Andrist 2022 Asymmetrical descriptions

Sonderdruck aus/Offprint from BIBLIOTHEK UND WISSENSCHAFT 55 · 2022 Bibliothek und Wissenschaft Herausgegeben von Cornel Dora, Claudia Fabian, Monika Linder, Elmar Mittler, Wolfgang Schmitz und Antje Theise 55 · 2022 Harrassowitz Verlag · Wiesbaden Faszination (Buch-)Handschriften im Jahr 2022 Tradition und Zukunft ihrer Erschließung in Bibliothek und Wissenschaft Herausgegeben von Claudia Fabian Harrassowitz Verlag · Wiesbaden Bibliothek und Wissenschaft Bibliotheken sind wichtige Institutionen des kulturellen Gedächtnisses. Sie überliefern gedruckte und handschriftliche Dokumente und Medien aller Art und stellen sie der Wissenschaft als Quellenmaterial zur Verfügung. Geschichte und Organisation der Bibliotheken sind ebenso Objekt der Forschung wie die Bestände, die sie bewahren. Das Jahrbuch Bibliothek und Wissenschaft publiziert Untersuchungen zu einzelnen Texten, Sammlungen und Quellengattungen sowie kulturund wissenschaftshistorische Beiträge zur Geschichte und Methode der Bibliotheksarbeit und zur Bibliographie. Bibliothek und Wissenschaft ist das fachübergreifende Forum für den Prozess der kulturellen Überlieferung durch Bibliotheken. Manuskriptangebote werden an einen der Herausgeber von Bibliothek und Wissenschaft oder an den Verlag erbeten. Redaktionsschluss ist jeweils der 31. März eines Jahres. Prof. Dr. Elmar Mittler, c/o SUB Göttingen, Papendiek 14, 37073 Göttingen, E-Mail: [email protected] Dr. Cornel Dora, Stiftsbibliothek St. Gallen, Klosterhof 6 D, 9004 St. Gallen, E-Mail: [email protected] Dr. Claudia Fabian, Bayerische Staatsbibliothek, Ludwigstr. 16, 80539 München, E-mail: [email protected] Dr. Monika Linder, Staatsbibliothek zu Berlin – Preußischer Kulturbesitz, Unter den Linden 8, 10117 Berlin, E-Mail: [email protected] Prof. Dr. Wolfgang Schmitz, Universitäts- und Stadtbibliothek, Universitätsstraße 33, 50931 Köln, E-Mail: [email protected] Antje Theise, Bibliotheksdirektorin, Universität Rostock, Universitätsbibliothek, Schwaansche Str. 3 b, 18055 Rostock, E-Mail: [email protected] Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über https://www.dnb.de/ abrufbar. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the internet at https://www.dnb.de/. https://www.harrassowitz-verlag.de/ Schriftführende Herausgeber: Claudia Fabian © Otto Harrassowitz GmbH & Co. KG, Wiesbaden 2022 Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen jeder Art, Übersetzungen, Mikroverfilmungen und für die Einspeicherung in elektronische Systeme. Gedruckt auf alterungsbeständigem Papier Satz: Bärbel Kröger, Göttingen Druck und Verarbeitung: Memminger MedienCentrum AG Printed in Germany ISSN 0067-8236 eISSN 2749-9162 ISBN 978-3-447-11924-5 eISBN 978-3-447-39335-5 Inhalt Claudia Fabian Vorwort ......................................................................................................................................... 1 Carolin Schreiber Handschriftenerschließung im digitalen Zeitalter: Stand, Herausforderungen und Perspektiven ........................................................................................................................ 5 Robert Giel Halb durch? Das DFG-Projekt »Handschriftenportal« zu Beginn seiner zweiten Projektphase ................................................................................ 27 Nathanael Busch, Diana Müller Normdaten als Forschungsaufgabe. Der »Handschriftencensus« und die Normdaten für deutschsprachige Werke des Mittelalters ............................................... 39 Christoph Rauch Alte Kataloge in neuem Gewand: Das DFG-Projekt »Orient-Digital« und der Verbundkatalog orientalischer Handschriften »Qalamos« .................................................. 53 Patrick Andrist Asymmetrical Descriptions of Biblical Manuscripts: A key to the success of the »Paratexts of the Bible (ParaTexBib)« project and its database ................................... 63 Monika Studer Unter neuen Vorzeichen. Vom Verbund »Handschriften, Archive, Nachlässe (HAN)« zum Metakatalog »swisscollections« .................................................. 79 François Bougard The »Institut de recherche d’histoire des textes (IRHT)« and libraries. From microfilm to digitisation, from paper cards to authority files ............................. 95 Anne-Marie Turcan-Verkerk Ancient written cultures and digital technology: »Biblissima(+)«. Preparing tomorrow’s research and providing access to the written heritage ........... 103 Gottfried Heinz-Kronberger Die Musikhandschriftenerschließung des »Répertoire International des Sources Musicales« an der Bayerischen Staatsbibliothek ................................................. 121 Erwin Frauenknecht Das »Wasserzeichen-Informationssystem WZIS«. Vom digitalen Format zur Forschungsinfrastruktur .......................................................... 135 VI Inhalt Philipp Lenz Handschriftenkatalogisierung als methodischer Impuls für die Mediävistik ............. 147 Patrizia Carmassi Between human and digital networks. Neue Erkenntnisse und Überlegungen bei der Katalogisierung der mittelalterlichen Handschriften aus der SUB Göttingen ... 163 Edina Zsupán, Katharina Kaska, Maria Theisen Gemeinsam zum Ziel – zur Erschließung der Handschriften aus der Bibliothek des Matthias Corvinus in der Österreichischen Nationalbibliothek ..................................... 185 Caroline Zöhl Buchschmuck in Inkunabeln – Perspektiven einer kooperativen Exemplarerschließung .............................................................................................................. 203 Wolfgang-Valentin Ikas Die Ordnung des Paradieses – Versuch einer tektonischen Analyse des Handschriftenbestands der Bayerischen Staatsbibliothek ................................................ 237 Sven Limbeck Neuzeitliche Handschriften. Konzeptionelle Überlegungen zu ihrer Erschließung ............................................................................................................... 253 Thomas Haffner Erfahrungen bei der Erschließung neuzeitlicher italienischsprachiger Handschriften in der SLUB Dresden ................................................................................... 271 Peter Sjökvist »Resource Description and Access (RDA)« – Transcription and the Dilemma of a Rare Materials Cataloger ................................................................................................. 279 Hans Brandhorst Meaningful patterns. Can a mix of pattern matching and computer-assisted classification produce a paradigm shift in iconography? .................................................. 287 Günter Mühlberger Text- und Layouterkennung mit »Transkribus« ................................................................ 307 Abstracts ....................................................................................................................................... 323 Neuerwerbungen der Bibliotheken 2021 ............................................................................. 333 Verzeichnis der Autorinnen und Autoren ............................................................................ 