ASOR/EPHE-PSL European Symposium

Paris, La Sorbonne, September 4-6, 2018

Download this project as a .zip file Download this project as a tar.gz file

Abstracts on digital practices: Collaborative Projects among European and International Scholars for the Study of Ancient Worlds Using Digital Approaches

Session description

Since the Digital Era, computational methods and practices have produced new evidence for our understanding of Ancient Near Eastern and neighboring civilizations from Neolithic to Late Antiquity. However, unfortunately, international collaborations are still limited. Just as CenterNet helps to build a network of “digital humanities,” our aim is to collectively initiate discussion for building collaborative opportunities among European and international scholars. No matter how advanced your Ancient World digital project is, we invite you to introduce your methods, in order to either find a partner, or build a network related to digital practices within our fields of research.

Note: See also the forum-round table on digital practices!
In addition, a practical example by Heidi Jauhiainen regarding Digital Practices vs Traditional Methods for the Study of the Ancient Near East and its Neighboring Regions

Day 1: Tuesday

Modelling Mesopotamia – Agent-Based Modelling for Early Agriculture

Maurits Ertsen (Water Resources / Delft University of Technology)

Small-scale activities many thousands of years ago in southern Mesopotamia changed the capacity of such regions to sustain large populations over long periods of time. Mesopotamia’s early anthropogenic irrigated landscapes seem to have emerged from short-term activities, but long term effects were massive. Water and irrigation can bring wealth and stability to communities and nations, but can also harm landscapes and food production on the long term. A narrow environmental threshold may separate stable, irrigated landscapes from unstable, over-used ones. Ancient Mesopotamia would have ‘collapsed’ because of salinization due to over-irrigation. However, as did the emergence of Mesopotamian society, its ‘downfall’ would have unfolded over centuries too. It is very likely that full scale effects of interventions are only visible by humans after some time has passed, possibly only in the next generation – which should make it very difficult for these next generations to relate the effects back to actions of their predecessors. The archaeological record of Mesopotamia is rich, and as such allows building mathematical agent-based models within which all different kinds of (human and non-human) agents ‘act’ and ‘link’ in building a computer-based society. This contribution will show the results of first modelling efforts for early Mesopotamian agriculture. These results are based on close cooperation between modellers, irrigation experts and archaeologists. Therefore, this paper will also discuss how such cooperative efforts can be shaped in order to maximize productive outcomes.

Keywords: Agency, Emergence, Interdisciplinarity, Irrigation, Mesopotamia, Modelling


A Digital Approach for an Ancient Digital Technology: the Case of Textiles and Weaving Techniques

Flavia Carraro (Institute for the history of science and technology Deutsches Museum)

Digital technology and simulation for reconstructing and understanding ancient techniques and technologies are particularly relevant in the case of technologies from the past which are both intellectual and material. Among these, textile technologies and weaving practices constitutes one of the most ancient and paradigmatic cases. In particular, weaving is since its invention a digital technology and testifies for algoritmical procedures dealing with both arithmetic and practical thought which become concrete in the woven artefact. The ERC Consolidator Research project PENELOPE investigates these issue by implementing digital machines and computer programs in order to understand and describe the particular features of weaving as technology and the conditions for its very reconstruction and/or simulation by a combined approach based on archaeology, philology, anthropology, mathematics and computer science. What does a digital approach to ancient textile technology tell us about the specific as well as exemplary knowledge and knowhow implicit in the woven artefacts from the past? And, crucially, how this technology and artefacts can illustrate the constraints, conditions, methodological choices implied in a digital approach? This paper will address these questions from the point of view of the anthropology of technology and science.

Keywords: Algorithmic Thought, Epistemic Tools, Knowledge, Reconstruction, Simulation, Tacite Knowledge


The Creation of a Prosopographical Website for Ancient and Late Antique History

Elena Gritti (University of Bergamo)

Two years ago, I’ve started to realize a prosopographical research about people who were native in north Italy as well as Ancient Near East and they have travelled respectively towards one or another pars Imperii, from 2nd to 7th century CE. I decided to develop this project also in digital format, so thanking to the cooperation with a webmaster (Dr. Lazzari), I have created a website, which is titled “Human Mobility between Oriens and Transpadana 98-604: A Prosopographical Research” (HUMOT). The website is structured into six sections among which a prosopographical databases: this part is structured into two different sub-sections, the first one for people north Italian native and the other for Eastern. Each of two parts includes a set of pre-formatted indexes, based on six categories of information (“name”, “textual sources”, “epigraphies”, “time”, “career titles”, ”places”), that permits the selection of sub-sets of persons based on various search criteria. Furthermore every sheet is endowed of a tag-system, based on key elements, to improve the search. Every category of two databases is further linked to detailed study or image (name: its biographical sheet; textual source: scanner copy, epigraphies: scanner corpus, carrer title: definition and significance, place: GPS localisation and map). This digital solution will allow an uninterrupted update from other primary and secondary sources and continuous exchanges with other interested people and scholars.

