Digital Kartvelology

Reading Georgian Manuscripts Automatically on the eScriptorium Platform

Jost Gippert — 2024-12-27

The article outlines the development of means for an automatic reading of Georgian manuscripts on the eScriptorium platform and the first results achieved with them. After an overview of the efforts undertaken in applying Optical Character Recognition (OCR) to Georgian printed books since the late 1980’s and a short introduction into the basics of the eScriptorium approach to Handwritten Text Recognition (HCR) and its functionalities, it exemplifies the application of the three core procedures of eScriptorium, which consist in the automatic segmentation of text-covered regions and lines, the automatic transcription of the detected lines based on manual input and the training of appropriate models, and the alignment with existing electronic texts in order to provide reliable ground truth for further training. With a total of 292 manually transcribed pages and 3812 pages with aligned (but not yet always corrected) text that have been processed so far, there is a strong material basis for further improvement of the models and the reading results depending on them.

Digital Humanities and the K. Kekelidze Georgian National Center of Manuscripts

Maia Matchavariani — 2024-12-27

The recognition of Digital Humanities as a scientific discipline emerged at the K. Kekelidze Institute of Manuscripts by the late 20th century. In the early 1990s, a first project in the field of Digital Humanities was initiated − the development of an automated database of Georgian historical documents. The database was designed to facilitate access to these invaluable resources for both Georgian and international scholars while contributing to the broader mission of promoting Georgia’s cultural heritage. In addition to cataloguing and preserving historical documents, the project had a broader, strategic goal, namely, to illustrate the transformative impact of emerging technologies on Humanities research. It aimed to foster a paradigm shift by encouraging scholars to adopt digital tools as integral part of modern research methodologies. Since then, the Institute – later renamed the Korneli Kekelidze Georgian National Center of Manuscripts – has continued to lead numerous digitally-based projects supported by the Rustaveli Scientific Foundation and in collaboration with various academic institutions. Some of these initiatives are outlined in the present paper.

Building a Dialectological Lexicological Database of Georgian Cognates for Digital Analysis

Hélène Gérardin — 2024-12-27

This paper examines the challenges involved in creating a lexical database for Georgian dialects. It begins by outlining the methodological approaches to data collection and formatting, followed by an overview of the current version of the database, and its applications for linguistic analysis. Designed to facilitate a wide range of research, the database supports investigations such as Levenshtein distance calculations and diachronic and diatopic comparisons. The long-term goal of this project is to develop an open, accessible resource that can be gradually enriched with new data, advancing computational tools and deepening our understanding of Georgian and Kartvelian dialects.

The Linguistic Model of the Georgian Language (Synthesis of Nominal Word forms)

Ketevan Datukishvili — 2024-12-27

In this paper we present a unique linguistic model designed for the morphological synthesis and analysis of the Georgian language. The morphological structure of the Georgian language is fully covered by this model. Based on it, we have created a software with a special tool, namely, a processor enabling the generation and analysis of word forms. We would like to emphasise the fact that no particular grammatical theory is used in our linguistic model. Rather, the language data is provided in a different format and in a structured way, taking into account the theories currently in use. Our model as presented in this paper is a collection of morphological equations (often around 3,000 units) required to generate every word form from a single stem. Currently, more than 266 million word-forms can be synthesised utilising the Georgian language's morphological processor. While not all of these forms can be found in electronic texts or extensive corpora, they are all viable. These words have been referred to as “potential forms”. The possible forms are crucial for studying a variety of topics pertaining to natural language processing. Additionally, they aid in the resolution of specific artificial intelligence challenges. Beyond, we intend to pinpoint the exact frequencies of the data produced by the processor and specify the domains in which they are utilised.

Research into Equivalence in Multilingual Parallel Corpora

Manana Tandaschwili — 2024-12-27

Shota Rustaveli’s epic “The Knight in the Panther’s Skin” is an exceptional monument of aphoristic style. The complexity of Shota Rustaveli’s worldview, which is mirrored in his epic, is a result of the mixing of mediaeval and Renaissance thinking. It combines Christian (and, in general, religious) thoughts with antique philosophy (Plato, Dionysus the Areopagite, Aristotle) and mythical and transcendental visions with analytical thinking. Rustaveli’s worldview, as well as several religious-philosophical doctrines – Christianity, Areopagitica, Neoplatonism, and Aristotelianism – are all reflected in the aphorisms used in the epic, which plays a special role not only in the cultural memory of the Georgians but also in the history of world literature. The epic has been translated into 58 languages and is included in the registry of the world cultural heritage of UNESCO. The variety of languages into which the epic has been translated provides a unique opportunity for the creation of the multilingual parallel corpus ‘Rustaveli goes digital’, which is an outstanding resource for translation studies to examine the quality of translations, analyse translation methods, pinpoint translators’ strategies, and contrast them. But how can we judge the quality of translations, especially in the case of aphorisms? Determining the degree of equivalence between the source and target texts is not sufficient for a qualitative evaluation of the translations of the epic; e.g., the translation quality of the aphoristic style in the translations is one of the challenges that we believe needs to be addressed. In the present paper, we discuss our research model on how to evaluate the translation quality of aphorisms.