367 Vorwort Die in diesem Band zusammengestellten Beiträge spiegeln die Vielfalt und die Dynamik der Handschriftenerschließung im Jahr 2022, mit einem Schwerpunkt auf Buch-Handschriften. Sie illustrieren und reflektieren verschiedene Facetten, Fragestellungen und Interaktionen dieses traditionell schon immer, heute aber besonders betonten, gemeinsamen Anliegens von Bibliothek und Wissenschaft. Dabei wird auch erkenntlich, wie die Erschließung von Handschriften immer »nur« Mosaiksteine, einzelne Aspekte eines weit größeren »Wissensraums« oder »semantischen Universums«, in den Blick nehmen kann. Die Katalogisierung beziehungsweise Erschließung von (Buch-)Handschriften ist immer an entsprechende Bestände und spezialisierte wissenschaftliche Kenntnisse gebunden. Sie bewegte sich in vielen Aspekten außerhalb der bibliothekarischen Verfahren, Normen und Standards. Wenn nun in der Reihe Bibliothek und Wissenschaft ein ganzer Band der Handschriftenerschließung gewidmet wird, so zeigt schon dies, wie sehr sich die Lage verändert hat. Die Handschriften sind heute – im digitalen Zeitalter mit seinen Möglichkeiten und Anforderungen – in ganz anderer Form in die bibliographische Erschließung integriert und selbstverständlicher Teil des bibliographischen Universums. Der Weg dahin ist natürlich bedarfs- und nachfrageorientiert. Er führte – wie bei anderen Überlieferungsträgern auch – über eine Vielzahl von Definitionen, Vereinheitlichungen, Annäherungen und Normierungen, wurde gefördert durch die internationale Vereinheitlichung und vor allem die stets verbesserten technischen Möglichkeiten. Sie erlauben, auf der Ebene der Daten zusammenzuwirken und zu interagieren. Es ist kein Zufall, dass der Standardisierungsausschuss eine Arbeitsgruppe »RDA (Resource Description and Access) und Handschriften« schon 2015 ins Leben gerufen hat. Diese konnte ein Kerndatenset definieren, das die zentralen Elemente der Handschriftenbeschreibung benennt. Langfristig nachhaltig aber geht es darum, die Handschriftenerschließung in den Terminologien und Standards von RDA, die sich ebenfalls sukzessive weiterentwickeln, darzustellen und zu verankern. Dies ist wichtig und zukunftsweisend, denn die Handschriftenerschließung funktioniert nicht mehr in dem geschlossenen Kontext der Publikation von einzelnen Handschriftenkatalogen in gedruckter Form, die in die Regale der Bibliotheken eingestellt und bei Bedarf konsultiert werden. Die Handschriftenerschließung und die Handschriftenforschung spielen sich heute in dem großen, von Daten bestimmten semantischen Universum ab. Es gibt verschiedene Datenbanken, Portale und Forschungsunternehmen. Es gilt, die Daten hier so zu integrieren, dass sie ihre Aussagekraft behalten, ja verstärken. Dabei ist es ein Vorteil, auf die Errungenschaften der bibliographischen Erschließung und ihre Standards zurückgreifen zu können, gleichzeitig aber eine Aufgabe, diese nicht nur anzuwenden, sondern so weiterzuentwickeln und neu zu definieren, dass die Besonderheiten der Handschriftenerschließung hier ihren Platz finden. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 2 Vorwort Die Überlieferungstradition ist nicht nur bestimmt durch den Medienwandel um 1450 von der Handschrift zum Druck und heute vom Analogen zum Digitalen. Sie kann auch in einer neuen, übergreifenden Einheit gesehen werden, die der digitale Raum verspricht und eröffnet, denn in vielem geht nicht länger mehr um Handschriften oder Drucke, es geht um Besonderes, Eigenes aber auch Verbindendes, um Daten und ihre Interoperabilität, aus denen sich immer mehr und immer neue Erkenntnisse für die Wissenschaft und das Verständnis unserer Welt ableiten lassen. Die hier versammelten Beiträge können und wollen nicht den Anspruch erheben, das Thema Handschriftenerschließung in Bibliothek und Wissenschaft im Jahr 2022 vollständig oder gezielt systematisch in den Blick genommen zu haben. Das zeigt sich bereits in der Konzentration auf Buch-Handschriften, aber auch in der schwierig zu gestaltenden Abfolge der Beiträge im Inhaltsverzeichnis. Zwar wurde mit Beiträgen zu den großen Nachweisinstrumenten der Bibliotheken und der Wissenschaft begonnen, dann folgen in loserer Gruppierung Beiträge zu einzelnen Erfahrungen und Reflexionen, zunächst mittelalterlicher, dann neuzeitlicher Handschriften, Überlegungen zur kunsthistorischen Erschließung bis hin zu Verfahren der Texterkennung, zu Herausforderungen und Chancen von RDA (Resource Description and Access) für eine bei Handschriften immer auf das Objekt, d.h. das Exemplar bezogene Erschließung. Schon bei der Lektüre des Inhaltsverzeichnisses erkennt man nicht nur, was alles noch hätte gesagt werden können, sondern auch, warum Zugang und Ordnung dieses Universums vor ganz eigene Herausforderungen stellt. Man erahnt, dass Handschriften weder in der Bibliothek noch in der Wissenschaft weder in einem noch in mehreren Erschließungs- und Digitalisierungsprojekten abschließend bearbeitet werden können. Man sieht, wie viele Wege und Verfahren sich um das eine Objekt bemühen. Die Handschrift als Kulturobjekt hat eine Tiefendimension, die sich in den Jahrhunderten, ja Jahrtausenden des Erhalts und Erschließens bewahrt und bewährt hat, und die nun in die digitale Welt der Daten umgesetzt wird. Dabei kann und wird etwas gelingen, was die Welt der Daten heute besser ermöglichen kann, und wofür so viel bibliothekarische und wissenschaftliche Arbeit bereits im analogen Zeitalter tragfähige Fundamente gelegt hat: die Vernetzung, die Interoperabilität von Informationen (= Daten), mithin der Aufbau einer nicht länger zwischen Buchdeckeln gefassten, geschlossenen Erkenntnis, sondern eines offenen und weiten, auf qualitätsvollen Aussagen basierenden Wissensnetzes. Der Blick auf die neu entstehenden Portale macht die Potenziale deutlich, aber auch die Handlungsbedarfe. Nicht nur große Konzepte sind heute nötig, sondern vor allem nach wie vor viele Einzelschritte, die umzusetzen und in geeigneter Form zu integrieren sind. In allen Beiträgen werden verschiedene Öffnungsperspektiven erkennbar, alle Beiträge zeigen, dass Handschriftenerschließung heute nicht nur Vergangenheitsbewältigung, sondern vor allem Zukunftsorientierung bedeutet. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 3 Vorwort Ich danke ganz herzlich allen Autorinnen und Autoren aus Bibliothek und Wissenschaft, aus dem In- und Ausland, oft langjährigen Weggefährten, aber auch spontan zum Mitwirken Bereiten, die zu diesem Band ihre – wie ich finde – lesenswerten Beiträge verfasst haben. Ich danke dem Verlag Harrassowitz für die erprobt gute Betreuung und erwähne hier eigens den erstmals für »Bibliothek und Wissenschaft« realisierten digitalen Druck. München, August 2022 Claudia Fabian © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. Patrick Andrist Asymmetrical Descriptions of Biblical Manuscripts: A key to the success of the »Paratexts of the Bible (ParaTexBib)« project and its database* The creation of a scholarly database about the paratexts of biblical manuscripts in less than five years was one of the challenges I faced when Martin Wallraff invited me in 2012 to help him prepare the ERC project ParaTexBib (Paratexts of the Bible: Analysis and Edition of the Greek Textual Transmission)1. How could our future team analyse the paratexts in hundreds – potentially thousands – of manuscripts in a systematic way and present the results effectively both for our needs and those of a broader audience? From the very beginning – even as we were writing the