Keywords: Cartography, Database, Expandability, History, Portability, Prosopography


Towards a Collaborative (and Multidisciplinary) Annotation of Ancient Greek Texts: the Euporia System

Gloria Mugelli (Università di Pisa and Ecole des Hautes Études en Sciences Sociales) and Federico Boschetti (Institute of Computational Linguistics, CNR Pisa)[1]

Euporia is an annotation system originally created to study the ritual dynamics in ancient Greek tragedies from an anthropological perspective. The system is designed to be flexible enough so that it can be easily extended in other directions of multidisciplinary research. The system combines the simplicity of a web interface pared down to its essential elements with the expressivity of a domain-specific language parsed with ANTLR, that avoids the verbosity of general-purpose markup languages (such as XML-TEI) during the annotation phase. In this way, the user is focused on domain-specific tasks by writing concise annotations. Upon exportation of our data, interoperability is ensured by two measures: references to the annotated text are translated from a system based on progressive word numbers to the Canonical Text Services (CTS) system, and the annotations are translated into XML-TEI. An annotation is constituted by a text reference, a condition related to variant readings and interpretations of the same text, as well as by a sequence of tags. Tags are created by following a bottom-up approach: they are progressively introduced and reused by the domain-expert during the annotation process. During revisions, tags are grouped and mapped onto an ontology, in order to enable and to exploit the identification of relations among the tags in querying the annotated corpus. Being designed for interoperability, our approach can be extended to other research fields (e.g. philology, archaeology) through the creation of new domain-specific languages and domain-specific tagsets, in order to improve the functionalities of the Euporia system.
[1] Additional co-authors: Angelo Mario Del Grosso (Institute of Computational Linguistics, CNR Pisa), Anas Fahad Khan (Institute of Computational Linguistics, CNR Pisa), and Andrea Taddei (Università di Pisa)

Keywords: Ancient Greek Tragedy, Annotation, Anthropology of the Ancient World, Domain-specific Languages, Folksonomies, Ontologies


Social Network of Scribes in the Neo-Babylonian Sippar

Paulina Pikulska (University of Warsaw)

Scribes in the Neo-Babylonian Mesopotamian city had their hands full of work. They wrote all kinds of texts - dowries, leases, debt notes, and many more. Most of them were recorded in the presence of witnesses, dated, and signed by sides and/or witnesses. Thanks to them assyriologists can reconstruct economic history of region and period. So far, the group of scribes as a single community of practise has not been analysed. In this paper I would like to present the first results of the research on the social network of scribes from the Sippar, ancient city to the north of Babylon, based on the corpus including almost thousand cuneiform economic texts. The investigated texts were written during Neo-Babylonian period and in Sippar. Out of almost 400 scribes, whose name were legible, only 106 could be identified as authors of two or more texts. Three groups could be clearly distinguished among them - one, the scribes who recorded transactions involving temple’s matters and properties, two, the scribes who wrote texts about individual transactions for inhabitants of Sippar, and three, the scribes who have their fingers in both pies. This paper aims to focus on the last group and to contribute to the research of local ancient communities thanks to the Social Network Analysis digital methods.

Keywords: Ancient Near East, Ancient Society, City Sippar, Neo-Babylonian Period, Scribes, Social Network Analysis


Cult Epithets and Networks of Divine Powers. On the First Steps of an Ongoing ERC Project

Fabio Porzia (Université de Toulouse - Jean Jaurès) and Elodie Guillon (Université de Toulouse - Jean Jaurès)

The five-years ERC Advanced Grant “Mapping Ancient Polytheisms. Cult Epithets as an Interface between Religious Systems and Human Agency” (MAP – 741182) aims at producing a cartography of the divine world in the North-West Semitic and Greek areas from the perspective of thousands of epithets found in epigraphic and literary sources from the first millennium BCE to the fifth century CE. Divine epithets, someone shared by two or several gods, someone specific or exclusive, shows that pantheons functioned as a complex, fluid and relational network of divine powers. The main goal of the project is the creation of an open access database of all divine epithets considered in their own contexts. This is particularly important for the Levantine world, whose epigraphic religious documentation is up to now scattered in different specialised publications. Given the huge amount of documentation and its geographical and chronological extent, this project aims, for the first time, at studying ancient religions thanks to Big Data processing methods. Borrowing tools from the Digital Humanities, such as statistical, spatial and network analysis, we intend to propose a fine-grained and dynamic understanding of the evidence at different scales of time and space. In order to produce such a complex picture of ancient religious systems and of human strategies for communicating effectively with the gods, we are looking for collaborations with projects dealing with the study of networks and mobility in ancient worlds, Levantine and Mesopotamian religious systems through epigraphical, literary and archaeological approaches.

