Eythan Levy
I am a researcher specializing in digital humanities and quantitative methods in archaeology. My main research interests are in the fields of epigraphy and chronology of ancient Egypt and the southern Levant in the Late Bronze and Iron Ages. I have obtained a PhD in computer science from Brussels University (ULB) in 2009, and a PhD in archaeology from Tel Aviv University in 2022. I am currently a postdoctoral research at the University of Bern, in the project "Stamp Seals from the Southern Levant: A Multi-faceted Prism for Studying Entangled Histories in an Interdisciplinary Perspective", funded by the Swiss National Science Foundation (see https://www.religionswissenschaft.uzh.ch/de/forschung/projekte/SINERGIA-project.html).
less
InterestsView All (51)
Uploads
Papers by Eythan Levy
Samaria III lists two bullae with alphabetic inscriptions, apparently dating to the days of the Northern Kingdom, uncovered in the 1933 campaign. They are not illustrated in the volume, however, and this is probably why they have essentially been overlooked in the literature so far. One of the bullae bears an Egyptian name written in the Hebrew alphabet, while the attribution of the signs on the other remains questionable.
Free download link at: https://authors.elsevier.com/a/1c5IN15SlTogcL (until January 6, 2021)
Permanent link at: https://doi.org/10.1016/j.jas.2020.105225
Samaria III lists two bullae with alphabetic inscriptions, apparently dating to the days of the Northern Kingdom, uncovered in the 1933 campaign. They are not illustrated in the volume, however, and this is probably why they have essentially been overlooked in the literature so far. One of the bullae bears an Egyptian name written in the Hebrew alphabet, while the attribution of the signs on the other remains questionable.
Free download link at: https://authors.elsevier.com/a/1c5IN15SlTogcL (until January 6, 2021)
Permanent link at: https://doi.org/10.1016/j.jas.2020.105225
Our second set of results consists in an integrated framework for modelling chronological networks in archaeology. Based on the premise that chronological information forms a network of interconnected units, such as reigns, strata, ceramic types and archaeological periods, we show how this information can be modelled using mathematical techniques. Our models comprise chronological units bearing termini post/ante quem and duration estimates. These units are arranged into sequences and are connected with each other through synchronisms. We show how such a model can be queried to check the internal consistency of the data, and how the tightest possible range estimates can be computed for the dates and duration of each unit. We provide new software called “ChronoLog”, that enables users to easily encode a chronological network, check its consistency, obtain the optimal range estimates, and test chronological hypotheses. ChronoLog is available online at no cost on chrono.ulb.be. As a case study to show the strength of the software, we have shown how the chronology of the Egyptian 26th dynasty can be entirely reconstructed by based on primary epigraphic and literary sources. We have also shown how the tool enables us to easily check the specific chronological impact of each piece of data, and to select a minimal subset sufficient for reconstructing the whole chronology of the dynasty. These results were published in the Journal of Archaeological Science (vol. 127, 2021).
We then proposed a new modelling framework for the concept of archaeological cross-dating, defined as the way by which archaeological contexts of unknown age are dated based on parallel contexts of known age. For this matter, we formalized the notion of chronological propagation, showing how chronological information, defined as termini on the start and end of a chronological unit, propagate from one unit to another in different fashions, according to the precise type of synchronism connecting these units. We discussed the intrinsic limitations of cross-dating between contemporary units, and ways to overcome them. We showed how ChronoLog can be used as a cross-dating tool, to efficiently automate the propagation process between chronological units. Finally, we applied this modelling framework to an important chronological case-study: Tomb 4 of the Late Minoan cemetery of Sellopoulo (Crete), which provides a chronological anchor linking the chronologies of Crete, Mycenaean Greece and Egypt. These results were published in Archaeological and Anthropological Sciences (vol. 13, 2021).
