Challenges for ISMI CIDOC model
Text vs witness vs codex
In the ISMI data model the TEXT (or title) is an abstract, ideal text: an intellectual concept formulated as a sequence of words.
The WITNESS (or manuscript) is a concrete, singular exemplar: a bunch of folios covered with writing. The WITNESS is an exemplar of a TEXT (WITNESS is_exemplar_of TEXT).
The TEXT has an author (TEXT was_created_by PERSON). The WITNESS has a copyist (WITNESS was_copied_by PERSON).
The WITNESS is bound (possibly together with other WITNESSes) in a CODEX (WITNESS is_part_of CODEX). The CODEX is part of a COLLECTION (CODEX is_part_of COLLECTION). The COLLECTION is part of a library/REPOSITORY (COLLECTION is_part_of REPOSITORY). The REPOSITORY is in a city (REPOSITORY is_in PLACE).
The usual identification of a codex "Staatsbibliothek Berlin, Sprenger 1855" is spread over multiple objects: PLACE("Berlin"), REPOSITORY("Staatsbibliothek"), COLLECTION("Sprenger"), CODEX("1855").
In the course of historical research a TEXT may have been erroneously attributed to have been written by a certain author when later research shows that the text was actually written by another person. For historical consistency it is important to record and display both the correct authorship attribution and the misattribution.
For example the Text "al-Kitāb al-Qiwāmī fī al-ḥisāb al-hindī" (TEXT 110211) that was written by Samawʾal (PERSON 24831) has been misattributed to Maḥmūd al-Jaġmīnī (PERSON 7437).
In the course of historical research a WITNESS may have been erroneously identified to be a certain text written by a certain author when later research shows that the witness was actually another text written by another person. For historical consistency it is important to record and display both the correct text identification and the misidentification.