Keywords: Database, Epithets, Networks Analysis, Religious Systems, Statistics


Common Babylonian Religious Praxis: Statistical Comparisons of Religious Theory and Practice

David Danzig (New York University, Institute for the Study of the Ancient World) and Michal Bacon (New School, New York City)

Access to common Babylonian religion through texts is very limited. By comparing the dates of ordinary transactional texts with religious calendars, or hemerologies, we aim to gain entry to the realm of non-elite religious practice. Elaborating upon cultic calendars, hemerologies list days of the year with positive or negative associations, allowing adherents of Babylonian religion to best manage their activities and thereby maximize their success (Livingstone 2013). We compare over 1,000 dated texts from the region of Nippur, Central Babylonia, from the first millennium BCE with select hemerologies in order to assess for adherence to the calendars. We look for potential difference in adherence by dividing results along socio-economic and ethnic lines, the latter based mainly on the language of the names of persons in the texts. This study follows on the heels of a similar one by Waerzeggers (2012), but uses a different data set, aiming to corroborate her results, while also varying some of her methodology to improve general results. In a more particular analysis regarding persons of one ethnic group, Judeans, we assess the religious development of Judaism in the context of the Neo-Babylonian and Persian empires. In search of potential adherence to the Judean/Jewish Sabbath, a crucial component of early Judaism (Hallo 1977), we compare the subset of dated texts related to Judeans to the seven possible series of seven-day-skips by means of statistical analysis (Rani and Sikka 2012). Visual representations of these two investigations will be developed to enhance understanding of our results.

Keywords: Babylonian Religion, Cultic Calendar, Cuneiform Economic Texts, Data Visualization, Ethnicity, Statistical Analysis


Implementing a spell-checker for cuneiform sources: experiments, tests and results within the frame of the Ebla Digital Archives project

Massimo Maiocchi (Università Ca’ Foscari Venezia)

On-line repositories of cuneiform texts are flourishing these days, providing transliterations of thousands of ancient inscriptions. Notwithstanding the game-changing nature of current projects, incoherent readings and input errors may happen. Such issues obviously reduce the research potential of any database, as output queries may provide either incomplete or wrong results. This situation may be due to several factors: typos, legacy transliterations being used as input, personal preference of the individual scholars involved in the digitization process, etc. In the case of big-data projects, human control is often insufficient, as manually reviewing the whole dataset definitely requires too much time. The paper tackles the problem of making a database of cuneiform sources as coherent as possible, borrowing different methodologies commonly used in Natural Language Processing, such as frequency lists, metric distance, and language models. The ultimate goal is to implement an open-source spell-checker (in Python and/or Perl) for the analysis of cuneiform sources in transliteration, as a useful tool for all scholars working with ancient Mesopotamian documents. The corpus of cuneiform texts from Ebla – available through the [Ebla Digital Archives project](http://ebda.cnr.it) – will be used as input to test preliminary scripts. In the spirit of the ASOR/EPHE-PSL European Symposium, collaboration with other projects will be seek in order to further expand the spell-checker, as to deal with other text corpora from ancient Mesopotamia as well.

Keywords: Cuneiform, Ebla, Internal Coherence, N-grams, Predictability, Text Restoration


Day 2: Wednesday

Semantic Domains in Akkadian Texts

Tero Alstola (University of Helsinki)

The “Semantic Domains in Akkadian Texts” project (funded by the Academy of Finland in 2016–2020) aims at enhancing the understanding of the Akkadian language by generating contextual semantic domains for Akkadian lexemes. Quantitative methods from language technology, especially word sense induction, are used to analyze an electronic corpus of Akkadian texts. The text corpus used in the project is primarily derived from Oracc (the Open Richly Annotated Cuneiform Corpus), the number of available texts being roughly 17,000. The project also aims at developing language technological methods and software tools which can be employed to other extinct languages or small and fragmented text corpora. The methods and software tools developed in the project will be made available online. The project has started its work by analyzing the text corpus and testing several language technological methods on the data. An early outcome of this work is a forthcoming article on the semantic fields of the Akkadian lexemes “horse”, “to speak”, and “power”. These lexemes were analyzed by using two existing language technological methods, Pointwise mutual information (PMI) and Word2vec. The results of this study show that these language technological methods can improve our understanding of an ancient language and its concepts when combined with careful philological work on the primary sources.