proposal – we realized the scope of the project was ambitious and that we would need a cataloguing methodology which would allow us to systematically target the parts of the manuscripts which were relevant to the project without losing sight of the broader manuscript context. We also wanted to summarize these (often very detailed) findings and present them in a way that was both accessible and uncluttered. The process of decision making, trial and error and course correction to adequately address emerging challenges and complexities was a constant part of our work. The resulting methodology is one of the most valuable results of the ParaTexBib project as a whole. It is my hope that in explaining our methodology and why it worked for us, it may inspire reflection in, and perhaps save time for, other projects which are facing the same or similar strategical questions. Initial Questions and Options The ParaTexBib project can be seen as a first step towards an overarching goal of systematically exploring, describing and publishing the paratexts in biblical manuscripts, i. e. (in summary) all the pieces of content – viewed in a broad way, including images etc. – except * I warmly thank Saskia Dirkse (who was part of the project since it started in 2015), who carefully corrected my English and suggested some important improvements. 1 Andrist, Patrick/Wallraff, Martin: ParaTexBib: an ERC Project Dedicated to Paratexts in Greek Manuscripts of the Bible. In: COMST Bulletin 2 (2016), pp. 63–68. Wallraff, Martin/Andrist, Patrick: Paratexts of the Bible: A New Research Project on Greek Textual Transmission. In: Early Christianity 6 (2015), pp. 237–243. – The project began at the University of Basel in January 2013 and ended at the University of Munich in September 2020, after a move in January 2017. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 64 Patrick Andrist the biblical text itself2. For reasons given below, ParaTexBib concentrated on the Gospels. A few years later, a sister project was launched for the paratexts of the Book of Revelation, in collaboration with Martin Karrer at the Kirchliche Hochschule Wuppertal/Bethel from 2019 to 20213. Since the conception of the initial project, it was clear to us that the information we were to collect about the paratexts and the manuscripts should be stored in a database format. This would allow us not only to retrieve it when necessary and link it with other online resources (such as images of the manuscripts) but also to share it with a broader audience. One of our first implicit decisions was to aim to produce a complete description of each manuscript selected for the project, and to stay away from the practice of certain specialised or thematic catalogues, which only give information directly related to the catalogue’s main topic, isolating it from its material and textual context. For example, it would not make sense for us to gather data on biblical textual paratexts only and totally neglect their biblical context or the other types of paratexts around them such as miniatures. On the other hand, it would not make sense either to describe the latter in detail, since the decision was made early on to concentrate on textual paratexts4. Like many other projects dealing with manuscripts and manuscript descriptions, we were well aware that our limited time and funding would not allow us to create full descriptions of all the manuscripts which would be of interest to the project, including their content, all their material features and their history. Even with a team of five to eight people working on a five-year project, it is not feasible, through budget and time constraints but also for practical reasons, to view a significant number of biblical codices in their repositories, especially since gathering manuscript data was not the main goal of the project but one 2 For a more technical definition and discussion, see, as an entry point to a quickly growing bibliography, Andrist, Patrick: Toward a Definition of Paratexts and Paratextuality: the Case of Ancient Greek manuscripts. In: Bible as Notepad. Tracing Annotations and Annotation Practices, ed. by Liv Ingeborg Lied/Marilena Maniaci. Manuscripta Biblica 3. Berlin 2018, pp. 130–149; Idem: The Limits of Paratextuality: Old and New Questions. To be published in COMST Bulletin 8 (2022). See also Andrist, Patrick/Maniaci, Marilena: The Codex’s Contents: Attempt at a Codicological Approach. In: Exploring Written Artefacts: Objects, Methods, and Concepts, ed. by Jörg B. Quenzer. Studies in Manuscript Cultures 25, 2 vol. Berlin, Boston 2021, vol. 1, pp. 369–394. 3 This project was titled »Die Paratexte zur neutestamentlichen Johannesapokalypse in griechischen Handschriften« and ran from 2019 to 2021. The three co-Principal Investigators of the project were Martin Karrer, Martin Wallraff and the author. We gratefully acknowledge here the support of the Fritz Thyssen Stiftung, which financed the project. See Müller, Darius/Malik, Peter: Rediscovering Paratexts in the Manuscripts of Revelation. In: Early Christianity 11 (2020), pp. 247–264. 4 We want to mention, however, that in the framework of our project Georgi Parpulov studied painted paratexts and the resulting monograph, Repertorium of Evangelist Portraits Miniatures from the Mid-Ninth to Mid-Thirteenth Centuries, has just appeared (Parpulov, Georgi: Middle-Byzantine Evangelist Portraits: A Corpus of Miniature Paintings. Manuscripta Biblica 7. Berlin 2020). © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 65 of the means towards accomplishing it. This is why, from the beginning, we set some limitations as far as the working method and the resulting descriptions were concerned. For the working methods, we decided: – to essentially base our work on images of the manuscripts; only in exceptional cases would an in-person viewing of a manuscript be planned. This decision was also facilitated by the fact that a large number of digital reproductions of biblical codices are available online not only on many websites of the repositories themselves but also in two specialized databases: – the New Testament Virtual Manuscript Room (NTVMR) a website dedicated to the study of Greek New Testament manuscripts from a text-critical perspective, established and supported by the Institut für Neutestamentliche Textforschung in Münster/Westfalen (‹https://ntvmr.uni-muenster.de/›, last accessed 30.05.2022); – the database developed by the Center for the Study of New Testament Manuscripts (CSNTM), in Plano, Texas, under the umbrella of The Center for the Research of Early Christian Documents (CRECD) (‹https://www.csntm.org/›, last accessed 30.05.2022); – in these conditions, it would be difficult to describe the physical features of the manuscripts systematically. Since paratexts are at the heart of the project, we decided to focus on the description of manuscript content. Only the most important and basic physical features would be included. At this early stage, we did not spend much time developing a detailed description model (understood as the conscious choice to describe a series of features in a certain way and how the overall description is organised in the output document or screen in order to produce a series of coherent scholarly online or printed descriptions). But we did have