The writing of Sumerian in the third millennium BC. A computer-basedanalysis of the use of phonographic cuneiform signs and itsdiachronic and regional variation

虽然一个楔形文字符号可以表达的多种可能的含义通常是众所周知的，但对于在特定时间，地点和背景下实际使用的这些含义，没有多少可靠的数据存在。虽然几个例子足以建立一个迄今为止未知的意义的标志，一个语料库的大部分内容必须经过系统的评估，以确定它不能被用来呈现给定的含义。然而，随着CDLI和BDTNS等大规模语料库项目的创建，这使得苏美尔文字材料的大部分都可以用数字形式获得，这项工作现在至少可以在现代计算方法的帮助下部分自动化。因此，第一次可以用相对较少的努力对楔形文字符号的使用进行全面的研究。本项目旨在尽可能全面地分析公元前3000年苏美尔文本中表音符号的使用。在第一步中，将在机器学习算法的帮助下识别和分类所有的表音符号实例。表音符号分为五类：词缀链中的表音符号、词干中的表音符号、用来书写闪米特语术语的表音符号、特别拼写的表音符号和表音补语。此外，拼写变体的phonographically书面茎和专有名词收集。在此基础上，建立了各重要地区、时期和符号类的音节表和符号表。G.作为文本阅读的资源，这些数据作为进一步研究的基础具有重要意义。因此，第二步，它对理解当地楔形文字传统及其发展，苏美尔楔形文字与早期闪米特文字系统之间的关系，苏美尔语的音韵学，和美索不达米亚的表音文字的发展将被探讨。这个项目的一个明确的目标是进一步创建一个基线，为使用计算方法在苏美尔和创建一个广泛的注释文本语料库，可以作为进一步研究的重要基础。

Whereas the multitude of possible meanings that could be rendered by a single cuneiform sign are usually well known, not much reliable data exists as to which of these were actually in use in a given time, place and context.There are practical reasons for this: While a few examples are sufficient to establish a hitherto unknown meaning of a sign, large parts of a corpus have to be systematically evaluated to determine with due certainty that it could not be employed to render a given meaning.However, with the creation of large-scale corpus projects like the CDLI and BDTNS, which have made the better part of the Sumerian textual material available in digital from, this work can now be at least partly automated with the help of modern computational methods. Thus, comprehensive studies of the usage of the cuneiform signs are possible for the first time with relatively little effort.This project seeks to analyze as comprehensively as possible the usage of the phonographic signs in the Sumerian texts of the third millennium BC. In a first step, all instances of phonographic signs will be identified and classified with the help of a machine learning algorithm. Five classes of phonographic signs are distinguished: phonograms in the affix chain, phonograms in the word stem, phonograms used to write Semitic terms, phonograms in ad hoc spellings, and phonetic complements. Additionally, spelling variants of phonographically written stems and proper nouns are collected. On the basis of this data, syllabaries and sign inventories for all important regions, periods and sign classes are established.Apart from its immediate value, e. g. as resource for the reading of texts, this data is of great significance as a basis for further research. In a second step, therefore, its contribution to the understanding of local cuneiform traditions and their development, the relationship between the Sumerian cuneiform and the early semitic writing systems, the phonology of Sumerian, and the development of phonographic writing in Mesopotamia will be explored.An explicit goal of this project is furthermore to create a baseline for the use of computational methods in Sumerology and to create an extensive annotated text corpus, which can serve as an important basis for further investigations.