Keywords: Akkadian, Assyriology, Language Technology, Natural Language Processing, Semantic Domain, Word Sense Induction

Language and Dialect Identification of Transliterated Cuneiform Texts in Oracc

Tommi Jauhiainen (University of Helsinki)

Automatic language identification is the task of determining a language of a piece of text from the clues in the text itself. Language identification for long texts in well resourced languages is not a difficult task, but it becomes increasingly more challenging when we target short, fragmented, and multilingual texts in languages where the amount of training material is severely restricted. The computational methods used in language identification vary from simple wordlists to state-of-the-art deep learning methods. In “The Finno-Ugric Languages and The Internet” project we have developed a language identification method called HeLI, which has proven to be generally efficient in different kinds of situations. Using HeLI, we will evaluate how reliably language identification can be done for texts in the Oracc corpus (the Open Richly Annotated Cuneiform Corpus). We have also previously developed a method to infer the set of languages used in multilingual documents, which we will use to find such texts in Oracc. HeLI is a supervised learning method, where the language models are created from a correctly tagged training corpus, but we will also be researching the possibilities to use unsupervised methods to form the language clusters from untagged data. The research is carried out in the contexts of the “Semantic Domains in Akkadian Texts” project and the Centre of Excellence in Ancient Near Eastern Empire of the University of Helsinki (ANEE), both funded by the Academy of Finland.

Keywords: Akkadian, Assyriology, Corpus, Cuneiform Texts, Language Identification, Language Technology


Digitalization and the literary history of Assyrian royal narrative texts

Johannes Bach (University of Helsinki, Faculty of Arts)

The paper will address the efforts made by team 1 of the Helsinki-based research project Ancient Near Eastern Empires towards the digitalization of narrative elements of Assyrian royal narrative texts, and demonstrate the structure and applicability of this upcoming database. The aim of the database is to provide an annotated overview of the literary history of Assyrian royal narrative texts in the form of an ORACC-integrated database. Based on the structuralist theory of Gérard Genette, the genre of Assyrian royal narrative texts can be viewed as a multi-complex intertextual discourse that can be analysed by meticulous registration of the historical developments of its constituting parts (form, mode, content). Although mode and especially form play a considerable role, the database will only focus on the segment of content and its literary analysis, for that very subset allows the most substantiated research into the historic processes of royal identity building and maintenance. The database will also be able to support intertextual research, for all entries in the database are annotated and outfitted with various links to trace the literary history of a poetic element up to its oldest attestation. Besides that, short overviews and articles focussing on specifics of Assyrian history and especially literary history both in broad and narrow perspectives are included. The paper will conclude with a few demonstrations of the database’s abilities as established by August 2018.

Keywords: Assyrian Literature History, Assyrian Royal Inscriptions, Digitalization, Identity Building, Intertextuality, Structuralism


Integrating Knowledge on Ancient Near Eastern Seals and Sealings

Anna Kurmangaliev (Ludwig-Maximilians-Universität München)

Ancient Near East seals and sealings on cuneiform tablets from more than three millennia provide us with a huge but complex set of visual and textual data. More than a century of separate analyses and publications by archaeologists, art-historians and philologists led to coexisting bodies of knowledge, almost impossible to approach in a synthetic manner. In 2017, the Institute for Near Eastern Archaeology of Ludwig-Maximilians-University at Munich launched a BMBF-sponsored project to develop a concept for “Digitizing and Labeling Ancient Near Eastern Seals and Sealings” (DigANES). Creating persistent identifiers, linking them with existing repositories, splitting complex images into pictorial elements, and integrating artefact-, image-, date-, context- and inscription-related data in a useful manner, is crucial to our approach, and will ultimately enable us to integrate ANE seals into a semantic web. In this paper, we present the current stage of our project, the structure of our SQL-database, and the principles applied when labelling pictorial elements with the help of a controlled vocabulary. Many of the problems we have struggled with revolve around bridging the gap between “traditional” or “conventional” ways of thinking in our fields of study with Digital Humanities methodologies and objectives, and will thus contribute to the overall-goal of this session.