a clearly defined idea of the criteria we wanted our descriptions to meet. In particular, we outlined the following principles: – fundamentally, the manuscripts to be observed and described are the manuscripts in their present state. Reconstruction of the previous states of the manuscripts is important, and particularly their original state, but this comes at a later stage of our analysis; – even if the descriptions could not answer to the best practices of manuscript cataloguing for the reason given above5, the gathered data should not only be congruent with the reality of the described features, this data must also be gathered as coherently and systematically as possible and then presented in the framework of systematic and coherent electronic manuscript descriptions. This allows for a more relevant appreciation of the corpus as a whole, for meaningful comparisons and more precise search possibilities; 5 About the quality of an ideal catalogue, see the still relevant remarks by Canart, Paul: Consigli fraterni a giovani catalogatori di libri manoscritti. In: Gazette du livre médiéval 50 (2007), pp. 1–13, in particular 8–10. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 66 Patrick Andrist – the descriptions should show the historical layers (that is to say, the production units) of the manuscripts6, in order to help reconstruct their genetic history as accurately as possible (and to avoid chronological errors in the analysis). Among the several ways to reach this goal, we wanted the descriptions to respect the »syntactical model«; this refers to a kind of description model based on the historical layers of the manuscripts, which makes them more visible and facilitates a historical assessment of the manuscript. According to the syntactical model, each main production unit is described fully and in order, within a wider description of the codex7. We believe that the study and description of manuscripts should always be a scholarly activity, even if the gathered data is not going to be published in a traditional catalogue format. As a result, data »published« in an electronic format should receive the same rigorous and scholarly treatment as data published in more traditional cataloguing formats, no matter how easily electronic data can be modified. As such, the database should maintain high scholarly standards at the levels of the data model, the description model and the data itself. As a corollary, the people who are responsible for analysing manuscripts and preparing the descriptions should be recognised as scholarly authors, and their name be mentioned whenever their work is being used. Getting Ready – The Structure of the Data and the Description Model The time between the approval and the start of the project was crucial. Our goal was to set up the main inputting functions of the database before new team members were brought on board, so that the description data could be gathered into the database from the first day of work on the manuscripts. After a period of intensive thinking and discussions, we concluded that it was in our best interest to avoid creating a new database, providing we could find another good solution. At this time, »NO NEW DATABASE« was our motto. After a series of exchanges with 6 A production unit is defined as »all the codices or parts of codices which are the result of one and the same act of production«. For example, if two scribes join efforts to copy a series of texts and produce a book, this is only one production unit, including the original binding, since it is the result of a simple »book project«. But if, some years later, someone adds a new text to the first ensemble, or annotates it, or repairs the binding, etc. this is a new production unit each time, which represent a new layer in the genetic history of the manuscript. See Andrist, Patrick/Canart, Paul/Maniaci, Marilena: La syntaxe du codex. Essai de codicologie structurale. Bibliologia 34. Turnhout 2013, pp. 59–60 (§ 2.2.1 in the upcoming second edition, in English; see also § 3.1 about the typology of the production units). 7 See Andrist, Patrick: Syntactical description of manuscripts. In: Comparative Oriental Manuscript Studies. An Introduction, ed. by Alessandro Bausi /Marilena Maniaci /Caroline Macé et al. Hamburg 2015, pp. 511–520. See also Andrist/Canart/Maniaci: La syntaxe du codex (cf. footnote 6), pp. 135–169, including a cataloguing example. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 67 various projects and institutions, we saw the highest degree of »philosophical« convergence with the Pinakes database, developed and maintained by the Section grecque de l’IRHT. This MySQL database was conceived to gather information about the content of every preserved Greek manuscript on earth (excluding most papyri and archive artefacts). Today it contains more than 73,000 manuscript shelfmarks, with a very variable amount of information, ranging from only the shelfmark number to a detailed description of the texts in the manuscript. The people of the Section grecque diligently keep the database up to date by systematically inputting information from catalogues and secondary literature, but they are generally not in a position to check or complete this information by consulting digital images of the manuscript (except in exceptional circumstances)8. In fact, we recognised a high degree of convergence both between the Pinakes data model and our goals, and with the people in charge of the database, as far as the approach to manuscript description in a database and the potential for collaboration are concerned. A few months later, this led to an agreement signed between our respective institutions9. As a result, the data which make up ParaTexBib manuscript descriptions are a subset of Pinakes (see Diagram 1). It can be conceptualised as a database dedicated to the paratexts Manuscripta biblica (ePTB specialised public interface). Pinakes’ general public interface Database (Pinakes) server in Orléans server in Munich Common Back End (IRHT+ ePTB) Diagram 1. The ePTB database and its relation to Pinakes. 8 See Binggeli, André/Cassin, Matthieu: Recenser la tradition manuscrite des textes grecs du Greek Index Project à Pinakes. In: La descrizione dei manoscritti: esperienze a confronto, ed. by Edoardo Crisci/Marilena Maniaci and Pasquale Orsini. Studi e ricerche del Dipartimento di Filologia e Storia 1, Cassino: Dipartimento di Filologia e Storia, Università degli Studi di Cassino 2010, pp. 91–106. In summary, Pinakes (public interface = ‹https://pinakes.irht.cnrs.fr/›, last accessed 30.05.2022). 9 We are very grateful to the IRHT, in particular to André Binggeli and Matthieu Cassin, for having opened the doors of Pinakes and embarking on this adventure with us. