Periodic Reporting for period 1 - ArchBiMod (Agent-Based Modelling to assess the quality and bias of the archaeological record)
Période du rapport: 2021-09-01 au 2023-08-31
The project focused on addressing how different bias affecting the archaeological can condition the hypothesis building framework of the discipline and, hence, our knowledge of the past. By focusing on the spread of farming throughout the Western Mediterranean, the project has tried to assess how the fragmentary characteristics of the archaeological record, the research biases when gathering information and/or the frequent gaps in specific areas and chronology might hinder the way we understand the path that has led us to the present. To do so, it has relied on intensive computational simulation and statistic modelling, thus positioning itself at the vanguard of current methodological approaches in archaeological inference.
This can transfer to society in several ways. First, it constitutes a meta-scientific project which not archaeology, but every discipline where biases and problems on the data are important, can benefit from. This virtually includes several research disciplines studying the past (e.g. Paleontology) but also current disciplines where the recovery of data is an additional challenge. Additionally, and by properly assessing how these data are affecting our current knowledge of the past, the project has been able to refine certain theoretical aspects affecting the life and interactions of the last-hunter gatherers and the first farmers of the Western Mediterranean, while also extending its methods and inferential background to other global areas (e. g. Scandinavia or Japan).
The overall objectives of the project were, first, determining the most relevant biases affecting archaeological research. The project has been successful in this regard by developing large amount of computational simulation, and the consequences of the different biases considered, modelling from the moment in which the record is produced to how this is lost and how archaeologists condition our potential hypothesis building through their different methodological approaches. Second, the project was interested in understanding how tactical simulation could help us in this regard. The extensive deployment of this technique has allowed the project not only to obtain a better informed knowledge of how different parameters interact and condition our theoretical framework, but it has also given place to some enhanced methodological solutions that can be exported to different scientific fields.
Finally, the project has been able to reassess and challenge (as will be shown in upcoming publications) some currently standing theories, regarding the spread of farming, that were relying on under-assessed data. By focusing on the process that define the archaeological data and the methods that we used to extract the most information out of it, the project has been able to bring further archaeological hypothesis building as well as developed bespoke methodologies that can be exported and used in different research problems.
This can transfer to society in several ways. First, it constitutes a meta-scientific project which not archaeology, but every discipline where biases and problems on the data are important, can benefit from. This virtually includes several research disciplines studying the past (e.g. Paleontology) but also current disciplines where the recovery of data is an additional challenge. Additionally, and by properly assessing how these data are affecting our current knowledge of the past, the project has been able to refine certain theoretical aspects affecting the life and interactions of the last-hunter gatherers and the first farmers of the Western Mediterranean, while also extending its methods and inferential background to other global areas (e. g. Scandinavia or Japan).
The overall objectives of the project were, first, determining the most relevant biases affecting archaeological research. The project has been successful in this regard by developing large amount of computational simulation, and the consequences of the different biases considered, modelling from the moment in which the record is produced to how this is lost and how archaeologists condition our potential hypothesis building through their different methodological approaches. Second, the project was interested in understanding how tactical simulation could help us in this regard. The extensive deployment of this technique has allowed the project not only to obtain a better informed knowledge of how different parameters interact and condition our theoretical framework, but it has also given place to some enhanced methodological solutions that can be exported to different scientific fields.
Finally, the project has been able to reassess and challenge (as will be shown in upcoming publications) some currently standing theories, regarding the spread of farming, that were relying on under-assessed data. By focusing on the process that define the archaeological data and the methods that we used to extract the most information out of it, the project has been able to bring further archaeological hypothesis building as well as developed bespoke methodologies that can be exported and used in different research problems.
From the beginning of the project, the main focus has been on assessing what are the research biases and data gaps affected our understanding of the past and to find out how could we overcome those with the use of tactical simulation and a the development of bespoke statistical modelling. This has been reflected in the development of specific models, which will be put to use as interactive web applications for the broader archaeological communities, as well as technical publications which contain innovative approaches able to give response to these questions. Additionally, part of the project has also involved the acknowledgement of certain limitations imposed by archaeological data, which demand a refocus of certain theoretical aspects of current archaeological practice.
The project has achieved satisfactory results. It has been successful in raising a strong awareness on how biases in archaeological data condition our current hypothesis building framework, and it has done so through the participation on several international conferences and publications (see below). In this regard, the project has made use of some of the most advanced statistical techniques (e. g. simulation, Bayesian computation) as well as introducing techniques not yet spread within the archaeological main methodological toolkit (e. g. t-SNE). Additionally, in the publications in their final stages of preparation, the project will introduce developments spread within the ecological modelling framework, but not within archaeology (e. g. dynamic models) as well as publish bespoke solutions aimed to be future research tools for the wider archaeological community (e. g. modelling of record loss and disturbance on archaeological sites through interactive ShinyApp websites). All this provides a good quality of the results, positioning the project at the vanguard of archaeological methodological development, while contributing to raising awareness of archaeological data deficiencies.
All these has been translated in the attendance to several scientific conferences in different places such as Kiel, Amsterdam, Budapest or Belfast and 8 dedicated publications (plus two currently under review), all of which are mentioned in the technical report. Finally, the final results should include at least three more publications, which capture in detail the advancements produced by the project, all of which are on their final stages, and should be sent out for review between the next days and the end of the present year.
The project has achieved satisfactory results. It has been successful in raising a strong awareness on how biases in archaeological data condition our current hypothesis building framework, and it has done so through the participation on several international conferences and publications (see below). In this regard, the project has made use of some of the most advanced statistical techniques (e. g. simulation, Bayesian computation) as well as introducing techniques not yet spread within the archaeological main methodological toolkit (e. g. t-SNE). Additionally, in the publications in their final stages of preparation, the project will introduce developments spread within the ecological modelling framework, but not within archaeology (e. g. dynamic models) as well as publish bespoke solutions aimed to be future research tools for the wider archaeological community (e. g. modelling of record loss and disturbance on archaeological sites through interactive ShinyApp websites). All this provides a good quality of the results, positioning the project at the vanguard of archaeological methodological development, while contributing to raising awareness of archaeological data deficiencies.
All these has been translated in the attendance to several scientific conferences in different places such as Kiel, Amsterdam, Budapest or Belfast and 8 dedicated publications (plus two currently under review), all of which are mentioned in the technical report. Finally, the final results should include at least three more publications, which capture in detail the advancements produced by the project, all of which are on their final stages, and should be sent out for review between the next days and the end of the present year.
The project has worked to bring the problems of the data towards the center of archaeological inference, to raise awareness on the difficulty and uncertainty of archaeological hypothesis building and, ultimately, to bring solution to that uncertainty when possible. In doing so, it has/will challenge some well established theories regarding population dynamics during the Middle Holocene, as well as it will bring additional tools for a more rigorous assessment of the analysis of archaeological sites and record.
With the work already done, the project has brought methodological reception, and also creative ways to face the challenges posed by the archaeological record through reproducible and quantitative methodologies. Additionally, with the culmination of the final publications, still under production, and which should be sent for review shortly the project will finally close its circle and present how this new focus on the record itself can ultimately lead us to a further insight into potential past processes.
With the work already done, the project has brought methodological reception, and also creative ways to face the challenges posed by the archaeological record through reproducible and quantitative methodologies. Additionally, with the culmination of the final publications, still under production, and which should be sent for review shortly the project will finally close its circle and present how this new focus on the record itself can ultimately lead us to a further insight into potential past processes.