Issue6

[paddymcall]

On Tue, Apr 28 2020, pchiou6932 wrote:

@pchiou6932 commented on this pull request.

-@Book{east:5100,

• date = {0710},
• title = {觀所緣緣論釋 // {Guan} suo yuan yuan lun shi},
• language = {zh-CN},
• series = {Taishō 1624; {Koryŏ} taejanggyŏng 625},
• translator = {{Yi Jing}},
• keywords = {canonical}, -}

I've replaced the text-relation of east: 5100 to Dharmapāla's commentary on ālambanaparīkṣā/vṛtti with east: 5029. But there comes up another problem. https://east.ikga.oeaw.ac.at/buddh/ind/5/177/#1431. On the page, the author name is followed by 710, which is wrong, since 710 is the year of translation, not the year of composition. I think this problem would also come up with other pages for other classical Chinese translations after I supplied the author name.

Since this is a bibliographic record, the date in that position should indicate the date of publication, not the date of composition or translation. So it’s an error in the sense that neither Taishō nor Koryŏ taejanggyŏng was “published” in 710. We’ll have to resolve this kind of issue, which concerns all of our canonical works, at a later point (when we’ve cleaned up the current bibliography). In terms of the bibliography, what you’d want to record is how each work is contained in its respective collection, and then combine those records into a set of information that can be presented as if it were a single entry. Ideally, we’d record at least three dates: date of composition (usually unknown), date of translation, date of “publication” (in the earliest, actually available edition of a canonical collection), plus perhaps a date when the collection was compiled.

[kellner]

A comment, perhaps worthy of being stored elsewhere for future discussion.

Let's separate this analytically: the date of composition belongs to the work, which is an entity in the EAST database. It does not belong to bibliographical records, and it is an open question whether it makes sense to record it at all (conventional assignments might be useful to define some kind of relative chronology, but would be really tricky -- but a worthwhile research endeavour!).

The date of translation is a property of the translation, and "translation" is a relationship between a published e.g. Chinese version of a work and a Sanskrit work recorded in EAST. This relationship is part of the EAST data model, not part of the bibliographical record.

The date which belongs to the bibliographical record is the date of publication -- and if I understand Paddy correctly, this needs to be cleaned up later when records are split up into records for each canonical version. (Here the problem is that the records are currently "contracted" with multiple canonical editions in "series", and each edition has a different publication date.)