Abbreviating transliterated Arabic first names that start with ʿ (ayn) or ʾ (hamza)

edited February 16, 2018
Using apa.csl, Zotero/citeproc-js/LO renders the name [al-Aswānī] [ʿAlāʾ] as “al-Aswānī, ʿAlāʾ”.

pandoc-citeproc, on the other hand, outputs “al-Aswānī, ʿ”.

The same effect is seen with hamza at the start of the first name, with both programs.

I feel neither of these is correct: citeproc-js does not abbreviate at all, and pandoc-citeproc treats ayn and hamza like regular letters (ayn is one in Arabic itself, but both ayn and hamza are commonly treated as diacritics in transliterations; see, e.g., Hedden 2007).

Arguably, either “al-Aswānī, ʿA.” or “al-Aswānī, A.” should be preferred.

Any insights, from the resident APA experts, for example?

Hedden, H. (2007). Arabic names. The Indexer, 25(3), 9C–15C. Retrieved from
  • Apparently these characters are used in several languages. U+02bf transliterates Arabic ayn and also Hebrew ayin. U+02be transliterates Arabic hamza, Hebrew alef, and Armenian apostrophe. It would be good to first confirm the use cases for the other languages. If necessary, behavior can be bound specifically to Arabic items (or, in Juris-M, Arabic fields).

    A second issue would be the potential use of U+0060 (back-tick/grave accent) and U+0027 (close single quote/apostrophe) in these roles. If that occurs in data, we would need to step carefully to avoid conflicts with tussenvoegsel and the like. Again, potential conflicts could be avoided by binding the transform to language, but we should know we need that before stepping into the coding.
Sign In or Register to comment.