Below are the abstracts for the CAAA Digital Archaeology Conference 2021. You can find the schedule for the conference here.
Links in the title will take you to each talk uploaded to our YouTube channel. Please note, talks marked with a * will not be uploaded to our YouTube channel at the end of the conference.
N.B. Affiliations shown are those of the first author. Titles marked with * are posters. Links in the title will take you to each talk uploaded to our YouTube channel. Please note, talks marked with a ^ will not be uploaded to at the end of the conference.
Luis Alberto Alcántara Chávez
Independent
ArqueoData. A radiocarbon results database from archaeological sites in Latin America.*
The international archaeology is in a field where disciplines such as computer science can provide tools to make the collection of data more efficient for its proper archaeological interpretation. The digitization of information in databases allows consultation and registration to be faster, more efficient and standardized. Despite this, Latin American archeology does not have many tools that focus on the research needs of the new continent. In this work I present ArqueoData 1.0, a database where I compile many radiocarbon dates of Mexican archaeological sites. Which can be consulted through a web interface (HTML5). The database can be updated by the same users, adding new records. Likewise, it allows you to create collections, where you can save and establish groups of the radiocarbon results that are of interest to the user. The system is a web application programmed in JavaScript by VueJS and PHP. The application does not require an installation, however for its use an internet connection is required.
Matthew Barrett, Joshua Emmitt, Rebecca Phillipps, Thegn Ladefoged, Simon Holdaway
University of Auckland
@mbarrett2312
Archaeological and paleoenvironmental studies as part of the Ahuahu Great Mercury Island Project investigate a dynamic contemporary coastal landscape currently farmed with domestic stock. Over the last decade, many researchers working on the project have recorded geospatial data using total stations, GPS units, laser scanners, and digital cameras during archaeological excavation and survey but also palaeoenvironmental sampling. At different stages of the project, excavation units and sample locations were opened multiple times, with some extended over different periods of fieldwork, and new units opened adjacent to old ones. All project locational data were and continue to be recorded in a GIS however the active contemporary environment combined with the involvement of multiple researchers employing a range of recording methods poses challenges ensuring geospatial data are accurate and comparable within and across field seasons. We present examples of two such challenges, as well as the solutions adopted, and lessons learned. The first involves attaining consistency in recording geospatial data with total stations, limiting the error associated with different machines, different users, and spatial ‘drift’ across field seasons. The second involves maintaining data integrity, capturing data entry errors, and ensuring the consistent recording of metadata, both important considerations when publishing open data.
Simon H. Bickler, Benjamin Davies, Garry Law
Bickler Consultants Ltd
Virtual vaka: a computational tool for thinking about seafaring.
Seafaring has had a profound impact on human history, especially in the Pacific. While ancillary signals of seafaring are widespread in the archaeological record, direct evidence for ancient seafaring technology and knowledge is scarce, leaving a great deal of uncertainty about the influence seafaring had on the exploration, colonisation and continued communication networks across water. This uncertainty lends itself to a model-based approach, which is manifest in a long history of computer simulation studies of ancient seafaring across the world. We discuss our efforts to develop computational models of seafaring in the past using agent-based simulation as ‘tools to think with’: technologies that can be deployed in the iterative process of theory-building. We describe the latest version of our agent-based simulation, Virtual Vaka, with a new on-line capability can be used to send vessels out with a variety of navigational and technological capabilities anywhere on the globe. It draws on available environmental data, allowing for exploration of voyages using real weather conditions from the last 70 years, while also maintaining the capacity to use modelled data for both weather and bathymetry to simulate past journeys as far back as the Pleistocene. Finally, we simulate inter-island voyages between Aotearoa and distant neighbours such as Lord Howe and Australia to examine how potential interaction spheres were created while others were not.