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 68 Patrick Andrist of the Greek Bible, which we nicknamed ePTB. Technically, both the normal Pinakes and the ePTB data are entered into the same tables with the same working environment. However the ePTB data can be visualised not only on the Pinakes website with all the other Greek manuscripts in a standard Pinakes format, but also on Manuscripta Biblica, a website dedicated to biblical manuscripts10. One consequence of hosting our data in Pinakes is that we were sharing the same data model, which, to a large extent also determined our description model11. The people in charge of Pinakes kindly agreed, however, to have the data model evolve, so that we could add new fields and develop new tools and functions, in order to address specific needs and reach our goals in a more efficient way. Among these, it is worth mentioning the tool that allows team members to mark data in NTVMR, including folio and content information, and import it into ePTB. Here it is inserted at the correct place with a link to its source; one can then click on this link from ePTB to immediately see the image of the corresponding page12. Additionally, thanks to a proof-of-concept grant from the ERC, we also developed another electronic tool called StruViMan, which allows to visualize the structure of the manuscripts13. As already mentioned, Pinakes is not intended for the full description of manuscripts, but for a thorough description of their contents. As visualised in Diagram 2, the data for a single manuscript and the corresponding online description are structured in three (main) hierarchical levels, which we could fit into our stratigraphic approach. These three levels are: 1. ›Cotes‹ (= ›Shelfmarks‹): this is the entry point for any manuscript description. It corresponds to the shelfmark or inventory number of the object in its current repository14, and it is uniquely identified by a »diktyon« number15. At this level, only general information 10 ‹https://pinakes.irht.cnrs.fr/›; ‹https://www.manuscripta-biblica.org/›, both accessed 30.05.2022. 11 About the link between the data model and the description models in most databases, and the need to disconnect them in order to develop a new generation manuscript databases, see Andrist, Patrick: Toward a New Generation of Databases and Database Applications for Describing Ancient Manuscripts. In: Digital Scholarship in the Humanities 36 (2021), pp. 9–16 (‹https://doi.org/10.1093/ llc/fqab068›, last accessed 30.05.2022), cf. § 2. 12 Through an API written by Troy Griffith in 2015. 13 Dirkse, Saskia/Andrist, Patrick/Wallraff, Martin: Structural Visualization of Manuscripts (StruViMan): Principles, Methods, Prospects. In: Open Theology 6 (2019), pp. 249–258. It was programmed by Caroline Strolz and Frank Percival, and is now available on both public interfaces. 14 With some exceptions. For example, the bifolios of codex Vat. gr. 1209 (diktyon 67840), which contains the celebrated Codex Vaticanus (GA 03 for the NT, see below) are now carefully kept separately in the Vatican Library, but they are described in ePTB as they used to be before the codex was dismounted at the end of the 19th century. 15 Binggeli, André/Cassin, Matthieu: Le projet Diktyon. In: Greek Manuscript Cataloguing: Past, Present, and Future, ed. by Paola Degni/Paolo Eleuteri/Marilena Maniaci. Bibliologia 48. Turnhout 2018, pp. 201–206. It will eventually be replaced with ISMI numbers, see Bougard, François/Cassin, Matthieu/Duba, William/Fabian, Claudia/Flüeler, Christoph/Turcan-Verkerk, Anne-Marie: Inter- © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 69 The »Paratexts of the Bible (ParaTexBib)« project and its database external resources Pinakes structure of the data describing a single manuscript (simplified) Shelfmark Production Units (Codicological units) shelf number pictures CU 2 CU 1 etc. (physicality) (scribe 1) (scribe 2) catalogues etc. Pieces of content (Témoins / Witnesses) bibliography content 1 content 2 etc. etc. detail 1 detail 2 etc. Diagram 2: The three hierarchical levels of Pinakes data structure about a manuscript. on the manuscript, links to related manuscripts and external online resources (identifiers) are given. One might be surprised that the shelfmark or inventory number is used as the entry point for the descriptions, rather than the Gregory Aland number (henceforth GA)16, which are commonly used by New Testament scholars to designate the New Testament content of a manuscript as it was originally produced. As is well known, there is not a one-to-one relation between GA numbers and shelfmarks/diktyons, which encompass the whole manuscript in its current state. The basic diverging situations are: – when a current shelfmark contains sections of the New Testament produced in different places and at different times, the individual sections are assigned different GA numbers, because they are different witnesses to the NT. On the contrary, working with the complete manuscript including its several historical layers allows us to perceive national Standard Manuscript Identifier (ISMI). In: DigItalia 15 (2020), pp. 45 52. ISMI are more universal and stable than Diktyon numbers, because they are not limited to Greek manuscripts and do not change if the shelfmark changes, for example if the manuscript goes into a new collection. 16 GA numbers, maintained in the NTVMR database. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 70 Patrick Andrist and study its paratexts, including their possible changes, within the manuscript over time, in a diachronic approach17; – when a shelfmark contains more than New Testament material but from the same production, one can better understand the New Testament paratexts in the broader context of all the paratexts of the manuscript18; – when paratexts circulate without the biblical text that they were previously connected to (usually due to a manuscript suffering mutilation) and thus don’t have a GA number, it is difficult to use GA numbers as an entry point19; – in contrast, when remains of the original New Testament content of an ancient manuscript are spread over several libraries, where each fragment has its own shelfmark/ diktyon number, GA numbers play a unifying role20; in this case we describe each diktyon but then study the paratexts as a whole21. As we see, GA numbers are an excellent tool for identifying manuscripts as text witnesses, but not manuscripts as material objects or overall containers22. In accordance with a more general Pinakes practice, and due to the importance of these numbers for research in general (and for visibility), a link has been systemically created at the shelfmark level. Furthermore, in the Manuscripta Biblica interface, GA numbers can also be used to quickly access a description. Each shelfmark can be then described through one or several »Unités codicologiques«: 17 For example, in codex Vat. gr. 1209 (diktyon 67840 see above), one finds the Vaticanus (GA 03 for the NT, see above) as well as its 15th c. restoration (GA 1957 for the NT). 18 For example pandects such as the Vaticanus or the Sinaiticus also contain the Old Testament. Furthermore, in the latter the NT is designated by GA 01 (= »‫)«א‬, while the OT is designated with »S« by the OT scholars; see also below footnote 21. 19 Such as the Canon Tables in the Venetian part of Codex Basilianus (diktyon 69472). Thanks to an initiative of Martin Wallraff in the framework of the ParaTexBib project, manuscripts in this kind of situations should also receive a GA number in the near future. 20 For example, parts of the Tetraevangelium Purpureus Petropolitanus (GA 022, = »N«) are held in various places under nine different shelfmarks today. The main part is in St. Petersburg at the Rossijskaja Nacional’naja biblioteka (RNB), Ф. № 906 (Gr.) 537 (Granstrem 18; diktyon 57609). 