Thanks to Ian Fairclough (who responded off-list), I have one more possibility to add to the list (actually, a variation of 1). Ian suggested recording the vernacular title in a 670 with NCRs, so that the data does not get lost. It avoids creating a reference field (suggestion 1), thus side-stepping the character situation in 1xx/4xx/5xx fields. I will call this possibility 1a:

(1a) Record the vernacular title with NCRs in a 670 field:
670  Islām kā ak̲h̲lāqī nuqt̤ah-yi naz̤ar: ǂb t.p. (title in vernacular, with NCRs: اسلام کا اخلاقی نقطہ نظر)

This example, incidently, illustrates the necessity of having the original-script data in the record. Before 2013, the letter romanized as "z̤" was romanized as "ẓ". Without knowing which version of the romanization table was used, one cannot recover the original-script title.

Richard

On Thu, Dec 17, 2015 at 4:25 PM, Lammert, Richard <[log in to unmask]> wrote:

I see three possibilities in this situation:
(1) Avoid putting the title in the vernacular script into the record at all.
(2) Substitute other Arabic letters for the two Urdu characters, giving:
اسلام کا اخلاقي نقطه نظر
(3) Substitute NCRs for the characters that will not validate, giving:
اسلام کا اخلاق&#x06CC; نقط&#x06C1; نظر
(That mixture of Roman and Arabic characters will likely come out in the wrong order.)

--

Rev. Richard A. Lammert           e-mail: [log in to unmask]
Technical Services Librarian       mail: 6600 N. Clinton St.
Systems Librarian                     Fort Wayne, IN 46825-4916
Walther Library                         phone: 260-452-3148
Concordia Theological Seminary