Caitlan Butler, Thegn Ladefoged, Alex Jorgensen
University of Auckland
@caitlanbutler
Understanding patterns of technological organisation, their indicators, and how they relate to raw material procurement has been the subject of many archaeological studies. In New Zealand this can be a challenge as metrics commonly used in such studies are not always reliable. This study conducts a use-wear analysis, along with a technological analysis, of obsidian from the sites of Pouerua, Tamewhera, Waipirau Pa, Oneroa Beach, Te Mataku, Waitetoke, and Kohika, located in the North Island of New Zealand, in an attempt to identify whether use-wear intensity reflects procurement. The use-wear analysis identified the location and extent of use-wear allowing for the intensity of use to be identified. The technological analysis likewise allowed for reduction intensity to be identified through flake measurements and dorsal scarring. The results of this analysis suggested temporal differences in the intensity of use-wear and reduction across the sites, and high proportions of Kaeo and Mayor Island obsidian, suggest direct access to these sources. The results suggest that there may be links between obsidian use-wear and procurement that are linked to temporal differences in how obsidian was obtained by Māori.
Goran Đurđević, Emiliano Gallaga Murrieta
Beijing Foreign Studies University
@gdjurdjevich
Excavation online. Social networks and archaeology in postcorona time.
In recent months, we have faced the coronavirus that affect archaeologists in their regular work (excavation, travelling, research in museums and library, participation in congresses and conferences). The project Thursday Twitter Thoughts (TTT) tried to bridge the gap in COVID and post COVID time. It explores the possibilities of Twitter and tweets for the promotion of archaeology. The project started on August 20 2020 and finished on August 19 2021 with a tweet per week. The aim was: a) promotion of archaeology as an academic field and discipline for the broader audience; b) opening new topics for archaeologists (contemporary archaeology, gender issues, racial identities, disability), c) using of digital space for archaeology as interdisciplinary academic fields; d) promotion of diversity; e) understanding archaeology as global discipline and field; f) dealing with archaeology in pandemic and post-pandemic times; and g) encourage young archaeologists to do public and digital archaeology. Global aspects of archaeology have been reflected on two levels: 1) The worldwide perception and perspective of topics, which has been overlapping various time frameworks from early humans to posterity; and 2), is multilingualism. Although tweets were published in English, the plan is to translate into additional 12 languages and present them into a book. Paper deals with statistical analysis of tweets (regions, time framework, topics) and visitors (total number, reactions). Authors compare TTT projects with other Twitter and digital archaeological projects for popular archaeology and a broader audience.
Joshua Emmitt, Patricia Pillay, Matthew Barrett, Stacey Middleton, Tim Mackrell, Bruce Floyd,Thegn Ladefoged
University of Auckland
@joshemmitt
Volumetric reconstruction of archaeological deposits using point-cloud data.
The collection of 3D data in archaeology is a long standing practice, but a more recent application of such data are its quantification and display in 3D. In particular for archaeology, excavation areas are often recorded in 3D, with deposits, features, and artefact points recorded by a variety of methods. Moving on from solely visualization, these data are now used to calculate the volumes of bound entities such as deposits and features. The construction of these volumes presents challenges that originate in computer aided-design (CAD), but have implications for how the data are used in archaeological analysis. We demonstrate the 3D construction process on excavation area 231 from Waitetoke, Ahuahu Great Mercury Island. Point clouds collected from Simultaneous Localization And Mapping (SLAM), photogrammetry, and total station collection methods are compared, as well as different methods for generating meshes. The differences between the methods are compared and contrasted from visual and analytical perspectives. Recommendations for best practice are made from data acquisition, workflow, and end-use perspectives.
Ying Tung Fung
University of Oxford
Reconsidering the chronology for the Shimao region in China with the application of Bayesian modelling.^
The recognition in 2011 of the extraordinary site Shimao (ca. 2300–1800 cal. BC) as a late Neolithic stone fortification has completely changed our understanding of the development of complex societies in the North Loess Plateau (hereafter the Shimao region). The unusual combination of stone fortifications, herd animals, millet, ceramic tripods discovered at the 400-hectare Shimao and other sites in such a climate-sensitive region, has led to the question of how the climatic, demographic and economic factors came together and eventually gave rise to Shimao and more broadly to the southern part of the region. A reliable chronology for the Shimao region is first required in order to underpin this bigger research.
The presented work focuses on the chronology analysis, which reviewed all the published radiocarbon dates in the region with the application of Bayesian modelling using OxCal 4.4. In the past decade, Bayesian modelling has become commonly used for establishing and refining the chronology in different parts of China, but it has not yet been applied to the Shimao region. The results are therefore meaningful as in generating a firmer chronology that provides the context for other related demographic and socio-economic analyses. Meanwhile, these results also identified the changes of settlement distribution over time and space. This further gives us a clue about how population may have moved across the region during the period when Shimao was developed.
