About the project
Astral sciences were cultivated, since antiquity, in a great variety of historical contexts wherein different types of transmission and interaction were often established. Thus, research into the development of these sciences represents a vast endeavour transcending cultures and eras. In order to try to better understand the interactions and influences between the different astronomical traditions, a sufficiently large and encompassing corpus must therefore be constituted; similarly, tools and methods capable of analysing this great variety of sources must be implemented.
Digital humanities transform the availability of historical sources gradually, along with the means to analyse, edit, and relate them. These changes should be addressed, anticipated, and fostered by various research communities in history. DISHAS (Digital Information System for the History of Astral Sciences) is a collective enterprise that addresses these changes in the field of the history of the astral sciences.
What? Object of study
For centuries across Eurasia, the astral sciences were fostered, manuscripts were compiled and copied, instruments designed and early printed books composed to meet a wide range of religious, ritualistic and political needs, to make calendars, to predict the future astrologically, and to observe and understand the natural world. Practitioners of the astral sciences, their written sources, and their astronomical ideas circulated among local cultures. Sources and ideas were appropriated and transformed by a long and far-ranging transmission history that connects a broad corpus.
DISHAS aims at providing tools to the community to edit and analyse the different types of sources usually treated in the history of the astral sciences.
Texts and genres of text, diagrams, table formats, models and parameters, have, at least from the eighth century onward, a shared history in Eurasia for sources found in Arabic, Persian, Hebrew, Sanskrit and Chinese corpora. The temporal and geographical dimensions of the project reflect an essential aspect of the history of the astral sciences, which are distinguished by their long-term transmission across diverse, and frequently cosmopolitan, local contexts. In the long run, DISHAS aims, in collaboration with partner projects, at providing tools to the community to edit and analyse the different types of sources usually treated in the history of the astral sciences: tables, texts and diagrams. As a pilot attempt, DISHAS focuses on astronomical tables and, more precisely, on their numerical and mathematical content.
A focus on tables
While texts and diagrams are resources used broadly in many different intellectual disciplines, numerical tables are vastly used in, and quite characteristic of, the astral sciences. Astronomical tables share certain qualities with one another, especially with respect to their numerical content, which make them easier to distinguish, align and compare than do texts or diagrams. Thus, at least since the pioneering critical edition of al-Battānī's tables engaged by Nallino and Schiaparelli at the beginning of the 20th century, historians of the astral sciences have developed tools to study and edit them. In the last 50 years, approaches to the numerical content of astronomical tables with computers and modern mathematical tools have been developed. The recent explosion of data analysis is also promising in this area.
Astronomical tables circulated among many cultures, and were appropriated and transformed by a great diversity of actors. Thus, the numerical data conveyed in these tables provide rich evidence for ancient scientific practices. For example, in tables we can recover the complexity and sheer mass of numerical computations which were handled in the ancient world and which thus contributed to the history of computational mathematics. Tables also reveal how astral phenomena were modelled, and how reasoning and prediction were shaped. As written documents of a special type, which fall somewhere between computation per se and data storage, astronomical tables expose epistemic writing practices in their layouts on the page and their frequent (re)configuration into “sets” of tables. Created by complex computation with often-interlinked algorithms, circulating tables generally were adapted to new contexts and purposes rather than recomputed from scratch. Thus in addition to their individual contents, astronomical tables, viewed more generally, can provide unmatched sources for studying the transmission of computational know-how, writing technologies and layouts, theoretical models, and numerical parameters. With enhanced digital, editorial and analytical tools, scholars will be able to chart previously unrecognized paths of circulation, learn how large collections of tables were shaped, and track the spread and appropriation of particular computational practices.
Tables can provide unmatched sources for studying the transmission of computational know-how, theoretical models and numerical parameters.
DISHAS will offer a platform to the scholarly community where this corpus can be assembled, edited and analysed in order to trace transmission on a Eurasian scale, and will restore numerical tables as the computation tools they often were for historical actors. The simple and ambitious objective of DISHAS with respect to numerical tables, leads to an in-depth reflection on the modelling and structuring of data as well as on the central practices of the history of the astral sciences, e.g. around the notions of critical editing, traditions, schools, transmission. It also implies to specify the type of numerical we consider and our approach to them.
Tables, as non-discursive elements of scientific texts, are of many different kinds and formats. Most tables in the astral sciences can be described as “numerical tables” because the core information they convey is expressed with numbers. They can be approached in many different ways. DISHAS relies on specific choices adapted to its central scientific objectives with respect to the analysis of the numerical content of astronomical tables.
DISHAS aims to propose a platform where a Eurasian-scale corpus can be assembled, edited and analysed in order to trace transmission on a new level.