21 Codex Sinaiticus is an example of a manuscript whose parts are held in various repositories and which contains non-NT material; about its paratexts, see Andrist, Patrick: Au croisement des contenus et de la matière: l’architecture des sept pandectes bibliques grecques du premier millénaire. Étude comparative sur les structures des contenus et de la matérialité des codex Vaticanus, Sinaiticus, Alexandrinus, Ephraemi rescriptus, Basilianus, »Pariathonensis« et de la Biblia Leonis. In: Scrineum 17 (2020), pp. 5–106 (‹http://dx.doi.org/10.13128/scrineum-11466›, last accessed 30.05.2022), see pp. 24–25. 22 Using GA numbers to conveniently designate the whole manuscript is not a problem as long it is entirely a NT manuscript resulting from a single production. In ePTB, GA numbers in multiple GA manuscripts are best described at the next hierarchical level (CU level, see below). © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 71 2. »Unités codicologiques« (= »Codicological Units«23, = »CUs«): this level is meant for the description of a physical part of the manuscript. In ePTB they correspond to the main production units of the manuscripts. Most importantly, this is the level where the date and the place of a manuscript’s production are given. For example, the codex München, Bayerische Staatsbibliothek (BSB) Cod.graec. 59424 has five CUs, in the following order: 3 CU describing restorations mainly at the beginning of the codex, named »1.R«, »2.R« and »3.R«; then the main part, from the tenth century (»4.A«); then a CU for the subscriptions, which were added at a later time (»5.Q«). From the CU level, one can enter information on: – a limited number of the physical features of the CU: its book form, material support, main scripts and main layout; – copyists, and owners if the CU circulated independently in the past; – the content of the CUs, through a description of each piece they contain, at the next hierarchical level: 3. ›Témoins‹ (= ›Witnesses‹): each of the manuscript’s pieces of content can be described separately. The main fields concern the location of the content in the CU, and its identification through a link to an »oeuvre«(»work«)25, described in another area of the database. A feature called ›details-contenus‹ (= ›details about the content‹), allows for specific details about a piece of content to be recorded such as the transcription of its initial title, incipit, or to attach it to one or more predefined markers. To continue with codex BSB Cod.graec. 594 as an example: the folios 158r–259r are described through a clickable link to the oeuvre »Testamentum nouum, Euangelium sec. Lucam«; two markers are attached to it, telling users that there is an initial title but no final one. 23 About Codicological Units, see Gumbert, J. Peter: Codicological Units: Towards a Terminology for the Stratigraphy of the Non-Homogeneous Codex. In: Il codice miscellaneo, tipologie e funzioni. Atti del convegno internazionale (Cassino, 14–17 maggio 2003), ed. by Edoardo Crisci and Oronzo Pecere. Segno e testo 2. Cassino 2004, pp. 17–42. See also Andrist/Canart/Maniaci: La syntaxe du codex (cf. footnote 6), pp. 41–44. Unfortunately, this concept is used with different understandings in the scholarly literature, leading to some confusion. 24 Diktyon 45044; GA 652; description by Agnès Lorrain available in Manuscripta Biblica and in Pinakes (accessed in May 2022). 25 This distinction between what the author(s) created and what is found in the manuscript is fundamental to all the disciplines working with ancient texts, even if the scholars express it using different terms, sometimes with strongly diverging understanding of what »work« should mean. See for example Eggert, Paul: Securing the past. Conservation in art, architecture and literature. Cambridge 2009, pp. 234–236; Robinson, Peter: Towards a Theory of Digital Editions. In: Variants 10 (2013), pp. 105–131, see 118–121; Gabler, Hans Walter: Text Genetics in Literary Modernism and Other Essays. Cambridge 2018, chap. 5 »Editing Text-Editing Work«, pp. 111–119, see also 171–172. See also Andrist/Maniaci: The Codex’s Contents (cf. footnote 2), p. 372. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 72 Patrick Andrist It quickly becomes clear that in this model, all the textual and physical features of the codex are attached to one CU entry. The CU level plays a crucial role in the organization of the description. The many possibilities and limitations given by a specific data model should not be confused with the way these possibilities are used in an electronic description project and so we also reflected on how to organise the data and describe each feature. As mentioned, one key decisions was that in standard ePTB descriptions, CUs should correspond to the main production units of the codex26, with the exception of small added elements or notes, that are either not described or described within their hosting CU through a special marker. We tried to maintain high standards while also addressing practical matters and special cases as we encountered them during the course of the project. After six years of work, the result was both a document counting 137 pages … and a fairly coherent set of data in ePTB descriptions. The Encounter of Best-practices Intentions with Sub-optimal Realities In January 2015 the improved internal interface was ready for the new team members to be able to work with it as soon as they joined us in Basel27. Simplifying the Rules Our original intention was to prepare a full description of the content of each New Testament manuscript according to our rules and to use electronic tools to optimize the time needed to prepare a description. It soon became clear, however, that our goal was too ambitious for the project’s timeline. Certain tasks were found to be more time consuming than we had thought, in particular: – the task of marking entries in NTVMR and importing them in such a way that the links to the specific images are preserved in Pinakes takes time. This makes it impossible to set up a tool which would create all the usual basic witness entries (including the common paratexts) in a Gospel manuscript at once when beginning work on a new codex, as we had originally planned; – transcribing Greek text with accents and punctuation marks as they occur in the manuscripts resulted in interesting but very time-consuming discussions; 26 In agreement with our initial goals, the ePTB description model is really a simplified Type A syntactical description model; see Andrist: Syntactical Descriptions of Manuscripts (cf. footnote 7), p. 516. 27 I would like to mention and honour the memory of our programmer Jenny Goude, whose dedication and hard work helped us complete this challenge and who unexpectedly and tragically passed away on March 28, 2016. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 73 – even for »standard« Gospel books, searching for secondary literature turned out to be more useful but also more demanding than foreseen; – our initial hope to also describe a fair number of Praxapostoloi28 was not realistic, since they contain many more pieces of contents than Gospel books29. We found that if we wanted to maintain a good level of coherence in the database, we had to narrow the objectives and lower certain standards. After a period of tests, we made a number of decisions in two phases. The first phase already happened in the first year of the project. In summary: First we excluded certain types of manuscripts, such as lectionaries and commentary manuscripts; codices where the biblical text is »framed« by commentaries would still be surveyed but not described in detail; further Praxapostoloi would be passed over altogether. Secondly we simplified the rules for the standard descriptions, in order to limit to a minimum the time-consuming inputting of information that is not central to the study of the paratexts in the project. In particular, we decided to describe the content of the non-Gospel sections of the manuscripts in a compact way, in imitation of what was done in the printed catalogue of Greek Judaica at the Vatican Library30: non-Gospel sections were described with generic information31. Production units which do not contain Gospel material were sometimes described only by specifying their location in the manuscript and a descriptive title32. Most of the post-production paratexts (such as marginal notes) were excluded, unless they contain important information for the history of the codex (for example, an owner or a place name). This decision allowed us to keep on describing all the contents of all the parts of the manuscript, but with a different level of details. We also decided, from that point on, to produce transcriptions without accents. In the following months, we developed several research tools in the internal interface, in order to more easily filter in manuscripts and witnesses matching several specific contents and/or other characteristics. 28 The Praxapostolos is a traditional byzantine biblical book containing the Acts and the Epistles; the Revelation is sometimes also copied with them. 29 At the end of the project, only seven Praxapostoloi without Gospels or the Revelation were fully described according to our rules. 30 Andrist, Patrick: Les codex grecs adversus iudaeos conservés à la Bibliothèque vaticane (s. xi–xvi). Essai méthodologique pour une étude des livres manuscrits thématiques. Studi e Testi 502. Città del Vaticano 2016, see the methodological explanations on pp. 119–123. 31 For example, the OT section could be described at once as »Septuaginta, Opera omnia«. Similarly, following the practice of the IRHT, the Praxpostolos part of these production units were covered by the three following generic Oeuvres: »Actus Apostolorum« + »Epistulae catholicae« + »Epistulae Pauli« + if necessary, »Tabulae liturgicae«. In all the cases, the generic entries implicitly include both the biblical text and their paratexts. When necessary, additional information are given as remarks. 32 No further information would be given at the witness level (see above) for non-Gospel production units; for the project codified ways to give titles, see below footnote 35. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 74 Patrick Andrist Fast-Tracking After the second year of the project we made another series of decisions. At that time we had both a clear idea of the paratexts we wanted to study and the information we wanted to gather in the database. Furthermore, based on our experience, we also had a good idea of the time it took to create a description. One of the major decisions was methodological: since we were not going to be able to describe all the Gospel manuscripts in detail, we needed to concentrate on those which contained important paratexts for the project. In order to identify them, we decided to scan through all Gospel manuscripts from the beginning of the tenth to the end of the 15th century in a short period of time (internally we named this process »fast-tracking«)33, where only the beginning and the end of the manuscript, as well as the beginning and the end of each Gospel were checked for relevant paratexts34. Thanks to a list of paratexts to look for and their main characteristics, it was possible to obtain homogenous results, even though the scans were done by all the project members. Every time a manuscript was scanned through, three pieces of information were added to the corresponding database entry: – a very short public description limited to a descriptive production unit title35; – an internal report in which the location of important paratexts is specified. This report, done in a structured way with a very idiosyncratic tagging system, allowed each »scanner« to pinpoint the presence of a paratext in a manuscript with a minimum of effort, leaving to the future editor the responsibility to analyse it and refine the identification; – the modification of the manuscript status, either as a manuscript to be further described, or a manuscript which will remain in fast-tracked status – i. e. which is not going to be described with more details within the ParaTexBib project. Secondly, once this action was completed, only the manuscripts marked »to be described« were dealt with. We also prioritized the descriptions according to the date of the manuscripts. The idea was to keep a broad time span (i. e. to the end of the 15th c.), but to give 33 Manuscripts older than the tenth century were not fast-tracked because we decided to describe all of them in detail. 34 In practice, each manuscript was scanned through by two different people in order to reduce the risk of making mistakes. 35 Since the beginning of the project, we paid a lot of attention to both shelfmark and production unit titles, which can be seen as a series of searchable keywords, giving information both about the main content and the paratexts. For example, the title of the first production unit of codex Athêna, Mouseio Benaki, T. A., 320 (GA 2658, diktyon 8430), reads »Tetraeuangelium cum tabulis liturgicis, canonibus (sola epist.), capitulis, prologis, subscr., imag., epigr.«, where each word clearly designates a distinct piece (or set) of content(s). © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 75 more weight to the older manuscripts36. Unfortunately, in spite of our best efforts, many manuscripts in the database which currently only have a very short descriptions, still await a more detailed treatment. Thirdly, we decided to limit the transcription of characteristic data to non-standard paratexts and standard paratexts with non-standard characteristics. As explained above, each witness is linked to a work within the database; we then supplemented the work entry with its standard characteristics37, so that standard paratexts with standard characteristics could simply be referred to by the work they are linked to38. The vast majority of these decisions were made in long and (sometimes) animated but always collegial discussions with the ParaTexBib team, whose contribution to our reflection should not be underestimated39. The choices explained above were primarily made for the project about the Gospels, but they were successfully carried over into the Revelation project40, with some minor adjustments. In particular, all the manuscripts were fast-tracked at the beginning of the project, even though it was decided fairly early in the project to concentrate on the Revelation manuscripts until the end of the 12th century. In these descriptions, Revelation paratexts are described with more details than Gospel paratexts41. Asymmetrical Results Our final methodology, including the improvement of the rules during the project, resulted in an asymmetrical ePTB database in terms of the manuscript descriptions, in the following ways: Firstly, four basic types of manuscript descriptions coexist in the database with basically four different levels of detail and accuracy: 36 Concretely, the manuscripts were divided into four categories with the following goals: a) through the end of the ninth century: to describe all the available (and workable) Gospel manuscripts, as mentioned earlier; b) 10th‒12th c.: describe most of them; c) 13th‒14th c.: describe some of them; d) 14th‒15th c.: describe only a select number. 