One of the ways meanings of words can be understood is based on their distributional properties. Such methodology offers an interesting quantitative viewpoint on the study of the lexicography of long-extinct languages. This chapter explores the use of Pointwise Mutual Information (PMI), a well-known statistical word association measure used in collocation analysis. PMI is applied to the data in order to gain insights on the semantic nuances of Akkadian verbs of seeing (amāru, naṭālu, palāsu, dagālu, ḫiātu, barû, and subbû). To evaluate the data-driven results, the findings are compared to previous philological work by Ainsley Dicks. The analysis of the top-ranked PMI-extracted collocates provides a good overview of the typical semantic differences between the seven verbs of interest.
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.
The following study examines Canaano-Akkadian, the unique cuneiform system used in the southern Levant during the Late Bronze Age (1550-1150 B.C.E.), as a diplomatic scribal code used in contexts of mediated diplomacy with Egypt. The methodologies presented draw upon recent work on the sociolinguistics of writing and script choice that best elucidate the genesis of this scribal system and its role in Egypt's eastern empire. The classification of the language of the Canaanite Amarna Letters is still a matter of contention. The primary debate is whether or not the mixed Canaano-Akkadian forms are a reflection of a local dialect(s) of Akkadian, or a written scribal code––one that was quite distanced from the actual language underlying such messages.Recent petrographic and paleographic analyses further complicate the correlation between language and writing in this corpus. Certain cuneiform scribes worked for multiple polities and, moreover, many tablets were created at quite a distance from the political centers generating this correspondence. For example, a number of letters were written at Egyptian administrative centers across the Levant and not at the local courts "sending" these messages. The Canaanite scribe emerges as the central figure in discussions of linguistic classification, as the language of these letters is a better reflection of scribal training during this period than what was actually spoken at local Canaanite courts. As such, there is a need for a reassessment of the scribal and administrative landscape of this period, and in particular, the system of scribes and messengers. This system of communication was a dynamic, complex process that entailed at the very least four linguistic layers: the spoken dialects of the original messages; Canaano-Akkadian, the scribal code of the written versions; the mediated reading of these letters once delivered, along with any additional socio-political or metapragmatic subtext; and the final translation into Egyptian.The present study considers the metapragmatic, linguistic, orthographic, and rhetorical strategies employed by cuneiform scribes to bridge the geographic and cultural gulf between Canaanite polities and the royal court at Tell el-'Amarna, Egypt. The Canaanite glosses and scribal marks resurface as unique evidence for how Canaanite scribes approached the problem of translation and linguistic "border-crossing" in these cross-cultural and multilingual exchanges. Such strategies added nuance and a metapragmatic commentary to guide the translation and interpretation of these letters and to ensure that they received a positive reception. The cuneiform script was not limited to a technological tool in such interactions, but entailed participation in a much larger cultural horizon—one shared by cuneiform scribes in Canaan and Egypt, who were the gatekeepers facilitating diplomacy throughout this period.
Building upon the traditional methods of philological analysis, this dissertation incorporates emerging technologies in text-mining and social network analysis as a new approach for analyzing large blocks of cuneiform text corpora. Working within the Classical period of the Old Akkadian dynasty, the height of Empire's reach and influence, these digital tools are deployed to ascertain the level of administrative similarity or difference between the major urban centers. The cities of the Diyala are used as a baseline specifically because of their peaceful relationship with the Akkadian Empire. These parameters explore whether the political relationship (peaceful or rebellious) affected the degree or extent of the Empire's administrative presence in its various territories. Overall, the results indicate that the Akkadian kings practiced similar policies throughout Mesopotamia. The imperial administration was only minimally involved with the daily administration of these cities; they sought mainly refined or finished goods and left the local government to manage the means of production.
CyberResearch on the Ancient Near East and Neighboring Regions provides case studies on archaeology, objects, cuneiform texts, and online publishing, digital archiving, and preservation. Eleven chapters present a rich array of material, spanning the fifth through the first millennium BCE, from Anatolia, the Levant, Mesopotamia, and Iran. Customized cyber- and general glossaries support readers who lack either a technical background or familiarity with the ancient cultures. Edited by Vanessa Bigot Juloux, Amy Rebecca Gansell, and Alessandro Di Ludovico, this volume is dedicated to broadening the understanding and accessibility of digital humanities tools, methodologies, and results to Ancient Near Eastern Studies. Ultimately, this book provides a model for introducing cyber-studies to the mainstream of humanities research.