We proposed an elaborate case study showcasing the modelling frameworks described above. We have looked at the question of the date of appearance of Philistine Bichrome pottery at Megiddo. Our goal was to compute an estimate of this date in an automated way using ChronoLog, under a diverse set of hypotheses. Our model includes stratified datable Egyptian material from Megiddo, correlation between three stratigraphic sequences at the site (Area H, Area K and the University of Chicago sequence), broad duration ranges for each stratum, affiliation of each stratum to a given archaeological period (Late Bronze I to Iron IIB) and the encoding of strata containing Bichrome pottery. Under this set of hypotheses, ChronoLog computed a terminus post quem of 1183 BCE for the date of appearance of Philistine Bichrome pottery at Megiddo. A variant model, featuring a stronger hypothesis regarding the relation of Ramesses III to Stratum VII, yielded a terminus post quem of 1124 BCE for the same event. Adding recent radiocarbon results from Areas H and K to the model yielded a finer estimate of 1111-1086 B.C.E. for this event. Finally, we showed that the ChronoLog results, with radiocarbon dating, suggest that Megiddo Stratum VIIB ended later than usually thought, namely during, or only slightly before, the reign of Ramesses III. These results have been published in the Bulletin of the American Schools of Oriental Research (vol. 387, 2022).
Finally, we have proposed a new methodology for easily creating complex Bayesian radiocarbon models with the help of ChronoLog. More precisely, we augmented the tool with a feature allowing it to convert any ChronoLog model into an OxCal Bayesian model containing the same data. ChronoLog connects automatically to the OxCal website and runs the generated Bayesian model, with no need for the user to write a single line of OxCal code. Users only need to build a ChronoLog model representing all the Bayesian priors, using just a few clicks with ChronoLog’s graphical user interface, then to encode their radiocarbon determinations directly into ChronoLog. They then click a button to run the OxCal model and just wait for the OxCal results. This approach enables users with no familiarity with computer programming to easily create complex Bayesian models. It has several additional advantages: (1) it helps avoiding coding errors in OxCal models (many of which might otherwise go undetected), (2) it provides significant time saving, as building complex models with ChronoLog’s user interface is usually faster than direct OxCal programming, (3) it enables users to detect any inconsistency among the Bayesian priors before running the OxCal script. These results have been published in Radiocarbon (vol. 64, 2022).
Supervisors: Profs Israel Finkelstein, Eli Piasetzky, Alexander Fantalkin
Finally, we look at the CONNECTED VERTEX COVER (CVC) problem,for which we proposed new approximation results in dense graphs. We first analyze Carla Savage's algorithm, then a new variant of the Karpinski-Zelikovsky algorithm. Our results show that these algorithms provide the same approximation ratios for CVC as the maximal matching heuristic and the Karpinski-Zelikovsky algorithm did for VC. We provide tight examples for the ratios guaranteed by both algorithms. We also introduce a new invariant, the "price of connectivity of VC", defined as the ratio between the optimal solutions of CVC and VC, and showed a nearly tight upper bound on its value as a function of the weak density. Our last chapter discusses software aspects, and presents the use of the GRAPHEDRON software in the framework of approximation algorithms, as well as our contributions to the development of this system.
corpus complet des stèles égyptiennes privées découvertes au Levant figurant une divinité locale. Notre troisième chapitre est ainsi consacré à une stèle de Beth Shean, qui dépeint l’égyptien Imenemipet en adoration devant la divinité locale Mekal. Le quatrième chapitre est consacré à une stèle d’Ougarit dépeignant un personnage nommé Mamy en adoration devant le dieu Baal-Zaphon. Le cinquième et dernier chapitre présente la stèle d’Anat de Beth Shean, qui dépeint Hesi-Nakht en adoration devant la divinité levantine Anat. Finalement, une conclusion met tous ces éléments en perspective, relevant l’unité organique profonde de ce corpus et discutant de ce que ces inscriptions nous révèlent de la nature de la présence égyptienne au Levant durant le Nouvel Empire.