Frederick Hardtke, David Johnson
Macquarie University
@fhardtke
The vulnerability of rock art: the digital methods deployed at an Egyptian rock art site.^
We live in constant fear that our rock art and other heritage sites will one day be subject to destructive interference or outright destruction. The recent destruction of the Juukan Gorge caves in the Pilbara region of Western Australia has highlighted the need to protect our precious heritage sites for future generations, but also the need to exploit modern technologies to the maximum extent for their recording and preservation. Rock art is particularly vulnerable in this respect as it has an enduring presence in the landscape with usually minimal safeguards. It is also a particularly challenging medium to capture and requires special techniques to facilitate its research as compared to other archaeological pursuits. This paper will present the techniques deployed at a vulnerable Upper Egyptian rock art site including on-site database development, 3D capture and photogrammetry, electronic tracing and other digital capture techniques currently being deployed at the site which assist in “bringing the rock art home” for the benefit of other researchers and audiences as well as assist in preserving its associated data for future generations.
Greg Hil, Susan Lawrence
La Trobe University
@Greg_Hil_
Remade ground: modelling nineteenth century landscape change with GIS.
The nineteenth century was a formative period for much of our contemporary cultural landscape. From the establishment of urban centres, to the environmental legacy of the gold rush, it is hard to overstate how influential this era was in producing the landscapes we know and live in today. As modern development continues to reshape the world around us it is now often the task of archaeologists to make sense of those changes as they relate to what came before or after. If archaeologists can determine how pre-colonial ground surfaces were shaped into their present form, they may be better equipped to manage and interpret uncovered cultural heritage materials. In this paper, I present a means to model, visualise, and ultimately interpret historical landscape change through the use of nineteenth century topographic maps, LiDAR, and GIS. This paper is based on ongoing PhD research at La Trobe University in Melbourne.
Liam Johns, Joshua Emmitt, Rebecca Phillipps, Thegn Ladefoged, Simon Holdaway
University of Auckland
Auditing archaeological material for data integrity.
Research data that are accessible by all for free enhance the reproducibility and transparency of archaeological analyses, providing the opportunity for re-analysis, and the integration of data from different sources to create new knowledge. Public funders of archaeological research are increasingly mandating access to data consistent with open access but also indigenous data sovereignty – the FAIR and CARE data governance principles. However, both these sets of principles require consideration of data integrity and quality where integrity refers to the completeness, consistency, accuracy, trustworthiness, and reliability of data with quality referring to the match between the intended use of data and data generation. To assess the integrity of part of the data collected during the Ahuahu Great Mercury Island Project, specifically an inventory of ~80,000 artefacts, we conducted an audit. This involved observation of each artefact in its container, matching the form of the artefact against its database record, and assessing the artefact and container condition. Where inconsistencies were identified these were recorded and corrected. Metrics for data input and interpretation accuracy were generated, providing an avenue for exploring data trustworthiness and developing methods for error reduction both in field and laboratory contexts. Integrity checks like our audit are one way of ensuring data quality and will form a key aspect in assessing the integrity and quality of open access data.
Ben Jones, Simon Bickler
University of Auckland
@thepastfromabo1
Scaling up deep learning to identify earthwork sites in Te Tai Tokerau, Northland.
Northland has over 12,000 archaeological sites recorded in the NZAA ArchSite database with around half, including earthwork features, related to Maori history. These include pa, pits, and terraces. Other sites such as stone structures, sod walls, tracks, ditches, and drains representing both more recent and longer-term landscape history are also present, identifiable, and recorded using LiDAR data. Although it is relatively easy to identify many of these archaeological sites and features using LiDAR, as we have previously discussed (Jones and Bickler 2017, 2019), the challenge is to scale up this process to search regions to allow for a landscape interpretation and reconstruction. As the Northland LiDAR is now becoming available, we have been working on using GIS to develop Machine Learning training data specific to the identification of these earthwork sites. We present our latest attempts to scale up the identification of archaeological sites in the forested areas of Te Tai Tokerau and determine the most effective Machine Learning tools for New Zealand archaeologists (Bickler 2021).