Paul’s 1st Epistle to the Corinthians ch. 13. - A Comparison of Georgian and Abkhaz Translations (Taking into account the Greek Original)

George Hewitt — 2024-12-27

The Bible is universally recognised to be the most translated book in the world. The whole Bible (Old & New Testaments) is, unsurprisingly, available in Georgian, which has a centuries-old literary tradition. In Abkhaz, however, only one version of the whole New Testament (NT) currently exists, and that is the privately published translation (from Russian) by the late writer Mushni Lasuria (2004). I spent several months a few years ago on behalf of The Institute for Bible Translation (once based in Stockholm but now in Moscow) going through Lasuria’s translation of the four Gospels (together with a different translation, also done from Russian, by Zaira Khiba) and suggesting changes to align texts with the Greek original – Khiba’s version of the Gospels is available online. I therefore had the idea that it might prove to be an interesting and appealing exercise to take another extremely well-known part of the NT, namely Paul’s discourse on the theme of love, and examine how this has been treated by Lasuria and Georgian translators (in both Old and Modern Georgian), contrasting the results with the ultimate Greek source. Given that the 13^th chapter of Paul’s First Epistle to the Corinthians is (one might say) universally regarded as a description of love, it might come as a surprise that not all English translations actually include this word, preferring charity instead. We shall touch on the reason why this choice caused such a furious controversy in 16^th-century England.

The Theoretical Grounds and the Practical Value of Distinguishing the Markers of Evidentiality in the Verb Paradigms of the Kartvelian Languages

Maia Lomia — 2024-12-27

Evidential verb forms of the Perfect are a common Kartvelian phenomenon. Expression of evidentiality in this case is a secondary function on the synchronic plane. Perfect evidential verb forms do not have a special morphological marker. They use lexical means to express that something has not been seen. In this way, they contrast with aorist verb forms expressing an action that was seen. Evidential verb forms of the Imperfect appear only in the non-literary Kartvelian languages (Megrelian, Laz and Svan). Out of these, in Laz, evidentiality is expressed lexically (in a descriptive way), whereas in Megrelian and Svan, evidentiality has a special morphological marker. In Megrelian, the Present and Imperfect verb forms express a seen action. They are opposed to evidential verbs in the Imperfect, which denote an unseen action. In Svan, the neutral Imperfect contrasts two forms: the Evidential Imperfect I in the superessive version and the Evidential Imperfect II. Originally, the Megrelian Evidential Imperfect I and Evidential Imperfect II, and the Svan Evidential Imperfect II were descriptive. As a result of transformation, these forms turned into organic formations and developed special markers of evidentiality. The Svan Evidential Imperfect I in the superessive version seems to have been organic from the very start. In Svan and Megrelian, evidentiality has developed as a morphological category. The segmentation and glossing of evidential verb-forms in the non-literary Kartvelian languages and the distinction of the markers of evidentiality serve important practical aims that are thematised in the present article.

Unknown Pages of English-Georgian Lexicography (George Ellis and his Comparative Lexicon of Kartvelian Languages)

Tinatin Margalitadze — 2024-12-27

In 1788 a book was published by the British author George Ellis in London under the title Memoir of a Map of the Countries Comprehended Between the Black Sea, and the Caspian, with an Account of the Caucasian Nations and Vocabularies of their Languages. The book describes the history, culture, and religion of the peoples of the Caucasus, and provides information about the languages spoken in the Caucasus region. While discussing the countries of the Caucasus, a certain space is dedicated to Georgia with geographic, political and ethnographic information, which is followed by a small comparative lexicon of the Kartvelian languages. It is noteworthy that the author is interested not only in the Georgian language, but also in the other Kartvelian languages, and the book contains material of Megrelian and Svan along with Georgian. It can be said that the lexicographic material included in this book is the first example of describing the material of Kartvelian languages for English readers. Our article provides short information about the book itself, and discusses George Ellis's Comparative Lexicon of Kartvelian Languages and its sources. Special attention is paid to the transliteration rules of Kartvelian words into English, which are mainly based on the rules of Modern English spelling. G. Ellis’ Comparative Lexicon of the Kartvelian Languages is important for the study of early stages in the history of English-Georgian lexicography; it is also interesting in the context of linguistic affinity, as the Kartvelian languages are presented in it exactly from this point of view.