Overview of Computer Supported Medieval Slavic Manuscript Studies in Bulgaria

Milena Dobreva
Institute of Mathematics and Informatics
bl. 8, Acad. G. Bonchev St.
Sofia, 1113


The medieval Slavic manuscript heritage is rich and widely spread. The exact number of Slavic manuscripts is unknown. Only in Bulgaria, the state-owned repositories store about 8,000 manuscripts of Slavic origin.

These texts have been created in an interesting cultural setting. The use of a vernacular language with many regional differences at the orthographic, lexical and phrase-structure levels makes the medieval texts an important source for the study of the diachronic and synchronic development of Slavic languages. The fact that a comparison of 150 variants of a sentence from the Gospel does not result in two ideal matches is a brilliant illustration of the variety scale.

The application of computer tools to the medieval Slavic text studies could facilitate research tasks, especially those, which require analysis of text variants. This explains the interest of the specialists to the IT applications.

This paper aims at presenting the state of the field in Bulgaria. It also points to the basic unsolved problem - the development of proper text encoding system, which will enable the encoding of the medieval Slavic texts close to the originals.

First endeavors in the field

The first applications of IT to the medieval Slavic studies have not been aimed at collecting and processing texts in electronic form, but rather to collecting structured information about the manuscripts [Geurts et al. 87].

The creation of formal models was the basic difficulty in the beginning. Due to the unique researchers' profiles and interests, the views on the data to be stored and processed were significantly different.

Initially, the samples of original texts were encoded in Latin transliteration. This was a practical solution but it was unsatisfactory for the specialists who would like to be able to represent the texts as close to the original as it would be possible.

Current state

With the further development of information technologies and the spread of Windows-based applications, a new stage of the work has been achieved. The specialists' dream -- to see the text in a form close to the original, already was viewed as easily reachable. The specialists were concerned with the development of fonts presenting accurately the paleographic features characteristic for different periods, schools or scribes. However, a major problem -- that of creating adequate text encoding standard -- remained unsolved. The problems with the creation of an encoding system are presented thoroughly in [Birnbaum 96], but his encoding suggestions are still not implemented in practice.

The difficulties in creating widely accepted encoding standard are caused by several reasons:

  1. The sets of graphemes appearing in different manuscripts are different. In some cases the difference of graphemes represents character differences; in other cases these were variants of the same character.
  2. The encoding of specific textual features (e.g. superscript, subscript, inscript letters and abbreviations) is still debatable. Some of the specialists insist on encoding normalized texts where all these features disappear. For others, the encoding of the text in a form, which represents the original as close as possible, is a must. But even if we have a satisfactory encoding standard, we will need to build tools enabling search within encoded texts. The 'normalization' approach leads to better solution of the problem with text search, paying the price of data loss.

A brief overview of 70 publications from 1995-1998 in the field of computer processing of medieval Slavic manuscripts shows that 40 publications treat text representation and processing, including TEI issues. Articles on data base applications form the next largest group (10 publications). Multimedia, AI applications and preservation issues appear in isolated cases. This study was done on the material of [Birnbaum et al. 96], [Dobreva 98] and publications by Bulgarian authors published in other editions (the complete bibliography is published on [KNIGCHIJ-SCRIBE].

The major projects which were undertaken in Bulgaria up till now, include:

  1. Experiments with data base applications for cataloguing manuscripts [Geurts et al. 87].
  2. Computer Repertory of Old Slavic Manuscripts and Letters based on TEI-conformant description of medieval manuscripts [Miltenova 98].
  3. Quantitative study of orthographic variety [Dobreva, Dobrev 98].
  4. Lexicographic study of the Psalter using the DBT system [Camuglia 96].


Although the Bulgarian specialists already have practical experience in different computer applications to the medieval Slavic studies, the basic problem of developing internationally recognized encoding system is still unsolved. Under these circumstances, the efforts to collect digital resources are prone. This situation is unpleasant in general, and in countries in transition with many economic problems is a real disaster.

Important characteristic of the work in the field is that the most considerable effort is oriented towards text encoding. Real digitization work is still not undertaken. This can be explained with the economic difficulties of the Bulgarian institutions working in the field of medieval manuscript heritage.

With the above in mind, we could expect that Slavic materials would still remain underrepresented in electronic form compared to manuscripts belonging to other written traditions.


[Birnbaum et al. 96] D. Birnbaum, A. Bojadzhiev, M. Dobreva, A. Miltenova (eds.), Proceedings of the First International Conference Computer Processing of Medieval Slavic Manuscripts, July 1995, Blagoevgrad, S., 1996, 336 pp.

[Birnbaum 96] D. Birnbaum, "Standardizing Characters, Glyphs, and SGML Entities for Encoding Early Cyrillic Writing," In: Computer Standards and Interfaces 18 (1996), pp.201-252.

[Camuglia 96] M. Camuglia, The Psalter, its Tradition and the Computer: a New Method of Textual Analysis, In: Palaeobulgarica, XX(1996), 1, pp.3-13.

[Dobreva 98] M. Dobreva (ed.) "Text Variety in the Witnesses of Medieval Texts," Proceedings of Int. Workshop, Sofia, September 1997, S., 1998.

[Dobreva, Dobrev 98] M. Dobreva, D. Dobrev, "Orthographic Variety in Medieval Slavic Texts: How to Study and Model It?" In: ALLC-ACH'98, Conference abstracts, July 5-10 1998, Debrecen, Hungary, pp.36-38.

[Geurts et al. 87] A.J. Geurts, A. Gruijs, J. van Krieken, W.R. Veder, "Codicography and Computer," In: Polata knigopisnaja, Vol.17-18(87), pp.4-29.

[KNIGCHIJ-SCRIBE] The Website on Digitizing Slavic Manuscripts in Bulgaria <www.math.bas.bg/~teleling>

[Miltenova 98] A. Miltenova, "Computer Repertory of Medieval Literature and Letters," In: M. Dobreva (ed.) Text Variety in the Witnesses of Medieval Texts, S., 1998, pp.138-149.