Numerical tables can be presented in many different formats. The grid type of format is probably the most familiar to us now, but numerical tables could be presented in prose or verse for instance. DISHAS is interested in the numerical content of those numerical tables whatever their format. Thus graphical aspects of tables, in this first stage of development of DISHAS, are of less interest to us than their mathematical aspect. Even more specifically, we have begun by focusing on those numerical tables that instantiate a mathematical relation between two (or more) quantities in the form of what can be described in modern mathematical terms as a discrete function.
We have favoured this more mathematically oriented view of astronomical tables for a variety of different reasons, and thus we will limit ourselves to commenting on our main considerations here. First, DISHAS is designed to study the transmission history of astronomical tables. This requires a comparison of table contents across different traditions of templating tables which were supported by different modes of writing and media. A model focused on the mathematical aspect of astronomical tables produces objects where these layout differences are secondary, and offers the opportunity to look beyond these differences. A second aim of DISHAS is to propose different kinds of mathematical exploration and tools of analysis, including statistical methods of squeezing astronomical parameters from table content, for which only the numbers are taken into account. Finally, we also want DISHAS to restore tables as the computation tools they were for historical actors, and thus it is necessary for us to make the table content easily retrievable.
Who? A collaboration of projects
DISHAS is a collective undertaking fostered and coordinated mainly at PSL-Paris Observatory in the team of history of astronomy of the SYRTE laboratory (CNRS UMR 8630). It relies on a network of international projects covering Chinese, Sanskrit, Arabic, Latin and Hebrew corpora. The most important of these partner projects are:
ALFA, Alfonsine astronomy shaping a European scientific scene (ERC CoG 723085, 2017-2022, PI Matthieu Husson);
PAL, Ptolemaeus Arabus et Latinus (Bavarian Academy of Sciences and Humanities, PI Dag Hasse);
HAMSI, History of Astronomical and Mathematical Sciences in India (University of Canterbury, PI Clemency Montelle)
DISHAS meets the common needs of these different projects by critically editing and analysing sources in the astral sciences. Together, these initiatives have pooled their expertise on specific domains of sources, tools and resources, to develop DISHAS.
DISHAS is a collective undertaking fostered and coordinated at PSL-Paris Observatory by the team of history of astronomy.
DISHAS is also supported directly by SYRTE, PSL and the DIM Sciences du texte, connaissances nouvelles, especially with respect to the engineering team and servers needed to develop this digital humanities platform.
The project development is regularly reported to the broader community of the history of astral science through the Commission for the History of Ancient and Medieval Astronomy (CHAMA), a committee that aims to promote study in the field of the history of the astral sciences on a global scale.
The DISHAS platform is the result of an interdisciplinary process where researchers of different research projects, engineers and data designers from the Paris Observatory collaborate to conceive of the mains component of the platform and their articulation. From this collaboration has emerged the data model, the edition and analysis tools, and the publication tools.
The platform itself has been primarily developed at the Paris Observatory by the technical team of ALFA and the DIM. Several significant elements, however, were developed, either as proof of concept or in their implemented form, by partner projects. For instance, HAMSI developed CATE, a tool for the automatic generation of critical apparatus for the edition of astronomical tables. CATE is now embedded in DISHAS. The PAL project is also collaborating in fundamental ways by developing a tool enabling DISHAS to manage historical calendars as they are used in the sources, and by providing the proof of concept for programs useful for the statistical analysis of astronomical tables.
How? Ambition and achievements
In the long run, DISHAS will be an open data repository for the study of mathematical astronomy sources from different periods on a Eurasian scale where research data from different projects (primarily, but not limited to, ALFA, PAL and HAMSI) can be retrieved and analysed with common and state-of-the-art mathematical and editorial tools. We expect that this will foster research in the history of astronomy, and produce new questions and results in the future.
DISHAS seeks to become an open repository for the study of sources in astronomy where research data from different projects can be retrieved and analysed with common mathematical and editorial tools.
Today, DISHAS is offering a research environment focused on astronomical tables where multiple steps of the research process can be performed and benefit from the use of state-of-the-art tools, and, as the data set grows, of insights produced by the power of the surrounding data set. These research areas include:
Analysis of the primary sources with a refined description of the tabular content of manuscripts or early printed material;
Production of transcriptions, critical editions and recalculated astronomical tables;
Astronomical and mathematical analysis of tabular content in order to study underlying models and parameters;
Publication of the research results in a powerful public interface.
Every step of the way, from the primary sources to the publication of research results, DISHAS will be able to assist researchers, and help them to frame their work within a wider historical context. DISHAS is also a collaborative tool - intellectual property of each edition is documented and public - while possibilities of interaction between contributors are implemented in the administration interface.
In order to achieve this, DISHAS is composed of different layers, which we will describe briefly below. These are, namely, a database, a set of edition and analysis tools, and a public interface to provide access to these resources.