Thomas J. Keep
University of Melbourne
The Hellenic Museum Digitization Project was a recently completed initiative to create an online-accessible catalogue of 3D digital surrogates of the material heritage collection of the Hellenic Museum of Melbourne. Using structure from motion photogrammetric modelling and digital modelling within Blender, over 50 models were created from the Museum’s collection of vases, sculptures, and arms and armour of Cypriot, Magna-Graecia, Roman, and Greek provenance ranging from the Archaic age through to Imperial Rome. Facing challenges resulting from the imperfections of photogrammetry software and the unique challenges created by particular object properties, editing within Blender was adopted to create models more representative of the original materials. The introduction of this process raises theoretical questions of ‘surrogacy’ and the legitimacy of a digital model as a replica: if models are manually edited or marred by digital artefacts, to what degree can they still be considered a valid replica? How much objectivity is there, or ought there to be, within photogrammetric modelling and digital reconstructions? These questions will be explored via the theoretical perspective of ‘simulation and simulacra’ offered by Jean Baudrillard, and the notion of ‘aura’ as discussed by Walter Benjamin. A discussion of the project as a whole will also be offered, explaining the motivations, outcomes, and potential future directions for the created models.
This paper was later published as: Keep, T. 2022, Digitization of Museum Collections and the Hellenic Museum Digitization Project. Iris: Journal of the Classical Association of Victoria, 32-33(1): 16-28.
Ilhong Ko, Woojin Shim, SooJin Park
Seoul National University Asia Center
The study of ancient pathways is essential to understanding the nature of movement and exchange in the past. Evidence relating to pathways consists primarily of artifacts and place names, expressed as ‘point data’; by drawing plausible links between the point data, ancient pathways are reconstructed. Least-cost models have widely been in studies on the reconstruction of ancient pathways. This study also utilises a least-cost approach but attempts to overcome the limitations of previous studies by adopting three strategies. Firstly, in estimating the least-cost link between two nodes, three types of cost-surface maps have been used: 1) slope degree, 2) Catena soil concept based slope classification, 3) slope degree + Catena soil concept based slope classification. Secondly, node-link base maps in which the least-cost link between two given nodes has already been calculated have been used. Thirdly, the nodes used to produce the node-link base maps, as well as the starting point and end point of the reconstructed pathway, are selected from a database of all surveyed and excavated sites in South Korea spanning from the Paleolithic to the modern period, which have been classified according to period; this database consists of over 140 thousand point data. As a result, it This multi-evidence based approach has allowed the algorithm to be used as a tool allowing researchers to explore the various factors that may have contributed to the formation of ancient pathways.
This paper was later published as: Shim, WJ, Ko, I and Park, SJ. 2022 ‘Benefit Maximizing Routes’: Development and Evaluation Using the Historical Roads of Korea’s Joseon Dynasty (1392–1910). Journal of Computer Applications in Archaeology 5(1): 96–111. DOI: https://doi.org/10.5334/jcaa.97.
James Frances Loftus III
Kyushu University | Japan Society for the Promotion of Science
@JamesFLoftusIII
Recent advances in open-source availability of statistical models in archaeology have propelled modern morphological studies of material goods into a trajectory of quantifiable and reproducible analysis, of which can be utilized in a multitude of socio-cultural contexts. These statistical methods have continued to shed light on previously unrecognized complexities in material form across multiple periods and regions; However, in modern studies of the transition from hunter-gatherer to wet-rice agricultural society during the Yayoi period of Japan, previous works often rely on un-quantifiable, visual assessments of material similarity, such as “attribute analysis”, from which typo-chronological analysis is connected. As such, many theoretical investigations based on these methods fail to quantify, through statistically significant means, complexities in morphological form, and often fall short of creating a convincing approach to modern anthropological theory. This study seeks to apply a micro-regional ceramic-based case study of Northern Kyushu Yayoi period earthenware to clarify the benefits of expanding methodological practice to include novel advances in geometric morphometric, and 3D morphometric analysis. Results show that during this period, waves of previously unrecognized ceramic standardization can be statistically quantified, with direct connections between increasing rice-cultivation intensity, population density, and ceramic production contexts apparent. Despite intensive standardization practices during production, novel 3D morphometric mapping is able to illuminate idiosyncratic micro-styles within these increasingly “uniform” assemblages.
This paper was later published as: Loftus III, JF and Seguchi, N. 2022 Visualizing Variability in Ceramic Micro-Curvature though Novel 3D Morphometric Mapping and Sliced Segmental Extraction. Journal of Computer Applications in Archaeology 5(1): 57–77. DOI: https://doi.org/10.5334/jcaa.89.
Aleks Michalewicz, Elena Vella, Robert Walton, Goran Duric, Daniel Williams, Justin Green
University of Melbourne
Investigating standing stones and swarm robotics: a new approach to ancient and emerging technologies.^
An interdisciplinary research project guided by archaeology, robotics and theatre, SACRIFICE seeks to engage the public in an immersive performance art experience. At its core, the project’s gallery installation performance deploys a group of robotic vehicles (a “swarm”) cloaked as ancient standing stones, navigating autonomously through the venue and responding to human participants’ movements and attention. By replicating the appearances of standing stones, the project combines a trusted ancient technology with emerging paradigms of automation.
Through our research we are investigating human trust within the context of intercultural exchanges and cross-disciplinary relationships. The global pandemic prompts imagination of new methods for communicating archaeological knowledges at a distance. Though digital solutions are never perfect, we are forging a global network of cultural custodians and archaeologists, centred around ancient standing stones, together to create a mobile multiperiod megalithic monument.
The swarm’s motions are composed using a blend of classical control algorithms with modern machine learning methods, to study behaviours that harmonise with social expectations of co-located humans. Our approach interweaves elements of archaeology, machine learning, mechatronics, performance art and fabrication. In this way SACRIFICE uses the lens of technology as a personal phenomenon to re-appraise notions of human trust and engagement. The research will culminate in a public installation at the Melbourne Science Gallery in 2022, inviting members of the public to experience and interact with a world-first autonomous swarm of stones.
Stacey Middleton, Rebecca Phillipps, Joshua Emmitt
University of Auckland
The use of three-dimensional (3D) modelling in stone artefact analysis using laser scanning or photogrammetry has provided methods to enhance the accuracy of artefact quantification. Shape morphology and measurement are a central element to classic stone artefact analyses, and current studies focus on quantification at the individual artefact level through morphometric analyses. However, quantification using three-dimensional methods at the assemblage level has not yet been thoroughly explored. The volume ratio is a method developed by Phillipps (2012) that uses the loss of volume of an assemblage as a proxy for artefact transport and therefore human movement. This research uses experimental assemblages to test different methods of volume calculation including photogrammetry, mathematical formulas, and density.
This paper was later published as: Middleton, S and Phillipps, R. 2022 Experimental Improvements to the Volume Ratio and Quantifying Movement Using Stone Artefact Analysis. Journal of Computer Applications in Archaeology 5(1): 152–168. DOI: https://doi.org/10.5334/jcaa.93.
Gala Morris, Joshua Emmitt, Jeremy Armstrong
University of Auckland
Digitising minted memories: photogrammetry and 3D modelling methods for ancient coins.
First-year Ancient History and Classics students at the University of Auckland have traditionally been given access to select coins from the university’s numismatics collection as part of their tutorials. First-hand student interaction with ancient coins in the classroom, however, is limited due to time constraints and the requirement for staff-supervision. Digitised versions of these coins could circumvent such limitations. Photogrammetry and 3D modelling of ancient coins provide new opportunities for artefact interactivity and accessibility. The circular shape and small size of ancient coins, however, present specific challenges for image capturing and 3D rendering. Several Roman Republican coins from the University of Auckland’s numismatics collection were digitised using photogrammetric methods. This paper will discuss the current methodological processes, challenges, and limitations of creating 3D models of ancient coins, and discuss their use as a pedagogical tool for teaching Archaeology and Ancient History in the classroom.
This paper was later published as: Morris, G, Emmitt, J and Armstrong, J. 2022 Depth and Dimension: Exploring the Problems and Potential of Photogrammetric Models for Ancient Coins. Journal of Computer Applications in Archaeology 5(1): 112–122. DOI: https://doi.org/10.5334/jcaa.99.
