Ideas

A few ideas to initiate project

terminology and ids should follow a proper terminology

Gumbert 2004 ? Andrist et al. 2013 ? RBMS ? vHMML lexicon? https://www.vhmml.org/lexicon

Frameworks

1. eXist DB
2. Django
3. both

import and export

• EMML
• CIDOC-CRM

Main views

SEE PROJECT DESCRIPTION

• Manuscript
  • all info in catalogue record
  • links to other entities referred to
• Corpus
  • list of manuscript
  • description of institution
• Person
  • connection to manuscripts
  • places related to person
  • specific info on person
• Place
  • manuscripts related to place
  • people related to place
  • other resources linked to place
• Text
  • in which manuscripts it is contained
  • in which kinds of units
  • to which people it connects
  • map of distribution of versions through time
• Biblio
  • show in which records the bibliographic record is pointed to with type

Maps

Search

• index separate items for complex searches

Relations

I have read the very interesting article by prof. Orlandi and I have thought of the following option.

We currently have the following data entities: Manuscript, Work, Place, Institution, Person. While Manuscript and the manuscript parts eventually contained in it correspond in my understanding to the Codicological Units at different granularity levels, our Work entity can correspond to the Textual Unit at any given granularity. Nevertheless we have nothing for Narrative Units, and perhaps we should have such kind of Entity.

Am I correct to think that a Textual Unit can exist also without material attestation (e.g. a known work of which no manuscript survives) and that a Narrative Unit cannot exist without at least one attested account? What should then this Unit contain as information? a description an origin and the relation to the Texts in which it is used would be sufficient?

In my opinion it still makes sense that the Textual Units, even if abstract, intended as defined in Orlandi's article, are listed under the content of each Codicological Unit, but perhaps this is not enough and gives only part of the information we want to record. On the other side we have a dedicated part of the model for "similar manuscripts" which might be too generic.

We have included in Places and Person records a part of the model to state relations therefore I wonder wether extending this relation function would be an agreeable option.

We could state multiple controlled relations also in the Manuscript as well as in the Work and newly added Narrative entities.

Each entity would contain lists of different relations, where these cannot be inferred otherway from the structure of the entity. For example similar manuscripts could be listed as relations between manuscript. e.g. this MS is copy of this other MS or this MS derives from this other MS. a Narrative Unit could have different types of relations with other Narrative Units, Textual Units and Codicological Units according to the need.

While the body of the manuscript description would be the transcription of that manuscript and the body of a Textual Unit the edition of that work, Narrative Units could be linked directly to one of such texts once transcriptions are given so that we do not need to provide them a different body other than the actual attestations. I imagine an example use case this sort: the life of a saint contains a miracle accounted in a certain way. The Manuscript gets its identifier, as well as the whole text and the identifiable subparts. The same recognizable narrative unit is used for a part of the life of a king with changes and alterations of different types. Then the narrative unit gets an id and is referred by in both texts stating the type or relation. if only one of the manuscripts containig the life of the king contains such text, than the narrative unit might be linked to that directly and not with the abstract text and its edition.

Relation can be stated between any given entity (so also between a codicological unit, even partial, and a person, or between a single word and a person). The relations could then be used to visualize structured but extremely flexible and non hierarchical graphs of the information available which can be as complex as it needs to be. I have made an example with an easy visualization (but entirely made up content!), which you can find attached and open in your browser. You can click on the items and drag them around to change the view. I have put in it made up example items of each of our entities and some relations picked from the ontologies below.

Eventually the question would then turn to the names of the relations. Each relation will need to be defined but I believe a good starting point would be the practice of syriaca.org, which uses SNAP for relations among people. SAWS Ontology, developed for collections of arabic sentences could instead be used as ontology of relations between texts. There are many other ontologies to pick from to form an initial list of defined values, but specific project values can also be defined off course. I believe this are valid starting points, if this altogether is an acceptable option.

http://snap.dighum.kcl.ac.uk/ontology/ http://www.ancientwisdoms.ac.uk/method/ontology/

The information about a given abstract literary work or document could be then given starting from such relations, which do not force us to impose any hierarchy but rather give evidence to any known relation between the entities transparently. For example we could display on a timeline the relations between Works and Narrative units, or the relations between Manuscripts and Works or among manuscripts and so on.

I think that starting with such method from the beginning would be very beneficial. The visualization of such informations will instead have to wait.

Also for documents contained in the manuscript probably the same model could be used. For example an addition to a manuscript containing a donation of land could refer to a Narrative Unit for that type of document, but there would be no need for a further record as that specific occurrence will remain connected to the manuscript where it is found.

da ANDRIST: msPart = stratum describe each fully

production unit = class PU?

circulation unit = class CU?

EVERY STRATUM HAS ITS OWN DATE, NO DATE FOR WHOLE MS

immediate precise data about the main strata, time and place of origin

how many strata the book is made of, their extent, date and place of production

minimal> correct information on the primary and secondary strata and a brief note explaining the situation

quires quire types (collation) ancient sequence marks (foliation) writing support ruling technique layout scribes, hands and writing systems decoration principles and characteristics content esp. where content changes and where are full or partial blank leaves

introduction of the catalogue to summarize decisions taken

give dates in each msPart and as precisely as possibly linking them to the evidence and part they refer to, not to falsely expand the scope of a date

searches including a date include all RELEVANT data and only those (e.g. date list organized by context)

on model of IRCYR schematron make more tests and restrict schema as much as possible

CHECK Syriaca Exist APP and sample files

CHECK OUT WIKI FROM GIT

Check git code from Syriaca.org and contact them for Schema and framework Study how to produce a TEI schema for Beta mesaheft based on guidelines and project decisions