#freebassel

Bassel Khartabil / باسل خرطبيل

• Syrian software engineer
• leading advocate for open access and Creative Commons (CC) in Syria
• author of the Arabic CC licences
• detained by the regime since 15 March 2012
• moved to an unknown location and probably killed in Oct 2015

The journal al-Muqtabas between Shamela.ws, HathiTrust, and GitHub: producing open, collaborative, and fully-referencable digital editions of early Arabic periodicals—with almost no funds

Project URL: https://www.github.com/tillgrallert/digital-muqtabas

Slides: https://tillgrallert.github.io/slides/Dixit2

Twitter: @tillgrallert

Email:

1.1 Importance of mundane texts / periodicals

• They are at the core of various discourses
  • Modernity / -ism at the end of empire
  • Arabic renaissance
  • Arab nationalism
  • Islamic reform movement
• They form large corpora with an equal distribution along a temporal axis (al-Muqtabas: 12 yrs, al-Manār: 43 yrs, al-Muqtaṭaf: 76 yrs)
  • linguistic analysis
  • historical semantics
  • data sets for social history

1.2 A two-fold problem

• Preservation:
  • Active destruction of cultural artifacts: iconoclasm, neoliberalism
  • Neglect: fragile materiality
• Access:
  • Absence / destruction of infrastructure / channels of knowledge transmission: lack of access to institutions, hardware, software, internet connections
  • widely-dispersed collections
  • technologies: absence of reliable OCR
  • technical skills: lack of basic scripting skills

The consequence is a focus on “high” culture and canonical texts

1.3 State of digitisation

1. gray online libraries / “crowd”-sourced transcriptions, e.g. al-Maktaba al-Shāmila, Mishkāt, Ṣayd al-Fawāʾid, al-Waraq etc.
2. Digital imagery, e.g. Endangered Archives Programme (EAP), HathiTrust

1.3.1 state of digitisation: text

gray online libraries / “crowd”-sourced transcriptions, e.g. al-Maktaba al-Shāmila, Mishkāt, Ṣayd al-Fawāʾid, al-Waraq etc.

• lack of / faulty metadata
• unknown editing principles
• unknown quality
• very limited structural mark-up
• cannot be reliably cited

1.3.1 state of digitisation: text

al-Muqtabas on al-Maktaba al-Shāmila

1.3.2 state of digitisation: images

Digital imagery, e.g. Endangered Archives Programme (EAP), HathiTrust

• lack of metadata
• limited licences, paywalls
• no or very bad text layers

1.3.2 state of digitisation: images

al-Muqtabas 6 on EAP

1.3.2 state of digitisation: images

al-Muqtabas 6 on HathiTrust without US IP

1.3.2 state of digitisation: images

al-Muqtabas 6 on HathiTrust with US IP

1.3.2 state of digitisation: images

al-Muqtabas 6 on HathiTrust, state of OCR (only visible to US IPs)

2. Suggested solution: unite facsimile and transcription

1. aims
   • validate the transcription against the facsimiles
   • improve the transcription with the help of the “crowd”
   • make everything citable for scholars, linkable for machines
   • provide the new edition with the broadest possible licence to facilitate access and re-use
2. principles
   • re-purpose available and established tools, technologies, and material
   • preference for open and simple formats and tools

3. Test case: digital Muqtabas

al-Muqtabas / المقتبس

• “monthly” journal published by Muḥammad Kurd ʿAlī between 1906 and 1918/19 in Cairo and, from 1908 onwards, in Damascus.
  • 9 volumes, 96 issues (at least 2 double issues), c. 7000 pages
• Muḥammad Kurd ʿAlī (1876-1952): Ottoman bureaucrat, journalist, president of the Syrian Academy of Sciences, minister of education.
• availability at c. 30 libraries (North America, Europe, Middle East):
  • original prints (mostly incomplete)
  • some copies of a “gray” reprint
  • a number of microfiche copies from the same source

3. Test case: digital Muqtabas

1. Basis: Generate and share a TEI edition of all 96 issues (c. 7000 pages) of Muḥammad Kurd ʿAlī’s Majallat al-Muqtabas with a CC BY-SA 4.0 licence
2. Core feature: gradually improve the digital edition (text and mark-up)
3. Sugar on top:
   • Static web-view (doesn’t require a permanent internet connection)
   • access through bibliographic metadata in public Zotero group