Lily Nash, Josephine Verduci, Martin Tomko
University of Melbourne
Connecting the dots on the Budj Bim Cultural Landscape.^
The UNESCO World Heritage Budj Bim Cultural Landscape is a complex of aquaculture structures engineered by the Gunditjmara people to systematically store, trap and harvest kooyang eels as early as 6600 years ago. Remnants of stone structures are scattered throughout this rugged volcanic landscape including circular dwelling foundations, shelters, eel traps and canals. The complex topography and dense vegetation that characterises the landscape poses significant difficulties for the identification of stone structures and artefacts and the interpretation of their spatial relationships. Yet, such knowledge is key in understanding how this landscape has been settled and used.
In this research, we combine geomorphological analysis and computational GIS methods (least cost path analysis) to study the relationships between loci of shelter and loci of resource exploitation in their landscape context. We use a very high-resolution digital elevation model collected in March 2020 by DELWP following the summer bushfires to estimate likely interaction pathways in the landscape. Analysis at this level of resolution has not been previously undertaken in Australia. We expect to improve the understanding of the spatial relationships between sites, including the potential identification of highly accessible and central foci of community life.
This research thus seeks to further understanding about pre contact Aboriginal life at Budj Bim by focusing on the way that the topography and spatial structure of the rocky landscape shaped movement, building and engineering.
This project has been made possible by the Australian Research Special Research Initiatives Project SR200200227 “Indigenous Engineering: interpreting engineering foundations of Budj Bim”.
Corey Noxon
Ritsumeikan University
@archaeojapan
In this study, the boom and bust population pattern that appears to have occurred during the Middle Jomon period in the Kanto and Chubu regions of Japan was reexamined through the use of skeletal remains. Prior Jomon population estimations have been based on the use of site counts, dwelling counts, or summed probability distributions of radiocarbon dates, but these methods all share the similar weakness of being susceptible to changes in residential mobility. Paleodemographic analyses relying on skeletal remains should be more resistant to these changes and provide a way to rule out changes in residential mobility as a driver of the population changes thought to have been identified in previous studies. This project introduced an updated approach to site dating using Bayesian methods rather than date range midpoints, adjusted time scales to help counter edge effects, and attempted a reproducible analysis workflow using R, the rrtools and renv packages, GitHub, and the Open Science Framework repository. The skeletal data analysis in this project showed results similar to previous population studies, indicating that the population boom and bust pattern was not strictly a reflection of possible mobility changes, but indeed represented an actual shift in population levels.
Rebecca Phillipps, Joshua Emmitt, Sina Masoud-Ansari, Stacey Middleton, Simon Holdaway
University of Auckland
@beccaphillipps
Application of machine learning to stone artefact identification.
Stone artefacts are ubiquitous in archaeological contexts world-wide and their use in making inferences about the human past is wide-ranging. Stone artefacts also indicate the presence of archaeological remains therefore rapid identification of stone artefacts in comparison to natural lithic clasts is desirable. Conventionally this relies on the ability of observers to identify landmark features of conchoidal fracture on an artefact, but even if basic knowledge is acquired identification may be difficult due to raw material type or fragmentation of the object. Machine learning offers one solution. Here we apply machine learning in the context of flaked stone artefact identification to differentiate natural lithic clasts from flaked stone artefacts. We discuss critical issues related to error rate and confidence in identifications. We test this method on three different stone artefact assemblages from the Fayum, Egypt; Ahuahu Great Mercury Island, New Zealand; and Western New South Wales, Australia.
This paper was later published as: Emmitt, J, Masoud-Ansari, S, Phillipps, R, Middleton, S, Graydon, J and Holdaway, S. 2022 Machine learning for stone artifact identification: Distinguishing worked stone artifacts from natural clasts using deep neural networks. PLOS ONE 17(8): e0271582. DOI: https://doi.org/10.1371/journal.pone.0271582.