Keywords: Ancient Near East, Controlled Vocabulary, Cylinder Seals, Linked Open Data


The Wisconsin Palmyrene Aramaic Inscription Project: Reflectance Transformation Imaging and the Digital Humanities

Preston Atwood (University of Wisconsin-Madison) and Jesse Pruett (University of Wisconsin-Madison)

The Wisconsin Palmyrene Aramaic Inscription Project (WPAIP) at the University of Wisconsin-Madison is attempting to re-collate the full corpus of Palmyrene Aramaic inscriptions, providing detailed photographic records by means of a new photographic technology, Reflectance Transformation Imaging (RTI), and creating new editions of each epigraph. We investigate the evolution of and stylistic variation within the Palmyrene Aramaic script (using computer programming that measures the script accurately with its image), the languages inscribed on the epigraphs, the onomastic data and prosopography presented on each relief, and the modes and avenues of the inscriptions’ distribution throughout the antiquities market since the 19th century. Such a compilation of the corpus is significant for the preservation of Syrian cultural heritage in the face of recent Syrian political unrest. RTI has made possible the detailed photographic preservation of inscriptional materials. With this computer software, a photographer can manipulate the light source within the digital photographic environment so that various contours of the inscription are illuminated, exposing every angle of an inscription as the light moves, even features of the inscription that remain invisible to the naked eye. Our digital humanities collection comprises hundreds of conventional photos, many of which have been captured from various light settings. Moreover, many of the images include certain effects produced with RTI technology (see uwdc.library.wisc.edu). Uploading a full-scale inventory of our RTI photographs to the website is WPAIP’s most imminent objective. These images are the inception of what we hope will become a comprehensive database of high quality photos of Palmyrene Aramaic inscriptions.

Keywords: Digital Humanities, Epigraphy, Inscriptions, Palmyra, Photography, Reflectance Transformation Imaging


Virtual Access to an Unreachable Site, The Saint-Symeon Site (Syria)

Micheline Kurdy (Labex RESMED)

The Saint-Symeon site dating back to the fifth century is located in the limestone Massif on North of Syria, at the North West of Aleppo. Anemblematic figure, the early Christian sanctuary, pilgrimage site of Saint-Simeon (V-Xth centuries, mainly VIth), is spectacular for its archaeological wealth, its grandiose architecture and its part in history. The architectural analysis of monuments at different scales is one of the main axes of the French archaeological mission research in Qalaat Sem'an and Deir Sem'an. On that purpose, variety of digital tools was applied quite early on the entire site since 2000. What allowed us to have a 3D documentation of almost the entire site, which builds up actually a valuable testimony of the state prior to the war and a way to continue our research project in spite of the inaccessibility of the site and despite the damage that has occurred. With the digital methods development, new opportunities and larger perspectives are available for us for a better understanding and more sharing. Our project of architectural analysis of the site focus on the three-dimensional restitution of the whole site in order to have a virtual representation of the complex in its initial state of construction and the various modifications over the time. Moreover the objective of the archaeological analysis, the aspect of the knowledge sharing and the diffusion of our studies was always part of the research project of the French mission of Saint-Simeon.

Keywords: Data, Paleochristain, Saint-Symeon, Safeguarding, Syria, Virtual Restitution


Interactive 2D and Multispectral Imaging on the Crossroads of Archaeology, Egyptology and Assyriology

Hendrik Hameeuw (KU Leuven), Athena van der Perre (KU Leuven), and Vanessa Boschloos (Metropolitan Museum of Art)

The time of documenting objects with merely a camera, combined with a (often biased) description, lies already a good deal behind us. Today, the possibilities of documenting objects in an objective and extremely detailed manner are numerous. More than a decade ago, a dome-shaped device was developed at the KU Leuven to register the reflectance and 3D characteristics of surfaces based on the principles of photometric stereo, and that allows interactively visualising them in real time. Since, this Portable Light Dome has been used for the documentation and study of many different types of ANE archaeological artefacts. The multispectral version of this Portable Light Dome (MS PLD) innovatively combines the multi-light reflectance (MLR) visualisation of surfaces with IR, UV and separately defined RGB imaging. This contribution introduces the method’s approach by focusing on the objectives, opportunities and results of MS PLD registrations of objects in archaeology, egyptology and assyriology. More specifically, case studies will demonstrate the advantages for the documentation of delicate museum artefacts such as Egyptian polychrome wood (coffins, cartonnage and mummy portraits), inscribed unbaked clay figurines, coins, and stone stamp seals. The system thus offers an all-in-one tool with very high potential for advanced studies in conservation and material studies and in archaeological and art historical research.

Keywords: Inscriptions, Multi-light Reflectance, Multispectral Imaging, Non-invasive Analytical Techniques, Photometric Stereo, Polychromy