuming process of manually sorting microvertebrate remains from archaeological sites, we developed ARCHAEOVISION (Archaeological Vertebrates Imaging Systems with Intelligence, Optics, and Neural Networks), an imaging system designed to classify bone fragments into three categories: Amphibia, Mammalia, and Reptilia. It utilises machine learning for real-time classification and employs Gaussian splatting to create 3D models. High-resolution images of sorted bone fragments from Callao Cave in Cagayan Province were captured using an Olympus SZ61TR microscope equipped with a DP22 video camera mount and Olympus Stream software. This resulted in a dataset comprising 20,000 images of 10,000 unique fragments, which were utilised in various convolutional neural network models. The model that employed YOLOv8 consistently achieved over 90% accuracy on the labelled samples from Callao and reached 83% accuracy with validation by expert archaeologists when classifying unlabelled fragments from Guri Cave in Palawan Island. This imaging system improves the speed and efficiency of archaeological and paleontological research in field and laboratory settings.
Introducing the Museum of Archaeology Ōtautahi
Katharine Watson, Hayden Cawte and Jessie Garland, Christchurch Archaeology Project
Archaeological data is a rich and unique source of information about Aotearoa’s history, providing tangible insights into people’s lives, the processes that affected them and the landscapes they occupied. This is why archaeological sites are protected by law, and why reports are required to be written on archaeological excavations. But these technical reports are (a) not that accessible to the general public and (b) not that user-friendly for future research. At Christchurch Archaeology Project, we want to change that, to make archaeological data accessible to the public and researchers. To this end, we built the (online) Museum of Archaeology Ōtautahi (MoAŌ), to enable access to collections that would otherwise be hidden, lost or inaccessible. MoAŌ demonstrates the power of a relational database to bring the past to life by highlighting the connections between people, place and objects. We’ll outline the challenges we faced, the lessons we learnt, the solutions we arrived at—and the incredible opportunities that lie ahead.