DISHAS is fundamentally interested in the numerical content of astronomical tables. Our fundamental interest led us to define tables in terms of the data they contain. Conceived in this way, the "table as data" constitutes the numerical values themselves. For calculation purposes, a digital table contains both the quantity as read in the text (i.e. using historical conventions to write numbers) and their floating number translation. A single “table as a data” is linked to a series of contextual metadata in three main areas of description: historical, astronomical and editorial.
First, historical metadata qualifies the physical witness (manuscript, author…), and the intellectual tradition (Works). The central unit, the table, is described historically as a Table Witness, understood as a given table instance written in a particular manuscript. This level of granularity is fundamental to critical editions resting on the comparison of manuscript witness, and anchors the study of transmission in manuscript evidence.
The database will gather and render inter-operable sources from various traditions and periods.
Next, we store astronomical and mathematical metadata. Read from the table or calculated from its values, the “parameter sets” attached to the table are a common list of astronomical values that are shared, notwithstanding the historical context. This sharing will help us to describe the geographical and historical dispersion of astronomical parameters based on their source tables.
Finally, we store authoring and editorial information. Indeed, the “table as data” is never considered the strict expression of a given historical document; rather, it is the reading of a specific author for a given purpose! Defining different types of editions implied the creation of a norm for transcription and common practices.
Overall the modeling effort related to the conception of this database has offered our community an occasion to reflect deeply upon its practices. Included among these elements of reflection are questions touching on the complexities of manuscript sources; the definitions of such terms as ‘table,’ ‘work,’ ‘author’ and ‘tradition;’ the making of table transcriptions and critical editions; and the different ways to describe the mathematical and astronomical contents of tables. This fundamental work, accomplished in close partnership with the DH team of the project, has already produced clarifications and revisions in our practices, which are essential to producing a solid ground for future collaboration on a broader scale. In fact, the database itself is more than just a convenient way to store research data. Its smart design allows it to become more and more powerful in guiding research as its dataset is growing. It is a place where erudition and expertise can truly become a collective rather than individual quality.
Edition and analysis tools
DISHAS proposes many different embedded tools to assist all stages of edition and analysis of astronomical tables. The two most important ones are:
DTI: DISHAS Table Interface;
DIPS: DISHAS Interactive Parameter Squeezer.
The first allows the user to produce editions of table witnesses; the second guides users in the process of squeezing the model and underlying parameters from the table content. We describe them briefly below.
DISHAS proposes many different embedded tools to assist all stages of edition and analysis of astronomical tables.
DTI assists users in the production of digital editions of astronomical table content. Various types of grids for one or two argument tables can be specified along with the symmetries of the table, the type of numbers used for argument(s) and entry, their unit, the presence of a difference table etc. While entering the numbers themselves multiple tools will help users work efficiently by allowing different levels of commentary, the generation of automatic suggestions of readings based on previously entered values or production, with CATE, of critical apparatus according to edition strategies specified by the users.
Some of these features are already producing an analysis and understanding of the table content. DIPS, however, provides opportunities for a deeper analysis of the astronomical and mathematical structure of the table. DIPS is guiding users in a study of the table that will help squeeze from its numerical values the underlying models and parameters. This multiple=stage process involves the identification of the table’s outliers, the potential existence of interpolation grids and, finally, the extraction of parameters. For each of these key steps of analysis, state-of-the-art tools are provided in a user-friendly graphical environment. The information and documentation provided by DIPS also allows the user to evaluate the quality of the statistical analysis produced at each step and the level of confidence he can reasonably attribute to the obtained result.
Finally, DISHAS is also a publication platform, offering tools to search, visualize and manipulate this data. The public interface of DISHAS offers three main features:
The possibility of exploring table editions;
The possibility of discovering data that describes the context of production for these tables;
Access to resources for the study of astronomical tables.
The difficulty of pursuing paper-based editions of astronomical tables is a well-known issue in the field of the history of the astral sciences. Numerical tables in grid format already saturate the page with information, and it becomes very difficult for historians to include with clarity the different types of critical apparatus needed for a scientific edition. DISHAS gives a framework for new attempts to compose editions based on digital readings. This will provide an enhanced access to research data and results.
DISHAS gives a framework for new attempts to compose editions based on digital readings, thus providing an enhanced access to research data and results.
More generally the public interface of DISHAS is offering multiple means to explore and visualise the database. First, a search engine is included where users can quickly find specific information. Two main paths of data exploration are defined. The historical navigation will give insights about the intellectual content, the milieu of productions, the actors, the material aspects of the artifact. The astronomical navigation will allow users to browse the database through the lens of the astronomical objects, their related types of tables, models and parameters. In both cases, dynamic representations of objects display their distribution in time and space, within primary sources or works.
Finally the public interface is also a resources portal. It provides ample documentation on DISHAS itself, as well as additional resources for pursuing the analysis of astronomical tables beyond those aspects principally addressed by DISHAS.