Kai Tateuchi
Tohoku University
The northernmost area of the Japanese mainland (Honshu) was outside of the domination of the ancient dynastic state in the Heian period (9th -12th centuries). In the 10th century, a large number of people migrated to the northernmost area from areas in the state, leading to a drastic change in the composition of the local community. Previous studies tried to detect the specific origins of immigrants by focusing on technologies and styles of pottery employed by immigrant potters. However, due to the lack of rigorous quantitative methods to analyze morphological similarities of pottery, the assumed origins have not been statistically evaluated. This study aims to quantify and investigate the similarity of pottery shapes to understand the dispersal processes of new pottery styles and the origins of the immigrant potters.
In this study, an outline-based analysis using EFA is performed. Pottery outlines are extracted from published illustrations as a series of x-y coordinates and converted into Fourier function coefficients. To examine morphological similarities, principal component analysis (PCA) and other statistical analyses are performed on the coefficients. The author uses R packages for analysis to ensure all analytical processes openly available. The result of PCA shows an overlap of PC scores between pottery from the northernmost area and pottery from relatively close areas. This indicates a similarity of pottery between these areas. This research provides insight into the origins of immigrant potters and allows further discussion of the causes of immigration in a historical context.
Oka Tomoyasu
Nanzan University
@OkaTomoyasu
3D analysis of small stone implements in the Jomon period.
Stone implements are the important element on a study of the Jomon period, and these are thought to be used at rituals. They are one kind of lithic tools, which were basically made by flaking, knocking, and polishing.
The purpose of this presentation is to elucidate the rituals which were performed using stone implements in the end of Jomon period. Stone implements made in the middle Jomon period were big, but these became smaller in the late Jomon period. This is an interesting phenomenon, but in this presentation, I consider the rituals after this phenomenon. How can the rituals with small stone implements be characterized? I will consider this by analyzing 3D models.
I examine the techniques to produce the small stone implements in order to clarify the ritual’s character. It is necessary to examine the production traces in order to examine the production techniques, and the 3D model is effective for showing the production traces. I use SfM/MVS to make 3D models. Completed 3D models are analyzed by GIS which are originally used to analyze geographical features. I use GIS to study small stone implements by regarding the surface unevenness on them as geographical features.
Liying Wang
University of Washington
@LiyingWang10
Indigenous settlements in northeastern Taiwan have revealed evidence of complex social systems around the 17th century, at the same time as the European presence. For example, we see a small number of burials with impressive grave goods, indicating some level of social differentiation. To explore the relationship between the European arrival and social differentiation in local Indigenous societies, we examine burial data from an Iron Age site, Kiwulan (the 14-19th centuries), using network modeling. In network modeling methods, exponential random graph models (ERGMs) are an important family of statistical models that allow direct modeling for the formation of networks and investigate the underlying mechanisms. However, the computational difficulties and the sensitivity to uncertainties limits the practical applications of ERGMs to archaeology. In this paper, we employ a Bayesian approach to ERGMs that can incorporate prior information and empirical data to provide an efficient computational process and effective quantification of uncertainty. The results show that the post-European burial network has a strong tendency of centralization associated with the rarity of goods. In contrast, the pre-European network shows a decentralized pattern with more clustered subgroups linked by ritual practice. This case study suggests that European influences can be detected in the burial networks using a Bayesian network approach with ERGMs. This approach can also be applied to a wide range of relational data in archaeology.
Simon Wyatt-Spratt
University of Sydney
@SWyattSpratt
Patterns and possibilities: a review of 3D analyses of stone artefact technologies.
With over 150 peer-review papers published over the last 20 years, 3D modelling is no longer a gimmick or a toy but an established and increasingly common analytical tool for stone artefact analysis. Studies using the technology are increasingly global in scope and utilise the models in a variety of different ways to answer questions ranging from the spread of hominid populations, assessing functional efficiency, and re-assessing cultural taxonomies. However there have been no published systematic reviews of how 3D modelling has been applied to lithic analysis and with an ever larger number of papers published every year the field is due for a review. This paper will use a bibliometric approach to explore trends in the use of 3D modelling in stone artefact analysis, the impact it is having on the wider lithic analysis literature and methodological, regional and theoretical gaps which future research projects could explore.
This paper was later published as: Wyatt-Spratt, S. 2022 After the Revolution: A Review of 3D Modelling as a Tool for Stone Artefact Analysis. Journal of Computer Applications in Archaeology 5(1): 215–237. DOI: https://doi.org/10.5334/jcaa.103.
