Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
18
result(s) for
"Gordin, Shai"
Sort by:
Reading Akkadian cuneiform using natural language processing
In this paper we present a new method for automatic transliteration and segmentation of Unicode cuneiform glyphs using Natural Language Processing (NLP) techniques. Cuneiform is one of the earliest known writing system in the world, which documents millennia of human civilizations in the ancient Near East. Hundreds of thousands of cuneiform texts were found in the nineteenth and twentieth centuries CE, most of which are written in Akkadian. However, there are still tens of thousands of texts to be published. We use models based on machine learning algorithms such as recurrent neural networks (RNN) with an accuracy reaching up to 97% for automatically transliterating and segmenting standard Unicode cuneiform glyphs into words. Therefore, our method and results form a major step towards creating a human-machine interface for creating digitized editions. Our code, Akkademia, is made publicly available for use via a web application, a python package, and a github repository.
Journal Article
Urbanscape, Land Use Change and Centralization in the Region of Uruk, Southern Mesopotamia from the 2nd to 1st Millennium BCE
We produce results that bridge the gap between physical and textual study of the ancient Mesopotamian landscape in the region south and west of the city of Uruk (Biblical Erech, Modern Warka). A brief survey of gazetteers of Mesopotamia, volumes listing place-names drawn from translated and published cuneiform texts from the 2nd and 1st Millennium BCE, are presented. The various gazetteers were reviewed for relevant place-names, and the results were recorded and analyzed. These are described in detail below, as are their implications. The resulting data are then compared to the results of a recently completed archaeological survey of the same region. The synthesis of textual and archaeological surveys indicates a more exacting methodology to add geographic objectivity to textual results, while connecting physical results to the qualitative detail available within the Uruk textual record. More broadly, we demonstrate how long-term historical records align with archaeological data, delineating state-level and local land use efforts around a major Mesopotamian city. In the 2nd millennium BCE, settlements were generally small but more numerous, but in the 1st Millennium BCE there was a shift towards fewer and larger settlements connected to the city of Uruk. These shifts reflect deliberate central, government policy and local responses.
Journal Article
Hittite Logograms and Hittite Scholarship. By Mark Weeden
Hittite Logograms and Hittite Scholarship. By mark WeedeN. Studien zu den Boğazköy-Texten, vol. 54. Wiesbaden: Harrassowitz Verlag, 2011. Pp. xvii + 693. €128.
Journal Article
Restoration of fragmentary Babylonian texts using recurrent neural networks
The main sources of information regarding ancient Mesopotamian history and culture are clay cuneiform tablets. Many of these tablets are damaged, leading to missing information. Currently, the missing text is manually reconstructed by experts. We investigate the possibility of assisting scholars, by modeling the language using recurrent neural networks and automatically completing the breaks in ancient Akkadian texts from Achaemenid period Babylonia.
Journal Article
Can God Deliver His Servants? Two Theological Problems in the Daniel Narratives (Dan 1:9; 3:17-18)
Abstract
This article addresses two cases from the narratives in Daniel in which a similar theological question arises concerning the uncertainty of God's ability to deliver his servants: (1) The chief officer's denial of Daniels' request (Dan 1:10) despite the fact that God granted Daniel grace and compassion from the chief officer, and (2) the speech of Shadrach, Meshach, and Abednego (Dan 3:17-18), in which they entertain the possibility that God will not, or perhaps cannot, save them. Commentators and translators throughout the generations have struggled with these theological problems, and we can identify a clear trend seeking to read the relevant verses in a way that removes the uncertainty, replacing it with certain faith in God's deliverance.
In this article, we demonstrate how this interpretive trend surprisingly continues even with modern biblical scholars. Based on a literary analysis, we suggest that reading the MT version without altering it to conform with certain theological preconceptions may shed new light on the Daniel narratives, thereby exposing their deep and complex message.
Journal Article
The Cult and Clergy of Ea in Babylon
In late first-millennium Akkadian names, the imperial deities of the Neo-Babylonian state were given preference over the ancient revered triad of the \"great gods\" Anu, Enlil, and Ea. Among these three gods, the cult of Enki/Ea is the most perplexing during this late period. On the one hand, his importance is attested in cult and ritual: Babylon contained numerous shrines to Ea, including a temple, the E-kar-zaginna, in the precinct of Marduk's great temple, the Esagil. On the other hand, he nearly vanishes from the onomasticon of given names, with the exception of some fossilized ancient family names and the personal names of close to two hundred individuals, mostly from the city of Babylon between the accession year of Nebuchadnezzar II and the thirty-second year of Darius I. This article traces the different forms of Ea worship reflected in literary and archival sources, as well as in name-giving practices. Its first part deals with the socio-religious interplay between Sîn and Ea, which gave rise to the use of ancestral names in specific urban kinship groups that revered these deities. Beginning in the Old and Middle Babylonian periods, these families spread from Nippur and the south of Babylonia (i.e. the \"Sealand\") and eventually settled in Babylon and nearby cities, where we find many of them well established in the long sixth century (620–484 BCE). In the second part of the article I investigate several such families that exhibit consistent traits of Ea worship or a relation to Ea's cult across several generations.
Journal Article
DIGITAL APPROACHES TO INVESTIGATING SPACE AND PLACE IN CLASSICAL STUDIES
Imagine a student reading Odysseus’ Cretan tale at Odyssey 19.172–84. When faced by a string of unfamiliar names – in addition to ‘native Cretans’, there are Achaeans, Cydonians and Dorians, as well as the individuals Minos, Deucalion, Idomeneus and the speaker, Aethon (Odysseus in disguise) –, they use their digital edition to find out more about each of these people and their places of origin. A personal name opens an online encyclopaedia entry, while clicking on a place launches an emerging world beyond the single text – an online atlas that provides information about the place's toponymy, form and exact location as well as links to other resources (textual and archaeological, ancient and modern) about this place, including those to which our student has contributed. The year? 2023 (Figure 1).1
Journal Article
Translating Akkadian to English with neural machine translation
Abstract
Cuneiform is one of the earliest writing systems in recorded human history (ca. 3,400 BCE–75 CE). Hundreds of thousands of such texts were found over the last two centuries, most of which are written in Sumerian and Akkadian. We show the high potential in assisting scholars and interested laypeople alike, by using natural language processing (NLP) methods such as convolutional neural networks (CNN), to automatically translate Akkadian from cuneiform Unicode glyphs directly to English (C2E) and from transliteration to English (T2E). We show that high-quality translations can be obtained when translating directly from cuneiform to English, as we get 36.52 and 37.47 Best Bilingual Evaluation Understudy 4 (BLEU4) scores for C2E and T2E, respectively. For C2E, our model is better than the translation memory baseline in 9.43, and for T2E, the difference is even higher and stands at 13.96. The model achieves best results in short- and medium-length sentences (c. 118 or less characters). As the number of digitized texts grows, the model can be improved by further training as part of a human-in-the-loop system which corrects the results.
Journal Article