37 For example the title, incipit and desinit as in the reference edition, sometimes with the main options, for example when two or three »recensions« of the work circulate widely. 38 With an indication such as »tit. et inc. = ed.« (title and incipit as in the edition). 39 For the names of the team members, see ‹https://www.manuscripta-biblica.org/team/›, last accessed 30.05.2022. 40 See above, p. 64. 41 For example, work titles are systematically transcribed from the manuscripts; the place and content of the apparatus are given more precisely. In the internal part of the database, more textual information has also been noted. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 76 Patrick Andrist 1. entries with no descriptions, mainly due to the absence of pictures and detailed catalogues entries; 2. ›very short descriptions‹, or ›fast-track‹, with a few bits of information at the CU level, as explained above; 3. ›compact standard descriptions‹, i. e. descriptions with the simplifications which have just been explained; 4. ›developed standard descriptions‹, i. e. descriptions done before the above- mentioned simplifications were put into effect. The difference with compact descriptions has to do with the amount of provided details, not with the structure of the description (the overall description model). Quantitatively, if one focuses on the Gospel manuscripts, more than 2000 descriptions were produced, representing a total of about 2200 entries (for reasons listed below), distributed as following: Type 1. (no description 2. very short descriptions 3. compact standard descriptions 4. developed standard descriptions Total of descriptions (Total Amount ~ 140) ~ 1325 ~ 545 ~180 ~ 2050 ~ 2190) Amount of descriptions of Gospel manuscripts produced, per type. Secondly, the compact standard descriptions (=type 3), which reflect the development of our thinking after two years of works, are also asymmetrical in the depth in which the pieces of content are described: they give many details about the paratexts of the Gospels and are precise about the Gospels, but are much more generic about the other types of content. However, we strove to have no asymmetry in the congruence of the data is concerned, both by having two people check the data against the manuscript reproduction (and discussing until they agree), and flagging places where the information is not certain, for example, when the manuscript or the pictures are not legible. In this case, producing asymmetrical descriptions made it possible to gain an overview of all our relevant manuscripts, and gather important project information about them in a systematic searchable way, without losing sight of the whole content of the codex or getting lost in describing in detail hundreds of pieces of content or physical features that were not (or less) relevant to the project. © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. The »Paratexts of the Bible (ParaTexBib)« project and its database 77 Conclusion: a Successful Project For five years, I had the pleasure of being the driving force behind the creation and development of the ePTB database. With a background as a professional computer programmer and manuscript cataloguer, I very much enjoyed working with our programmers, devising database solutions with my colleagues for situations that were unforeseen, refining the description rules as new situations popped up, and teaching new team members how to use this great tool. This is not the place to present a global assessment of our successful project42. It suffices to say here that, even though we did not reach all of our goals, we successfully managed not only to analyse more than two thousand manuscripts, but also to prepare a series of editions which have begun to appear in the series Manuscripta Biblica43. We believe that the kind of cataloguing methodology and database descriptions we developed is one key to this success. Producing an asymmetrical database with asymmetrical descriptions is, of course, not satisfying as far as the best practices of cataloguing are concerned. Nevertheless, as I mentioned earlier, this layered approach proved successful in gathering relevant manuscript data for the project and sharing it for the most part with interested users, within the time span of the projects and without sacrificing the advantages of a systematic database. One can speak of a complex systematic approach, which, I believe, could also be beneficial to other database projects about manuscripts. I do not mean that the many points which are specific to the Pinakes database or to our project could or should be reproduced by others. However, the ideas of fast-tracking on the one hand, and of producing asymmetrical descriptions in a systematic way on the other, certainly form part of an approach to be considered by people working with other manuscript projects in need of a pragmatic, efficient but still scholarly and sharable database solution, no matter the underlying technology. Of course, some users may be surprised by this heterogeneity, and feel frustrated when they try to retrieve information. This is, however, a new approach to manuscript descriptions, whose potential still needs to be made more broadly available through more adapted tools. I hope to be able to contribute to that goal in the frame of new projects, which are 42 Martin Wallraff and I hope to be able to publish a more global presentation and assessment of the ParaTexBib project, including all the problematic around identifying and publishing the paratexts. 43 Wallraff, Martin: Die Kanontafeln des Euseb von Kaisareia. Untersuchung und kritische Edition. Manuscripta Biblica 1, = Paratextus Biblici I. Berlin 2021; Parpulov, Middle-Byzantine Evangelist Portraits (cf. footnote 4). © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited. 78 Patrick Andrist not limited to New Testaments in Greek, and which could also take advantage of the development of a new generation of manuscript databases, allowing a better study of all the features of manuscripts in general and Bibles in particular44. 44 See Andrist, Patrick: New Tools and Database Models for the Study of the Architecture of Complex Manuscripts. In: Ancient Manuscripts and Virtual Research Environments, ed. by Claire Clivaz and Garrick Allen. Classics@ 18, Washington 2021 (‹https://classics-at.chs.harvard.edu/classics18andrist/›, last accessed 30.05.2022). © Harrassowitz Verlag, Wiesbaden 2022 This PDF file is intended for personal use only. Any direct or indirect electronic publication by the author or by third parties is a copyright infringement and therefore prohibited.