3. Test case: digital Muqtabas

Project scheme

3.1 Basis: Generate the TEI edition

• crawl the digital text from shamela.ws
• transform it into TEI XML: semi-automatically (mostly XSLT)
  • documented URI design to reference all elements
• add structural mark-up (sections, articles, heads, authors …): semi-automatically
• link to digital facsimiles from the British Library’s “Endangered Archives Programme” (EAP) and HathiTrust or produce your own imagery: semi-automatically (mostly manually)
• host everything but images on GitHub
  • distributed version control
  • attribution of authorship
• provide a CC BY-SA 4.0 licence for all files: edition, tools, webview

3.1 Basis: Is this legal?

3.1 Basis: Is this legal?

Copyright depends on the jurisdiction of creators, distributors, etc.

1. text of al-Muqtabas
   • is in the public domain: transcription is legal
   • the transcribers do not / cannot claim copyright: copying is legal
2. images of al-Muqtabas
   • digital files are protected by copyright: use is subject to licence, linking is legal
   • download and redistribution: almost certainly illegal
3. digital edition of al-Muqtabas
   • all our work is licenced as CC BY-SA 4.0 licence

3.2 Core feature: Continuous improvement

3.2 Core feature: Continuous improvement

1. Improvements depending on human labour (probably a “crowd”)
   • correct the transcription
   • add structural mark-up
   • add semantic mark-up
2. Automatic improvements:
   • provide reliable bibliographic metadata based on the facsimile
   • mark-up of natural entities with link to external reference files (e.g. personal names, toponyms)

3.2 Core feature: how to contribute

• go to GitHub and register a free account
• fork the the edition’s repository: https://tinyurl.com/muqtabas
• edit the text
• send us a pull request
• changes will be reviewed and merged

3.3 Sugar on top: web-view

Display of al-Muqtabas 6(2)

3.3 Sugar on top: web-view

• Adaptation of TEI Boilerplate XSLT stylesheets
• human-readable and static web-view (either rawgit or gh-pages)
  • generated on-the-fly by the user’s browser using XSLT to transform the TEI XML files.
  • can be run without an internet connection and with local facsimiles.
• parallel display of text and facsimile
  • simple changes to display different facsimiles
• link to metadata on the article level (BibTeX)
• the code is shared with a CC BY-SA 4.0 licence on GitHub

3.3 Sugar on top: Zotero group

Zotero group “digital-muqtabas”: list view

3.3 Sugar on top: Zotero group

Zotero group “digital-muqtabas”: item view

4. To do, ongoing work

• Editorial decisions: TEI schema design
  • mark-up of some text features has not yet been decided
• Editorial work:
  • mark-up of page breaks
  • correcting transcriptions
  • add non-Arabic words omitted by shamela.ws
  • add footnotes
  • correct publication dates for all issues.
• Web-display:
  • needs some polishing
  • search functions beyond the Zotero group and individual issues (project can be searched on GitHub)

5. Experiences: simple, fast, sustainable

• Simple technologies and relatively little coding needed: Initial set-up took less than four weeks of after-hour labour
• Hosting with GitHub is free
• Core (but simple) features cannot be automated: all c.7000 page breaks must be manually tagged
• Code can be re-purposed:
  • We set-up the sister project Digital Ḥaqāʾiq as a digital edition of ʿAbd al-Qādir al-Iskandarānī’s monthly journal al-Ḥaqāʾiq (1910–12, Damascus) in a single day.

Summary

• open scholarly digital editions of [Majallat] al-Muqtabas and al-Ḥaqāʾiq providing
  • TEI XML files (transcription and links to facsimiles)
  • plain text files
  • BibTeX files for every article
  • customised version of TEI Boilerplate (XSLT and CSS) with stable URLs for every element
• within a framework (git and GitHub) that allows for
  • collaborative, open, version-controlled improvements of the edition
  • re-use of the text

Project URL: https://www.github.com/tillgrallert/digital-muqtabas

Slides: https://tillgrallert.github.io/slides/Dixit2

Twitter: @tillgrallert

Email: